*** UPDATED JAN 2021 ***
Updated to include views and opinions on the newly launched Tri-Spark MOSFET reg/rec.
Table of Contents
This article is broken down into sections – you can skip to them by clicking below:
- More Power Back Then
- Factory Standard Charging
- Shunting Regulator/Rectifier
- Short-Type SCR Regulator/Rectifier
- Short-Type MOSFET Regulator/Rectifier
- Open-Type (also known as Series-Type) SCR Regulator/Rectifier
How a charging system on a bike is specced out and works seems to be a really emotive topic – certainly on a par with wetsumping and engine breathing.
There are a lot of strong opinions out there – unfortunately many of them are incorrect as they have been formed over many years based on assumption, hearsay and sales spiel.
For the record, I am not a salesman – I am just a techie – I trained as an electro-mechanical engineer, before moving into electronics and communications and finally settling in computing.
I am a member of the IEEE (the Institute of Electrical and Electronics Engineers, who are the world’s largest professional association for the advancement of technology) and the IET (the Institute of Engineering and Technology who were formed when the Institute of Electrical Engineers were merged with the Institution of Incorporated Engineers in 2006). Importantly, I guess, I can drive a soldering iron and diagnose an issue rather than look it up on YouTube and declare I’m an expert. I’m not an expert, I’m learning every day!
I have never made a cent from the time I spend helping other people or from the hundreds of hours I spend documenting and drawing diagrams for our community and for individuals.
I do not sell products and am not linked to any manufacturer – any recommendation I make is based on my own research, findings and experience.
What you will find in this article is based on information I have obtained from several manufacturers, their specs, their documentation and discussions I have had with their staff. Sadly, some are no longer with us, such is the way with this industry.
Where I do share my personal opinion, it is based on my own findings either on a bike, on the bench or through helping others – so I would like to think I am quite well informed.
I have obtained information from the following, and pass on my sincere thanks for putting up with my incessant questioning:
- Lucas (Wassell)
- Lucas (LAP)
- Boyer Bransden
- Podtronics (rest in peace Bob Kizer who supplied me with a LOT of info)
In addition, I have used data from the following:
- Electrex World
- Rick’s Motorsport Electrics
- Rex’s Speedshop
- Lucas (the original 60s and 70s service books)
- Cycle Electric Inc
Hopefully these sources will help put together an interesting article, that clarifies some points.
This article sets out to provide some information around regulator/rectifier options for British bikes – where possible I have tried to break things down to make them simpler to understand, and explain terminology.
There is no information here that I haven’t presented or discussed at length across the various groups or forums before, but I guess this is the first time I have put it all in one place.
We are discussing charging systems that are built around permanent magnet alternators, so this does not cover alternators with field wound rotors or bikes that use a dynamo.
This also does not cover more modern (often Japanese) bikes where many are either using field wound rotors (so can reduce the alternator output depending on the demand) or are switching in and out additional alternator stator windings based on the power demand (just like Lucas used to do with their ET alternators on the light switch back in the early days)
The article is quite lengthy, but my guess is that most people will want to skip to the diagrams.
My drawings of regulator/rectifiers shows how the regulator and rectifier circuits work and how the components are configured to perform those functions.
The ‘control circuit’ that I reference in some diagrams will vary from one manufacturer to another, and also over different periods of time. This can range from a simple zener diode (in the case of something like a Tympanium, to quite an advanced integrated circuit in the case of the latest and greatest MOSFET designs.
Different manufacturers will use slightly different chips, some will use resistors and capacitors, others have RFI chokes to help with interference (some don’t but maybe should have) and some even have thermal protection circuits integrated.
In addition, some manufacturers offer units with a larger capacitor integrated into the unit, to allow a battery to be eliminated from a bike if you wish.
More Power Back Then
Most brit bikes of the era used the Lucas RM21 10-amp single phase alternator stator.
The majority of coil ignition bikes of the time consumed about the same amount of power, so this was a good fit in most scenarios – I will cover power, load and consumption in more detail further in to this article.
However, when bikes with Electric Starters started popping up, it was clear that a more powerful charging system was required.
It is worth mentioning a couple of higher power systems from the time:
The MK3 Commando had electric start, and with the starter solenoid using 3 amps and the Prestolite starter motor peaking at 250 amps, it meant that the charging system was in desperate need of uprating, otherwise you’d never fully recharge your depleted battery during the course of your ride.
So, the new and improved RM23 16-amp single phase stator was fitted (which Norton said in the factory manual produces 15 amps at 6,000rpm, but was 14.4 amps according to Lucas)
This was an interesting implementation, as Lucas were already running the zener diode very, very close to it’s maximum rating (100 watts) which meant that they needed to do something different for the MK3
The solution was to downgrade the rectifier to a half wave one LU49181 (2DV406)
They then used two zener diodes on the AC side to bring rectification up to full wave, but also to bring the voltage back down to something that wouldn’t boil the battery.
The benefit of doing it this way is that they didn’t have to perfectly match the two zeners, and that the regulation could be split between both zener diodes knowing that both would be used – the zeners were mounted on both z-plates to help with heat dissipation.
The drawback of this was the fact that although the alternator output was higher, the voltage was in fact a little lower than the older RM21 10-amp system it was replacing – 0.7 volts lower in fact.
Here is the diagram from the Norton MK3 Commando workshop manual which illustrates how the components are arranged:
Many people assume that because the MK3 Commando has two zener diodes, it is a three-phase charging system.
That’s a myth – it’s still just a single-phase system.
The next system to note is the Triumph T140E. In 1979 they introduced a slew of updates – including major changes to the charging system.
This was a little after the MK3 Commando, and on the grounds they were both under the same management at this point, I would suggest they took some of the lessons learnt with the Commando, and applied them to the Triumph.
The T140E used a three-phase alternator stator the 10.5-amp RM24 coupled to a three-phase rectifier the LU83539
Because they chose the lower output stator, it meant they could stick with one zener diode.
The T140E was negative earth though, so neither the rectifier nor the zener diode were useable on positive earth bikes.
Here is the diagram from the Triumph T140E workshop manual which illustrates how the components were arranged:
The final system to cover is the Lucas Powerbase kit.
This was an aftermarket kit put together by Lucas and Mistral Engineering and was sold as an upgrade often alongside the excellent RITA electronic ignition system (which craved a solid, stable and consistent 12 volts)
The Powerbase was a three-phase system, built around the high output version of the RM24 stator and they proudly offered 180 watts of power and 85% of full output being produced at only 2,400 rpm in the sales blurb.
The kit was supplied with a matched pair of Zener diodes – matched by hand at the factory to be within 0.1 volts of each other. There was a need for all the cables in the install to be exactly the same length and gauge (same resistance in the wires on both sides) so that both zeners would work equally.
Unfortunately, as the zeners burned in to service, there was often a slight discrepancy between the two, meaning that one zener would often take most of the load.
Also, with owners installing their own kits, they often shortened the cables to keep things tidy, not knowing they were affecting the way the kit operated. As a consequence, the kit was fairly short lived.
Here is the diagram from the Lucas Powerbase kit which illustrates how the components are arranged:
Factory Standard Charging
The factory standard charging system on a Norton Commando (and most other brit bikes of the time) was built around the 10-amp Lucas RM21 stator and consisted of a separate Zener diode (the regulator, that brings the voltage down to a safe level for charging the battery) and Rectifier (the bit that converts AC to DC).
Zener diode Lucas Part Number LU49345:
Silicon full wave bridge rectifier Lucas Part Number LU49072 (2DS506):
Contrary to rumour, Lucas stopped using selenium rectifiers in around 1964 – so they never made their way on to the Commando.
The old selenium rectifier Lucas Part Number LU47132B is very recognisable by it’s square fins.
This is a googled the pic, as I don’t have one of these:
Thanks to Classic bike electrician and all round guru Al Osborn at AO Services for keeping me honest and correcting me with this information, as I had posted the wrong picture previously.
It is the job of the rectifier to convert the AC output from the Alternator stator into DC.
Here is the AC output:
And here is the rectified DC output:
A battery or a capacitor or (both) are then used to smooth some of the peaks out – to give the bike a nice clean, stable supply of power.
The problem with the alternator being attached directly to the engine’s crankshaft is that the output will vary massively according to how many rpms the engine is doing – if this output was not regulated in some way, your lamps would blow, and your battery would boil.
So, the job of the zener diode is to make sure that the DC voltage that comes out of the rectifier does not get dangerously high.
The zener diode starts clipping at around 14.3 volts – it does this by dumping anything over that setpoint to ground.
The zener is attached to the bike’s z-plate – a nice lump of aluminium stuck out in the flow of air, so it acts as a superb heatsink.
Lucas were cutting it fine with the specs of the zener – it is rated at handling 100 watts. The standard RM21 alternator is rated at 120 watts (they always based their nominals on 12 volts) so the zener doesn’t have any spare capacity to handle an uprated alternator stator.
Here is what the rectifier and zener look like as a simple circuit diagram:
You should note that the diagram shows the rectifier and zener connection to ground.
Although on a Commando, the positive feed off to these components is technically by the red wire that is part of the harness, both are attached electrically to ground by their mounting bolts.
This means, you cannot use them if you change the polarity on the bike from the standard positive earth to negative earth.
This standard charging system is absolutely superb and actually very robust. The only real thing to watch out for is on the rectifier with the soldered joints on the diodes between the cooling fin ‘wafers’ – these joints can eventually shake themselves loose with vibration.
This can be easily repaired with a soldering iron, or alternatively you can fit a modern rectifier which has been encapsulated in resin – these are robust and are much better at resisting vibration.
These modern rectifier units are less than five dollars to buy and work perfectly!
In the mid-seventies and early eighties, a popular aftermarket “upgrade” was to swap the standard rectifier and zener for a combined regulator/rectifier.
The very early shunting reg/recs (and even now still the very low-cost ones) are basically functionally a direct equivalent of the original rectifier and zener setup. And there is nothing wrong with this – they work totally perfectly for most people and will provide many years of trouble-free service. Just don’t upgrade your stator to a higher output one, or introduce LED lighting. For good measure you may consider riding with your headlight on, as that continual 3-4 amp current draw acts as a great place for any excess charge to go.
These units are nothing more than a pair of thyristors that are clipping the AC input from the alternator at around 14 volts. The rectifier is then converting the clipped AC supply into DC.
In this example, the ‘control circuit’ that I reference elsewhere in my diagrams is nothing more than a zener diode – it is effectively switching the rest of the unit on or off depending on the voltage in the battery.
However, the DC output was very poor quality and choppy indeed, so over the years the subsequent designs, redesigns and iterations have been refined a little – although you still see these used extensively on lawn mowers, and small tractors to this day.
The simplicity of these is very attractive, and it is the nearest you’ll get to the original rectifier and zener setup.
You should note that a battery and/or a capacitor is an absolute must, as the output needs a lot of smoothing and conditioning.
It is also worth noting that many of these are polarity sensitive – i.e., they are grounded to their heatsinking enclosure electrically as well as thermally, so can only be used for negative earth (if the flylead is red) or positive earth (if the flylead is black) This is particialrly the case for the agricultural ones (by Agricultural, I am referring to the fact they were first used on tractors and farm/plant equipment), so be careful if you are repurposing one for use on your bike.
Also watch out – some of the agricultural ones are rectifiers only – they don’thave regulation built in. Confusingly, they look the same as a reg/rec, so please make sure you know what you’ve got before you fit it.
This type of unit would probably not be recommended (in my opinion) for sensitive electronics like some of the digital electronic ignitions and electronic speedos/tachos.
Short-Type SCR Regulator/Rectifier
SCR is silicon-controlled rectifier, and it refers to the type of technology used under the covers.
In fact, the original Lucas rectifier on the bike is based on the same tech, thus it is well proven, reliable and low cost.
The most common and most sold units these days are short-type regulator/rectifiers.
These are often identified as Shunt reg/recs, but that is a confusing and incorrect term to use because actually every type of unit available (including the factory standard zener) shunts to a certain extent, and as you’ll find when you read on, how these work is very different to the basic functionality covered in the shunting type described above.
I feel that it is good practice to refrain from using “shunt” when describing a reg/rec type to prevent further confusion.
A short-type regulator/rectifier works in the following way:
- The AC output from the alternator stator is rectified (converted) to DC
- When the battery reaches it’s target charge voltage, the alternator stator input legs are dead-shorted by the control circuit, effectively closing the gate to the introduction of more incoming power.
- Any remaining over voltage (i.e., if the DC voltage of the battery is a little over the target voltage setpoint) is then dumped out as heat.
The reason the AC input legs are dead-shorted is mainly down to keeping heat out of the reg/rec.
To be clear, at no point are the AC input legs shorted to ground – they are shorted to each other. The alternator stator winding still remain electrically separated from ground.
Think about the original setup:
- the zener and the rectifier are kept separate and away from each other
- the zener is attached to the z-plate which is acting to wick the heat away from the component
- the zener is positive earth only, so the mounting stud will always be connected to the positive feed – this can be used to dump excess voltage and for heat transfer
Now compare this to the new combined reg/rec:
- everything is in one enclosure
- they are (most commonly) manufactured to be dual polarity – meaning that the same unit can be connected to a positive OR negative earth bike. So the manufacturer cannot rely on electrically connecting the components to the heatsinking enclosure for grounding
- the gauge of the wires used would not take a single high amperage current dump
- although most instructions specify that these should be installed in cool airflow, most manufacturers are aware and accept that these will be hidden out of sight in an enclosed area with zero airflow
I feel it is really important to understand this mode of operation, and what is actually going on here.
So, with the factory original charging system, when the target voltage of the battery is met, the alternator is working like this:
The whole output from the alternator, independent of the engine rpm is being converted from AC to DC by the rectifier.
If that DC is not required because the battery is fully charged and at it’s target voltage, the excess power is dumped to ground via the zener diode. This means that the alternator is constantly working at it’s full potential, and the load is outside of the alternator stator.
The load (and hence the heat) is kept away from the stator windings, even though they are constantly running at full load.
When a short-type regulator/rectifier is used, the alternator will be working more like this when the target voltage of the battery is met:
The control circuit inside the reg/rec dead shorts the AC input, which is effectively exactly the same as crossing the two wires on the alternator stator together (or three wires, if you have swapped your stator from the standard single-phase model to a three-phase one). As a reminder, this is not shorted to ground (it still remains floating).
At this point, the alternator stator becomes it’s own load. The gauge of the stator windings and their interlinking wires are considerably lighter weight than the gauge of the wiring outside the unit, so it is the stator that gets hot, because it is dissipating the excess power as heat.
If the battery voltage is then still a little higher than the target voltage, the control circuit can dump any of that overvoltage out to ground (this is where it wrongly gets the name shunting reg/rec from) – it is this constant clipping action that makes the thyristors (the components that are used instead of zener diodes in these units) get hot – the better quality regulator/rectifiers use a thermal paste to attach these to the aluminium heatsink casing, prior to the whole thing being encapsulated in a decent quality thermally conductive epoxy resin. This will act to protect the unit against vibration, as well as assist with thermal transfer.
The cheaper, poorer quality units available from eBay usually cut corners in this area, as it is an obvious place to bring the cost down. Not using thermal paste at all, and using a lower cost potting compound like an acrylic or polyurethane are popular.
The analogy I often use here is:
Put a small lightbulb across the terminals of a 9-volt battery – the battery will remain cool to the touch and the bulb lights up and gets hot. The lamp is the load.
Now put a table knife across the terminals of the same battery – the battery will get hot and the knife will stay cool. The battery is the load.
Running dead-shorted, is not the same as running under full load – another mistake that is often made. These alternators are designed to run at full load, and provided the power has somewhere to go (an external load) it is no problem at all. However, when dead-shorted, the stator windings become the load – which is not a good thing, and not what the alternator stator was designed to handle.
This methodology is absolutely fine, and in normal circumstances no big deal – it was how a short-type regulator/rectifier was designed to operate and there are hundreds and thousands of units out there operating in this exact way with no issue at all.
The reason it is fine is because of the way the standard charging system was originally designed.
Let’s look at the power consumers on the bike:
|1||Ignition (running) – approximate||2.5|
|2||Main Beam (60 watt or 45 watt)||5.0 or 3.8|
|3||Dip Beam (55 watt or 40 watt)||4.6 or 3.3|
|4||Pilot Bulb (5 watt)||0.4|
|5||Tail Light Bulb (5 watt)||0.4|
|6||Speedo Bulb (3 watt)||0.3|
|7||Tacho Bulb (3 watt)||0.3|
|8||Ignition Warning Light + Assimilator||0.1|
|9||Main Beam Warning Light (¼ watt)||–|
|10||Turn Signal Warning Light (¼ watt)||–|
|11||Turn Signal Bulb pair (21 watt each)||3.5|
|12||Stop Lamp Bulb (21 watt)||1.8|
Now let’s consider that the standard RM21 alternator stator that is fitted to the bike is rated at 10 amps.
The 10 amps it outputs is at 6,667 rpm.
At engine idle (1,000 rpm) the alternator stator is putting out 1.5 amps.
With a wasted spark electronic ignition system, two 6-volt coils wired in series, and doing away with the condensers and ballast resistor, you are consuming a little more than the points ignition as quoted in the above table. In reality, it is nearer to 3 amps.
All this means you are actually consuming more than you are producing at idle (this is the reason the red warning lamp flickers on and off at engine idle). This is by design, and is a good thing – something that people struggle to get their head around.
During the course of your ride, you are busy recharging a depleted battery, which means there is no need at all for the short-type regulator/rectifier to be dead-shorting your alternator stator.
Happy days, and working as designed – over the course of your journey, the power you have produced roughly equates to the power you have consumed, leaving you with a battery that’s in a decent state of charge, and ready for your next outing.
The reg/rec was designed as a direct replacement for the zener and rectifier on the bike – and if you change nothing else, there is no technical reason at all why it won’t provide you with years of good, reliable service.
Here is a simple circuit diagram of a single phase short-type SCR regulator/rectifier:
And here is a circuit diagram of a three-phase short-type SCR regulator/rectifier:
Where the short-type regulator/rectifier can become an issue on a bike is when you mess with the balance of the charging system.
An example is when you fit a more powerful alternator stator to your bike.
Lucas have available the RM23 high output stator – which is rated at 16 amps (compared to the original RM21 at 10 amps)
You are now producing an additional 6 amps of power, but are not consuming any more than you were previously.
This means that your battery will be charged significantly faster, and the short-type regulator/rectifier will dead-short the AC input legs when that target battery charge voltage is met.
With the battery charged sooner into your ride, you’ll now be spending more of your ride passing more current through your stator windings compared to the original setup.
This can result in the temperature of the stator racing up, and failure occurring – this most commonly can be seen as the resin around the stator cracking out, and eventually breaking down.
A second scenario is when you fit LED lighting to your bike.
LEDs consume significantly less power than their filament lamp counterparts, which means once again you are upsetting the balance of the charging system.
Instead of consuming around 6.5 amps across your headlamp, tail light and instruments when you turn your lights on, with a full complement of LEDs onboard, you could be consuming less than 1 amp.
So much less depletion of stored battery energy, meaning that once again, your short-type reg/rec will be spending most of it’s time dead-shorting the AC input legs.
Unfortunately, I frequently see that as part of a restoration or upgrade project, a bike owner will follow the advice of a salesman, or the manufacturer’s websites and install a new short-type regulator/rectifier in lieu of the factory original rectifier and zener diode, and they will add a new high output 16-amp stator and they will upgrade all their lighting to LEDs.
It is at this point, I often see cracked out stators, bikes no longer charging and in some extreme occasions I have seen people reporting melted resin too – one in particular was a pretty catastrophic foul up inside the primary case (although I have a theory this is down to physical contact between rotor and stator).
Don’t forget that inside the primary case is not very well ventilated, runs really hot in there, and there is a splash of oil to lube the primary chain, but not enough to cool anything down. It’s a pretty hostile place for anything electronic.
I am really keen that people understand that bigger does not always mean better – despite what we are told by the manufacturers and the salespeople looking to sell their latest and greatest kit.
The charging system on a bike should be balanced – take out around the same amount of power that you are putting back in over the course of your ride. Making changes to produce more, consume less or both will upset that balance and cause problems.
Ride with your headlight on all the time – as I said with the shunt-type reg/rec that extra 3-4 amp current draw over and above the igintion electronics acts as a great place for any excess charge to go during a ride will make sure that you are constantly recharging a depleted battery. Your alternator stator will run a lot cooler, as will the regulator/rectifier itself.
Short-Type MOSFET Regulator/Rectifier
These are relatively new to the world of classic bikes, however have been available on Japanese sports bikes and the like for many years already.
MOSFET is an acronym, and it stands for Metal Oxide Semiconductor Field Effect Transistor – it is several decades more modern than SCR technology, so in comparison is considered state of the art.
The bit that is of interest to us is that MOSFETs switch much, much more quickly than SCRs – which means the quality of the output is much better.
Here is an example of the DC output of a MOSFET regulator/rectifier:
As a matter of interest, compare that to the DC output of an SCR-based regulator/rectifier:
You can see that the MOSFET-based regulator/rectifier is switched several times during each cycle (i.e., each revolution of the crankshaft) which means the output is a lot cleaner and more precise in contrast to it’s SCR counterpart.
These reg/recs were designed for modern bikes that have a lot more electronics onboard – complex electronic fuel injection systems with high pressure fuel rails, and a multitude of sensors monitoring engine temperature, inlet and exhaust gases, water and air temperature/pressure etc…
Each sensor is basically a point of analogue to digital conversion, so can be prone to radio frequency interference (RFI) and electromagnetic interference (EMI), plus these low voltage electronic systems require a very stable and consistent power supply for maximum reliability.
The MOSFET-based reg/rec mitigates against a lot of the issues that were linked to the older SCR-based technology and they are being fitted as standard on many of the newer Japanese bikes.
It is well worth noting that the MOSFET regulator/rectifiers that are currently on the market are all short-type.
That is to say, they deal with a fully-charged battery that has met the target voltage in exactly the same manner as the short-type SCR reg/recs discussed above.
The AC input legs from the alternator stator are dead-shorted by the control circuit and any over-voltage caught the other side of the gate (i.e., at the battery) is shunted to ground.
However, because the MOSFET reg/rec is switching so much more frequently, the reg/rec itself runs significantly cooler, a really nice benefit of using this technology, since we are all well aware that these are often not installed in the cooling airflow that the manufacturers ask for in the install guides!
In the dead-shorted state though, the alternator stator will still be getting overly hot, and exactly the same applies with regard to making sure your charging system is balanced.
Make sure that you produce as much power as you consume, and regulate it yourself by using your headlight as a load.
Here is a simple circuit diagram of a three-phase short-type MOSFET regulator/rectifier:
To this date, I have not come across any manufacturers making a single phase MOSFET reg/rec – therefore I have not drawn one.
Most manufacturers say you can use a single-phase alternator stator with a three-phase reg/rec with no problems.
Open-Type (also known as Series-Type) SCR Regulator/Rectifier
Like MOSFET, this is the new kid on the block for classic bikes, although the technology itself is certainly not new – it has been used in aeronautic applications for many years.
Some manufacturers refer to these as series-type reg/recs, but I personally try to avoid using that term, as I feel it can confuse people.
(It’s the same reason as I try to avoid describing short-type SCR reg/recs as Shunt reg/recs as discussed above, some terminology can make things unnecessarily confusing for others, and I feel it is my role to try and help people get their heads around it all)
Open-type reg/recs have been adopted on some adventure bikes, ATVs, jet bikes, snowmobiles etc. – vehicles that may have a lot of additional electrical load like big banks of lights, heated clothing, additional pumps and heaters. The reason for this is that these vehicle types will typically have bigger batteries and larger alternators to cope with the additional electrical load. However, this additional load may not always be required – no heated clothing and no big spotlights needed on a hot sunny day, could easily equate to needing 300 watts less power, and dumping that back through the alternator stator windings like a short-type reg/rec does would very quickly cause damage.
The most common way to deal with wide range of loads is to use a different type of alternator – one with a field-wound rotor instead of permanent magnets. That way, you can easily dial down the amount of power that the alternator actually produces, taking away the issue of trying to deal with excess power. However, this type of alternator is significantly less efficient than a permanent magnet alternator (especially when you factor in the compact dimensions that we need for our bikes), and will also put significantly more load on your engine (so less engine power gets to the rear wheel).
So, the open-type regulator/rectifier differs from the short-type one in the way it handle things when the target voltage of the battery is met.
Instead of the control circuit dead-shorting the input legs of the alternator stator, it actually opens the circuit (or disconnects the alternator stator input legs completely.
This totally takes the load off the alternator stator, allowing it to run a lot cooler, without the potential of causing damage to the windings.
A complete tangent, but there is also the fact that at 4,000rpm the effect of taking all load off your alternator will free up around half a horsepower at the rear wheel. That’s nearly one percent on a 750cc Commando (based on the 58bhp the Workshop Manual quotes and depending on whose numbers you are working to), but nevertheless it’s still worth having!
Norton engineering powerhouse Jim Comstock runs a non-standard alternator on his Commando as he has some additional, power-hungry electronics on his bike:
- data logging (from when he was experimenting with his fuel injection setup)
- heated vest and gloves
- 3 HID headlamps
Jim’s alternator is an Excelsior-Henderson external alternator rotor (no longer in production) and a homemade stator – the spec is 38 amp and 450 watts.
He runs the Compu-Fire open-type reg/rec and very interestingly saw a gain of around 1 ½ horsepower at the rear wheel on his dyno.
For me, this absolutely qualifies the claims I have read about in the past, which is why I am including it in this article.
If you take the rule of thumb permanent magnet alternator efficiency of 50%, Jim’s alternator would be taking 900 watts out of his engine.
Using the accepted figure of 745 watt equals 1 horepower, Jim’s dyno observation is about right.
There is also the matter of factoring in power loss through the drive train, which I haven’t done here, as it is far too complex for my little brain – but I have a feeling that if it were factored in, Jim’s numbers would be spot on.
I think there is also the matter of what you are comparing, as to what the horsepower gain actually is.
I have done some testing with the bike’s electrics running on a big, external, fully charged battery which meant I could rule out the variable of bad electrics affecting the ignition system from the equation.
As an example, when I was testing how hot an alternator stator gets when running dead shorted, I found that with the two AC legs of an RM21 stator dead shorted, it was nigh on impossible for me to kickstart the bike. The resistance it introduced was massive. I liken the feeling to having too long bolts holding the primary inner chanincase on, and it obstructing the flywheel (don’t ask me how I know)
If I set the idle to 1,000rpm with the stator wires disconnected, if found that when I shorted the stator windings with the bike running, I could very easily stall the bike each time. Don’t forget, the alternator isn’t electrically connected to anything on the bike so the stall was being caused by the braking effect introduced by shorting the stator.
If you compare this behaviour to that of an open-type reg/rec, then you can certainly appreciate where the numbers are coming from, as you can actually and physically feel the difference.
The open-type reg/rec still uses the archaic SCR technology, so is slow switching and gives a choppy output versus the MOSFET design, which means it is by no means the holy grail of regulator/rectifiers on the market.
However, for me personally, I feel it is the best of a bad bunch and in terms of being kind to the other components on the bike, it is my personal choice, what I use on our own projects, and what I feel comfortable recommending to others.
To date, I have never had an open-type reg/rec fail on our bikes, or any one that I’ve assisted over the years – plus it certainly puts an end to alternator-related issues which for me makes it the weapon of choice.
Here is a simple circuit diagram of a single phase open-type SCR regulator/rectifier:
And here is a circuit diagram of a three-phase open-type SCR regulator/rectifier:
Utopia would undoubtedly be an Open-Type MOSFET regulator/rectifier, as it would offer the best of both worlds. Watch this space, as sooner or later, I have no doubt that one will become available.
But they are not here yet, which means you have to make a choice between nice, clean and precise power or kind to your alternator.
There are a huge number of manufacturers in the market, so I am just going to pick out a few of the more well-known or notable ones and share my opinion.
You should note that many of the manufacturers have a few different models available. In particular, both Boyer Bransden and PODtronics offer units that are marketed as battery eliminator, batteryless or battery-free. These include an internal capacitor, which means you can run the bike without a battery.
If they are not marketed with this capability, it is safe to assume that the functionality is not there, and a battery MUST be used. This is important – the battery serves as a load, which makes the reg/rec work as intended. Also, the circuitry inside a reg/rec is driven by DC – they are a power consumer just like a headlamp, so they need to get their power from somewhere, be that a battery or a capacitor.
The other point to note is the compatibility of a reg/rec with the Warning Light Assimilator (WLA) also known as the Ignition Warning Light/Headlamp Warning Unit.
Whilst most units these days are protected against reversed polarity (i.e. accidentally connecting to the wrong battery terminals) every unit I have come across to date will not tolerate any of the AC input legs being grounded. Doing so simply fries the reg/rec.
The warning light assimilator, because it is monitoring the AC output directly from the alternator stator can have a similar effect to grounding an AC input leg. Boyer Bransden explicitly say in their fitting instructions “If your Norton has a charging warning light simulator it must not be used with the power box, remove it” for example.
Boyer sell a different model of PowerBox that includes it’s own circuit for the charge light, however I would recommend that you use a seperate Charge Warning Light instead, as it will provide you with a lot of useful information like state of charge, battery state, reg/rec failure etc…
The Podtronics units are short-type SCR regulator/rectifiers
The “POD” in Podtronics stands for Prince Of Darkness and was a quip on what some see as Lucas’ poor reliability by the founder Bob Kizer (an opinion I don’t share)
Bob is sadly no longer with us, but he was responsible for designing and making what has undoubtedly become the most popular reg/rec in the classic bike world – he was a great (and funny) guy, who kindly provided me with a lot of his time. This is also the most imitated reg/rec.
In Bob’s design of the Podtronics reg/rec his primary goal was that he “wanted to dumb it down so that any idiot could fit one” (his words)
Always bear this statement in mind, and consider that this design was not necessarily technically the best solution.
The Podtronics was designed in the late seventies/early eighties to directly replace the existing Lucas rectifier and zener diode on the bike.
The intention was for it to be fitted as is, using the existing alternator stator, with no other changes – no high-power stator upgrade and no LED lamp conversion.
In this scenario, the Podtronics works absolutely perfectly, and there are many thousands out there that prove it. Riding with your headlight on, as mentioned already, helps things even further!
I have noticed in recent years that there has been a seemingly higher failure rate of Podtronics units than there used to be – more inline with standard figures. Several years ago, production of the units was outsourced to Taiwan, and I cannot help but feel the higher number of DOA (dead on arrival) and ELF (early life failures) tallies with the change in production.
Although, most of the charging issues I help people with are running Podtronics, I must remind everyone that this is because there are far more Podtronics units out there – they have been sold since the eighties, and they are the most common, most well known and biggest selling brand.
Bob Kizer used to make them in relatively small batches, and test them as he went – I think the quality was higher back in those times – industry average for classic motorcycle electronics failure rate is about 2%, and I know that Bob prided himself on a much lower figure.
I have also noticed a ‘rationalisation’ in the product line – the single-phase unit is now identical to the three-phase one, it just has the third yellow wire cut off.
I do wonder if this is what the documented issue with the Tri-Spark electronic ignition is all about, since it apparently only manifests on the single-phase units – just thinking out loud on my side, but if only one phase is ever used on a three-phase rectifier is it electrically noisy?
For those that haven’t heard about the Tri-Spark issue, you can read more about this topic by clicking here:
The final thing I have noticed is that some Podtronics units I have been testing seem to hold a voltage internally over and above most others – these have not been labelled as battery eliminator “POD-1P-MAX” models, so this should not be the case. Definitely some inconsistency there.
The great John Healy from Coventry Spares Ltd in Middleboro, Massachusetts took over Podtronics from Bob Kizer some years ago, so I am hopeful that in due course we will see a resurgence in quality of these units again.
The Boyer Bransden units are short-type SCR regulator/rectifiers
Boyer are well known for their electronic ignitions; however, they have also manufactured the Power Box reg/rec for many years.
Just like Podtronics, they make a high-power variant, single-phase and three-phase plus a battery eliminator version to enable you to go batteryless if you wish.
A few years ago, these units didn’t have reverse polarity protection, and I saw a few people connect them up wrong, causing irreparable damage to the reg/rec units. To my knowledge, that issue has now been resolved.
The other thing to be aware of is that in all of their documentation, Boyer point out that the reg/rec units must not be used in conjunction with the Norton Commando warning light assimilator.
They point you toward their reg/rec that has a charging light – however, I would advise you use a battery status monitor or charge warning light from the likes of SparkBright or ICM instead.
I don’t see or hear much about Tympanium these days, but back in the day they were very common indeed!
Once again, they are short-type SCR regulator/rectifiers – however there are a couple of things to watch out for.
Some Tympaniums are shunt reg/recs – the most basic of the types available.
Tympanium made units for lawnmowers and small tractors and some were rectifiers only.
Some were regulators only.
Some were grounded to their heat sinking enclosure, so have only a red or black lead depending on whether they are negative earth or positive earth.
All of these different types look more or less the same, as they are in similar looking casings, so make sure that you know what you’ve got before you use it!!!
Shindengen are a massive manufacturer, and happen to be the manufacturers I side with.
They make short-type MOSFET, short-type SCR and open-type SCR reg/recs.
The FH020AA is their MOSFET model and it is used extensively on Yamaha, Honda and Triumph sports bikes – they have gained a very strong reputation for reliability over the many years that they have been available. It is worth noting that this reg/rec is a short-type unit. It still dead shorts your alternator stator when the battery setpoint voltage is met. A lot of people are of the opinion that MOSFET doesn’t behave in this way (and the marketers sure don’t help with their confusing messaging).
That is not the case – they are still short-type.
The SH775 and SH847 are their open-type models (30-amp and 50-amp rated) – the SH775 is my preference, and I have used and recommended them many times. The 50-amp version is of course total overkill for a Norton Commando, however sometimes that maybe the only model you’ll be able to get your hands on. It works absolutely fine, but is very slightly larger in size (don’t forget, that is it’s maximum rating – the actual output is governed by the size of the alternator stator)
The issue with the SH775 and SH847 is supply chain related. The designs and the rights of these models is wholly owned by Polaris, who make ATVs, snowmobile and watercraft. So, the units are only distributed as Polaris spares to Polaris parts distributors.
This makes it difficult (and expensive) to pick them up for use on classic bikes.
A good source in the US is Jack Flemming at Roadster Cycle – he has struck up a good relationship with a Polaris parts distributor and usually keeps a pretty good stock.
Here in Europe, MTP Racing are a good source.
One thing to watch out for is the huge number of counterfeit Shindengen units on the market – if you are paying less than $100, you can safely assume the unit is a fake one.
Jack Flemming at Roadster Cycle has done a great little YouTube video where he shows you what to look out for, and how to make sure you have the genuine article.
You can find it by clicking here.
Cycle Electric Inc
Cycle Electric is a US manufacturer that make a very high quality open-type regulator/rectifier (which they are keen to market as series-type)
These are aimed at the Harley Davidson market, so are available in all sorts of different shapes and sizes to blend in on a Harley.
Some are a little on the large side, which is great for cooling, however can be a challenge to hide them away somewhere inconspicuous.
Another make of open-type regulator/rectifiers is Compu-Fire
The Compu-Fire 55402 comes highly recommended, however one word of warning:
This would be suitable only for a negative earth application, since it is grounded through the heatsinking enclosure.
RMSTATOR are another company worthy of mention – I’m not sure whether these guys actually manufacture their own units, or if they source and white label others.
However, they have a huge selection to choose from – open-type SCR as well as short-type MOSFET (also traditional short-type SCR, if after this article you still choose to go up that route)
Rick’s Motorsport Electrics
These guys are quite prolific at the moment – lots of advertising and social media activity.
Rick’s Motorsport Electrics have introduced a regulator/rectifier that has been tweaked for lithium-based batteries.
Their voltage setpoint is 14 volts ± 0.2
It is nice to see someone has picked up on the fact that lithium-based battery needs to be treated differently to lead acid ones.
After all, with a petrol tank by our nuts, a battery under our butts, and a red-hot engine making sparks between the two, we need to be careful!
The final manufacturer I wanted to mention in this space are Tri-Spark
Tri-Spark used to sell and recommend Podtronics, however, when they saw quality issues and more recently the issue of their Tri-Spark Classic Twin electronic ignition units misfiring between 3,000 and 4,000rpm, they stopped selling and recommending them.
Tri-Spark are now selling a short-type MOSFET regulator/rectifier which has been very well received, and I am pleased to see it is becoming available through the likes of Andover Norton, RGM Norton and Rex’s Speedshop (as well as directly from Tri-Spark’s website)
I have always liked Tri-Spark – they are by far my preference on the electronic ignition side, and I think they get an unfair beating on the reliability side (as I mentioned in a previous article)
It is awesome to see Stephen Kelly (the owner of Tri-Spark) keep a keen ear on the community and innovate based on market demands and passionate discussions.
He has brought a new ignition model to market – the FireBox Pro which is a quality piece of kit:
- a separate electronics box to satisfy those that don’t like the idea of having everything in the heat of the points cover (although it has never been an issue for me)
- programmable ignition curves from your PC
- either wasted spark or separate triggers for each cylinder
- great cable management and strain relief
- plus retaining the anti-kickback and idle stabilisation benefits of the Classic Twin model.
He has introduced oil-filled coils – much better quality, better cooling and improved longevity over the resin potted ones.
And now, he has introduced his short-type MOSFET regulator/rectifier too.
Recently, a video of the new Tri-Spark reg/rec was posted on YouTube
You can find it here:
The video is a comparison of the new Tri-Spark short-type MOSFET regulator/rectifier “versus” a Podtronics short-type SCR regulator/rectifier.
It is very interesting, and clearly shows the superb, clean quality output of the MOSFET unit – indeed a major advantage of it’s counterparts.
However, in the interest of being fair, and trying to be impartial (which is hard, as I am a big fan of Tri-Spark) there are a couple of things I would like to point out:
- Although a load bank is used in the video to simulate the power consumers on the bike (lights, ignition etc.) there is no mention of whether or not a battery is used in the test. I don’t know about the new Tri-Spark, but the Podtronics MUST be connected to a battery, or a large capacitor at the very least, otherwise the output will be erratic. Podtronics have always detailed this requirement in their installation instructions.
- The RPM of the motor that the alternator is connected to in the test probably goes up to no more than 3,000rpm – it would be interesting to see higher speeds, as this single-phase alternator’s output is rated at 6,667 rpm.
- Most bikes will be using a 10-amp stator, not the 16-amp high-power one used in the video.
- We are very interested in temperature, and of course the MOSFET unit runs a lot cooler than the SCR one – a major benefit in terms of where the unit will be hidden, and probable lack of airflow. However, for me, the more important temperature reading would be one around the alternator stator, not the reg/rec – particularly with the reg/rec attached to a fully charged battery.
Thumbs up to Tri-Spark though, this is a major step in the right direction and as I said before, I commend their innovation.
So, there you have it – my overview, findings and opinions about charging systems on our beloved bikes.
I have tried to put as much data in here as I could, and as many diagrams and pictures to help explain things as possible.
There are two sister articles, which may also be of use to you:
- Alternator Issues – where I dig into how an alternator can be damaged, how to spot an issue and what it is usually attributed to.
- Choosing a New Alternator – if you are set on buying a new alternator, this will walk you through some considerations and hopefully help you make the right purchasing decision.
Hopefully this article has helped squash some of the rumours and bust some of the myths out there, while making it easier for people to understand how things work and what you should be thinking about when planning to change anything in and around your bike’s charging system.
Let me know your thoughts and of course reach out if you need any help!
This piece is part of a series that covers the charging system in general.
Hopefully these will fill in some of the gaps and questions you may have, but as always don’t hesitate to reach out if you need any help or advice.
Categories: motorcycles, Tech Articles
Thank you for this great article.
Far too much inaccurate/ misleading information on the web for classic bike owner do it yourselfer lay persons.
Petrol tanks, nuts, butts, sparks and hot engines…….we must be crazy:)
Just curious…what is your bike and ignition/ charging set up
Hi Des, glad to hear the article was of some interest to you!
We have a few “keeper” bikes here:
Modern BT-H magneto (it is actually a CDI self-generating ignition)
Lucas LU47244 3 phase alternator
Shindengen SH775 open type reg/rec
Dommie 650ss Cafe Racer
Tri-Spark Classic Twin electronic ignition
Lucas LU47244 3 phase alternator
Shindengen SH775 open type reg/rec
Tri-Spark Classic Twin electronic ignition
Lucas LU47244 3 phase alternator
Shindengen SH775 open-type reg/rec
Slimline Featherbed Commando 750 (Combat)
Power Arc programmable electronic ignition (I will probably change this to the Tri-Spark FireBox Pro in the future)
Lucas LU47244 3 phase alternator
Shindengen SH775 open-type reg/rec
This is an excellent article; probably the best I’ve yet come across! However, I do have questions regarding your apparent disdain for LED lighting. My first observation is that many older bikes were never fitted with indicators and by fitting LED variants, there would be very little additional load. Secondly, the light emitted from a correctly manufactured LED headlight is noticeably brighter than the vast majority of incandescent bulb type headlights, while drawing about a tenth of their power and I agree that were you to be riding your bike for significant periods at over 6,000 rpm, you would not be using all of the supplied power, these days the older bikes do not tend to be ridden so hard. I have a Triumph Daytona T100R which rarely gets above 4,500 rpm, even when I’m travelling long distances at say 60-70mph on dual carriageway roads! When travelling locally and around town it’s usually around 3,000rpm! Travelling at these speeds, particularly in town, I find that even my 45w bulb will drain the battery quite quickly and have considered fitting a purpose made LED headlight (not an LED bulb) to relieve the draw! One problem is that the indicators (fitted standard) will stop flashing at low speed in traffic and I’d like to try to prevent this happening.
So, my question is: is it possible or practical to incorporate SOME LED lighting into a classic bike’s electrical system? I am currently restoring a 1970, BSA A65L, from the bottom up and would like to fit an LED headlight. If practical, I’ll stick to the standard RM 21 alternator and a 10-12 amp/hour battery. I won’t be using LEDs apart from the headlight and indicators – all other bulbs will be as standard. I may, occasionally, ride fast, up to 70-80mph, but for the most part will stick to 50-60, or town riding at 20-30. I will be fitting a 21 tooth drive sprocket, which will contribute to slightly slower engine speeds. Am I doing anything of which you would not approve?
Thanks for a brilliant set of articles.
I certainly don’t have any disdain for LEDs on my side – I love them, have them all round the house, in the workshop, in the car and on the project bike I’m building at the moment.
What I do have a problem with though is so called “electrical experts” selling high output alternators, PODtronics reg/recs and a box full of LED lamps to people that are doing a resto job on their bikes.
It’s bad advice, and in the case of a short-type reg/rec (which is the most common type) it’s technically a really poor fit for our bikes.
Furthermore, in the case of LED turn signals, people are often not warned that their factory standard Lucas two-wire flasher unit is not suitable, and that the warning lamp in the headlight shell won’t work in the same way.
I totally agree with your point that LED headlamps often have a superior light output to an incandescent lamp (provided they are coupled up to the correct type of reflector).
However, a 60 watt incandescent lamp is a great switchable load to have at your fingertips – and an important aspect of managing your charging on the bike.
Without this, you lose a superb way of burning off excess power, which can be a major issue.
Couple this up to transplanting in a high output alternator instead of a standard 10 amp one, and upgrading to a PODtronics reg/rec because everyone has told you that the standard zener diode is rubbish, then people wonder why their new alternator melts after only a few hundred miles.
The one that was there before was working perfectly for 50 years.
So for me, it’s just the lack of education I don’t approve of, certainly not the technology itself, which I embrace and use wherever I can.
So far, I’ve probably read these articles five or six times, because there’s so much information. Can’t believe how much effort you’ve put into it! Thanks.
I’m still putting the A65 back together and will shortly be starting to wire it up in a completely non-standard way and also with negative earth.
One of the things I find most interesting/attractive, is the idea of an open reg/rec. Am I correct in thinking this means that when the battery is at full capacity, output from the stator is completely shut off? Is this therefore a safety feature that will prevent the battery ever boiling over? If so, then were I to install an LED headlight, I would be secure in the knowledge that on a long run, my charging system will not try to produce too much power and damage my battery!
I’m in the UK and try as I might I cannot even find information on which reg/recs are of the open type. Can you suggest a make/model and supplier?
If I can’t get an open type, I’m inclined to go for a solid state negative earth rectifier and zener diode and may have to drop the idea of an LED headlight as I do intend taking some long trips, although I have recently learned that my stator is kaput (wires broken) and has to be replaced and I may go for the 10.5 amp RM24.
Kindest regards again,
This sorted a lot of things for me, correcting some previously held beliefs and confirming others.
Very well written and easy to understand, thank you!
Great to hear from you Glen, and thanks for your kind words!
After reading your article I’m thinking of using the Shindengen 775 on a Norton 920 project. In addition to the stator benefit, the free 1.5 rwhp @ 4000 is a nice bonus.
If I have it right, 1 HP is equal to 746 watts of electrical power. 1.5 rwhp equates to about 1100+ watts rear wheel or 1280 watts at crankshaft.
Is this large number due to the inefficiency of the alternator or are my numbers off?
I can certainly recommend the SH775 from a reliability standpoint – I now use and recommend them exclusively, and have never had one fail.
The downside here in the UK is that they can be a real pain in the butt to get hold of.
I tend to order from the USA, as the UK distributors never, ever have stock.
I don’t know if it is the placebo effect or not, but I do feel that when I fit a SH775 to a bike, it feel more free revving – like a bike or car feels when you take weight out of the flywheel.
It’s not something I talk about much, as I don’t know for sure that it makes a difference in this way – I am perfectly willing to accept that this is “all in my head”
However, on the numbers side – Jim Comstock installed a Compu-Fire open-type (AKA series) reg/rec on his bike and ran it on his dyno.
The downside of these particular units is they are negative earth only, so not suitable for a Commando in standard trim.
Jim saw an additional 1 ½ hp at the back wheel – although do bear in mind that he was running a different alternator to the standard Lucas offering.
For me, this does reinforce the claims though and gives them far more credibility.
The permanent magnet alternators on our bikes are actually pretty efficient in comparison to the alternative – the field-wound alternator.
The general rule of thumb is that a permanent magnet alternator can put out 1 ½ times more power than a field-wound alternator of the same size.
As power is not used to energise a field, and there are no brushes, they are generally declared to be more efficient.
In terms of actual efficiency i.e., how much of the energy put in is gotten out, the generally accepted rates are permanent magnet alternators are 50% efficient versus field-wound alternators which are 30% efficient.
Using the published figure of 745 watts to one horsepower, I once tried to get my head around it all.
But being totally honest with you, I couldn’t make it all add up.
I could only assume at the time that I was not correctly factoring in any kind of loss via the drive train.
Great article Grant, has been a huge help working on my commando mk3, converting to lithium and tri spark r&r + cwl
Happy some of the info here has been of use!
Keep an eye on the reg/rec – it’s charge threshold is at 14.5 volts, which is perhaps a little higher than i’d personally like to see for a lithium-based battery.
I’d be much more comfortable with something a bit nearer to 14 volts.
However, using a decent battery status monitor or charge warning light from the likes of these guys will make sure that you keep everything in check.
Thanks for your input to the Commando Facebook page after I posted the question about LED headlights. I have a modern voltage regulator fitted so I could be in for trouble if I fit the LED headlight and do nothing else?
I think I will go back to the original system I had earlier. I have a good Zener Diode. I will get a rectifier locally to replace the one I have. It is probably fine but the modern equivalent looks a better item. Is this one I can buy in NZ OK? https://www.jaycar.co.nz/400v-35a-bridge-rectifier/p/ZR1324?pos=4&queryId=4b1d000999ad52edb8b1522f4bc3a74d&sort=relevance
Do I still need the capacitor? I still have that component too so it can also be refitted.
If I go back to 100% original, the lower load problem created by using LED bulbs should be alleviated?
Really appreciate your help. Like most British bike owners, electrics is still a bit hazy for me. Your article about rectifiers and regulators has been a great help in my grasping the basics!
That looks like it will work just fine!
The main course of failure of the zener diode is when they get dirty.
Keep it nice and clean, and I see no reason why it would be good for another fifty years!
…don’t hesitate to reach out if you get stuck, I’m here to help.
Hello, I have a T140es which is fitted with as standard a high out put alternator.
Which reg/ret would you recommend. These bikes are fitted with the rm24 alternator and three zenors .
I may have the wrong bike, but I was under the impression that the T140ES had the three phase RM24 stator, three phase 3DS5 rectifier, but still only one zener diode.
If I am not mistaken, this is the model that offered the Lucas RITA electronic ignition as standard circa 1979?
Other than the soldered joints inside the rectifier shaking themselves apart with vibration, this is actually one of the best and most robust charging systems of the lot.
A good idea is to replace the 3DS5 rectifier with a modern equivalent like the 36MT60 which should cost you under 15 quid.
The connections are the same as the original 3DS5, and technically it is an exact equivalent. It is just potted in resin so is more robust in terms of vibrations.
I did a short article about it here, if it is any interest to you:
The nice thing about these bikes, is they took a lot of the mistakes of the MK3 Commando, and sorted them for the T140 – so lots of good bits on there.
One thing to watch is that it is negative earth (pretty much all the other brit bikes of the era were still positive earth, but even with the cars, Triumph was among the first to change)
The good thing about being positive earth is that the zener diodes are still readily available!
Thank you for the detailed explanation and insight. It helped me avoid the mistake of switching to LED lights with the current setup I have on my ’71 Commando with podtronics reg/rec, Boyer electronic ignition, stock lucas single phase stator, and lead acid battery. If I was to want to switch to LEDs is there a viable configuration that does not stand the potential to damage the charging system components?
Thanks again for all you share on your site! Truly helpful.
Happy to hear that some of the content on the site is of use to you!
In all honesty, the lowest cost option is actually a good one if you want to replace your lights with LEDs
Fit the original zener diode to the z-plate – that big lump of cast aluminium is a superb way to get rid of excess power!
Swap the original Lucas LU49072 (2DS506) rectifier for a modern equivalent like the KBPC3504
One of these will set you back about £5 and is functionally a direct equivalent of the original.
However, the componets are resin encapsulated, so the joints don’t shake themselves apart like the origginal rectifiers do.
Couple of things to watch for.
1) keep the zener diode nice and clean – a good connection for heat and electricity is important.
It’s great they are out in the airflow, as that’s what keeps them nice and cool, but it does mean they will pick up road grime along the way.
Most ‘failures’ are simply down to a bad connection.
2) if you have turn signals and plan to swap these to LEDs, you will probably need to replasce the flasher unit at the same time.
The originals rely on using 4 x 21 watt lamps – one on each corner.
Lowering the load will make them flash slower, and will eventually weld the little bi-metallic strip inside, rendering it useless.
I did a little article on swapping in a modern flasher unit not so long ago:
Hope that helps!
Thanks Grant for demystifying some of the terminology and giving a simple clear and very informative appraisal of the alternator & reg/rect scenario. Highly insightful that reality differs to the hype around maximum charging output and minimum load that seems to have become the mantra and that I had mistakenly considered as the way to go.
This nice simple explanation of the different approaches to regulation now allows me assess some of the material out there on the web that I could never get my head around.
Thanks for your kind words Ray!
It’s always nice to hear that someone out there has found an article useful.
Reach out if I can help at all.
Grant, just regarding your comments on the Compu-Fire 55402 in your article and the warning:
“This would be suitable only for a negative earth application, since it is grounded through the heatsinking enclosure”
I purchased one for my 1970 Norton commando I am rebuilding along with new wiring loom and a 3 phase alternator and ideally, would like to stay +ve earth. So as I have it and made up a bracket to suit, would have liked to use it.
So does this mean the unit has to be earthed and will not like or fail to work when on a +ve earth system?
I would have thought it would work both polarity’s as like the rest of them and the finned part is only passively connected to the circuitry to radiate away the heat.
I enjoy and learn from the info you post, keep up the good work
Great to hear from you Steve!
Sorry to be the bearer of bad news, but unfortunately the Compu-Fire 55402 is definitely NOT suitable for positive earth applications.
It’s a real shame to be honest – they are superb units, and I really want to be able to use them on brit bikes, but cannot.
Unlike most of the aftermarket reg/recs available – the Compu-Fire 55402 IS electrically connected to the heat sinking enclosure, as well as being thermally connected.
If you turn the unit over and look at the back, you can see that the negative terminal is connected to the heatsink.
Sorry for the bad news,
Grant, well thanks for the info, I will save my Compu fire for when my street triple reg rec fails, or sooner as they apparently have a tendency to cook the alternator too and the Compu fire being a series unit, seem the go.
Im of to buy a trispark unit then, and wondering if the zenner and rectifier plugs can just be left empty and taped up on the Lucas loom and I just connect up my new trispark reg rec to the battery and alternator when I get to that stage of the rebuild
No problem Steve – I am just sorry it was probably not the news you were looking for!
It is important to be clear that the Tri-Spark MOSFET reg/rec is still a short-type device.
Although the unit itself will run a lot cooler (compared to the ‘legacy’ SCR devices, which are the same more-or-less as the original rectifier on the bike) it offers no improvement in respect of “cooking the alternator”
It works in the same way as most of the other aftermarket units in that respect.
When the target voltage of the battery is met, the reg/rec will dead short the input legs of the alternator stator.
The MOSFET unit however, does product a nice smooth output, so is a plus if you are looking to run more sensitive electronic devices.
Thanks, yes, could not find a SH775 easily, many USA companies don’t do post to Australia and a few copies around the place, so lets hope the new Lucas stator will hang in there with my trispark, all my old alternators lasted for many years, still have a 1971 complete system still running on my bsa, quite amazing actually, the loom is what is falling apart in places
Have you come across one of these? Are they comparable to a 200w Podtronic when it comes to voltage regulation quality?
Yes, the A REG ONE is made here in the UK in small batches.
It is a low cost SCR short-type reg/rec like the PODtronics, but that is where the similarities end.
Because they are made in very small numbers, the quality control is far superior, as is the quality of the parts used.
They are available from Alan Osborn of AO Services as well as Paul Goff.
Because they are short-type, it is still very important that you match the alternator stator output with the power consumption of the bike.
…for most cases, the standard 10 amp RM21 stator is a good match to a twin running electronic ignition and incandescent lamps.
In this scenario, a low cost short-type SCR reg/rec works absolutely fine, and should be no problem at all.
If you go up the LED route, or fit electric start, you should revisit this of course.
Thanks! Would one of these be advisable for use with a Tri-Spark ignition system?
The Tri-Sparks seem susceptible to electrical noise, which is one of the reasons why he stopped selling the PODtronics units.
He is selling a MOSFET one now which is probably your best bet, as they will guarantee compatibility.
I personally like the open-type Shindengen SH775, but the Access Norton crowd laughed me out the door for that.
I remember that on Access Norton… not cool.
I am trying to help a guy who is experiencing what seems like electrical noise with a Smiths electronic speedometer. I have hooked quite a number of them up over the years with the standard Smiths pickup on the speedo drive and never had an issue. This fellow is running a hall effect sensor on the rear wheel that requires power and ground from outside of the regular Smiths wiring. I have asked him to isolate the charging system to see if it makes a difference. I am a mechanical engineer that knows just enough about electrical to realize that I could be potentially dangerous – lol.
My thinking is the is likely conditioned power being provided to the by the speedo to the Smiths pickup, where as the hall effect sensor’s output could be affected by the quality of power coming from the main harness…
Waiting for him to get back to me with the verdict.
Yes, the Smiths tacho is really susceptible to spiking – I have seen a few nuked, and now recommend a 1/4 watt 22KΩ resistor inline on the input side to lessen the risk.
I have never had issues with the Speedo, but that doesn’t mean to say it’s not equally as susceptible to noise issues – after-all, the guts of the speedo and tacho are identical, other than the LCD display of course.
You need to be careful when using three wire speed sensors.
It is important to make sure that there is no voltage differential between the gauge and the sensor.
Using the frame as the positive feed for example, can be a contributing factor here!
The subject of stability used to be a massive problem on 60s and 70s cars, and it was commonplace to fit a voltage stabiliser for Smiths gauges.
I would be tempted to do the same again if he is experiencing issues and doesn’t want to seek out a MOSFET reg/rec.
Voltage Stabiliser – Positive Earth
Part Number: 070.362
The fellow got back to me – he confirmed with an oscilloscope that the output of the A Reg voltage output was pretty choppy. He replaced it with a TriSpark MOFSET and his instruments are now happy.
That’s great news!
Hopefully he’ll have no further issues.
Hello Grant. Thanks for the great article(s). They were particularly timely for me as I am in the final stages of resurrecting my 69 BSA A75 which has been sitting disassembled for 40 years. although the engine was intact and hence the alternator rotor remained in place and still would hold itself up by its magnets I decided to hedge my bets by reducing power demands by replacing all bulbs with LEDs and installing a boyer micropower ignition (a whole other can of worms!) . And to complete the “update” topped it all off with a Podtronics unit to manage battery voltage effectively. This all seemed to make some sense until I read your insightful article. I followed up by purchasing an Ebay (Chinese) potted rectifier as you recommended, turfed the new Podtronics and reinstalled the stock zener diode (back to the future!). My question to you is whether the little rectifier has to be grounded via a central bolt to the frame(or any good grd) and does it require an additional heat sink. As installed its just “floating”, being held up by its wiring and a zap strap in a space behind the tool storage. The positive terminal is wired to a common grounding post which connects to the positive battery terminal. As I hinted at the Boyer ignition has been an absolute pain in the ass and a huge source of frustration. I have only managed to start the bike once since installing the Boyer despite heroic efforts and a worn out right leg. It was in the course of this ongoing gong show that I came across your article and gratefully diverted my attention to reconfigure the charging system. Because it wont start I dont know how robust the alternator output will be and it may turn out that the rotor is too weak and needs to be replaced. In any event I am grateful for you sharing you expertise and perhaps saving me from burning up my alternator. Tony
Great to hear you have enjoyed some of my content Tony!
If your new rectifier is like this KBPC3504
…then there is no electrical connection between any part of the device, and the outer casing.
It can be used for positive or negative earth bikes, as it is “double insulated”.
However, they do get warm in use (not super hot)
So there is a definite benefit is bolting it to a metal panel, which will act as a heat sink and help dissipate any heat produced.
This will help you protect the longevity of the unit, and if looked after, it should easily last for the next fifty years!
Thanks great info, seems I’ve been on the right track for decades, always running the headlight with the idea that the energy is best used like that rather than wasted to the heatsink on the zener system. Particularly interested to learn that the MOSFET system is alternator shorting contrary to what everyone seems to tell me, also never considered the concept of the alternator acting against you as a brake when shorted making it harder to kickstart the bike.
Thanks for your kind words Stephen.
I think a lot of the information and articles out there are usually wrapped around someone trying to sell something.
Sometimes that unfortunately hides some of the points that should be considered!
I’m Evan, Head of Engineering for RMSTATOR. Thanks for mentioning us and putting together such an excellent article about voltage regulators! I’d love to talk with you more about our parts, or other charging system information. I’ve had a few old Triumphs myself and definitely put some time into upgrading their systems.
Great to hear from you Evan – and thanks for your kind words about the site.
You have my email address now – it would be great to chat reg/recs with you. Since posting this article, I have seen some RMSTATOR products first hand, and would be keen to chat!
Grant, many thanks for your helpful articles. I’m putting a 3-phase RM-24 stator and SH-775 open R/R on a 1974 Norton Commando and want to be sure I’m thinking correctly about the connections to the existing but now redundant rectifier and Zener diode (and for that matter, the capacitor — can’t think of why I’ll need it).
As I read your wiring diagrams in another article, it looks like I could A) tie the brown/blue connectors at those components together to keep the existing path from master switch terminal 1 to the battery negative post; or B) and better, ignore all those connections and run a new wire direct from master switch 1 to battery negative, reusing only the existing fuse holder at the battery. I’m not going to introduce any new problems, will I?
On the Commando, the alternator, main switch and battery (and my proposed location for the SH-775) are all very close together, so I want to run new wires directly among these components and avoid the 50-year-old harness completely for the charging system. Does this make sense?
Thanks very much. –k
Yes you are right!
You can either re-use the four cables where the original rectifier was, then run an extra to the stator.
Alternatively (and what most people do) is run three new cables from the reg/rec straight to the stator, then wire the new reg/rec directly to the battery.
I would take that option too, and have it on it’s own dedicated fuse.
Excellent! Thank you, and keep up the great work. –k
First and foremost, I want to thank you for all the work you put into your website, and the contributions you make to the motorcycling community. And all for free at that!!! As a person who struggles in the electrical department, your publications have helped me greatly.
I am finishing my 74 850 rebuild with a Lucas RM24 14.5 amp stator. I want to add LED lights for brightness and safety. After reading your article, I now realize that I may be producing too much power with this setup. As I live in a cold climate, I can extend my riding season by 2 months by wearing heated jacket and wearing heated gloves, which I can wire to my battery. With the heated gear connected, I think I am where I should be power wise for a balanced system. Is there a resistor of some sort that I could wire into the loom (perhaps where my capacitor would have gone) for the summer months to consume the extra power my stator would be generating?
Great to hear from you, and thanks for your kind words!
Sounds like you are on the right track with your upgrades.
The three phase stator is a great choice, as it’s output is still decent at a much lower RPM, so all good there.
What you need to do is consider they way that additional output is handled.
I like the Shindengen SH775 (or the SH847) as it is an open type reg/rec (some companies refer to this as a series-type reg/rec) and it is superior in applications where the power you are going to consume will vary massively.
It was designed for the Polaris ATV – where there is a high output fixed magnet alternator, and the potential to connect winches, big light bars, trailer electrics and heated clothing.
…however, when that additional load is not connected, the open-type reg/rec will effectively disconnect from the stator (opening the circuit), rather than try to dump the excess power to earth or to heat.
Most common reg/recs deal with scrubbing off excess power by shorting out the AC side, which is not good in high power applications like yours.
An electric start and heavy wattage filament lamps is a good way to deal with consuming excess power too, as the alternator will be always topping up a heavily depleted battery.
But in your case, having moved to LED lamps, and having heated clothing which of course you are not going to be wearing in the height of summer, I would definitly recommend that you go for an open-type reg/rec.
Many thanks for your reply. This will solve my issue. Now to find one!
Tour-de-Force, thank you
Thank you very much for writing such an informative and interesting article!
I found it whilst searching for the specification for the Shindengen SH650C-11 R/R fitted to my 2004 Yamaha FZS1000 (Fazer 1000). The Fazer’s charging system is known to be a bit marginal when, like me, owners have fitted accessories like heated grips, heated clothing, etc. Consequently, based on recommendations from the Fazer forum, I’ve just finished adding a separate, parallel, circuit to take some of the charge from the R/R straight to the battery. The reason I was looking for the specification was because I wasn’t sure what size of fuse I should use in this new circuit: my Yamaha workshop manual states that the nominal output of the alternator is 14V/365W at 5000rpm and that the ‘rectifier capacity’ is 18A – by ‘capacity’ do you think they mean maximum output? If so, given the fact that the ‘old’ circuitry will still be conveying some of the charge, I was intending to fuse the new circuit at 20A (the new wire is 3mm2, so can handle much more than that, obviously). By a strange co-incidence, given the point you made about the need to balance charging against load, I’ve also recently modified the bike’s electrics so that rather than the headlights (2 x 55/60W) coming on automatically when the engine fires, as is standard, mine can be switched manually – I’ll keep an eye on my (retro-fitted) voltmeter and juggle the load to keep the charging voltage below 14.4V (I fitted a Lithium Iron Phosphate battery for weight and space saving years ago).
Anyway, thanks again for taking the time to explain to novices such as myself how these things work.
Sounds like you are on the right track.
Based on your figures of 14 volts and 365 watts, you are looking at 26 amps max.
Always size up not down for reliability, so a readily available 30 amp fuse will cover your needs here.
Your 3mm² cable will handle that no problem, so you’re all good.
I would consider wiring in a SH775 or SH847 (which are both full blown 3 phase series type reg/rec units) direct to the battery positive and negative terminals, and doing away with the rest of the parallel circuit completely.
The beauty of these units is that they can easily handle big changes in load, as they actually open and close the alternator inputs as charge to the battery is required.
This is great when you are going from a hot summer’s day with no power consumers other than the bike’s ignition (you are probably drawing 5 amps at the most), through to cold winter riding when you have all your lights on, and heated grips and clothing cranked up to the max (where you could be drawing 25 amps).
It is a massive delta for a standard reg/rec to handle, but these series units take it all in their stride.
In my opinion the above would be a better choice in your situation rather than the FH020 MOSFET upgrade that is discussed on most jap bike forums as THE way to go.
Hope this helps!
Thanks very much for your kind comments and for your suggestions for further improvements – next winter’s project perhaps! It’s very reassuring to receive advice from someone who really knows their stuff – I’m forwarding a link to your website to a friend who’s had charging problems on his Triumph Sprint ST, which I’m sure will benefit him.
PS. Note to self: it’s cable, NOT wire!
Again, thanks for writing this piece, its becoming a standard reference. Two questions: Is there a way to search your many custom wiring diagrams by alterations made, or is there another searchable feature? May the Shindengen SH775 be used with the Lucas RM21 stator? Under the ‘Podtronics’ chapter there is a musing referring to electrical “noise”. I have access to both the Shindengen and the Tri-Spark Mosfet. Openning the stator circuit sure makes more sense to me. At this time I do plan to keep the R21 stator.
Thanks Douglas – that’s very kind!
There is a search box on the top right of the main page of the website.
It does a google search on the site, and isn’t perfect, but it’s OK.
A search for Tri-Spark will return any articles that has Tri-Spark tagged in it.
Shindengen is covered too.
Yes the three phase Shindengen SH775 or three phase Tri-Spark MOSFET will both work okay with a single phase RM21 stator.
I would certainly recommend the Shindengen over the Tri-Spark (in fact it’s the only reg/rec I recommend at the moment)
I really like the way a series-type reg/rec opens the AC inputs rather than shorts them out. It’s far kinder on the rest of the bike.
While the SH775 is SCR based, not MOSFET I don’t find it an issue at all. The output is very clean indeed and not peaky like many of the low cost reg/recs.
I have never had (nor heard of) a Shindengen unit fail, hence my recommendation.
Music to my ears!
Thanks and a Happy Easter to you and yours
First of all many thanks for putting this together, I just wish I had seen it earlier! I have built myself a pre-unit engined Triton and have had charging problems from the beginning. It is fitted with a Lucas RM23 high output alternator and 200W regulator rectifier bought brand new as a kit from Rex’s Speed shop on their recommendation, 9 AH AGM battery, wired negative earth, fitted with indicators and a Vape (Wassel) electronic ignition. The regulator rectifier is fitted to a large (roughly 250mm x 200mm) aluminium plate under the seat. All bulbs are normal incandescent. After about 50 miles it stopped running and I discovered the battery had 2 dead cells presumably from being overcharged. Rex’s Speedshop were most unhelpful and told me the battery had failed due to excessive vibration!
I replaced the battery and added additional earth’s to the regulator in case it may have been unable to dump excess current to earth?
Having done this, with a multimeter connected directly to the battery terminals I started it and readings seemed quite normal i.e. approx 12.7V at tickover, rising to about 13.8V at 3500 RPM. I then took the bike for a test ride and all was fine for roughly 15 miles at which point the Acewell electronic rev-counter / speedo began to go haywire constantly going through the start up needle sweep. My suspicion is that this is caused by it receiving excess voltage?
I managed to limp home with the headlights on and straightaway when I arrived I connected my multimeter to the battery terminals and restarted the bike, it now showed as much as 19V at 3500 RPM. My conclusion is the regulator rectifier is faulty.
What would your recommendation be to replace it? As I certainly don’t want to replace it like for like.
Hi Phil, and sorry you found this too late!
The good old zener diode originally on the bike, coupled to a 10 amp alternator stator was a great balance for the power consumers on the bike.
Typically with your engine running, you are drawing around 3 amps.
With your lights turned on, this will go up to 8 amps.
…so still under the 10 amps the original stator puts out.
Any excess power produced was converted to heat by the zener diode – on a Commando, this was mounted to the big chunky footrest “z-plate”
On the Trumpet, they had a cool space age finned heat sink hung under the headlamp in airflow.
Your new RM23 high output alternator is putting out 16 amps or around 100 watts of power.
The nature of a permanent magnet alternator as we have on these bikes means that output is produced whether the battery needs the charge or not.
So you can see, that you are constantly producing too much power for the bike, which means that excess power needs to be scrubbed off somewhere.
With a modern reg/rec it does this by dead shorting the stator.
They do not ‘dump excess current to earth’ so no need for extra earth on these.
A combined reg/rec is perfectly ok for a 10 amp stator that is balanced to the bike.
But this is terrible for a high output stator that is constantly producing more power than is required, and will result in damage to the electrical system.
In your particular scenario, and on the basis that you have wired your bike for negative earth, you can easily obtain a negative earth zener diode and fit it to your bike.
These are still readily available – it’s the positive earth ones that are like rocking horse shit.
However, you do need to drop in an RM21 10 amp stator, and do away with your RM23 16 amp high output stator.
The only reason you would need a stator so powerful is if you fit electric start, and are constantly needing to recharge a massively depleted battery.
Rather than the combined reg/rec – you then need to drop in a simply rectifier – these are around £5 to buy, and again readily available.
The modern ones are potted in a resin compound, so much more shake proof than the originals were back in the day.
The KBPC3504 would be perfect for the job.
Hope this helps!
Thanks so much for the reply Grant.
As it happens I also own a 1974 Mk2A Commando to which I have fitted an Alton electric start. That is still running the original rectifier and Zener but I have the original Lucas alternator sitting in a box doing nothing. If I were to fit the stator to the Triton would you recommend I use the original rotor which presumably will have demagnetised somewhat over the years or the new one that came with the RM23?
The rotors are the same, so no issue at all with you keeping your new one on there!
…the extra magnetism is always useful.
Just make sure the stator is concentrically spaced with a nice even air gap all around the rotor, and you’ll be good to go.
I’ve fitted the ex Commando RM21 stator and I can find the KBPC3504 rectifier easily. Can you give me a reference number for a suitable Zener diode please?
Most of the stud mount type I can find are 20v and I don’t know if they are suitable.
You bike is negative earth – so you are looking for Triumph/BSA zener diodes, not Norton.
The original Lucas Part Number for this was LU49589
These are very often for sale on eBay and on the forums!
Alternatively, you can buy a new equivalent.
The part number you are looking for is BZY91C15
15 volt / 100 watt is the spec you need.