Related Posts with Thumbnails

Last Updated: May 22, 2010

Motorcycle Voltage Regulator DIY

This project is a voltage regulator for a 3-phase, permanent-magnet alternator found on motorcycles. The same design can be used for single-phase alternators simply by eliminating some parts.  This project only describes the regulator, even though a rectifier is typically integrated into the unit.

Circuit Analysis:
Please refer to the circuit diagram during the analysis of the circuit.

The way this type of regulator controls the alternator’s output voltage is by simply shorting out the stator winding for one cycle of the ac waveform. This is called shunting. This is done because it is much easier to short-circuit an inductor (the stator winding is an inductor) than to open-circuit an inductor. Very high voltages are induced when an inductor circuit is opened. This may cause a breakdown in the winding’s insulator.

The device used to shunt one cycle of one phase is an SCR. A silicon-controlled rectifier acts as a diode when triggered at its gate. It stops conducting when the current drops below a very low threshold value. Once it stops conducting, it will not conduct again until it receives another gate signal.

A voltage detection circuit is used to trigger the gate on each of the three SCRs. The threshold voltage to trigger the gate is selected as 14.6volt (on the motorcycle’s system-voltage). This is when the gate would trigger if there were no capacitor as described in the next paragraph.

Since the voltage from an alternator fluctuates, the detection circuit will trigger during a peak in the waveform of one of the phases, but this causes the average voltage to be too low. To raise the average system-voltage without increasing the detection threshold, a simple capacitor is used to delay the triggering of the SCR gates.  The capacitor acts as a filter to reduce the ac ripple going to the detection circuit. If the system voltage was strictly DC voltage (with no ripple) the gate would trigger at 14.6v and the average voltage would be 14.6v. In reality, the average voltage is always less than the threshold voltage and only approaches it as the ripple reduces. The actual trigger voltage is higher than 14.6v, but the average voltage is lower than 14.6v. As the ripple reduces, the average and trigger voltages approach 14.6v.
R1, Z1, D1, D2, and R2 make up the detection circuit. R1 allows for an “idle” current to flow through the detection circuit so the Zener diode Z1 and regular diodes D1 and D2 are in the linear portion of their operating range. As the motorcycle’s electrical system-voltage increases past a set threshold, Q1 starts conducting. This sends current to the gates of the SCRs. D3-D5 act to isolate the gates from each other. R3, R5, and R7 act as current limiting resistors for the gates. R4, R6, and R8 act as drains for any possible leakage in Q1.

To determine the values of the components, it is probably best to start at the output and work toward the input. 
Selecting S1-S3:
S1 through S3 must shunt a large amount of current to reduce the alternators output voltage.  Typically this may approach 15 or 20 amps for short periods of time. The typical forward voltage-drop on the SCRs will be around 1 or 2 volts. This results in significant power that must be dissipated by the SCR. The SCRs will be selected as 25 amp devices that can surge up to 300 amps. They will have to be heat-sinked with thermal gel to handle the power.

The commercial part number is NTE5460 or ECG5460.  These are the most expensive components in the project and are not available at Radio Shack. They can be obtained through Newark or MCM.

Selecting D3-D5:
D3 through D5 are simply to isolate the three SCRs from each other. It is not known how significant the diodes are to the circuit, but they were put in just as a precaution. They are selected as small signal diodes. The commercial part # is 1N4148. There is no significant current or power to concern with.

Selecting R3, R5, R7:
These three resistors act to limit the current to the SCR gates. The specs for the SCR place .040 amps as the limit for the gate current. The resistors will be selected to limit the current to .019 amps. The gate voltage is approximately .8volt. The voltage on R3 can be found as follows:

R3voltage = system voltage – Q1 voltage – D3 voltage – gate voltage
R3 voltage = 14.6v - .1v - .7v - .8v = 13v

The value of R3 can be found now that the current and voltage are known:

R3 = R3voltage / R3current
R3 = 13v / .019 amp
R3 = 684 ohms

The power in R3 is found:
R3power = R3voltage * R3current
R3power = 13v * .019amp = .247watt

For safety, a .5watt resistor will be used.

R5 and R7 are found the same way as R3.
R3, R5, and R7 will be selected as 680 ohm, .5watt resistors.

Selecting R4, R6, R8:
These three resistors are to drain any possible leakage there may be in Q1. They will be selected such that they will drain about .001amp just as the gate is triggering. The gate’s voltage is the same as the voltage on R4. Therefore, R4 is found:

R4 = gate voltage / .001amp
R4 = .8v / .001amp = 800 ohms

Since it is not critical, for convenience, R4, R6, and R8 will be selected as 680 ohms to match R3, R5, and R7. 

The power in R4 is found:
R4power = R4voltage * R4current
R4power = .8v * .0012 = .00096watt

R4 can be a .25watt resistor safely, but to use the same part as R3, R5, and R7, it can also be .5 watt.

R6 and R8 are found the same as R4.
R4, R6, and R6 will be selected as 680 ohm, .5watt resistors.

Selecting Q1:
Q1 is a PNP transistor. Since each of the three gate circuits will draw .02 amps, the transistor must supply .06 amps.  For reasons of availability and reliability, Q1 will be selected as a TIP42 transistor.  The transistor can handle 10 amps and has a power rating of 65 watts. This far exceeds the requirements of Q1. Therefore, the transistor will not even need a heat sink at all. The minimum gain is 20 for a TIP42, but typically they have a gain closer to 100.  It will be assumed, therefore, the gain is at least 50 in this application.

Detection Circuit:
Z1, D1, D2, R1, and R2 make up the voltage detection circuit.  R1 determines what the idle current will be before Q1 conducts. The idle current ensures that Z1, D1, and D2 are operating in their linear range before Q1 starts to conduct. The idle current needs to be about .013 amps to accomplish this.

The emitter-base junction of Q1 reaches about .6v before it conducts. That voltage will be used to determine the value of R1 to obtain the desired idle current. The voltage on R1 will always be the same as the e-b junction on Q1, so it will never go much higher than about .6 volt.

R2 is the current limiting resistor in case the system voltage runs high or a battery charger or some other voltage source is applied to the system.  R2 should be made as small as possible so that the detection circuit’s threshold voltage will not be affected much by the current in the detection circuit.   R2 will be selected to make the final adjustment to the threshold voltage while still maintaining its current-limit function.

Z1, D1, and D2 should have a relatively constant voltage drop which is relatively unaffected by a change in current. Normally, one would use a single Zener diode to maintain the necessary voltage drop. However, in this case, the required voltage is not found in a readily available Zener diode, at least it’s not readily found at Radio Shack. Because of this, D1 and D2 are necessary to raise the overall voltage drop.  If Z1 were a higher voltage Zener, D1 and D2 could be eliminated.

Before any current can flow in Q1, the system voltage must be higher than the sum of the voltages on R1, Z1, D1, D2, and R2 while the idle current is flowing.

Selecting R1:
Knowing the idle current, the value of R1 can be determined. The voltage on R1 will be .6v and the idle-current will be .013 amps.

R1 = R1voltage / idle current
R1 = .6v / .013a = 46.2 ohm

The power dissipated in R1 can be found:
R1power = R1voltage * R1current
R1power = .6v * .013a = .0078

R1 will be selected as a 47 ohm, .25watt resistor.

Selecting Z1:
For availability, Z1 is selected as a 12v, 1watt Zener diode. The commercial part number is 1N4742a.  Since it is 1 watt, the maximum current in Z1 can be found:

Z1max current = 1watt / 12v
Z1max current = .083 amp

The maximum current allowed in Z1 is .083 amps.

The actual voltage on the Zener diode at the idle current (.013 amp) is about 11.7 v. This value can vary from diode to diode. The value can range from about 11.55v to 11.95v. Therefore, the threshold voltage may vary by that same range (about .4 volt).  However, the design will be built around 11.7v as the typical value for the Zener diode.

Selecting D1 and D2:
Diodes D1 and D2 are selected to give additional, constant voltage drop in the detection circuit. Diodes are used since their voltage does not vary (much) with a change in current. (As opposed to a resistor whose voltage varies proportionally with current).

D1 and D2 will be small-signal diodes. The commercial number is 1N4148.

At .013 amps, the forward voltage drop on each diode is about .7v.

There is no significant power dissipation in these diodes.

Selecting R2:
R2 is the one resistor whose value may be changed to alter the output voltage of the regulator. It also protects the Zener diode from an over-voltage condition caused by an external voltage source such as a car alternator or battery charger.

In order to protect Z1, it must limit the current to .083 amps. The protection must be provided up to a system voltage of 17 volts. At 17 volts, the detection circuit can have no more than .083 amps flowing.  At .083 amps, Z1 will have about 12.5 volts on it. D1 and D2 will each have about .8 volts on them. R1 will have about .8 volts on it.

At a system voltage of 17 volts and .083 amps in the detection circuit, the voltage on R2 can be found.

R2voltage =  system voltage – R1voltage – Z1voltage – D1voltage – D2voltage
R2voltage = 17v - .8v – 12.5v - .8v -.8v
R2voltage = 2.1v

From this, the minimum value of R2 can be found.

R2minimum = R2voltage / R2current
R2minimum = 2.1v / .083a = 25ohm

R2 can’t be any less than 25 ohms. Since R2 will be a higher value, as will be discussed below, the circuit will be protected to a much higher voltage.

The actual value of R2 will be determined by the requirements of the detection circuit.  The threshold voltage will be selected as 14.65 volts. The idle current was set at .013 amps. At .013 amps, Z1 voltage is 11.7v, D1 voltage is .7v, D2 voltage is .7v, and R1 voltage is .6v. The idle voltage on R2 is found as follows:

R2idleVoltage = System voltage – Z1voltage – D1voltage – D2voltage – R1voltage
R2idleVoltage = 14.65v – 11.7v - .7v - .7v - .6v
R2idleVoltage = .95v

The value of R2 can then be found:

R2 = R2voltage / idle current
R2 = .95v / .013a = 73 ohms

The power dissipation in R2 will be highest when the over-voltage of 17 volts occurs. Therefore the power factor will be calculated with this in mind.

At a system voltage of 17 volts, the voltage on R2 may reach 3 volts. The power is found as follows:

R2power = ((R2voltage)^2) / R2
R2power = ((3v)^2) / R2
R2power = (9v^2) / 73 ohms
R2power = .123 watt

For safety, a .5watt resistor will be used.

For availability, R2 will be selected as a 75ohm, .5watt resistor.
Most likely, R2 will be comprised of two 150ohm, .25watt resistors in parallel.

Selecting C1:
C1 is used to raise the average system voltage. C1 delays the triggering of the SCRs, thus allowing the instantaneous system-voltage to exceed the threshold voltage while the average system-voltage remains below the threshold voltage. As the load on the electrical system drops, the average system voltage will approach the threshold voltage. 

The value for C1 was determined experimentally rather than being calculated.

C1 is selected as a 10 microfarad, 35volt (minimum) electrolytic capacitor.

 It is possible that different alternators may require different values for C1. However, 10uF should probably be adequate. If the capacitor is too large, oscillations may occur. If the capacitor is too small, the average system-voltage will remain too low. A capacitor should be selected which gives a maximum average voltage between 14.0v and 14.6v.

It should be noted that the system-voltage could have up to 2 volts (or more) of ripple as measured on an oscilloscope depending on the engine RPM and electrical load.

To convert this regulator for single-phase use, simply leave off D5, R7, R8, and S3.


Anonymous said...

Aren t the D3, 4, 5 diodes in the wrong way?

LEI said...

sorry for the late reply...nope..d3 d4 and d5 are just for scr gate isolation...

rahil said...

can you please tell me exact power ratig for bridge diodes? i have used IN5408 for bridge and it was getting hot in few seconds. does it is having any circuit prob or diode incompatible? single phase stator for RD350 yamaha. ( )

LEI said...


Im not actually familiar with RD350 might not be compatible with your system...I'll check on this thanks.

LEI said...

ok found it....

is this the wiring diagram of your bike pls check the link

rahil said...

can you please tell me exact power ratig for bridge diodes? i have used IN5408 for bridge and it was getting hot in few seconds. does it is having any circuit prob or diode incompatible? single phase stator for RD350 yamaha. ( )

Anonymous said...

Aren t the D3, 4, 5 diodes in the wrong way?

Anonymous said...

hello LEI
i want to put this circult in my bike royal enfield electra 350cc
i have three phase stator. and using single phase oem regulater.

i want to accelerate current because of i have hid system and big horns so that's why i want extra more current

this diagram will work or not ?? i mean compatible with my bike or not plzz reply, m weighting

LEI said...

it will work...for three phase system can have the option of calculating the value to where how much current and voltage you'll be needing for your add-ons accessories...

motorcycle accessories said...

It's beautiful and nice blog. I am glad I visited here and come to know about it.. I have gathered a huge knowledge regarding the motorcycle parts and accessories.. I will share it out with my friends...

LEI said...

@motorcycle accessories, Thanks just visited yours, i say...great products out there, just wish your near me to see it personally,

old as dirt said...

Thanks for all your excellent work creating this post. I was able to follow all the electronic details and it all makes sense. I am waiting for delivery of a replacement regulator ( 1st one DOA) but if that doesn't work out I will build your design. I may build it anyway, I'm an ET with a long history of building stuff. Thanks again

Fazer9999 said...

recommend to everyone:

sorry pass the BTA26 to yours cicuit ??

LEI said...

hit there fazer9999, no problem, we are here to exchange something to be useful for everyone, carry on.

Fazer9999 said...

" S1-S3 Must be heatsink mounted. " is it correct? All of them on one heatsink? Does it must be isolated mounted?

TO-220 Heatsink is anode and in this circuit the alternator is here.

Then it is short circuit but not just in the katode site!!!

LEI said...

thats why whoever want to build such circuit like this one especially for motorcycles must have enough knowledge about circuitry. The tabs of those SCR is connected to 2nd pin and must used isolation such as mica insulator..thanks for the clarification...

Anonymous said...

can this be used in a sinski scooter 150cc?..
sorry for the noob question..
i'm a newbie to bikes/scooters..

fabriziolelli said...

HELLO. know "Ducati Scrambler 250cc in 1970?" often burns the regulator, 6v 70w. Who can help me? I
build it.'m a little help. can 'be good this Site
for me? help me? I can also turn to 12volt.non understand if my star has phases, sorry for google translator. and that 'the link, tell me if you see.
Thank you CIAO CIAO

LEI said...

hello fabriziolelli, did you chech the resistance of your coils, if there are short on the windings, you'll just fried the regulator with excessive current. check with multimeter.

LEI said...

oh by the way mr. fabriziolelli, your system is 6volts, and you cannot use the system here that requires 12 volts electrical system.

fabriziolelli said...
This comment has been removed by the author.
Anonymous said...

thanks so much for the answer, a coil was
burned, time and 'ok, and 'history,the regulator always burning the stator, and' evil out of the factory, burns and sends short one coil (diodes, too small), I ask, the phases of this regulator
are good for my ducati 250? it's good enough
for 12volt, from my stator,out more than 40volt,(0.8 ohm + 0.8 ohm),
Many friends have changed from 6 to
12volt, changing only the regulator, coil
candle,lamps,horn,battery, W=VxI 6Vx10amp=60watt 12Vx5amp=60watt, so 'I thought,He raised the VOLTAGE,less current, but I get the same watts, right? now I want to build it
I build many things in electronics!!!
I do not understand only the phases(my ducati),star? triangle?
your regulator and 'good for my stator, in this link? thanks,you're the only site in the world that talks about this thing that burns in many bikes, I ask, if possible, simple answer for google translator, thanks again. phases???? ciao ciao fabrizio

LEI said...

hello again, three phases coil is the best to use when it comes to generating AC to power electrical accessories for motorcycles, it can be used as single phase and three phase unlike a a single phase system...The disadvantage of does not have a lighting coil to power headlight..instead all electrical must be redirected to the battery...thus creating too much load on the battery.

A stator burns due to regulator malfunctioning, example are diode getting short, SCR triggering malfunction.

I used this circuit with one of my scooter that has a single phase stator, and it works, although i cant use my headlight as AC since the circuit above only regulates the battery charging, it lacks AC regulator for the headlight, but the circuit can be modified to have an AC regulator for headlight. CIAO

fabriziolelli said...

thanks again, but have you seen my stator on this link? In your opinion, so I can plug '? yellow-red-yellow directly of the scr s1-s2-s3? (also obvious with the rectifier diodes) in summary: OK 6 to 12volt (I edit), OK power (12volt battery-4amph think) I go? I connect so '? please, if you can, look at this link, here we say: you do from light a candle to the Virgin Mary, even if it does not work!!!!. see the link? and I connect it (mamma mia!!! fear)wishes CIAO CIAO fabrizio.

LEI said...


the red line must not be connected to your generator center tap if you will be using the schematic circuit above posted, based on your diagram, i cannot imagine why the battery positive terminal is directly connected to the generator, wherein its DC and must be isolated from the AC your generator will be generating when you rev your engine..that coil will be burned, if you need simplification, look at my other post about voltage regulator here at my site,

look at this page

you have to identify what type of generator you have, before planning to build the circuit.

Fazer9999 said...

Voltage Rectifier/Regulator malfunction:
A typical cause is !!NOT!! a bad regulator. The generator will be groundfault at high temperatures and at high transient voltage. CAN NOT BE MEASURED BY MULTIMETER only by Insulation Tester ! New Regulator alternates:
My facebook pictures of alternator rebuilding:

Anonymous said...

I made this one many years ago for my ZZR600 (with a few adjustments)and that was the last time I had to buy a new regulator.

Pete said...


I have some questions about my voltage regulator maybe you can help?

Pete said...


So I will go ahead an post my question and hopefully you can answer it.


It started when the negative terminal on the regulator/rectifier shorted out from corrosion. The system stopped charging. Parently very common with this bike...

I replaced the harness with a used one in good condition and it was fine for a view years. Then it did the same thing.

I decided to cut the plug going to the regulator/rectifier and put new pins on the wires and directly connect did'nt work. So I replaced the rectifier unit with a new OEM unit. Still did not work... Then I replaced the stator with a new oem unit...still did not work.

There is a catch here...(maybe not sure) when I say it did not work that means I tested it. I hooked up the rectifier/regulator to the stator but DID NOT hook up the dc output to anything. I figured I could use a volt meter to mesure the dc output and expect to get the require 14 volts. But when I do this, all I get is 3 or 4 volts. why????

I also tried hooking the battery up directly to the regulator/ rectifier and still the same thing happens, why????

I have checked the stator output, and it is good, it reads about 50 volts AC which is 3 phase. Another wierd thing is that whith the bike idling, the dc output is 3 or 4 volts, as soon as I increase the engine RPM the dc output voltage drops to 0. Also the regulator/rectifier unit gets very hot, very quikly.

The type of rectifier/regulator is a SH650-12A

From all the research I have done, this is a shunt type and not a MOSFET type regulator.

Sooo...MY QUESTION IS: Why is the unit not putting out the required 14 volts? Is it because there is no load across the DC side of the circuit? I would have though that the batter would trigger it, but apparently it does not. Or is it because the battery voltage is not high enough?

How would I trigger the unit to put out the correct voltage without hooking it up to the bike...The reason I ask is because i do not want to put the entire bike back together only to find that it doesnt work...

Any thoughts would be greatly appreciated . I have been having trouble with this for two years, and unable to ride my bike because of it.... :O(

LEI said...

hi, sorry for the late reply. have you checked your generator (stator) for resistance check. sometimes a faulty charging system is not directly responsible by the voltage regulator, when there is a shorted path between those windings the tendency of the ac voltage is that it will be loop again back to ground.

By the time you replaced the OEm with an aftermarket SH650-12 before, did it work or same as you did AFTER putting back the OEM.

I have posted somewhere here in my blog the block diagram of a 3 phase regulator to which i think is your R6 system..kindly look at it here.

kazuki_chan8 said...

hi..i hav honda wave 125...i want to half rectified the ac signal from alternator hich only hav 1 white not sure how many ampere it would im having trouble to select a suitable diode to do rectification...right now im planning to is still think it is not apropriate....if posibble email me if u hav any idea in this? my email is:

LEI said...

wave 125 is a single phase alternator, that uses only two wires for charging, the white wire is connected to your regulator which regulates the voltage of course to 14.4 which also converts the AC generated to DC, without the regulator, the voltage as you rev will be more than you can imagine near 60 volts AC, if you'll just put rectifiers on it like the one you mentioned, in what particular you'll be using that too much DC.

Abu-Hafss said...


Today, I stripped a shunt type Chinese R/R for 200cc three-wheeler.

It had BT152-600R for SCRs,
SMD button type rectifiers (white colored, amp unknown),
SMD 510 Ohms for R3, R5 and R7,
SMD 100 Ohms for R4, R6 and R8,
no D1, D2, D3, D4, D5
SMD transistor instead of TIP42 (could not read the markings)

Nathaniel Berdan said...

Can you do me a favor of posting the whole picture of the board that you had stripped there mate. thanks

Abu-Hafss said...

Why not? But, tomorrow with schematics.

admin said...

ave 125 is a single phase alternator, that uses only two wires for charging, the white wire is connected to your Các loại Ổn áp lioa 2 kva
Các loại Ổn áp lioa 3 kva
Chuyên Ổn áp lioa 3 kva
Bán Ổn áp lioa 10 kva
Giá tốt On ap lioa 3 pha
regulator which regulates the voltage of course to 14.4 which also converts the AC generated to DC, without the regulator, the voltage as you rev will be more than you can imagine near 60 volts AC

Cam puchia said...

Dịch vụ Tư vấn chứng nhận TQP Uy tín ,
Cung cấp tư vấn chứng nhận Iso trọn gói Chất lượng ,
Giải phápChứng nhận hợp quy chuyên nghiệp HCM Chất lượng ,
Công tyDịch vụ chứng hợp quy vậy liệu xây dựng HCM Uy tín ,

Chuyên Thi công biển quảng cáo tại HCM Giá rẻ ,
Đơn vịChuyên In uv giá rẻ tại TPHCM Chất lượng ,
Thi côngQuảng cáo trên bảng hiệu HCM Uy tín ,
LàmBảng hiệu quảng cáo HCM chuyên nghiệp ,

adminvip said...

chia sẻVideo giới thiệu căn hộ the easternQuận 9,

Ron Morris said...

I built this for a single phase system. It worked for a few days then stopped regulating!
Volt meter shows dc @ 17.5
Goes from 12.5 to 17.5

When working it would clip at 14.4

I replaced the tip42 and 12v zener, still not working