The Great KB Regulator & Rectifier project.

[davereid]

Allun and I have noticed that heaps of the threads on here are about dodgy recitifiers, and voltage regulators. And often, sourcing a replacement can be expensive.

We discussed the idea of designing a R-R on here, so that KBers can build one at home for a few $ rather than forking out heaps for a factory one, particularly for an older bike, where the electrics may be worth more than the bike.

All contributions are welcome, and if you have a particular problem bike, let us know , and we will have a crack at solving the problem.

Also, anyone who has the time and patience to build and test the dodgy designs we will produce should let us know.

Cheers


Dave

[davereid]

To start the ball rolling...

Design #1 - The burn-yer-leg off shunt regulator
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Use 2 x MJ150037 power transistors, wired as a darlington pair. In this configuration the collectors are connected together, the emitter of the first transistor feeds the base of the second, and the emitter of the second is ground.

These transistors can handle a whopping 20amps, and have a minimum current gain of at least 25, so in our confguration you could expect a current gain of at least 600.

So we need a base current of about 30mA to push these transistors full on.

Use a 12 volt zener, like a 1N5349, at 5 watts. this can handle 400mA, so is easily up to the job.

Wire it in series, cathode to positive with a 120 ohm resistor, connecting the base of the first transistor to the positive supply, and you have a simple shunt regulator.

Good for 20+ amps if placed on a large lump of steel as a heat sink and costing under $30.

But... relies on stator have at least 2 ohms effective winding resistance, and needs a separate rectifier....may need to try several zeners, even up to 14volt models to get effective charging.

[CookMySock]

Considering how much cheaper switching FETs are, it might be worthwhile looking at a switching design as well, particularly for the larger alternators (mine is 400watts - 30A.) There are other benefits, such as it will act as the rectifier as well, longer stator life, lower torque on from the engine at full revs, lower component count. But yeah there should be a basic one first.

edit: This search at linear.com shows two suitable devices for input voltages of 100V and outputs of 12V 20A. I wonder if many stators will exceed this voltage. The device should current limit at 20A - unfortunate for a 30A capable stator. Useful on race bikes, as it is nearly weightless, and it will unload the stator when no load is being used.


Steve

[avgas]

hmmmm build it yourself seems rather an interesting concept.
While not impossible - i'm gonna steer clear from it.

Old bikes it would be fine, but my experience with putting high power electronics into new vehicles says that the likes of Yamaha, Honda, Toyota..... spend far to much time letting BOSCH etc worry about the hamonics etc.
So be sure guys that you chuck in enough 3rd and 5th harmonic sinks (RC circuits will be fine) so that the poor bike battery etc doesn't bake trying to deal with it.

Also this thread is useless without schematics people

[davereid]

Old bikes it would be fine, but my experience with putting high power electronics into new vehicles says that the likes of Yamaha, Honda, Toyota..... spend far to much time letting BOSCH etc worry about the hamonics etc.
So be sure guys that you chuck in enough 3rd and 5th harmonic sinks (RC circuits will be fine) so that the poor bike battery etc doesn't bake trying to deal with it.

Also this thread is useless without schematics people

Ask and ye shall receive.. pdf schematic attached.

Transistor is rated to 1 Mhz so it will easily clean up any harmonics from the alternator, and as we dont switch anything, we arent making any.

Costs..
Jaycar has transistors each at $8.90, Zener $0.45, resistor $0.06 heat sink... up to you, total about $20.

[k14]

Ok, you brainy electrical types. How do I make me one of these (see attachment). At 60 quid a pop you could probably make some $$ if you worked out how to do it properly. I'm guessing the components are only worth $5.

[notme]

After Dave and I started discussing this, I was thinking that a better solution (note that "better means different things to different people!) would be a small 8 pin micro monitoring voltages and currents, and controlling the charge level.

Moving the control to a micro means that once a power circuit is set out, a lot of future improvements will be software only.

The thing could log battery charge and warn when your battery is on it's way out, or when the charging system suddenly isn't charging, and allsortsastufflikethat(tm).

Could also have different profiles selectable, e.g. "normal" maintains the battery to the best of it's ability, but "race" takes the minimum power possible from the engine.

[notme]

Ok, you brainy electrical types. How do I make me one of these (see attachment). At 60 quid a pop you could probably make some $$ if you worked out how to do it properly. I'm guessing the components are only worth $5.

It's a simple DC/DC converter. Looks like it's a step down (buck type) only, and it's probably not isolated.

So how you build one is you get a switch, inductor, and a diode and control them in the right way.

Rather than designing and building one - since this is a specialized application I would look for a cheaper version of the same thing. It's amazing how much of a price premium people will charge for a "compact electrochemical energy conversion device" just because most people don't know that all they are getting is an AA battery.....

In the same manner, this "RS125 voltage regulator" is a generic DC/DC converter, much like the ones linked below that are a weeeeee bit cheaper:

http://www.jaycar.co.nz/productView....Max=&SUBCATID=

Oh and the material cost of even the jaycar one? You don't want to know, it would make you

[davereid]

After Dave and I started discussing this, I was thinking that a better solution (note that "better means different things to different people!) would be a small 8 pin micro monitoring voltages and currents, and controlling the charge level.

Moving the control to a micro means that once a power circuit is set out, a lot of future improvements will be software only.

The thing could log battery charge and warn when your battery is on it's way out, or when the charging system suddenly isn't charging, and allsortsastufflikethat(tm).

Could also have different profiles selectable, e.g. "normal" maintains the battery to the best of it's ability, but "race" takes the minimum power possible from the engine.

Good idea Allun.

We will leave regulator #1 where it is. It wil work, it will be cheap, but its err, not exactly flash.

Say a PIC micro ? common, cheap, easy to program at home ?

[notme]

I prefer Atmel AVR but 6 of one, half a dozen of the other. Think of the possibilities once you have a thinking engine in the battery loop - all micro's worth thier salt have EEPROM nowadays so you can easily store profiles, battery characteristics, user prefs etc.

A smart charge controller could also allow the engine to be started with no alternator load on it, and only start chargin the batt after say 10 secs of engine running....but i think the early warning of a failing batt or RR or alternator would be far more valuable - could save you boiling your battery, or not being able to start the bike when you want to come home at the end of the day or whatever.

I have made a start on this sort of idea - my last few bike have had a small micro powered voltage monitor on the dash, it's just a bicolour LED that indicates low/normal/high electrical system. If it ever indicates anyhting other than green (apart from brief transients) I will know about it and save myself a boiled battery or a dead bike! Never mind burnt out wiring and so on.

[CookMySock]

At redline, it seems the open-circuit voltage from the stator will be as high as 400-500 volts. Search for "open circuit" on this page. So be careful using anything but a shunt regulator. This is impossibly high for an economical series or switching regulator. Note his "fingertips" comment as well.

Steve

[k14]

It's a simple DC/DC converter. Looks like it's a step down (buck type) only, and it's probably not isolated.

So how you build one is you get a switch, inductor, and a diode and control them in the right way.

Rather than designing and building one - since this is a specialized application I would look for a cheaper version of the same thing. It's amazing how much of a price premium people will charge for a "compact electrochemical energy conversion device" just because most people don't know that all they are getting is an AA battery.....

In the same manner, this "RS125 voltage regulator" is a generic DC/DC converter, much like the ones linked below that are a weeeeee bit cheaper:

http://www.jaycar.co.nz/productView....Max=&SUBCATID=

Oh and the material cost of even the jaycar one? You don't want to know, it would make you

You sure that is exactly the same? The thing that makes me think the jaycar unit is slightly different is because it says it needs a heatsink. To me that implies it is inefficient, which is the very reason for having the unit in the first place. I run my bike now with no voltage regulator and just straight off the batteries (14v down to 12.6ish) so would be worse off with the jaycar unit. Or do I have my wires crossed??

[notme]

You sure that is exactly the same? The thing that makes me think the jaycar unit is slightly different is because it says it needs a heatsink. To me that implies it is inefficient, which is the very reason for having the unit in the first place. I run my bike now with no voltage regulator and just straight off the batteries (14v down to 12.6ish) so would be worse off with the jaycar unit. Or do I have my wires crossed??

wires crossed....good one.....:-)


The jaycar one will achieve the same thing as the unit in your link. The one in your link most likely needs a heatsink but it isn't mentioned, or the mounting instructions will be very specific about mounting it to the frame (i.e. a heatsink....)

Plus the jaycar one mentions that it only needs a heatsink for very high transformation ratios - i.e. if you wanted full current out of it at 12V output with a 30V input or similar.

Any switching bypass regulator is going to be bazillions of percent (excuse the tech speak) better than a series shunt regulator.

A swithcing regulator only lets through the energy required, wheras a shunt regulator wastes all the energy that is not required as heat.

[notme]

At redline, it seems the open-circuit voltage from the stator will be as high as 400-500 volts. Search for "open circuit" on this page. So be careful using anything but a shunt regulator. This is impossibly high for an economical series or switching regulator. Note his "fingertips" comment as well.

Steve

Sigh. I'll take this one Dave but you get to answer the next one :-)

Steve - the link you gave is some guy posting his opinion on a forum. Not much different to what I'm doing right now you reckon?

Well, he has several errors in his statements, feel free to check what i am telling you against any references you like as long as they are "official" i.e. textbooks, other qualified people etc.

Oh, and also please bear in mind dude, I'm only interested in the correct information getting out there for everyone's understanding - not meaning to call you a liar or say I'm smarter than you or any such horseshit!

Here's what he says in his post and the issues with it:

"The stator (or generator, but its not an alternator) has an open circuit voltage around 45 volts. If you redline the engine, you will be testing the stator's insulation with about 500 volts. That can blow your fingertips off if you touch it. The voltage from any two phases should be the same. When you hook it up to the regulator/rectifier, it will clamp all three down to about battery volts. Its a shunt type regulator. That means the power you don't use is shorted to the coil."


1. He seems to be saying that the spinny thing that generates power in your bike is called a stator or a generator, not an alternator.
WRONG. A stator is an internal part of an alternator. A generator produces DC from a mechanical input. The thing in your bike that the engine turns to make power is an ALTERNATOR. Note - there is a shitload of complexity behind this, and if you really want to get into it an alternator can be dubbed a form of a synchronous, single source fed generator, but let's not get into that. For all intents and purposes, to 99% of people the thing on your bike is an alternator simply by virtue of the fact that it produces alternating current.

2. His comments about 500 volts.... You can measure the raw alternator output many ways, and get many readings. A common mistake is measuring the phase to phase voltage and going oh my god it's several hundred volts! The correct readings will be in the order of 80 volts, and even so this means nothing, load it down a bit and the voltage will drop.

500 volts from a bike alternator (humoring the guy for a moment) will not "blow your fingers off. Ever had a static shock? Well guess what - you might have had as much as 50,000 volts across your finger during that shock !!!!!!!! OH NO!!!!! MY GOD WILL SOMEONE THINK OF THE CHILDREN!!!!!!.

yeah.


I am far more worried in my day to day work about the "mere" 48volts that the battery banks i test my new designs on hold......talk about blow your fingers off - people have had big spanners vaporized by those things! And then later that day I go and accidentally stick my fingers across a 450volt PFC bus, go "ow, fukme!" and get on with it.

There's more to it than voltage, and the simple fact of the matter is that 500 volts is nothing to modern power hardware.

3. The power you don't use is shorted to the coil huh? OK.....which coil is this? There are many coils in your alternator......

In a shunt regulator the power that you don't use to supply the load (your battery and lights and so on) is shorted to ground and DISSIPATED partly in the alternator coils, partly in the power semiconductors inside your regulator/rectifier, partly in the wiring joining it all up, and so on. Hell, some of it is even dissipated into various magnetic and electric fields!


So - a shunt regulator is the cheapest way to tackle a typical bike electrical system's requirements, but it is DEFINITLEY the nastiest as well.

A series reg is much nicer but - and here's the reason it is unusual for a bike to have one - it costs the manufacturer more to put in.

HTH your understanding

[notme]

You sure that is exactly the same? The thing that makes me think the jaycar unit is slightly different is because it says it needs a heatsink. To me that implies it is inefficient, which is the very reason for having the unit in the first place. I run my bike now with no voltage regulator and just straight off the batteries (14v down to 12.6ish) so would be worse off with the jaycar unit. Or do I have my wires crossed??

Forgot to mention on other point - if you are happily running your bike off the battery directly, then don't introduce another loss!

The point of the reg is to....well....regulate.... :-) so the only downside of not having one is that your bulbs and ECU and coils and other electrics are exposed to a varying voltage as the battery runs down rather than a constant 12V or so. Bulbs may burn out sooner than they otherwise would (you might not even notice the reduction in life) and the coils will be ok - the thing I'd be worried about is the ECU or any other sensitive electronics.

If you've been running it battery - only for a few days and it seems fine, i'd suggest that the damage just hasn't shown up yet. If you've been doing this for a year or something then obviously everything is holding up fine!