Regulator Rant...

[Dualfuel]

Thank you Bruce,
Unfortunately, I am unable to purchase anything at this time. I would be very curious about what an AVR circuit looked like.

I have lots of junk. I am hoping to be able to learn how a voltage regulator operates and adapt the junk I have, to the application. I am by no means fanatic or crazed but am busy learning how this works. I am enjoying thinking about the magnetic fields and what the feed back loops would have to do.
thx BPJ

[Thob]

Bruce -  I'd like to see that schematic and PC design for the regulator.  Do you know what it would cost for the parts for it, roughly?

Thanks.

[thomasonw]

Thank you Bruce,
Unfortunately, I am unable to purchase anything at this time. I would be very curious about what an AVR circuit looked like.

I have lots of junk. I am hoping to be able to learn how a voltage regulator operates and adapt the junk I have, to the application. I am by no means fanatic or crazed but am busy learning how this works. I am enjoying thinking about the magnetic fields and what the feed back loops would have to do.
thx BPJ

Hi, so some background questions...

First, have you confirmed the AVR you have is the problem?  Doing the quick go/no go test in the shop manual link you posted early on?   Did you check the fuse?  (If only I could say the times I overlooked something simple were few...)

Assuming it is bad, what load are you driving with the generator?  And specifically, how sensitive it is to 'over voltage'?   Is it pumps and such, or just a heating element?   My 'creative thinking' is:  If you are driving a load that is really not too fussy about the voltage / frequency given to it, you might be able to do something quick-n-dirty.  Like, just use a bridge rectifier to get a high voltage DC, then run it through a series of auto light bulbs - maybe start out with 6 or so, to cut the Field voltage down to something usable.  Play with adding and remove light bulbs until the  produced AC voltage is close to what you need.   But also know, ANY change in the load will through this all out the door.  So, if you are driving a heater element that is cutting in and out - wow..   But if it is a constant load, and one that would not mind being over or under voltage say 20%, perhaps this will get you going?

If all you are trying to do is 'get the Hay in', and will fix things right over the winter  - - -  Tell us a bit more about what you are using the AC generator to drive...


-al-

[Dualfuel]

Thank you for reading my earlier posts....I will review the diagnosis of the generator


Originally it was in a motor home, and the engine would not run. I removed it from the motor home, and built an ignition circuit for the engine. I got the engine running, and then noticed that the generator worked...
I forgot about it and it rained on it. Then I noticed the generator was not making electric power anymore...testing led to the discovery that there was no current or voltage at the rotor.
I hooked up a 12volt dc wire to "flash" the rotor...the engine gave a grunt, and I heard a pop. I smelled the magic smoke. I couldn't see what had happened until I shut off the engine, looked at the .pdf manual and went looking for the voltage regulator....sadly, it was this component that exploded. Not sure why, as I put the 12vdc on the wire that it was supposed be on, to flash, or "build up" the rotor magnetic field. Regardless, I know that the regulator is a useless blob of melted plastic...but the underlying problem must have been in the decision maker.
These Kohler parts are very very expensive and cannot be purchased...yet the engine is superb..very quiet at 1800rpm. Also, the more I learn about the generator, the simpler it actually seems.

[BruceM]

OK, I'll look up my schematic and post it somewhere with a link here.  Maybe by tomorrow...
Bruce

Here's a PDF file with schematic, instructions, parts list, etc.

https://www.dropbox.com/s/8s60dog9ifskre6/ST5%20120V%20AVR%20Multi%20PDF.pdf?n=149234369

I have the schematic and layout in Expresspcb.com's free program form if someone that wants that. The board is their standard experimenters board size, last I checked it was $60 for three boards.

Not much interest in this since affordable AVRs became available.  

How it works:  The regulator rectifies the 120VAC into DC, with the positive lead connected to the brushes.  The negative side is switched through a logic level gate, N-channel Mosfet, whenever the AC voltage drops too low.  A single op amp comparator drives the Mosfet gate.  Since the rectified and R-C averaged AC voltage is lowest during zero cross time, the unit will tend to switch on during lower voltage times.  This is easy on the AC waveform shape, and easier on the Mosfet.  The circuit does NOT do an RMS conversion of AC voltage so if you apply very poor power factor loads, the voltage regulation will be crap.  I tried doing a true RMS version, but the analog RMS conversion chip introduces too much lag and the flicker correction is much worse, so I went back to the original circuit as shown.

Let me know if you have trouble accessing the file.

Bruce

[Thob]

Bruce  -  I was able to get the PDF from dropbox.  Thanks for sharing the circuit.  It looks like I could build it on a perfboard with no problems.  I wasn't able to find the transformers on digikey, they must not carry them anymore.  I might have to do some substi-tootin.  But then their web site is really slow for some reason.

Thanks!

[BruceM]

You're very welcome, Thob.  The Mosfet MUST be an N-channel logic level gate type comparable to the one on the parts list, or it will fail during the start up.

PM me for any assistance and my phone number you need if you do decide to build one.  I know of at least two others that did it on perf boards. This was designed as the simplest AVR possible.  

I'm still running my own perf board prototype, which also does high/low voltage detection, field polarity reversing via DPDT switch, flash field, allows "manual reversion" to the harmonic winding only, and does not require pushing on a start button.  I meant to upgrade it to a PCB but haven't gotten the "round toit".  Starting over now, I would not bother to support the stock harmonic setup, and instead would just make a backup board.  On my prototype, I used a couple Darlington power transitors (BU924 I think) to switch the low side, and slow the rise/fall time to reduce EMI on the AC power.  (fast switched Mosfets do cause some high frequency ringing)  This plus 0.1 uF metal film capacitors parallel to each diode in the bridge makes for pretty clean AC.

[Dualfuel]

Dear Bruce,
Thank you for posting the regulator circuit. I find it absolutely fascinating what can be done with Mosfets and modern electronics...

Would I be out of place, here, on this thread, if I asked to explore other ways to regulate voltage? I promise, sometime tonight I am going to go dig out my Stockel's and review the mechanical voltage regulator used on alternators. Also, I will dig for my GM service manual that gave the step by step build up of the solid state regulator in the 10si series. Til, then, something you wrote, that a mechanical regulator would be too slow...has me puzzled. The four pole alternator on the Kohler turns at 1800 rpm...its 60hrz. Are you saying that a regulator that operates on a much higher frequency and three phase, would be too slow to react to 60hrz?
Look, if these questions are inappropriate, I sure understand. I don't waste my time teaching people how to make cold weather biodiesel either...its just in this case, I sense there is a fundamental problem that I am missing and I better get a grip on it, so I can head in the right direction.
The problem as I define it right now, is that I have to take a signal from either the harmonic winding or the stator winding proper, convert it to a signal that the regulator is designed for, then convert the signal from the regulator to a signal the rotor can use.

What I know so far is the rotor is flashed with 12vdc. The original regulator was tested with 120vac, and so it stands to reason that that is the operating voltage of the harmonic winding. I have been repeatedly told the the rotor is using pulsed dc. I cannot prove the pulse is anything other then 60hrz, @ 120vdc or there abouts as its adjusted.

Anyhow, I am still very excited about this problem...and after walking around the place, I found another generator that is a four pole Kohler powered, and a two pole without regulation as well. So its readily apparent that I have got to get "up" on how this stuff works. So thanks for reading this, and thanks for any comments that you thiink will help.
BPJ

[BruceM]

My AVR typically gives the rotor a little piece of the full wave rectified AC (think back to back humps of 120V peak) during the lower voltage portion.  If you try to do that every 8.3 milliseconds with a relay, it won't work, because of the lag in mechanical relay response, and because even if you had a "miracle relay" the contacts will fail very quickly.  In addition,  the deadband for on/off on a relay is just too broad to be useful for this. 

Alternately as someone suggested, you could connect the harmonic to a bridge diode, and put an adjustable resistor in line with the output.  Guess at 20 ohm, 200 watts.  If the harmonic winding regulates fairly well, this might be barely adequate.

Perhaps you can find an AVR off of a scrapped generator head. 

[Dualfuel]

 8 cyl engine turning 3600rpm cycles its breaker points at 240hrz. I have several of these engines whose hour meters have clocked over 4000 hours.
Sorry, I just couldn't help it.
I won't bug you about this anymore...BPJ

[Dualfuel]

Okay, I have been having a blast studying up on voltage regulators. This morning I was enjoying reading about Diactor regulators...they seemed super simple. The drawback is that they are from the last century, and I don't have any of them laying around.
I did manage to find my...Auto Mechanics Fundamentals by Stockel. and my Delco Remy Training Manual DR5133S. On page 390 and 391, of Stockel, he describes a single unit mechanical voltage regulator...whose operation is as follows: At start up, the lower contact set is engaged offering the field 2amps @battery voltage. Rising voltage creates a stronger magnetic field in the pole piece drawing the lower contacts apart. This sends 3/4amp to the field. As the alternator speeds up even more the pole piece engages the top set of points. This grounds the field and no current flow is offered to the field. Three positions, 2amp, 3/4amp, and no amps. What I found fascinating is the amperage offered the rotor of a Delcotron is the same as that offered to the rotor of the Kohler. Hmmmm.
I thought about the difference in stator outputs, less then 1200watts for the Delcotron and 4500 watts for the Kohler. I suppose the difference is how long the field is offered the two amps before stator voltage rises enough to need regulation. Its probably very quick in the Delcotron. In fact, the training manual says the regulator can switch between 10hrz and 7000hrz.
The training manual gives the build up of a 10SI voltage regulator...which is basically two transistors and a zener diode, with an accompanying group of resistors and a capacitor. The manual suggested that the zener breakdown voltage was 8vdc.
Another remarkable thing about both these regulators was that both can regulate the current offered them.
The one major difference between these two types of alternators is, the automotive alternator operates at varying high frequencys while the Kohler's is fixed. Not sure what it means to the regulator....
So this research is allowing me to design the test I want to perform on the Kohler...if I can find out what the harmonic winding offers in the way of voltage and power, I think I can adapt it to either a single unit mechanical or a 10SI regulator. The other test is to see what the harmonic winding output is compared to the stator winding's output.
DF