Georgia Generator AVR installed on my ST-5 - 6/1

[Jedon]

Thanks much to Tom at Georgia Generator, I got his AVR hooked up and running.
Good price, fast shipping, great support, can't say enough good things.
The AVR has fixed the issue with my ST-5 where load on one side would drop the voltage on the other causing the first side to stop drawing load and then the over voltage on the other leg would turn that side off also, rinse, repeat. The input was to my inverters which seem to do as they please when it comes to drawing power. I have each one set to draw 8A @ 120V, right now it's locked in at 125V, 60Hz and working like a charm. Flicker seems to be reduced.
I'll report back with more info as I get it, it's only been running a couple hours now.

[BruceM]

That's great, Jedon.  Did you set it up with one leg of AC as the excitation or as a series Harmonic regulator? 

[Jedon]

The instructions said to hook one part of the AVR between the DC excitation and the rectifier, and then each 120V leg also went to the AVR.
So maybe that's a series harmonic regulator? That sounds cool anyway :-D

[BruceM]

I'd have to see a wiring diagram to know for sure. Sounds like Tom dropped the ball on the 120VAC excitation wiring diagram for customers.  Perhaps he just blew me off.  

Flicker reduction using the harmonic for excitation is going to be pretty poor.  There isn't enough extra power to compensate for rpm/voltage sag during the compression stroke.  Over voltage on the power stroke it can correct, and you might see some modest improvement from that.

You won't have to strain to see the improvement if you switch to 120VAC excitation, but I'm glad it's working for you as is.  Also, with 120VAC excitation, most appliances are happier and don't make noises, as the AC waveform is cleaned up significantly.

Edit-  I did not mean to imply that there is something wrong with these regulators- only that in my opinion they would perform better if EXTERNALLY wired to use 120VAC instead of the harmonic winding.  There is NO CHANGE required in the design. 

[mobile_bob]

Bruce:

i got a few questions, an observation, and maybe a theory of operation if i may

1. have you seen a schematic for Tom's AVR, the inner workings of the unit?

2. is it possible the harmonic winding cannot deliver enough current/voltage based on how it is generated and used?

3. is it possible that the AVR in this case does not follow conventional AVR design, in that it was designed to work with the
harmonic winding a source of power?

what i am thinking as a possible explanation for the AVR's reported success is it is engineered to work with the deficiencies of
the harmonic winding.

for instance, the harmonic winding generates power of course, but it delivers it in pulses, some very strong ones as the load increases

what if the harmonic winding instead of being asked to supply power in pulses, generated a continuous relatively level power, that is stored in
a capacitor and that stored power is delivered to the field via high speed switching mosfets or similar. basically the capacitor is the equivalent
to a battery that can deliver relatively high amounts of of intermittent power and be charged continously at a much lower average current.

done right the result should reduce the heavy surge pulse on the harmonic winding and the transformer effect transposing that onto the 60hz
sine wave resulting in that dreaded harmonic dip.

this type of AVR would definitely be different than the classically designed AVR that probably all other generators use to some extent or other.

all i am asking is, is it possible that the unit Tom is selling has a different architecture than what we are used to seeing?  i think it is certainly possible
one thing for sure, before the st head i had never heard of harmonic winding excitation, and quite frankly i know of no other generator heads that use that method, and i can't find any text outlining the theory of operation for that damn winding either,,so it seems likely to me that
the AVR might well be designed to make use of the harmonic winding effectively reducing the flicker and getting the job done.

it would seem that it is possible, given jedon is happy with the results?  so far i have not heard of anyone being upset that the avr didn't perform as expected?

i got lots of questions, and few answers, but

i for one wanna know

maybe before we summarily dismiss the AVR, and make Tom out to look like he is selling something that cannot work, we discuss this more?
maybe do some testing?  it just seems to me like, here we have a supplier of st heads, and others, parts, at a reasonable price, that for all reports
is an honest guy that has treated his customers well, should be given the benefit of the doubt.

especially given the void in suppliers of genheads these days. and the fact Tom is a member of our group.

like i said, i have lots of questions, and few answers

bob g

[rcavictim]

I just placed an order with Tom for a pair of big boy 24 kW, 277/480 volt STC 3-phase, 12 wire heads.  They contain AVRs.  This represents a huge purchase for my budget and was well researched, especially since one basically has to pay double the price to land goods into Canada from the US.  Tom was friendly and very helpful to deal with.  One head is going into my wind turbine, the other onto my Changfa 1115.  I am eager to learn how the AVR's work and how I can control the one which will get variable speed in the wind turbine application.  I suspect on that one the AVR will be completely bypassed.

[BruceM]

Bob,
I sure didn't mean to imply that Tom's AVR's didn't work.  They obviously do.  It's just that they would work better if wired for AC excitation, and Tom indicated to me that they had been tested in this configuration, and that he would be providing a wiring diagram for this.  This is an EXTERNAL wiring change- no modification to the unit is needed.  I've added this to my original post, and I'm sorry for the confusion.  

Based on size and cost, I would assume that the units are a basic design similar to other commercial AVRs.  I don't believe there is any practical way to solve the harmonic distortion problem other than use line AC for excitation; you seem to think it can be inductively filtered (and stored) but I remind you that the stock configuration uses the gigantic rotor inductance as a filter, yet there is still significant AC distortion.

Regarding flicker, as I stated above, some reduction will be apparent from the units using the harmonic, as they will limit the power stroke over voltage.  Could these units be storing "excess" harmonic energy for compression stroke voltage boost? Not at all likely given their size.

Tom is very personable, and is offering them at what I think is a bargain price.  Perhaps someone will do some real tests with both wiring configurations (harmonic vs 120VAC) and give them a review.  Tom offered to give me a unit for this, but I had to turn him down due to health problems.  Maybe Ronmar or someone else with good technical knowledge would give it a whirl?

RCAvictim-
All an AVR does is limit some excitation source to the rotor in order to get the desired output voltage.  I'm sure Tom's units will work fine for your application.  Not as much flicker reduction is possible on the larger heads, because the larger inductance of the rotor prevents fast changes in it's magnetic field strength.  This shouldn't be an issue for your Changfa, anyway.  Also, I suspect the 3 phase heads don't have harmonic distortion (it has to do with winding placement), so you should be all set.
I think for your 3 phase  application using the harmonic excitation is going to work as well as the AC.

[Jedon]

After running it for a while, I don't think flicker is reduced much. I didn't think it would, just wanted it to stop my voltage drop problems which it did.
I think to really get rid of flicker I'll need to use it to just charge the batteries like Geno is doing.
Here are a couple pics of it wired up before I mounted it.
http://picasaweb.google.com/jedon13/MetroLister61#5436010277450688226
http://picasaweb.google.com/jedon13/MetroLister61#5436010315715874450

[rcavictim]

RCAvictim-
All an AVR does limit some excitation source to the rotor in order to get the desired output voltage.  I'm sure Tom's units will work fine for your application.  Not as much flicker reduction is possible on the larger heads, because the larger inductance of the rotor prevents fast changes in it's magnetic field strength.  This shouldn't be an issue for your Changfa, anyway.  Also, I suspect the 3 phase heads don't have harmonic distortion (it has to do with winding placement), so you should be all set.
I think for your 3 phase  application using the harmonic excitation is going to work as well as the AC.

Thanx for the comment Bruce.  It would be nice to discover your hypothisis to be true.  Part of the decision towards this particular head was based on the fact that it can be rewired to be a 120-0-120 single phase 16kW head as well as the high voltage 3-phase head that I need right now.  That gives me all the versatility I might need down the road. 

[BruceM]

Jedon, If you'd like to see more flicker reduction, ask Tom for a wiring diagram for using 120V excitation. This only requires external rewiring of your AVR setup.  He may only provide it for people who ask for it.

On my ST-3 with my own simple AVR, with 120VAC excitation,  I got from not being able to use the generator AC for any lighting (almost instant onset of complex partial seizures), to be able to use my shop heat lamps (250W) without any problems, and 100 watt bulbs are only slightly "weird" feeling. (I can work with them for a couple hours.)  I have epilepsy and bad flicker sensitivity.

I think with a "smart" AVR specially designed especially for flicker on slow speed engines, the correction could be even better, by leading the compression stroke with a voltage boost, and perhaps briefly shunting the field during the power stroke. Alas, I think the market is too small to attract any attention..   

RCAVictim, If you have a o'scope or can borrow one, we can find out if the "harmonic hump" AC distortin is present in the 3 phase ST's in a matter if minutes.  It's plainly visible if present. Depending on your application, the "harmonic hump" may be of no concern.  I found that it made my appliances groan and buzz, especially the AC timer on my washer. You can always add an AC driven AVR later if needed. You have lots of options!



 

[mobile_bob]

Bruce:

thanks for the comment, and sorry i missed it till this morning

i am not convinced the AVR could be done by filtering, but am thinking something along these lines

what if one were to take the harmonic winding power, (apparently it fluctuates depending on power drawn from the stator)

and use this power to charge a buck/boost sort of controller, wherein
on the combustion stroke the increased power could be absorbed in the controller (either a cap or inductor?) which might effectively
shave off the voltage spike or surge, and
on the compression stroke as the voltage starts to sag the controller could return the stored power to the field on top of it normal power
with the effect of propping up the sagging voltage.

what i am wondering is, perhaps the harmonic winding being loaded more evenly across the full rotation rather than coming on and off
more effective power could be recovered/stored and returned to the field than it is able to do intermittently?

one  would probably have to use a crank position sensor or a speed sensor, a micro controller, a buck/boost converter, perhaps some
large caps or inductor for absorption and storage, and some clever coding?

probably well beyond the markets willingness to pay?

i guess to do it right, reliable and marketable the thing would probably be as costly as an mx60 or the equivalent in price to the st head one
it trying to clean up?

probably why we haven't seen something like this yet, huh?

what i am describing is an electronic and digital version of the oil point breaker type used by kohler or onan years ago
the old system worked remarkably well in limiting flicker on those old heads and they had some pretty high inductance in the field as well.

maybe it might be interesting to see if the old method could be appropriated and used on an st head, that might be interesting
to someone with the motivation to get into the weeds and sort it out.

always many questions, and few answers from me!



bob g

[mike90045]

Do any of the AVR's eliminate the need for the field winding/brushes ?  Or are any styles better at preserving brush/commutator life ?   Or is that getting into the realm of the PM alternators ?

[BruceM]

I think there is no point in trying to wring stored energy of the troublesome harmonic, which is barely able to meet the excitation demand. You can't get blood out of a turnip, even with fancy electronics, no matter how fast you switch. And given the distortion caused by the harmonic despite inductive filterng of the rotor, I'd say it was a hopeless engineering endevor.

Fortunately the AC side typically has lots of power to spare, even on my lowly ST-3.  So there is no good reason not to use the AC as the excitation source.  A compact design for special flicker reducing AVR would use a PFC front end, perhaps 350VDC bulk storage, along with acoustic sensor for the HP injector line, embedded processor, and I think a field coil shunting ability.  

How many could I sell at $1000 each?  

[BruceM]

Mike-  

No AVR eliminates the need for the field coils and brushes.

Some measurements I have seen indicate that the inductance of the field coils smooth out the current draw across the brushes so that pulsed excitation shouldn't be an issue for brush life.

The stock harmonic setup pulses 4 times per 16.67ms (60 Hz cycle), each pulse is only a couple of ms wide at the base, very sharp at the peak.  On a surge load, the ST3 harmonic peaks over 300 volts.  (I think I captured a 400V peak once.)

My simple AVR design posted here and on LEF normally does switching during the lower voltage portion of the rectified AC used for excitation.  (120Hz pulses)  It switches at higher voltages only during surge loads and during the engine compression stroke.  But from what I know now, I don't think high voltage pulsing is a big deal for the brushes, anyway.  

[rcavictim]

RCAVictim, If you have a o'scope or can borrow one, we can find out if the "harmonic hump" AC distortin is present in the 3 phase ST's in a matter if minutes.  It's plainly visible if present. Depending on your application, the "harmonic hump" may be of no concern.  I found that it made my appliances groan and buzz, especially the AC timer on my washer. You can always add an AC driven AVR later if needed. You have lots of options!

I have an o-scope for every day of the week. Plenty of test equipment, HV lab and RF/microwave as I try to build and operate an astronomical radio observatory.  I used to make Tesla coils commercially (longest terminal to terminal streamer in air achieved = 18 feet), and generally dick with many ideas based on electrical/electromagnetic energy.  I sold off most of my HV lab though because of the extreme danger to life, working alone and no accidents allowed nature of the work since I was compromised in a bad highway head-on auto collision. I am now dependant on powerful painkillerz to go about my daze.  I spend many more daze in bed than up doing stuff.  That is why it is taking me so many years (3 so far) to build my very ambitious wind turbine.

The photo shows a prototype twin Tesla coil operating in my lab about ten years ago. It is a time exposure so it has recorded where the discharge paths were going over a period of perhaps 8 seconds or so.  That was a very impressive display of electricity to be with in the same room. Clear air distance between terminals in this picture is 11 feet.  Distance from floor to grounded sheet metal ceiling is 10 feet.  Those were the good old days.  

Sorry for the slightly OT but this IS the generator forum and these ARE electric sparks!