Three phase ST gen question

[tlfrantz1]

Hey Race Fans,,

I have a three phase ST generator and in the doghouse there is a fairly big (4""  x  6") potted circuit board that I believe is the automatic voltage control (AVR?)---can I eliminate this board? I've heard some folks eliminating what I thought was this board. Whaddya think?

Thanks
Tommy

[matt]

can ya post a pic?

[tlfrantz1]

Here's one:

http://listerenginegallery.com/main.php?g2_itemId=1777

Its the large black potted circuit in the upper left hand corner of the box. I thought I heard (maybe from Utterpower) that you can eliminate this thing to make the gen more reliable.

2nd question: Can I place a switch in the rotor winding circuit (exciter) so I can kill the gen's output while its still turning? Will this hurt the residual magnetism? I want to be able to kill the gen's output without clacking on and off the three-pole circuit breaker I'm installing.

Thanks
Tommy

[trigzy]

Tommy,
          It looks like you have a full blown electronic regulator, powered from the main output windings.  You generator may not even have a Z winding, but I cant tell based on the picture alone.  Can you tell for sure where this box get's it's input and output?  Is there anything hooked up to the brushes other than this box?  If you dont have a Z winding, you will need an AVR of some sort.

          As for these AVR's, they aren't as unreliable or as uncommon as Utterpower might lead us to believe on thier website.  Most Honda generators have an AVR of some description, and pretty much any commerical generator (larger than 10kW) will have one.  Ronmar evaluated some AVR's that I imported from China, and found they reduced the flicker quite a bit, and I've sold quite a few, and no one has reported any problems.  There were some questions about the brush life due to the manner in which it excites the field, but the current to the field can be leveled off with some high voltage capacitors.  The AVR would also give you the ability to choose a different output voltage (say 120/208V) if that is more popular in your area.

As for the switch, if the assumptions I made above are correct, DO NOT put a switch on the output of the AVR.  It really should have a load on it at all times, so that the output voltage doesn't soar sky high due to no-load on the regulator and falling input voltage when you open the field circuit.  Instead, put a switch on the INPUT of the AVR.  The output of the AVR might try to spike for one excitation cycle due to the loss of input voltage, but it will only have the energy stored in the AVR's capacitors, and then will immediately stop exciting.  As for the residual magnetism, many of these AVR's will begin to self excite at around 3 volts, so hopefully you wont have any problems in this regard.

Let us know what you find.


Steve

[mobile_bob]

finally an opportunity to ask a few questions of someone that is in the know on AVR's



first of all the stc 3phase heads likely do not have a Z winding, i don't think there is room left for one on the stator.

you mention that the avr allows the possibility for 120/208 output?

most of the stc heads i have seen are 230/400 Yconnected, so
how do you use an avr to get down to 120/208 without losing roughly half of the heads kva capacity? and
do the avr's have this much adjustability?

as for switching, how about just using a 3phase contactor on the alternator output?
you can get 3 phase contactors cheap on ebay all day long, and if one were to get one with a 220/240vac coil
we could get the power for the coil from one phase of the alternator?
to get the switching voltage down to a reasonable and safe level the use of an iso 12 or 24 volt relay could control
the switching power easily i would think?

thoughts?

bob g

[tlfrantz1]

I don't think the gen has a Z winding, although the instructions said something about a harmonic winding. I went ahead and wired in the regulator, and I'm not going to install a switch to turn off the rotor winding. I'm new to this, so I'm going to wire it just as it came to avoid a second 'smoke'. 

[trigzy]

Tlfrantz1,
           A harmonic winding and a Z winding are the same thing.  Without a unit here in front of me, I make no gaurantees to the presence or lack thereof on the 3 Phase STC units.  The head I have is WTW Brushless, so it's not much good for comparison.  Was there a first smoke?? (from the genhead?)

Bob,
       Yes, most of the AVR's are adjustable over a wide voltage range.  For 230/400V, most of the regulators are usually connected between one of the phases (X/Y/Z) and Neutral (N).  Therefore the input of the regulator sees (and regulates to) about 230V.  Since 120/208V is much more common around these parts, I would hook the regulator up between two phases (X-Y, X-Z, or Y-Z)  Then with a small adjustment on the regulator, you can dial the voltage down to 208V across the two phases, leaving 120V between any of the phases and neutral.

       As for the KVA rating of the head after this adjustment, it depends on what the weak part of your genhead is.  The head I imported, I spec.'d to be rated for 120/240V at full KVA.  If I raise the output voltage, I can increase the KVA amount to a point, but I reach a limitation where I would start to supply to much power to the exciter (field on an ST).  You'd have to consult the company that made the generator head on this to be sure.  If possible, placing windings on the same phase in parallel instead of series would double the current  and halve the voltage ratings, but generators with this kind of reconnectability (12 lead) are usually quite pricy.

As for the contactor, sure, as long as you can find one sufficently rated.  Make sure to put a fused disconnect switch or similar before this contactor, and put it in a proper enclosure.


Steve

[mobile_bob]

Steve:

just to get this clean in my muddled mind


if my stc 3phase head is 230/400
that makes it 400volts across any pair of phases xy, yz, or zx

and therefore 230 volts across any phase to neutral

so i would want to regulate the 400 volts down to 208 in order to get 120 from neutral to each leg, but

in doing so wouldn't i be cutting the kva capacity of the head roughly in half?

seems like one would do far better keeping the oem excitation system and using a 3phase transformer
or set of single phase setup for 3 phase
that way you get full capacity (kva) or nearly so, and
by virtue of resolution better voltage regulation anyway?

another question if i might,

can you explain to me how an electronic avr is going to reduce flicker?
its hard for me to get my head around how any avr is going to actively compensate for flicker when it
is always behind the event?

i suppose some sort of microcontroller and some very good coding, so that samples could be made, averaged
and compensation made on the fly,, but
such a controller certainly is not going to be easy or cheap in my opinion.

even at best there would have to be some flicker to sample so the puter could make the computations, and set
the control... which might work ok at a fixed steady load that never fluctuates very often. something most folks
don't have (a fixed load)?

more questions than i can find answers for

bob g

[trigzy]

if my stc 3phase head is 230/400
that makes it 400volts across any pair of phases xy, yz, or zx

and therefore 230 volts across any phase to neutral

That's correct.

so i would want to regulate the 400 volts down to 208 in order to get 120 from neutral to each leg, but

in doing so wouldn't i be cutting the kva capacity of the head roughly in half?

Maybe yes, maybe no, it depends on your heads design limits.  If you cant find out how many KVA your head would be rated for from the mfg, then yes, you would have to assume that it's half.

seems like one would do far better keeping the oem excitation system and using a 3phase transformer
or set of single phase setup for 3 phase
that way you get full capacity (kva) or nearly so, and

If you you knew, or had to assume, that the heads output would be lowered by 50%, and you have the engine power for a higher power rating, than yes, a transfomer is a good idea.

by virtue of resolution better voltage regulation anyway?

This point is not necessarily correct.  I believe the voltage regulation to be linear, so it's not like you have the resolution issues of an A/D or D/A converter.  Also, when using the 230/400V setup, you are regulating phase-ground (230V), and when using a 120/208 setup you would be regulating phase-phase (208V), so you're not moving the measurement voltage by more than 10%.

can you explain to me how an electronic avr is going to reduce flicker?
its hard for me to get my head around how any avr is going to actively compensate for flicker when it
is always behind the event?

i suppose some sort of microcontroller and some very good coding, so that samples could be made, averaged
and compensation made on the fly,, but
such a controller certainly is not going to be easy or cheap in my opinion.

even at best there would have to be some flicker to sample so the puter could make the computations, and set
the control... which might work ok at a fixed steady load that never fluctuates very often. something most folks
don't have (a fixed load)?

more questions than i can find answers for

bob g

I think you're looking to far in to it.  The harmonic winding on a ST head produces an interesting voltage/current pattern, and that varies with the frequency (and therefore the power pulses) of the engine.  The AVR's output (down to the low freq. roll-off point) is independant of the frequency.  This has been shown (by other people - giant thread on this topic) to reduce flicker in incandescent lights.  Others have obtained the same result by supplying the field with a DC votage from another source, or by using capacitors.  It's the harmonics that cause the flicker, the AVR isn't removing it, it's just not introducing the harmonics to begin with.

Steve

[mobile_bob]

Trigzy:

assuming a typical ST head, or specifically an STC 3phase head
10 kwatt at unity power factor, equates to 10kva, so

if you regulate down from 240 to 120 you have effectively cut the heads capacity in half
reason being the stator is limited by the ampacity of its windings. we can't double the ampacity of the
stator 240 windings after lowering the voltage by reduction of excitation to 120vac.

so i am not sure what if any benefit you have in mind?

as for resolution

if we regulate off of one phase, (reference/sample) at 400vac, and the regulator can achieve +/- 20 volts
and then feed the product through the transformer to step down we can get approx +/- 10volts (actually a touch
worse, i don't have my calculator handy)  regulation is always measured in voltage and then expressed as a percentage
of the voltage you are regulating.  so it is generally beneficial to generate at the highest voltage possible and regulate at this
voltage, then use a transformer to step down the controlled/regulated product to the voltage required. this works out in practice
very well in my experiments and is easily demonstrated.

on the subject of flicker:

not sure i agree that the reason for flicker is based on the Z winding, but am willing to listen to your reasoning

how does the harmonic Z winding interact with the stator winding to not only alter the waveform but
how does it make the output flicker? (alteration of the waveform is easily explained)
i can see how the engine slowing on compression lowering the generator voltage causing the light to dim, and
after ignition the engine speeds up raising the generator voltage causing the light to brighten, thus the flicker.

from what i understand even remote dc excitation does not eliminate flicker, not sure if it even reduces flicker?

so how does an AVR reduce flicker?

thanks
bob g

[trigzy]

reason being the stator is limited by the ampacity of its windings. we can't double the ampacity of the
stator 240 windings after lowering the voltage by reduction of excitation to 120vac.

You've made the (likley) assumption that the limiting factor is the ampacity of the stator windings.  That may not be correct for every generator.  Some generators have heavier stator windings, but the total KVA might be limited by the field current.  That's all I was saying - that its a possibility you might not have to derate your head, although I do agree that in many cases it's likely.

so i am not sure what if any benefit you have in mind?

Lower KVA at a voltage that you can use, is better than all the KVA in the world at a voltage you cant use..........  unless you have a transformer.  I wasn't covering all the benefits initally, just the fact that the generator may not have the capability of functioning without an AVR or other excitation system if there is no Z winding.

as for resolution

if we regulate off of one phase, (reference/sample) at 400vac, and the regulator can achieve +/- 20 volts
and then feed the product through the transformer to step down we can get approx +/- 10volts (actually a touch
worse, i don't have my calculator handy)  regulation is always measured in voltage and then expressed as a percentage
of the voltage you are regulating.  so it is generally beneficial to generate at the highest voltage possible and regulate at this
voltage, then use a transformer to step down the controlled/regulated product to the voltage required. this works out in practice
very well in my experiments and is easily demonstrated.

You cant regulate off one phase at 400V, the regulator would have to be across 2 phases.  These regulators cant take that input voltage.
All of the spec sheets I have quote the voltage regulation in % only.
In the example you give,   +/- 20 Volts on 400V is 5%, and most regulators I have been offered quote 0.5% regulation.  So say our regulator is +/- 2VAC.
90% of the AVR's I have been offered have a rating of 180-240V, so that means they dont regulate the 400V, they regulate the 230V.  So that means you would get a range of 228-232V.  When reattached to regulate to 208V you'd get a range of 206-210V.  Nothing even worth mentioning IMHO.  Just to be clear, I'm talking in a situation without a transformer, where you are getting the 208 natively out of the generator head.  I wouldn't recommend regulating from anywhere than the generators main lugs.

on the subject of flicker:

not sure i agree that the reason for flicker is based on the Z winding, but am willing to listen to your reasoning

how does the harmonic Z winding interact with the stator winding to not only alter the waveform but
how does it make the output flicker? (alteration of the waveform is easily explained)
i can see how the engine slowing on compression lowering the generator voltage causing the light to dim, and
after ignition the engine speeds up raising the generator voltage causing the light to brighten, thus the flicker.

I dont know I'm qualified to explain the "WHY", I know that in Ronmars tests, published on this forum, and other comments sent to me by users, a substantial flicker reduction was seen.  Remember the harmonic winding works on the 3rd harmonic of the frequency, so that the frequency changes seen during power/coast cycles of the engine can cause a change in excitation, and therefore voltage output.  The AVR eliminates that possibility.

from what i understand even remote dc excitation does not eliminate flicker, not sure if it even reduces flicker?

I dont know for sure, but I was under the understanding it greatly improved the situation or eliminated flicker all together.  Some more discussion here: http://listerengine.com/smf/index.php?topic=3569.0

Steve