Generator with ferrite magnets

[bob g]

in an effort to clarify what i have written earlier as it relates to efficiency

(i wrote this all earlier but either forgot to post it or it didn't post?)

1. my need to measure efficiency may well be different than others needs, and i accept
that.

2. my measurement of efficiency is "relative" to either comparing modifications or one unit against another.

3. my use of a resistor bank as a load is only to satisfy the need to remove as many variables as possible, i can think of few thing more variable than battery bank charging, and conversely few things less variable than resistance load banks.

4. all my testing is done with an engine drive as a prime mover, something infinitely more stable and less variable than windpower on its best day.

having stated all that, i would also like to add the following

i have long been a proponent of load matching the blades and available wind resource to the battery bank, beit via mppt or buck converter or some sort of smart control such as a controlled rectifier and microcontroller based system, or some combination?

at the present time it is apparent that the most efficient alternator likely is not the best match for the typical wind generator in the typical installation, therefore getting all knotted up about getting the most efficiency from an alternator is not only a difficult proposition but more likely not even desirable.

at which time some sort of electronic load matching becomes available, "then" folks will likely start to see the need for a more efficient alternator. seeking the highest efficiency and having the controller match to the load would provide the most return on what is presented to the turbine.

so this is why i work on efficiency and measurement of, it is from a long view and the fact that i am designing around an engine as the prime mover.

hopefully this makes sense

bob g

[ghurd]

I ran it a bit to make sure everything is working on my test setup.  I found out the generator is actually running around 42 volts with the battery bank at about 29-30.  I have an old silo unloader cord hooked up to the rectifier - about 100 feet of 8/4 wire.  And two #2 welding cables about 150 feet long from the rectifier to the battery bank.

I have to figure out why I'm getting that much voltage drop with that big wire.

Why?  
Partly, because of those "pulses", IMHO.

"test setup"?
Man, I have been waiting for somebody with this setup, except it needs a few wire and rectifier rearrangements!
Whats the chances you want to try something (almost) new?

The 2-ph has 4-wire output?  
Connect the 4 wires to the 4 rectifier terminals.
Move the 8/4, and connect it directly to the 4 rectifier terminals.
Keep the 4 wires of the 8/4 separate!  Meaning to NOT connect pairs of Pos or Neg together at the rectifier end.
At the other end of the 8/4, connect them together normally with the two #2.

What I think is the magic part.
Hang a BIG FAT cap on each rectifier output.  Something like maybe electrolytic 5800uF 100V (some on ebay now for cheap), or maybe 100,000uF 100V?

Then test the output.

Might want to try it with the battery at the end of the 8/4 too.  (no long stretch of #2)

I doubt those caps would last long...  I don't really know.
I think they would last long enough for a test run.  Want to keep an eye on the cap temps.  And wear safety glasses.
Eventually or permanently speaking, probably need some fancy caps to deal with it without overheating.

I tested it on 1-ph of a 3-ph PMA on a much much much smaller scale and got a 7% increase IIRC, and it started with nowhere near the voltage losses you have.
I do not have the equipment to do a test with much resolution.

A friend has about the same resolution equipment as me, but larger PMAs.
He tried it with motor run caps and the output went down.  
Not sure if it was the measuring devices or the caps, though I'd lean toward the caps or a combination of both.
Then again, it could be the increased efficiency deceased the RPM from overloaded prime mover?  
I have no solid idea.  Betting on the caps were the problem.

This is nowhere near the Aussie's use of caps to gain power from F&Ps.  Not even a similar concept.
The only thing I think it does is smooth out the losses from the wire losses produced by the pulses.
I have a feeling it will be most effective at about 2x ~ 4x cutin RPM... just a wild guess.  I can not accurately vary the RPM to my choosing.


I do believe 3-ph is more efficient if everything is matched close enough to compare apples to apples.
But 90% of mine are 1-ph or 2-ph.
G-

PS- Let me know if you want to give it a whirl.  I have some (enough?) caps that may be OK for a test but they'd need a big pile of them in parallel, twice.

PPS-  This meets Bob's #2.

[ChrisOlson]

The 2-ph has 4-wire output?  
Connect the 4 wires to the 4 rectifier terminals.
Move the 8/4, and connect it directly to the 4 rectifier terminals.
Keep the 4 wires of the 8/4 separate!  Meaning to NOT connect pairs of Pos or Neg together at the rectifier end.
At the other end of the 8/4, connect them together normally with the two #2.

What I think is the magic part.
Hang a BIG FAT cap on each rectifier output.  Something like maybe electrolytic 5800uF 100V (some on ebay now for cheap), or maybe 100,000uF 100V?

It is four wire output.  Two wires from each stator going to a bank of 35 amp block diodes.  I got four of those block diodes on each stator with 10 gauge wires from the DC outputs to the positive and negative studs where the #2 welding cable hooks up.

But I'd have to see some sort of diagram to understand what you're talking about with this hookup.  I got some big electric motor starting and run capacitors, but I think they're AC only.  I probably got a dozen of those, of different sizes, laying around here.

I did make a test run tonight with my AVR hooked up to measure amps.  With the Honda at full throttle (It's got some reserve power) I can get the input shaft to 400 rpm and the generator to 900.  Going thru the 100' 8/4 silo unloader cord plus 150 feet of #2 welding cable to the battery bank, the AVR measures 102.7 amps into the bank with the bank at 30.3 volts.  The generator was running at 44.7 volts.  According to my calculations, that's 3.1 kW output into the battery bank.

Not too shabby for a two phase, dual stator, with ferrite magnets.  All it takes is a good prime mover driving the pump.  The stators warmed up pretty good within five minutes but they didn't smoke or burn out.
--
Chris

[ghurd]

Like this...

[ChrisOlson]

Ah, now I see.  So you put the capacitors across the DC outputs - will those motor capacitors that I got work?  My camera went tits up but I'm holding one in my hand right now that says 220-250 VAC, 216-259 MFD, and it's about 4-5" long and probably 2" diameter with two spade terminals on it.
--
Chris

[Flux]

Capacitors will have to he huge to even touch the ripple in those cables.

Flux

[ghurd]

That cap is something like the ones my friend tried, and the output went down.  I think they probably leak too much.  (but that kind would take the heat).

I have some 1,000MFD (4X bigger) 63V.  Only about 1/2" dia x 2" long?  Like I said, a BIG stack of them may be good for a test, but I don't know enough about caps off the top of my head to say if they would deal with that kind of amps (heat).

You have a PM.
G-

[fabricator]

What would be the purpose of this experiment? Smoothing out the output?

[Flux]

I haven't follower the argument so forgive me if I missed the point but as I see it the common cable has fairly smooth dc. The individual cables after the rectifier up to the common junction have a fairly heavy ripple and the peak volt drop will be much more than the mean if carrying smooth dc. This is for test purposes so it may be worth looking at to get alternator figures. in real operating life I suspect all you will do is throw the losses into the alternator windings at the expense of reducing line loss.

Flux

[TomW]

What would be the purpose of this experiment? Smoothing out the output?

Fab;

I am not Glen or any kind of expert but...

Probably to capture the peaks of any power spikes? Caps can collect them better than a battery. I doubt the ripple is an issue to a battery but I don't think they can absorb them like a capacitor?

If you put a cap on a line with spikes the cap will charge up to the voltage of the spikes.

Tom

[ghurd]

My thought was it would smooth out the current pulses.  Smooth out the current pulses, lower the line loss during the pulses.

The 7V line loss you have will charge the cap up almost 7V higher than the battery.
When the PMA goes into the part of the cycle where the amps are lower or 0, the amps stored in the cap start moving into the battery.
Kinda Sorta like those crank flashlights that store some power in a cap while the LEDs are on.  They have a voltage dropping resistor instead of a voltage dropping wire run.
Makes perfect sense in my head.   

I think it would help on IRP too, if someone wanted to run all 6 wires all the way to the battery.

Like I said, I tried it on single phase and it worked for me.  I posted about it several years ago with less favorable than expected responses.
I don't have the stuff to try it large scale or with any accuracy (especially regarding accurate PMA RPM), but the expensive and cheap meters both said the amps into the battery went up for me.
G-

[ChrisOlson]

I'm getting my head wrapped around the concept here.  I don't know for sure yet, but I think the power transmission characteristics of the two-phase IRP generator is what makes it run so well on the turbine.  In other words, the less efficient power transfer (as compared to three-phase) is allowing high enough losses in the line so the blades don't stall.

This is where testing this could get tricky.  I'm winding two new 12 AWG stators for my test gen today, but when I test it, it will be tested at different RPM increments and output measured vs input to determine efficiency with a real world wire run to a battery bank.  Any time you increase the efficiency of one component (like line loss) you will improve the output on a bench test holding the generator at that same rpm.

What happens on the turbine is another issue because increasing the efficiency of that same component may load the generator harder.  This in turn causing the blades to run slower and make less power, in the end making your efficiency improvement on the component have a negative effect on overall output.  Increasing the power transmission efficiency on the bench will show higher output with the same input power.  But on a turbine, efficiency increases downstream from the gen generally reduce operating voltage of the generator and that lowered voltage translates to reduced rpm.  So this is why I feel that bench test results only show what the generator is capable of, but are not indicative of it's actual performance on a real turbine unless you can achieve that elusive "match" to blade power with it.

So again, my gut feeling is that getting that "match" requires the reduced transmission efficiency of the two-phase setup and that's what makes the turbine run while a three-phase just stalls it and makes less power even though it might be more "efficient".
--
Chris

[ChrisOlson]

I didn't get the 12 AWG stators done today.  So while one of them was setting up in the mold I put the pressure and flow gauges on the hydraulic motor and made a test run on the original 13 AWG stator shown in this thread.  I had never tested this generator except on the turbine.  Below is the chart I made of the results - I plugged them into a spreadsheet I got that figures out the shaft power from the motor using lookup tables.  The gen output was measure with a Fluke volt/ammeter and the output at the batteries with a Sun AVR DC tester.

My wife was running the electric clothes dryer at the time I ran the test.  That dryer pulls about 4 kW and I could not get the bank above 28.0 volts, which hurt the efficiency some on the top end.  I feel this 13 AWG generator is good for 60 amps but I ran it up to 2 kW output briefly, then backed it off to 50 amps and ran it a steady 50 amps for most of the afternoon, trying to keep up with my wife's power using problem in the house (we had a poor power day with no wind and no sun here).

The REF column on the right is for my reference to compare difference in output watts vs input, due to the way the spreadsheet calculates shaft power.  If the number is positive then the motor was making a less watts than what it shows.  If the number is negative then the motor is making a few more watts than it shows.  I put the REF column on there to make the comparison between actual measured data and calculated data to make sure it jives.  If the number is close to zero then it's pretty close.  If it's 100 or more either way then I got a measurement problem and need to re-do the power measurements to fix the discrepancy.



--
Chris

[ghurd]

I haven't follower the argument so forgive me if I missed the point but as I see it the common cable has fairly smooth dc. The individual cables after the rectifier up to the common junction have a fairly heavy ripple and the peak volt drop will be much more than the mean if carrying smooth dc. This is for test purposes so it may be worth looking at to get alternator figures. in real operating life I suspect all you will do is throw the losses into the alternator windings at the expense of reducing line loss.

Flux

Flux,
I see what you are saying.

Given 2-ph, I do not follow "in real operating life I suspect all you will do is throw the losses into the alternator windings at the expense of reducing line loss".
Each phase will not operate below the battery voltage.
The cap will only charge when above the battery voltage and line loss.
The cap charges more efficiently than the battery (maybe?).
The 1A the cap absorbed while the line loss was 7V (7W loss), is supplied to the battery later at a 100ma increase with exponentially less losses (Delta 1V, 0.1A, and a 6.9W gain?) .

I know that is not well articulated.
I tried.
G-

[ghurd]

"I think" my idea kind of splits the difference.

The caps can never get lower than the battery voltage.
The less the line loss, the smaller the caps look, and vice versa (*).
It is kinda sorta the same way you are thinking with the whole "less flux and semi-load-matched with ceramics", but inside out and backwards?

(*) Like Flux said, to slightly/effectively reduce the ripple would take caps the size of refrigerators or VW Bugs in your system.
I am not proposing a completely filtered, constant, DC current.
I am proposing a very slightly filtered current.

(**)  Like Flux implied, it may not work with 2-ph.
I do not have the equipment to suitably test it with 1-ph.
Therefore I don't have the equipment to adequately test 3 lines (ph #1, ph #2, and combined) at once.


"my gut feeling is that getting that "match" requires the reduced transmission efficiency of the two-phase setup and that's what makes the turbine run"
I am NOT talking about increased PMA efficiency.
I am talking about less line losses.
I am talking about shifting the large losses (at the time they occure) to be stored in a cap, then later send them to the battery when line losses are much less.
I AM talking about more amps into the battery, and very cost effective amps.
G-

[ChrisOlson]

I am talking about less line losses.
I am talking about shifting the large losses (at the time they occure) to be stored in a cap, then later send them to the battery when line losses are much less.

How much capacitance does 300 feet worth of 2 AWG welding cable have in it?  It was always my understanding that big long cables like that already act like capacitors but the problem of capacitance is more pronounced with AC in the cable because the current switches directions and causes the cable to have to "charged up" 120 times per second.  With DC capacitance is not an issue, even with ripple, because the current never changes directions and the capacitance of the big long cable smooths it out anyway.

That's why, if your gas or diesel powered portable welder is stuck in the back of your pickup, and your job is 200 feet away from the welder so you have to stretch out a bunch of cable to reach the job, you use DC.  With DC you get a nice smooth steady arc.  With AC the arc "flickers" and is not smooth and steady.
--
Chris

[Flux]

There will be capacitance between cables and from cables to ground, both will be tiny in this case. Things are different with oil filled underground power cables when the run is long.

Inductance is more likely to be a factor for the sort of cables you are describing and the effect may be noticeable on that sort of length at high welding current especially if the cables form a loop. running them parallel or twisted together  would help.

I don't want to comment on the power curves at present except to say that they make basic sense. As far as blade matching goes you need to include the line loss and the alternator loss. At full power this is below 50% and that is typical of what is needed to get the blade speed to rise to keep the prop efficiency up.

With a fixed voltage after the rectifier, direct connection and no means of changing winding ratio this is inevitable. Even a small increase in curt in speed will let you run higher efficiency at the high wind speed end and it won't be very noticeable at cut in. I still believe most people aim for too low a cut in speed and either live with stall or end up opening the air gap ( in which case less turns would have given a stator with less losses and heating.)

The prop power curve in low winds is very flat, an efficient alternator has a steep curve. I must do some tests on single phase and see what the curve looks like for an axial, it is essentially linear for 3 phase, 6 phase does have a flatter tail at the low wind end and may help. I always rejected it because the absolute efficiency is lower but in real life matching is much more important than efficiency and anything that reduces the slope at cut in is probably very desireable.

Flux

[ChrisOlson]

I must do some tests on single phase and see what the curve looks like for an axial

I think that's what I've tried to get across, Flux.  Even though single phase looks "bad" from an efficiency standpoint, the loading characteristics on the generator, and the resulting power curve are good.  And that's what makes it work.  It's really hard to explain when the general consensus is that three phase is the "only thing to use".  Stacking two single phase generators increased my power output to more reasonable levels for the rotor size.  But the IRP configuration retains the inefficiency of the single phase power transmission in the line, without heating the stator, which makes the rotor have to turn more rpm's to get the power up.  So ti's not as "stiff" as three phase.

I don't know how to explain it very good.  I think Jerry discovered this same thing even with three phase IRP quite awhile ago.  It works, that's all I know.  And it works good.
--
Chris

[SparWeb]

Does anybody mind if I take a stab at the capacitor subject?

(I can hear the chorus of "no" already)

I've tried them and they seem to work.  I've never pictured it in terms of line losses, but I have long AC runs...  hmmm...  I used to think I was counteracting stator reactance, but it doesn't seem to be that either.

I've put capacitors across AC lines on my 3-phase PMA and seen a small increase in total power to the battery.  Impossible to see on spot measurements, but with a few 1000 data points from the logger, I can see the higher average.

Each capacitor is put across two phases, so you end up with a "Delta" of caps when the stator is in "Wye".  Each capacitor gains its charge at times that the Wye voltage across the two phases is coming up to the battery voltage.  For the time that current flows on these Wye phases the capacitor does nothing but keeps its charge.  At the moment the phase voltage drops below battery voltage, the capacitor now has the chance to deliver its charge instead of the two Wye phases it's on.  At the same time, there are two other capacitors doing roughly the same thing and one more Wye phase that I haven't mentioned.  That Wye phase is high when the others are low and there is one more fully charged capacitor in the mix on its phase, so the two capacitors have plenty of charge to send into the battery while the high Wye phase "holds the door open" (the rectifier is "on").  The cycle repeats for each capacitor in turn, once forward and once backward, for each of the 3 so there are a total of 6 little discharges into the battery for each complete 3-phase cycle.

The effect is 100's of mA on a PMA running at 10's of Amps.  The effect is stronger at low wind speeds but this is the result of the caps taking a rather constant amount of charge (so many Joules) and in low wind speed it's larger compared to the PMA's output.  It does little for the high power range...  The benefit of capacitors will therefore vary depending on your blade matching and weather conditions.  They won't fix a bad match, but they can make a good one better.

I see no way for this kind of interaction to work in IRP, and my experiments also bore that out (I didn't try Delta).  That's not to say that a different configuration of capacitors wouldn't help - I just haven't tried a different configuration, and if I did, my explanation above wouldn't apply.  Can't say much about the 2-phase PMA system either, sorry.

With the usual cheap ~1000 MFD electrolytic capacitors you have to pair them back-to-back to make a non-polarised set, so in total 6 are required.  I collected some 2200 MFD capacitors and they're the ones that actually showed an effect.  I suppose bigger ones will do more, but Flux pointed out you can get bigger and bigger but maybe there's not much to gain.

[ChrisOlson]

I suppose bigger ones will do more, but Flux pointed out you can get bigger and bigger but maybe there's not much to gain.

I went to the motor shop at noon to get some 12 AWG magnet wire.  I asked my friend there about big capacitors.  He has five big capacitors that are about the size of a Folger's coffee can.  It says on them that they are 1 farad, 16 -20 volt DC.  They have 1/4" studs on the top with nuts and it looks like you can hook battery cables up to them.  He said I can have them if I want - he thinks they came out of a UPS.  Would those work?
--
Chris

[ghurd]

A quick note about SparWeb's circuit and those caps,
They would NOT work for SparWeb's circuit.
Not enough of them for one thing.
And I think the voltage rating would have to be a lot LOT LOT higher.

They would need to be 2 in series for my circuit in a 24V system, and then the rated voltage is still questionable at best.

They would be dandy for my circuit, for a 1 or 2 phase, in a 12V system... unless they leak a lot.
If they leak a lot, they would be draining the battery in below cut in winds, and make a test look bad (thats what I believe happened when my friend tested it with motor caps).

Probably want to get the caps charged up through a car headlight or something before connecting to a battery.

Free is Good.  I like Free.
I'd charge them up to the battery voltage through a car headlight, give it a minute or 2 more, then watch how fast the voltage drops when the battery power is removed.  If it drops a 1/4 volt in a minute, maybe they ain't so great?


Lightning strike thought!
I have 2 online videos that show why I think what I think.
Start with the amps are flowing, then not flowing, then flowing......
The voltage loss is there, then not there, then there.......
The current is stored when there is a voltage (power) loss.
It is sent to the circuit later, when the voltage (power) loss is lower.
I^2*R

Microscopic display,
(pic)
http://i701.photobucket.com/albums/ww20/ghurd1/Stray%20Pics/PICT0398.jpg
(video)
http://s701.photobucket.com/albums/ww20/ghurd1/Stray%20Pics/?action=view&current=PICT0410.mp4

Bigger, but small.
The resistors are the same as your LONG wires.  When I touch the power wire to the cap, it 'simulates' the time of the cycle when the PMA is sending amps to the battery.  Consider it the ultra slow motion version.
I said the cap is 10,000uF, but it is 15,000.
(video)
http://s701.photobucket.com/albums/ww20/ghurd1/Stray%20Pics/?action=view&current=PICT1355.mp4

G-

[Flux]

Sparweb
Capacitors on ac lines, is very different from having them on the dc side.

You don.t say whether you have this on a motor conversion or an axial either. With machines running normal ac loads much of this is understood. The armature reaction of a machine reacts on the main field and the effect depends on the load power factor. With inductive loads the eventual effect without going into too much detail is demagnetising and the load regulation will be worse than the resistive load.

Capacitors work the other way and increase the main field mmf and in extreme cases the voltage will rise with load rather than fall. The effect will be much greater with an iron cored machine and the worse the leakage reactance the more beneficial it may be. Digressing slightly it is even possible to neutralise the leakage reactance with capacitors in series with the load.

What happens with rectifiers is probably less well understood but shunt capacitors will likely be beneficial in increasing the main field mmf ( may be quite small effect with neo). The rectifier presents a lagging pf load but pf correction in the normal sense is not possible here ( too complicated for now) but there is little doubt that the waveforms will be modified and the conduction pattern in the rectifier may change.

At low voltage it needs large capacitors and I haven't tried much in the way of experiments but I didn't find it very beneficial with axials with neo, but I didn't have enough large capacitors. With wound field alternators that I used to play with the benefit was considerable.

My immediate instinct here is that with neos and axials it's not cost effective even if useful. it probably is with motor conversions and for those who play with salient pole synchronous motors ( F & P and opther similar things) then I think shunt and series capacitors are necessary to get much out of the things, but these run at higher frequency and the capacitor values are a bit more sensible ( but still very large at low voltage)

Probably room for investigation here as it is outside the normal electrical engineering field where things have been studied in detail.

On the dc side, personally I see it as a dead duck except for interesting investigation.

Flux

[ChrisOlson]

I think I see what Glen is saying.  And you'll have to forgive me because I'm pretty slow with understanding electronic stuff.

With the single phase circuits the diode is only conducting when the sine wave is above the battery voltage level.  Then when the sine wave goes below battery voltage it unloads that phase, which is good because it lets the turbine run.

So the capacitor will charge up during that peak time when the diode is conducting.  But what I don't understand is why would the power want to go to the capacitor to charge it up when it can go to the battery instead?  Anyway, now the sine wave drops below battery voltage so no current is flowing in the wire and the phase is unloaded.  The capacitor now injects its stored up power into the line, which goes into the battery, assuming it didn't go to the battery in the first place and went to the capacitor instead.

I still don't see how it works, I guess.  The battery is already a capacitor.  A damn big one.  So why stick another one in there?
--
Chris

[electrondady1]

I've been trying to   understand  capacitors for a long time
i may be mistaken ,
but i think caps will accumulate charge even if the  wave form is below battery voltage.

[Flux]

If there was no line resistance then the capacitor couldn't make any difference.

The idea is based on the fact that the voltage at the alternator terminals is higher than at the rectifier. The capacitor will charge to battery volts plus the peak line drop and when the alternator volts fall below the capacitor volts then the capacitor end will still be above the voltage needed for the rectifier to conduct ( above battery volts). The idea is that some of the energy can be recovered and fed to the battery.

At one time I believed that there was an advantage to having a battery at both ends of a long line so that the nearest one would take the full alternator charge and gradually charge the one at the remote end until the voltages equalised.

I now think this is mistaken but it is a slightly different situation from the capacitor before that rectifier. Capacitors don't' suffer from the surface charge that batterers do and there is the blocking by the rectifier.

In the end It may be difficult to prove and even if it does work as intended you are forced back to the old issue that removing some of the line loss, although beneficial electrically will still bring the prop off its curve and you will need to add the loss somewhere else.

There is a small confusing issue that this may change the load characteristic of the alternator with speed and this might be beneficial or otherwise  and would very much depend on how the cables are run from the two alternator phases to the rectifier.

Chris did you ever try your two stators in phase so that the load was single phase rather than 2 phase when rectified?

Flux

[Flux]

I've been trying to   understand  capacitors for a long time
i may be mistaken ,
but i think caps will accumulate charge even if the  wave form is below battery voltage.

Yes if the capacitor is connected to the alternator before the rectifier then it will certainly charge below battery volts. Its voltage will follow the alternator volts but with a charging current leading by 90deg.

This then comes under the discussion I had earlier regarding the effect of capacitors on alternator windings, it is different from this discussion on line loss on the dc side. If the capacitor is after the rectifier than any charge before diode conduction will be from the battery.

Flux

[SparWeb]

Flux,

My posting was getting long but ironically I did leave some things out:  I have only tested these capacitors on a motor-conversion, and only with them across phases - not as shunts.  I realize now that this is a big digression from the original topic - somehow I though there were questions about capacitors on the AC side of the lines but now I realize that this wasn't the point at all.

The "re-magnetizing" effect of capacitors in series on the line is very interesting (you've mentioned it to me before in fact).  However I've deferred these experiments until I can buy more complex measuring equipment.  My head is swimming in theory but I have no experience to relate to it.  Once I learn how to use an oscilloscope and plot several traces on it simultaneously I'll have the technology to go deeper into the subject.

A footnote about capacitors on the DC lines:  I've found that the combination of small battery, mismatched rectifiers, and a poor multimeter can give erratic voltage results if you don't include a capacitor on the DC line.  This experience comes from crank-testing these motor conversion on a bench to get early V/RPM estimates.  All of my more serious tests, and of course up on the tower, the numbers are steady.

Sparweb
Capacitors on ac lines, is very different from having them on the dc side.

You don.t say whether you have this on a motor conversion or an axial either. With machines running normal ac loads much of this is understood. The armature reaction of a machine reacts on the main field and the effect depends on the load power factor. With inductive loads the eventual effect without going into too much detail is demagnetising and the load regulation will be worse than the resistive load.

Capacitors work the other way and increase the main field mmf and in extreme cases the voltage will rise with load rather than fall. The effect will be much greater with an iron cored machine and the worse the leakage reactance the more beneficial it may be. Digressing slightly it is even possible to neutralise the leakage reactance with capacitors in series with the load.

What happens with rectifiers is probably less well understood but shunt capacitors will likely be beneficial in increasing the main field mmf ( may be quite small effect with neo). The rectifier presents a lagging pf load but pf correction in the normal sense is not possible here ( too complicated for now) but there is little doubt that the waveforms will be modified and the conduction pattern in the rectifier may change.

At low voltage it needs large capacitors and I haven't tried much in the way of experiments but I didn't find it very beneficial with axials with neo, but I didn't have enough large capacitors. With wound field alternators that I used to play with the benefit was considerable.
...
Flux

[ChrisOlson]

Chris did you ever try your two stators in phase so that the load was single phase rather than 2 phase when rectified?

Hi Flux - no, I never tried that.  I'd have to drill different mounting holes in the second stator to line it up with the first one.  And then (I think) with two of those things in-phase and in parallel it would probably vibrate really bad.
--
Chris

Edit:  I picked up those capacitors they had at the motor shop.  I got four of them.  My motor rewinding friend told me that if I hook them in series (the studs are labeled + and -) that they should take 32 volts, 40 surge without blowing up.  I'm going to hook it up to the 13 awg stators that I tested the other day, as per Glen's wiring diagram he posted earlier, and do a low output test to see if it makes any difference.  I'm scared to push it too hard because my motor friend said I should have those capacitors in a bomb-proof shelter in case one blows up.

He did NOT recommend doing with them what I'm going to test.  I think that's bad.

[DamonHD]

You should attempt to keep the voltage across the caps balanced by putting some (equal-value) resistors in parallel with the caps, which is what I think is done (roughly) in big grown-up supercap banks.

As this is an experiment you can aim lower than you might otherwise want to for the resistance (ie incur some extra losses) to avoid percussion.

I'd guess picking a value that would half discharge a cap in peak-to-peak time, just as a first approximation.

But Googling will probably reveal better methods.

Rgds

Damon

[ChrisOlson]

Don't get too complicated here, because I know NOTHING about electronic stuff.

I looked at Glen's wiring diagram and now I'm confused already.  He's got the capacitors ahead of my 8/4 cable, then to the #2.  That's not the way I got it hooked up.  The 8/4 is my AC cable.  Why does the wiring diagram show it on the DC side?  I don't have another 8/4 cable laying around here to get power to the rectifier, and I don't want to cut that one off because I might use it for a tower drop someplace.  If I move the 8/4 cable to the DC side, ahead of the #2, then that will skew my test results I did the other day

Why can't I just hook the capacitors up on the DC output of the rectifier bank, with the #2 in between the capacitors and the battery bank?  Then I could put all four of these coffee cans in series for 64 volt capacity.

I wish my camera would work so I could take a picture of this before I wreck something.  The camera got knocked on the floor the other day and it hasn't worked since.  More electronic stuff that don't work.
--
Chris

[DamonHD]

If you're going to put supercaps (like I think you have just acquired) in series, in order that the voltage gets shared reasonably equally across them and one doesn't go pop trying to carry an unfair chunk of it, you should put something like a resistor in parallel with each cap, and for equal caps, equal resistors.

At a very wild guess, if you have anything in the hundreds of ohms range then you might start there.  I just don't know.

It's quite likely I'm talking through my backside.

Rgds

Damon

[bob g]

the only resistors i have seen across caps are "bleeder" resistors, meant to bleed off the charge upon power down so that they don't retain lethal amounts of power long after the unit is shut off.

not sure what you want to accomplish with resistors to balance voltage across the caps?

how does that work?

bob g

[ChrisOlson]

I hooked the four capacitors in series and got them hooked up at the DC studs on the rectifier heat sink, right at the end of the #2 cables.  This is not the way Glen had it in his wiring diagram but either he misunderstood my hookup, and I don't want to do anything different than how I tested the gen the other day.  I figure if these capacitors make more power, like it is claimed, then it would have to work in the real world.  And in the real world you don't hook up 100 feet of 8/4 to the DC side of the rectifier and then run it to 150 foot  #2 cables.  At least, that ain't the way I hook them up.  The 8/4 carries AC power from the four wire output two-phase generator to the rectifier, so I didn't change that.  And it's about 100 feet long, which is a typical length tower drop cable.  So I figured it's best to test it that way, with a real wire run.

I used the carbon pile in my AVR to limit the power to them to charge them up like my friend at the motor shop told me to do.  Nothing bad happened.  The bank is at 28 volts because we got good wind tonight and they are not getting hot or doing anything weird.  No explosions, no smoke, no sparks.  The AVR says they are not drawing any power from the bank.

It's raining outside and only 38 degrees, so I don't feel like setting up my hydraulic power unit to run the generator tonight and see what happens.  Tomorrow is another day and I'll try it then.
--
Chris