48V charging via ST generator and solar charge controller (no ST mod)

[BruceM]

Yellowhead sent me a note questioning my comment on power factor correcting circuitry.  I did some research on PFC chips and applications this morning and I think he is right-  a PFC front end on a switcher should be helpful for low frequency operation (bulk storage cap is less affected), and is not inherently problematic as I thought.

I had assumed most implementations of PFC would be AC zero cross and time based, but I was wrong. Of the three most common schemes, none would seem to care about AC frequency.

Thanks, Yellowhead, for your message and for forcing me to get a better understanding of PFC circuitry.

[Geno]

I finally got around to putting the metal film caps on my rectifier that Bruce recommended. I also put the negative battery connection all the way to earth. I've been putting 33-36 amps @ 54 vdc into the batteries for the past 4 hours with no display issues. I need more time on things for sure but this is a very good indication that Bruce's fix is going to work.

Bruce: There is no way I would have ever known about these caps and what they could do. Thank you very much for your help here and everywhere else on the forum. You truly deserve all the thanks I can give.

Thanks, Geno

[mbryner]

That's good news!  ~2000 watts is a good charging load for a Listeroid.   Now can you write a simple wiring schematic of what you did, and post it for posterity and to sum-up your config?   Thanks.

Marcus

[BruceM]

You're most welcome, Geno.  I really enjoy your creative projects and respect your contributions here greatly.  You are all signal and no noise, and that sure is a pleasure!

A cheap AM radio is a dandy tool when troubleshooting EMI problems.  You tune it between stations, and then it will squeal if a strong (primarily magnetic field) source swamps the receiver. A really nasty source will show up at high, medium and low positions on the AM tuner, even a foot or more away from the source.  The readings do actually correlate fairly well to spectrum analyzer measurements, often. For much troubleshooting and quick and dirty fixes, the $15 tool in the right hands is faster and  more effective than the $15K tool.

Radio Shack presently carries a model 12-586 ($15 online)  that has been tested by a friend at Los Alamos to be quite sensitive to a wide range of frequencies.  Some AM radios are lousy for this purpose.  The former Radio Shack 12-467 was also pretty good as a "near field sniffer".

The use of 0.1uf metal film caps around the bridge diodes is a cheap fix and dramatically effective in reducing that EMI.

I hope that your setup will continue to be a solid performer, and I'm glad my EMI guess seems to have done the trick.

Best Wishes,
Bruce

[Geno]

Thanks for the AM radio trick. I'd guess it would make a lot of noise near that old UPS. With its pathetic efficiency #s it will probably be out of there soon anyway. I'm losing ~10% (edit: It's more but I don't know how much yet) with the transformer, rectifier and charge controller. The other ~30% losses are all in the UPS. A few dozen more hours and I'll pretty much have proven the concept reliable though.

Thanks, Geno

[BruceM]

Thanks for providing the diagram Geno! 

[mobile_bob]

just so i am clear with what this schematic is for...

i assume it is for the st head?

is it for the excitation supply?

or is it for rectification of the output, so that it can be fed to an mx60 or flexmax80?

do the caps smooth the waveform? or do they clean up some harmonic crap superimposed onto the dc current?

maybe a short writeup where the problem is clearly defined, the solution presented, schematic posted
and the results... all in a nice "white paper"
it doesn't need to be very long, but would be nice to have it somewhere, where it won't get buried and is easier to find
later?

bob g

[Geno]

My ST head supplies 120 vac to a transformer which steps it down to 70 vac which feeds a rectifier which in turn feeds my Flexmax 60. At charge rates above 30 amps the display on the Flexmax went to garbage characters but the unit was still charging. Bruce suggested the caps in the pic be added and that I put battery ground to earth to reduce probable emi. The Flexmax is open between battery negative and ground for unknown reasons. (any ideas why?) Several hours of testing last night at 33-36 amps with no display problems. I'm about to fire it up again right now for a little longer test.

I'll put it in a white paper write up after a few dozen more hours run time.

Thanks, Geno

[mbryner]

all in a nice "white paper"
it doesn't need to be very long, but would be nice to have it somewhere, where it won't get buried and is easier to find
later?

Yes, exactly, with a link to to the white pages section would be great.   Thanks for the schematic.  Helped me visualize much better.   I assumed it was for cleaner rectification of the output before entering the charge controller.

Also, BruceM, the 3 toroidal transformers arrived.   No doc's with them.   Just what the e-bay listing had:  "dual primary for both 100 -120 VAC or 200 - 240 VAC 50/60 Hz operation.  The secondary is 40 VAC center tapped at 30 ampere current rating.  It comes with two rubber pads, a top mounting plate and mounting hardware."   The transformer has a simple label:

I/P:  AC 110 V x 2 (GRN --- BLK)
O/P:  19.9 V (RED) -- 0 V (BLK) -- 19.9 V (RED)

You recall I was going to transform down 240 VAC from the ST head before a rectifier setup like Geno, then a choke.   Assuming I treat each green wire as one hot AC leg of 240 V and bond the 2 blacks to neutral?   How do I get up to 80 V?   Can't just stack transformers.   Do I put 120 V in one primary on one transformer and the other 120 V leg on the other transformer, and put the red leads in series since they would be out of phase?   That would seem correct.   Of course if I get the red leads backwards, there will be smoke, or hopefully just the breaker will do it's job.

Marcus

[mbryner]

Thanks, Geno.   You posted while I was replying.

[mobile_bob]

i don't see why you couldn't put the two transformer primaries seperately across each leg of the st head
or put them both in parallel connection 240volt windings across the 240 output of the st head
and then take the secondaries and connect them in series for 80vac

if you have the secondaries out of phase, the resultant voltage will be near zero, instead of 80vac
simple enough to check for and no chance of smoke as long as the the output legs are not connected together
which they should not be in any case anyway.

bob g

[BruceM]

Geno-

It's a nice feature that the Flexmax has ground and negative isolated- that lets the system designer decide how grounding should be done. (Normally only in one place.)  In some systems, the positive rail is grounded.

I wanted you to do the grounding as I assumed that any input filtering on the Flexmax (small ceramic caps) would be to ground, so without the system grounding done, those filters would be ineffective.  I've also found that most digital systems are less EMI susceptible when not floating relative to ground.   I suspected the bridge as a problem EMI source because the spike or transcient type partial failure of the digital electronics happened only at higher currents- and bridge diode noise also increases with higher current.  I also expected that there was little input filtering on the Flexmax, since it would normally be connected to a PV array.  The designer was not expecting a "noisy" source.

Marcus-  I'll draw you a wiring diagram and will post it later tonight or tomorrow AM.  I'd wire the two transformers each for 240V (primaries in series), and then all the secondaries (4) will be put in series to get you to 80VAC.  As Bob notes, you will have to check the phase of the secondaries before connecting them. All red wasn't a big help, but you can manage anyway, and the price was great.

[BruceM]

Marcus- Here's my wiring diagram from the wiring info you gave me.

Make sure to test the phase of the primaries and secondaries before connecting them. If you need help doing that, I'll be glad to try and help.  I'm afraid my writing skills suck.

[mbryner]

Thanks both mobile_bob and BruceM.   Sounds like you both are saying the same thing.   The wiring schematic is especially useful -- I have a much easier time with a schematic than a description.   I tried hooking up 120 V to one primary this afternoon and got about 44 VAC across the 2 red leads, which gives 88 VAC in series. You are correct on your diagram about the input to each primary as 1 black and 1 green wire.   I tested it with an ohmmeter and have no resistence btwn the appropriate black and green wires forming a coil.   Pardon my ignorance, but how do I know which green/black pair to connect together?   Maybe the order they exit from the transformer is the same as the schematic (blk-grn-blk-grn).    And also, on the secondaries side, how do I check phase with just a simple multitester?   Thanks.  Oh, and I found a huge old transformer to use as a choke.   The guy said I could have it for free, but it must be 100lbs+.

Marcus

[mobile_bob]

the secondary side is easy,  connect the two in series and check the voltage output on the two open end wires
if you have ~88 volts ac, you are in phase, if you have something very close to zero you are out of phase.

series connecting outputs are the easiest kind, they result in no possibility of smoke and sparks, provided you leave
the two output leads open for testing

paralleling outputs is where you can get into trouble, in which case you connect one of each coil set together and then measure the
voltage on the two remaining, if there is anything other than nearly zero, you will get smoke if you connect them, because they
are out of phase with one another, if the voltage is near zero they are then safe to connect together and use.

if you ever have any doubts use a small fuse holder and a 5amp fuse to connect with, a blown fuse makes a little flash, might spook you a bit
but won't fry anything other than very small transformers (much smaller than what we are working with here)

the transformers i am working with are all polarized color coded leads, the secondary has red and black 8 guage leads, the primary has a black and a white pair of 14 guage leads and another pair of 14 guage leads that were used to some sort of trigger or feedback to the inverter they came out of, in my application they are unneeded and will simply be taped off.

bob g

bob g