ST head modification for 48 volt battery charging

[BruceM]

Here's a rough schematic for 120VDC feeding the MX80, with over-voltage protection.

[bschwartz]

I'd think 50 amps at 60 V (3000 w) should work well as the 6/1 wouldn't put out much more than that anyway.  As to what the inverter can handle, well I don't even have one yet.  I wanted to make sure I could power one before purchasing it. 
While working on this I also ordered a 24v 160a Leece Neville from ebay for $125 and two 600w (claimed) 20v-50v grid tie inverters for $150 each on ebay.
When I'm done hopefully I'll have 120vac,240vac, 60vac, 60vdc, 24vdc capabilities.  That way when a 'real' grid tie inverter (Outback,Xantrex etc.) comes along cheap, I'll be able to power it.  In the meantime, I'll have some cheapie inverters to play with, and I'm having fun with the ST.

I'll get a snapshot of the transformer when I get home in a few days.  Then you can tell me if you think it will do the filtering job, and which winding/s you think I should use.


On a side note, as I have the coils separated, and the individual wires will come out of the ST head, several possibilities come to mind (some already mentioned)

1. wire as normal for two 120vac or bridged for 240vac
2. wire as 4 60vac coils used in parallel as one large 60vac supply
3. wire as 2 60vac coils in parallel and 2 60vac coils in series for 120vac
....... the interesting thoughts.......
4. rewire the coils for 180 degrees as per 1 or 3


I wonder if any of these would reduce groaning even if left as a standard 120/240 unit.

[mbryner]

Bruce,

Thank you very much.   That's not too complicated.
So you mention filterconcepts.com.   Looking at their website, would you recommend something like the LH45 series for an inductive filter?

And this bridge rectifier?
http://cgi.ebay.com/50-Amp-1KV-50A-1000V-Bridge-Rectifier-KBPC5010-x-2_W0QQitemZ250553096161QQcmdZViewItemQQptZLH_DefaultDomain_0?hash=item3a5620d7e1

Need to make sure there's a good heat sink for the rectifier.

Doubling the circuit, I could use both hot legs of the 240 V setup.

Marcus

[mobile_bob]

here is the reality as i see it, and the reason for modification of the stator

for a typical 6/1 driving an st5, if we modify it we can only get about 3kwatts output anyway
whether it is making 120/240 or 60volts, so the decision comes down to, do you want to use
an mx60 in this application? maybe? maybe not?  its rather expensive to be feeding power
that might go overvoltage and cause damage to it? 

now if we go to something like a 10/1 driving an st 5 we can get the full 5kwatts, or about 86amps
at 57.6vdc, this is still close to the capability of the mx 60 and very close to the newer version iirc

its not until we get up to about the st 7.5 that we see the benefit of a modified st head for 60volt output
where we can get upwards of 130amps at 57.6vdc, this is clearly outside the capability of a single mx 60
so it would take two?  probably not something most would want to do, given that a purpose built regulator could
be built or modified to control the st head at that voltage anyway.

while this experimental st5 of Brett's might not be for everyone, the lessons learned on it will be applicable to the larger
heads where the gains clearly favor the modification to any other alternative in my opinion.

another thing we may all learn from this exercise is the reconnection of 1 and 3, along with 2 and 4 to get 180 degree displacement
may well eliminate or reduce the growling or whining we hear about with these heads?

bob g

[Crofter]

Hard to say what all implications of 180 coil placement would do to the waveform or efficiency compared to the 90 deg location. I see the 90 degree placement causing a mechanical side load on the rotor when only one leg of the 240 is loaded. guessing here, but could that be causing a resonant vibration of the whole rotor?
No amount of glyptal or other winding stabilization would ever cure that! Mobile Bob could be bang on in suggesting the ST5 style head was primarily a 240 unit and having 120 available was an afterthought.

[BruceM]

Marcus-   First and foremost- if you have a 240V setup, you'll have to do a step down transformer.   While you're doing that, pick one so that you don't need overvoltage protection.  Toroid transformers have the highest efficiency, if that's important.  You can parallel matching transformers to get more current, but make sure each is fused separately. 

I would not use bargain rectifiers from Hong Kong, with specs that look like bullshit to me.

On the thread about smoothing capacitors you'll see what happens to these.  See Digikey, Mouser, Allied, FutureElectronics, etc. for reliable suppliers and specs. 

A primary inductive filter takes a big choke-  Filterconcepts stuff is only for high frequency filtering, the inductors are tiny uH units.  Since the MX60/80 will reduce current as the battery approaches full, you need a pretty big choke-  I'd be looking for a surplus transformer with a big current secondary (1.5-2x the max DC current) , and hopefully still 0.1 Henry or better. Multiple smaller inductors in series can be used, if the current ratings are high enough.

Marcus's idea of having both PV and ST head feeding the MX80 seems like a good way to get your money's worth out of it.  IF both sources are diode protected from backfeed, that is sufficient to do the job. 

I think that changing the ST winding configuration to 180 degrees may very well solve the growling 240 with neutral problem.  It will be interesting to see if it improves the imbalanced regulation problem somewhat, too.  Brett's experiment may be a boon for ST 240/120VAC users.

[BruceM]

Marcus's idea of using a MPPT charge controller for both PV and AC charging made me look for something comparable to the MX-60/80 but more affordable.

The Morningstar TriStar TS-45 and TS60 MPPT charge controllers are only about $200 at windsun.com,  the 60 amp unit will do 48V battery banks and will also handle a high voltage input (150V open collector) , and do temperature compensation, 3 stage charging.  One of these with a filtered DC input from either 120V head direct for from transformer stepped down 240 might make a good charge controller that could do both PV and AC sources, I think.  At 60amps x 56 volts peak cyclic charge voltage, that would be over the full output of a 6/1.  Obviously you'd need something bigger for a bigger engine, and at that point I'd start looking at a separate AC charger, or something custom like the mod Brett is doing with Bob's help.

It's a "buzz bomb"- (EMI spewing switcher), but you guys don't care about that

BruceM

[Geno]

Marcus, my 80amp/1000volt rectifier (like the kind in your link) smoked last night after about 15 minutes when I put my system on line. My max power through it was 32 amps.
http://www.microcogen.info/index.php?topic=411.msg5171#msg5171
I'm deciding on whether to go with the stud mount type or to parallel a few of the type I smoked.

Thanks, Geno

[mbryner]

Hi Geno, thanks for the input.  I was following that thread, too.   Parallelling a few of those rectifiers may work.   I'll eagerly follow your progress.

Bruce, why would I need to use a transformer if I use two circuits like you diagrammed, but in parallel?  One for each hot lead?   The output of each would be 120 V DC.   It's the difference between the 2 hots that is 240 V, not btwn hot and neutral.   Or am I trying to make it too complicated?   It would be nice to keep the 240/120 from the ST head for emergency power, because my well pump is 240 because it is 800 ft from the house.   Otherwise I'd change the ST head to 120 V only config and have the inverters parallelled instead of stacked.

The Morningstar charge controllers are a good idea....I'll research them some more.

(Trying to buck the stereotype of guys with initials behind their names not having any practical experience!!!)   

[Halfcrazy]

If one needs a charge controller the new Classic from Midnite solar should be on the market real soon and comes in 3 voltages there first is 150 volts operational safe to 200 volts the second is 200volts operational safe to 250 and the third is 250 volts operational safe to 300. My plan is to rectify the 120 volt output from a st head and feed the 250 volt unit. no over voltage protection necassary. Oh the classic will also be able to boost as well as buck voltage

[mbryner]

Great idea, halfcrazy, but it costs about $700!

[BruceM]

Marcus, no, you can't just bridge rectify the two sides of a center tapped 240VAC configuration and connect them in parallel.  They aren't floating, since they have a common neutral.  You will have created a (brief) smoke generator.  So no big deal, just step down 240 to 90-100VAC.  

Here's one: ebay item- 360217315683 (you'd need 2 or 3 with 120V secondaries in parallel at $39 ea)

Here's another:     330385136785  (2 each with secondaries in series would give you 80V @ 30amps)
(This is the one I'd use- no overvoltage issues.)  Toroidal transformers are typically over 90% efficient.

A MPPT charge controller like the MX-60 or Tristar-60 will covert that 2400 watts with good efficiency into your 48V battery bank, the peak charge current will be near 50 amps, these are buck converters, often over 93% efficient.

If you need more charging watts, the go to 4 of the 330385136785's.  60 amps of 80VAC.

If you care less about efficiency, there are plenty of conventional E-I core step down transformers, cheap - just don't push them to spec like you can a toroidal transformer. Oversize in watts by 50% or more. Expect 83-85% efficiency.

My Dad is also an MD Radiologist (now retired), a highly skilled woodworker, furniture maker, including antique reproductions, a decent mason, framer/builder, plumber, electrician, sailor, musician, all around handyman, inventor and not too shabby with computers and basic electronics.  It's just not true that advanced degrees mean no practical, hands on skill.

[mobile_bob]

before i spent 700 bucks for a buck converter, i would look to build one myself
the genhead itself probably has enough inductance to work against for the converter design

i would also integrate a standard delco regulator at its core, because it is cheap and readily available.

Brett's experiment is interesting in that in the end he can get the charge voltage and current without
spending likely more than half of the cost of an mx60 and without the 7% hit in efficiency (conversion efficiency
of the mx60 on its best day)

that 7% doesn't seem like much until you factor for what that actually equates to in fuel consumption, to cover that
7% loss (off the top of my head, and without doing the math) i would expect the fuel consumption to be 10% higher
than a modified ST head for 48 volt charging.

i might be wrong, that has happened at least once before...  ahem... choke

ok, maybe twice!



bob g

[BruceM]

The outback Mx-60/80 is rather grossly overpriced, that's why I suggested the Morningstar Tristar-60.  At $200, and the same features, it seems a lot better deal!  It's not just a buck converter, it does a proper, temperature compensated, 3 phase smart charge.  That you won't be whipping up on your own, Bob.

Regarding direct charging- via down regulated 60VDC;  Sure, you could do that, added to my basic AVR circuit, and perhaps adding the TI 3 phase charge controller chip, as I have previously suggested.  Then you'd have the ultimate in charging efficiency.  I don't see to many guys here who are going to build a proper temperature corrected, 3 stage charger from scratch.  If you don't the efficiency you save may be lost, many times over not just in much equalization charging but also in battery life.

Auto alternator regulators are a notorious engineering joke- they are typically not even temperature compensated, fixed voltage chargers.  That is of little use in a solar system battery bank, and results in chronically undercharged, sulfated batteries, unless frequent, protracted (fuel waste!) equalization charging is done.  It's barbaric.  A fixed voltage charger just shouldn't be used, if you value your battery bank.

[mobile_bob]

Bruce:

all good points, but 
don't think that i haven't given this subject a lot of thought

as for the automotive regulators, yes they are not temp compensated, but
the delco regulator, or i should say some of them can have the link clipped on the two spade terminals and have
varing resistance inserted to control the charge rate to anything you want,

we could therefore use a microcontroller and one of the chip resistor networks, i forget the proper name, but they step
in value with pulse commands from the microcontroller, all you need then is two sensors to supply the micro with battery temps
and with alternator temps and the micro can crunch the numbers and pulse the resister network to the proper level which in
turn dictates to the regulator what the voltage set point is or is commanded to be.

much of what the mppt's and other pwm controlled regulators such as the balmar, xantrex etc, do is overkill in my opinion
because they build them to be all things to all applications, beit flooded cell, gell cell, agm, or whatever, and they got to build it
in such a way as to make it programmable from an idiot interface so that the user can't get access directly to their coding.

also of note, none of these controllers include the cost of the temp sensors with the price of the control unit, so you can easily
add another 75 bucks or more to the cost, and maybe much more with some products.

i don't like dismissing out of hand any off the shelf mass produced item, before i fully investigate whether or not it can be made to operate
in a manner that does the job i need done.

the fact that the little internal regulator typical of the delco alternator can be bought for 10 bucks more or less, makes it attractive to see
if it can be incorporated into a larger control scheme.

in my opinion of course.

as for automotive regulators not being temp compensated, those that are internal are certainly temp compensated, but
not for battery compensation but protection of the alternator, this is one major reason automotive alternators drop voltage
like a rock as they get hot.  if you remove them from the heat source they will certainly keep the voltage at whatever level
they are designed for, and in the case of certain delco's , certain leece nevilles and a few others, that voltage setpoint can
be adjusted over a fairly large range. 

certainly high enough to keep from anything like chronic undercharging, or allowing normal sulfation a chance to harden and crystalize
from undercharging.

i would agree not to use a fixed voltage charge system, and would not support the use of one save for very specific circumstances
where they are advantageous and appropriate.  one appropriate use for a fixed voltage system is the 50/80 charge regime
where a fixed voltage charging from 50to 80% state of charge is certainly acceptable and in most cases desirable.

bob g