ST head modification for 48 volt battery charging

[BruceM]

We'll have to agree to disagree on this one, Bob.  I do agree that a well designed, custom AVR on the 60V ST head mod (directly charging the 48V battery bank) would be the highest efficiency possible. Even better with some speed control!   But I'm not holding my breath for someone to implement it.

Implementing a proper 3 phase charge is more than just adjusting the regulation voltage. You have to monitor the charge current to watch for a decline while holding the cyclic charge voltage.   I can't say how well the various MPPT manufacturers do it, but I'd bet fairly well; they have both battery voltage and charge current, and that's what you need.  Since the available current for charging may vary (PV) or even be suspended (clouds) at any time, it can get to be a bit of a software chore to sort out.  Likewise, load may suddenly exceed charge current, some more software gyrations to sort out.  I've been working on this software for my 120VDC Battery bank charge controller (AC, PV, Wind charge sources, running SOC).  It's not trivial.

Unless the Tristar-60 is unreliable, I can't imagine wanting to replicate that function, not for a measly 200 bucks.  I can't use the thing because of my extreme EMI-related health issues, but if I could, I sure as hell would, rather than rolling my own, at greater cost.  (No way I could make such a beast for $200!)  Besides, my main battery bank is 120VDC.

[mike90045]

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. 

Huh?

Morningstar has 2 60A Tristar lines, one, the old PWM line (~$200). 

The other is the new MPPT style, at about $550.  That is the down-converting, 3 stage controller.
  http://store.solar-electric.com/motr60ampmps.html

[mobile_bob]

Bruce:

don't get me wrong, we agree more than we disagree
i too cannot see reinventing the wheel if tristar does it for 200 bucks, that makes sense

what is not clear though at least to me is, what do i get for 200bucks?  what losses are there?  7%? 10%?
and yes a proper controller should keep track of volts and amps into the batteries, but upon serious investigation
the various oem's do that with some rather crafty if not dishonest means

i would go into graphic detail on that, but would alienate more folks than i would convert, people pay big bucks for controllers
and aren't keen on being told that they don't work exactly as advertized.

there is a huge fudge factor that they all use, because it is almost impossible to build an multipurpose controller that will suit the needs
or everyone, so they calculate things like current rather than measure, some allow you to set bulk and absorption times, etc

in the absorption and bulk phases it is desireable to control voltage tightly, and allow the batteries to absorb current at their natural rate
of acceptance, once you get over ~80% SOC however you must start to reduce both voltage and current, that is where it gets a bit tricky
if you are going to do it right.

i suspect those oem's that do it right, make and sell units that are bloody expensive and not commonly used by the common user
but rather very large or high dollar military or nasa applications

all i am proposing here is this

if we modify an st head as described and set it up to produce 60vac, rectify it, then build a regulator or adapt one to provide for
a three step temp compensated charge control, we ought to have a fairly highly efficient charger.

in my opinion, building such a regulator would be many times easier than the total system automation system we have also been
discussing, if for no other reason than it is easier to conceptualize, parts are widely available to work with, oem units are available to study
how they do it, and there are many folks that have built various regulators for specific uses,, we are just adding a few bells and whistles
that are best controlled by a microcontroller.

ymmv, but this is something that seems like a useful project

also, when one considers the most efficient way to charge flooded cells with an engine driven alternator is the 50-80 regime
very tight control of anything other than voltage is of secondary importance, leaving the remaining charge from 80-100% to a smaller
charge source that is more in line with the reduced needs which can be controlled via an off the shelf controller that is relatively inexpensive
that can control amps, volts, temp compensate accurately etc.

maybe we should start another topic/thread and explore the possibilities?

what a group we have here!

where else can a guy go to get this kind of detail on so many facets of offgrid power production?



bob g

[BruceM]

Mike-  Thanks for correcting me,  I got the different Tri-Star models mixed. I was looking at the non-MPPT unit price.  I recant the recommendation, I'll have to find the specs on the $200 unit.  It might be crap.  The MPPT function per se, I think is unnecessary, but I'd have to see the details on their charging algorithm on the $200 unit.

Edit-  I did check the manual on the basic PWM Tristar-60.  It looks like a decent unit, capable of doing a good job of battery charge management but its max input voltage is 125V, which rules out direct rectification and inductive filtering  from 120VAC.  A step down transformer would be needed for this unit, or Bob's 60V ST mod.

Bob,
I like the concept of the solar charge controller doing both PV and AC charge regulation, because it will not only get you the bulk AC charge control (typically fixed current charging), but also then do your PV topping off charge (fixed voltage absorption phase).  A decent unit at a good price would be a boon for guys with less interest and experience with custom electronics.  The PWM Tristar60 (the cheap one) doesn't look bad for this role.

I'm trying to implement a smart charging scheme for my 120VDC BBCC unit;  as you suggest, finishing a charge at reduced current via generator is grossly inefficient-  if you have sufficient estimated time for PV or wind available that would be vastly better.

At some point I'd like to be able to upload a simplified 5 day PV and wind forecast to my controller, to thus minimize fuel, while still giving my battery bank life (keep it topped off as much as practical)  a significant priority.  Alas, right now I'm still doing hardware debug. Darn that PCB designer- oh yeah, that was me.  

[mbryner]

Bruce and Bob, you guys are amazing.   This forum is a great well of knowledge and experience, and I'm thankful.   

So for my purposes, 2 toroidal transformers, like your recommended ones on e-bay ($50 each), each using 240 VAC input on primaries?   Then run the output through appropriate bridge rectifiers, then connect resultant output in series?   Or I guess since I'm connecting to the 240 V input, I can connect the toroidal secondary output in series before rectification.  Then past, say, 10k uF capacitor for smoothing?   Gives 80 V DC at 30 amps (with 5-10% loss) = 2400 Watts.  Appropriate for a Listeroid 6/1.  Then into the Outback MX80 or Morningstar Tristar TS60.    If using the Outback MX80, I'd have to provide appropriate isolation since it's connected to the PV array already.   If I bought a TS60, it could just be hooked up to the batteries w/ appropriately sized breaker inbetween.   All correct?

Brett's ST mod is still cool and best for efficiency, but this works better for me.

Marcus

[BruceM]

Bob, should we move Marcus's thread to a new topic, like "48V charging via generator and solar charge controller"?  

Marcus-  

You didn't mention your generator type and capacity- or your battery bank capacity;  that's important for getting things right overall.  I could be leading you down a primrose path by assuming Listeroid 6/1.

Yes, the secondaries of the toroidal transformer I liked best on ebay should be connected in series before the bridge diode.  The primaries can be connected either in series (120VACx2) or parallel (240). When combining transformers you should put an inline fuse for each transformer's primary.  

The rectified, combined (2 or 4)  transformer output must still be filtered for the charge controller- at least some inductance as primary filter is wise from a PF efficiency factor.

If you are already using an MX-80 already on your PV array, then by adding a diode in the AC chargers filtered DC, you should be fine connecting both to the same MX-80.

The Tristar-60 PWM units can be paralleled, if you need more than 60 amps for charging, I wouldn't mix different units unless I had documentation that assured me that was going to be OK;  things like disconnecting the charge controller from the battery, while still connected to PV, are potentially bad.  The outback units may be able to be parallel'd also, I haven't checked the documentation.

BruceM

[mbryner]

Yes, maybe my questions should be in a different topic.   I didn't mean to hijack bschwartz's ST mod thread.

My equipment:
    JKson 6/1 Listeroid, propane canister muffler, no heat recovery yet.
    Powersolutions 7.5 kw ST head w/ doghouse already removed (price same as 5 kw from Joel Koch at purchase time, so I upsized)
    Outback 3600 watt inverters x 2, in stacked config for 240 VAC
    16 x L16 type batteries (actually currently 8 of them, because the house isn't built yet, but it will be upsized to 16 in a few months)
             (I don't remember exact AH of battery bank)
    Outback MX80 charge controller with input from 18 x 190 watt panels (3000+ watts solar total)

So sizing for 2400-3000 watt from a 6/1 is appropriate.

[BruceM]

A nice setup, Marcus. I just wanted to be sure I understood your setup.   L16's are 350 amp hrs.  Adding a second new string in parallel, that's not the same age, is a no-no.  Don't do it.  Better to sell the used first set, replace all with new matched set.

With your setup, you should be fine using your existing MX-80 for both PV and ST sourced DC charging. 

I'll see if I can find the calculations for the minimum inductance DC choke for filtering, then we can look for  surplus toroid(s) with high current secondaries.

BruceM

[Jedon]

I am very interested in all these methods, my setup is pretty much like Marcus's
16 L16H's at 48V, 2 Xantrex SW+5548's for 120/240 ( well pump ), 800W of PV, a Metro 6/1 with ST-5 ( currently not used until I replace the water pump and fix governor issues or make it a DC charger ). I have a Lister SR2 with 5400W Lima head that I currently am using for all charging.
I don't have a charge controller ( Well I have a 12V Trace but that would seem to be worthless right now ).
Since I'm not currently ( ;-) ) using the ST-5 I could mod it since the SR2 is a good fallback.
Right now since I have a lot of other stuff going on like building the house I'm looking for the cheapest easiest solution that could perhaps be improved ( efficiency? ) later with a suitable application of time and money.

[bschwartz]

Hope no one minds me changing topic back for a minute.....
I picked up fly fishing line to re-tie the ST head after finishing the wiring.  Does anyone have any suggestions for a varnish or other coating that is available at a Home Depot, Lowes, general hardware type of place?  Glyptal isn't something I can readily find around here.  I gave up on the bubble gum when I calculated how much chewing I was going to have to do. 

[mbryner]

I'll see if I can find the calculations for the minimum inductance DC choke for filtering, then we can look for  surplus toroid(s) with high current secondaries.

Please, and thank-you.

Adding a second new string in parallel, that's not the same age, is a no-no.

String 1 has been very lightly used, just so that I would have some power temporarily.    It was actually my system installer's idea to finish the battery bank when the house was constructed, to keep the costs down now and get the tax credits before the gov't can't pay any more.   The current bank has never been below 80% charge, mostly fully charged at all times.

bschwartz, continue....

[mobile_bob]

what i might be tempted to use for varnish i probably would hesitate to recommend

perhaps some shellac if you can find that, it was what was used years ago, but i have
no idea if modern shellac has the same characteristics.

because the stator does not run hotter than hell i would expect just about any clear coat varnish
laquer or urethane probably would work, basically all you need is something to dry and lock everything
together nice and tight.

there might even be some sort of aerosol sold at the box stores for coating motor windings?

maybe you can order a can from somewhere, glyptal would surely be the best and probably worth the
wait to order some?

just lace it up really well, and coat it with whatever you can find, spray in light coats and let dry between each

probably work just fine

i gotta believe there are tons of motors and generators out there still running decades later that were made
and coated with god knows what long before glyptal

bob g

[BruceM]

Marcus,
I did some checking and ebay searching; the 40V secondary windings of the same transformer ($50 ea) I recommended for your step down transformer would also be your best dealt for a filter choke.  You'll  need two wired in parallel to handle the current, so now you're up to four total. If you proceed with this project, and order the transformers,  let me know and I'll make you a transformer, bridge and choke wiring diagram. 

Bruce

[mbryner]

Sounds like the best option.   Except the guy on ebay only has 3 available.   There are 6 of the other ones available but that would give me less then 2000 watts.   hmm.   Also, how would the current be limited?   Didn't think of that until now.   

I started a new thread: 48 Volt charging from ST head through ST head
http://www.microcogen.info/index.php?topic=500

We can move this discussion over there.   

Brett, you can continue with your 60 volt mod, which is still quite interesting.   Sorry to divert this thread so much.

Marcus

[bschwartz]

Marcus, no problem with the diversion.  This whole forum leads from one idea to the next.  Rarely is it a solitary idea, thought, or experiment.  It wasn't really my thread to begin with anyway, it was Bob's.  I'm just the first fool to jump into ripping his ST apart. 

And Bob........ You have my curiosity peaked!  Hypothetically speaking, if someone were to ask you for a varnish type of material, what might you recommend if no one took you seriously or paid any attention?