ST head modification for 48 volt battery charging

[mobile_bob]

something else that will work very well is a flat shoe string, cut to length they are basically a flat
hollow tube that you can sleeve the connections with, then dope it good with your favorite varnish
after laceing it into place.

the insulation class of the stator is class B, so it is unlikely you will really need kevlar or nomex to handle
very high heats generated by some other generator/alternators like the automotive units.

the idea is you want an excellent mechanical joint, very good solder job, insulated/sleeved well and tied down tight
because if there is any wiggle room over time the splice being thicker will have a tendency to wear through and short.

so good work will make it something you don't have to revisit again later.

bob g

[bschwartz]

Where is the fun in not getting to revisit it later?  This is all about the fun.  I live in the middle of a town with $.15 per KWH electricity.  I'm not doing this to save money, that's for sure.  I'm not doing it for convenience.  I'm doing it purely because I can.  The road of experience and adventure is what my journey is all about.  I could pay someone to do some of this stuff, but where is the fun in that.  I want to know how it all works, even it it means breaking some things in the process.  I don't have a lot of money, or time, but what I do have is all for the adventure of learning.  These are the lessons I'm trying to teach my son.  Live for the journey, not only the destination. 
Of course I want to do things right, and carefully.  That is why I ask questions.  But working or not, I plan on revisiting most of these things later just to see how they worked in the long run.

Enjoying my life to it's fullest.  Successes, failures, and all in between. 
I hope we can all learn from the process.

Happy Holidays everyone!!!!

-Brett

[mobile_bob]

Brett:

hopefully my comments didn't offend you, i share the same philosophy as you, in that
the journey is worth more than the destination.

i like doing things to learn, and just as you related, sometimes the results are less than spectacular
but i figure to chalk it up to the cost of an education.

what you are doing with your ST head has been something that has been rolling around my head for several years now,
but for some reason i have more rocks rolling around up there than health will allow me to do.

this is what fascinates me with your jumping in there and doing it, i can live vicariously through you, and learn right along
with you and others.

this forum is turning out to be quite an interesting thing, what other forum have any of us heard of that is actively engaged
in the many facets of system design, control, integration and manufacture?

for being a touch over 3 months old, this place is blazing right along the cutting edge of the diy experience.

so far we got a pretty fair start on the following

1. producer gas, we had a good start, perhaps we can get that up and going again, at least there was on working unit
described on our forum.

2. several types of automation systems, some of the best minds around the diy world are actively working several angles
on this one

3. heat exchangers, we have several folks actively working in this area, with some really good stuff coming out almost daily

4. inverter/charger/battery systems,  most specifically work that is being done using automotive alternators
which most forums dismiss out of hand without having given them a hard look in my opinion

5. now your move toward the modification of the stator of the st head for 48volt battery charging

6. we got fuel engineers, chemists, and other petroleum professionals and other practical folks with much experience.

and the list goes on.

seems really interesting to me, and i for one am really curious to see how things progress over the next year

its going to be very interesting to see the results of your modification of the st head, this may very well be the answer to
the problem for many folks with 48volt battery banks.
and quite possibly an option to eliminate the flicker and other wave form issues associated with the st head driven by slowspeed
single cylinder engine's, without taking a serious hit in overall efficiency.

thats going to be really cool to see come to reality

btw, i have another idea on how best to apply this altered generator when you are done with it, there might be a way to get both
battery charging off of one set of coils (one leg pair) and 120 ac off the other set of coils.  might not be the cleanest 120 vac
because of the transformer effect/coupling in the stator, but might be just what is needed for some applications.. in any event
it would likely be a welcome option to have handy.

i am thinking maybe a buck converter taking the rectified 60vac (60 x 1.414 = ~85 volts DC) and down converting it to the needed
57.6 vdc for a 48volt bank. the mx60 might be just the ticket for that because the ampacity of the unit would not be overrun
or the voltage limit. you would drop the voltage and increase the amperage in proportion and gain proper battery charge control in the process
this would leave the other 120volt leg available to service AC loads to some extent depending on the capacity of the engine.

but that is a test for another day

baby steps as they say



bob g

[mobile_bob]

Brett:

i just wanted to add the following

if i came off as condescending or somesuch it was not intentional, i don't know you personally or know how capable
you are, and perhaps more importantly how capable others might be that are following this thread on the modification
of an st head.

what i don't want to be responsible for is understating the requirements for a proper job, specifically when it can cause
problems and/or serious damage to the machine or perhaps a shock hazard to a user.

its not widely appreciated why those stator coils are so tightly bound, and in a good core tightly sealed with varnish.
as you likely already know there is more at work on those wires than just vibration, you have the periodic magnetic tugging
at the wires as well, and if they are not bound tightly they might over time either work your connections loose, or abraid and short.

this sort of failure might be easily fixed, or in a more likely event (in my case it is always the worst that happens) the connection
abraids and shorts into the coil conductors burning off a turn or more deep within the coil. this  often results in a seriously damaged
stator coil that will require a rewind.

so if i over emphasize the need to attention to detail in this area, its only to hammer the point home to those that might take a shortcut
had i or someone not mentioned it, later having problems and/or a severely damaged generator and then having them pissed off at me.

i got a nasty mean exwife, i don't need more enemies



bob g

[mbryner]

mobile_bob,

Now you're talking holy grail:  1/2 rectified DC and 1/2 120 VAC.  That would solve some problems I was still deliberating with this mod.   I have an Outback MX80 (very similar to a MX60) which was what I would feed the rectified current into.

Also, I appreciate your detailed descriptions on what to do, since I've never taken apart an ST head.   

Marcus

[mobile_bob]

Marcus:

i offered that option only as a "possible" solution, however one caveat that might be insurmountable
is the close coupling of the two coil groups on the same iron core. if the mx60 or similar with its buck converter
injects all sorts of high frequency harmonics back onto the generator output for that group, they will end up on
the other coil group and be imposed onto the 60hz 120vac output as well.

now having said that we got to look at (at) least 4 maybe more outcomes

1. there will be little effect, because the mx60 doesn't put much back onto the dc buss, or

2. there is significant harmonics being backfed onto the  dc buss, but the resultant 120/60hz isn't coupled
close enough to have a marked effect,

3. there are significant harmonics being backfed onto the buss, and the coupling is very good, and we get
a bunch of crap onto the 120/60hz side of things, but for the intended loads being serviced at which time
that charging is taking place, it is determined that it is acceptable, or

4. there are tons of harmonics being dumped back and when coupled they are amplified, and delivered to the
120v/60hz output making it useless for anything other than resistance type loads

its going to be interesting to see which is the case, hopefully the results would be acceptable for certain types of
loads.

the benefit would be having something that would even allow this type of operation, that being the mx60 or other
buck converter, because it also eliminates the need for an expensive 3 step regulator for battery charging.

if i had to bet on the outcome i would expect to see option #3, however #2 is possible because of the 90 (physical) degree displacement
of coils for each leg/group. if they would have been 180 (physical) degree's apart the two coil groups would be much more coupled in my opinion.

hopefully option #4 is not the case, although it would not be the end of the world, as capacitive filtering to dampen the high frequency
component superimposed onto the 60hz output might be an effective option.

lots of questions,  no answers yet.

bob g

[BruceM]

I just looked at the specs for the Outback MX80;  this MPPT switching charge controller unit will do an operating input DC voltage of 145VDC.  So if you were going to use an MX80 to charge a 48V battery bank, you could just use the stock 120V ST head output, bridge rectified and filtered via inductive filter (120VDC), with a bit of capacitance (10K uF) afterward; there is no need or benefit to stepping down the ST head voltage first,  since the MX80 will do this for you with high efficiency. 

The inductive filter would be ideal for this, as it will improve efficiency  (PF) and also preserve the usefulness of and remaining 120VAC,  which otherwise will have the top "sawed off" by the switcher (gross waveform distortion).  A no load, high voltage cutout circuit (could be a SS relay) woud be needed to protect the MX80 from overvoltage.  This would be connected to the final capacitor, and would disconnect the mains, or the rectified DC until a load dropped the capacitor voltage. 

A good AC charger with a decent filter (filterconcepts.com) in front of it might do a better job, with less hassle.  Seems like it might be available for less than the price of an MX80. 

Just my 2 bits.






.

[bschwartz]

Goodness Bob!  No offense at all!!! 
As much as you wanted to stress the importance of a well thought out, planned, safe job, I wanted to stress how NOT important this to my energy needs.  Many people reading and participating in this forum NEED these power plants to work.  I would caution any of those to not jump in as quickly for fear of messing something up.  I'm working on this not to charge batteries, but to hopefully directly run a 48v grid tie inverter to run my electric meter backwards.  This ST head was hooked up to my metro 6/1 with unsatisfactory results.  A heavy load on one leg would cause a bad voltage droop on that leg (expected) with a large rise (139v) on the less loaded leg (unexpected).  I hooked the legs through the primary windings of a 3KVA transformer with some degree of success, but still not what I was hoping for.  I then purchased one of Tom Osbornes 6.5K permanent magnet heads.  WOW what a difference!  I can pull all the engine can deliver out of one leg and the gen head doesn't mine at all. No moan, no groan, no huge voltage droop or rise.  Soooooooo I had this extra ST head sitting in a corner begging to be played with.

After I get this wired in place, and producing the ~60vac, and rectified through a full wave bridge, how do you think using the primary windings from earlier mentioned 3kva transformer would work for filtering?  What do you suspect output voltage would be, and would it be clean enough for a grid tie inverter, or would it need to be cleaned up more with some large caps or a small 48v battery bank?

Wooo Hooo !!!!! What fun!!

[bschwartz]

As for a glyptal or varnish, what would you recommend that is available at a home depot type of place.  Heck this thing is already going to be held together with shoe strings and fishing line.  Maybe some bubble gum? 

[mbryner]

If the coil sets are 90 degrees apart, it seems like we would also need a dump load at the 120 vac coils if used for DC charging on the other set?   Otherwise the groan and vibration would be terrible?

BruceM, I already have a MX80 for solar arrays, but for others, yes, that would be a significant expense.

Brett, you may be onto something with the bubble-gum idea! 

[BruceM]

Thanks for getting me sorted on the MX80, Marcus.  Nice unit when on hand for free.  If you decide to do a 120VDC setup, and need help with overvoltage protection, let me know.  A simple SS relay and 120V zenier circuit would do the trick.  Resistor after zenier, goes to NC SS relay, opens AC or DC supply.  A different arrangement could be done for a NO SS relay, that would require a small DC supply.

Regarding Brett's application of 60VDC to grid tie inverter:

60VAC with inductive filtering will result in 60VDC.  Caps probably not needed at all, or some (10K uF) after the inductor only if needed to keep the grid tie inverter happy. (not likely)

You have to look at not the VA rating but the current rating of  the various transformer windings, when using them for a choke.  Starting with a winding with about double the max rectified current is about right, but for some cores, you can go equal the rated without seeing saturation (no improved filtering).  I use a meter to measure inductance, and resistance of each winding. The DC resistance and ohms law gives you the voltage drop you can expect at a given current.  At 60V, I wouldn't cry over a couple volts.

The 3KVA transformer would likely have a primary or secondary you can use, do you have some specs or a wiring diagram for  it? If there are multiple taps, primaries or secondaries can be used in parallel. But not if center taps only.  What is your desired maximum current at 60VDC? 50 amps would be 3000VA.  What's the max you grid tie inverter can handle?

BruceM

[Crofter]

Very interesting to see that the coils for a 120 pair are not at 180 deg but adjacent instead. Perhaps that is significant as an issue in growl with unbalanced leg loads on the ST5. Probably a less costly way of interconnecting coils but maybe has some negative tradeoffs

[mobile_bob]

i think we need to keep in mind what the original design was for the ST head
it was likely setup to be a 230-240 single phase 50hz head, and as such the placement
of the coils make sense.

then when the need came up for a 120/240 generator, it was probably easier to divide the coil groups
in one place, and attach leads,

some of the cheaper units didn't even completely divide the groups and provide two seperate leads
rather they just attached a neutral lead and called it good.

once i got my head around this issue, it explains some of the problems with growling, and i suppose
apart from being a nuisance it probably won't cause problems.

unless of course the coils are not tightly constrained with varnish/shellac/waterbuffalosnot



makes sense to me

bob g

[mbryner]

Sort of what I was thinking, Bob and crofter, but I don't have any experience with this.    Makes sense.   So could the configuration be easily changed so that the coils opposite each other were joined with a simple mod?  It would seem so...

BruceM or anyone, I'm a little confused.  I'm sitting in the hospital reading x-rays and CT scans so don't have access to electrical texts:   If you rectify 120 Volts AC RMS with a bridge rectifier, is it 120 V DC or somewhere around 171 Volts?   Is the smoothed out ripple still 120 V DC?  Just double checking.    That's still close to the max open circuit voltage for a MX80, so I would definitely need a overvoltage protection circuit, in case the Listeroid overspeeds, etc.   But that would simplify the situation.  No transformers needed to charge 48 V battery bank, but still have 240/120 from the ST!

Sorry bschwartz for diverting this thread a little.   Many people probably have an MX60 or MX80 charge controller.   If you can breaker and keep the inputs from the generator (120 V DC) and DC in from the solar arrays separate, you can use the same charge controller.   When it's dark, charge using generator.  When the sun is shining, charge via solar.   It could even have an automatic transfer switch, but probably not necessary considering the manual start of the 'roid anyway.   And there could still be regular 120 V AC plugs next to the 'roid for emergency use.   I like....  

P.S.  Yes, BruceM, I'd appreciate your help on the overvoltage safety.   I think I understand but do you have a schematic (just quick sketch)?   I understand schematics when not *too* complicated, no circuit boards or anything! 

[BruceM]

The output of 120VAC, after a full bridge rectifier, is 120VDC, with a peak voltage of 1.4 times that. It's all RMS, remember.  If you add an inductive filter, of sufficient capacity, then the output will still be 120VDC, but the peak voltage will be 120V plus a few volts ripple.  Adding a capacitor after the inductor will just reduce the ripple a bit.

It's when you try to use a capacitors as the primary filter that output voltage increases to approaching 1.4 times the RMS AC voltage, and your generator sees big "gulps" of current as the AC waveform rises above the capacitor voltage.  Capacitors are cheap, compared to inductors, and for small low voltage DC supplies, the transformer buffers the "gulps" sufficiently, so inductors aren't used for filtering in these supplies. 

In my battery charger(120VDC battery bank) I zero cross- switch small amounts of (metal film motor run) capacitance before the primary inductor filter intentionally, to be able to dynamically adjust charge current.  (At some expense in power factor.) I don't have to use large amounts of capacitance as I also have two "boost" transformer windings to switch in.  Just an example of how both inductance and capacitance can be used to make a linear supply adjustable.

I'd recommend doing some reading on basic linear supply design equations.  I did a quick look at Wiki, and they don't have much 

For the overvoltage protection, I'll see what I can do.

Bruce