another st charger option

[mobile_bob]

Guys:

here is another option that might be fairly inexpensive to implement

the use of an mppt controller is pretty expensive in my opinion, and i am not sure of the benefit other than
the boost converter picking up the amperage from the st head.

then there are those that are using transformers to drop the st voltage to ~80volts or so, so that after rectification
the voltage does not exceed the mppt upper voltage limit, so

if one is to use a transformer to start with, why not just get on to drop the voltage down to perhaps 20 volts or so open circuit
and feed it to a controlled rectifer, which is triggered by a less costly regulator or a 3 step, via a mosfet driver.

using surplus parts, and maybe some adaptation of existing hardware one ought to have the ability to get full wattage from the st head
into dc charging to the batteries, assuming of course they need and will accept it,, in any case i am sure one could get far above
the 60-80amp limit of the common mppt, most especially working with 12 or 24volt systems, where the amperage would be twice or more
over the ability of an mppt controller.

attached is a concept sketch, and yes i probably have a few flaws included 

thoughts?

bob g

[Geno]

What the hell is a "Controlled Bridge Rectifier" and why haven't I heard of it before!!! Am I risking life, limb and pocketbook not knowing all the options!!!

What I meant to say was:
Please sir, may I have more information on this "Controlled Bridge Rectifier" thingy?

Thanks, Geno

[mobile_bob]

i think Bruce can explain the function of a controlled rectifier better than i, but here is a feeble attempt

basically we replace the bottom two diodes in the full wave bridge with something like a triac, or maybe an N-channel
mosfet

the rectifier then cannot rectify unless the controlled elements are triggered on, once they are triggered on they stay on
until one of two things happen, the trigger signal goes away or the sinewave crossover zero volts (then the trigger will either remain
on to keep the rectifier conducting or allow it to stay off.

the regulator/driver is basically a pwm unit that samples the battery voltage and adjusts the pwm which in turn drives the mosfet driver, which
fires the controlled elements of the bridge (which in this case are the lower rectifiers of the bridge)

this is all done at frequencies from a few hundred hz to several thousand hz, perhaps much more depending on various criteria.

it of course is the average on time that adjusts the voltage and keeps it regulated to the voltage you want, or program for.
depending on what kind of regulator you use.

i will defer to Bruce for the finer points of operation.

bob g

[Lloyd]

So Bob,

it of course is the average on time that adjusts the voltage and keeps it regulated to the voltage you want, or program for.
depending on what kind of regulator you use.

Kinda of like the a modern version of the vibrating dynamo regulator.

Lloyd

[mobile_bob]

Lloyd:

exactly, that is what the modern regulator does, it replaces the vibrator with an electronic pwm control


more on this subject

there are at least three control handles one could use with this system

one would be as described, the controlled rectifier

another would be a pwm control of the AC going into the primary of the transformer, the pulse width would be modulated based
on a feed back sample of the battery voltage, and

another which might be even simpler would be a regulator that controlled the st field and based the pulse width on the same scheme as
above, being a sample sense line from the battery bank

i suspect each has its pro's and con's, or compromises to deal with

but i am thinking that either of these methods would be less expensive, more robust, and higher output capability than the mppt controllers

i got a heck of a book that might be useful in designing one or all of these systems, i guess i better dig it out and start reading again.

bob g

[veggie]

Bob,

With this setup, would the ST head have to remain at 60HZ or can it be slowed as low as 40 hz ?

veggie

[mobile_bob]

upon some further thought, i am  thinking the third option might be easiest because it would require the lowest power electronics
of all the options, the field current is relatively low compared to the AC output and definitely dramatically lower than the DC product of the
step down transformer.

i can see conceptually a marriage of a microcontroller, a battery temp sensor, a driver circuit for the field, a couple look up tables and some
coding, and you would have a nice 3 step st battery charger.

maybe someone wants to jump on this project?  all i know is my hands are full!

but the more i think about it, the more i am liking the possibilities such a system would provide

its is a very large alternator that can put out 7.5kwatts of DC power into a 48volt, 24volt or a 12 volt battery, this system would make it possible
i think.

the deal is, if this concept shows promise, it would mean me being faced with scrapping a good part of my system for the lack of need, and a
huge leap in simplifying things.

so there you are, more questions, and fewer answers


bob g

[mobile_bob]

veggie:

i think a system could be designed with the capability to operate at 40hz, and still allow for operation at 60hz if there was a need
to do so.

you would certainly want to engineer for 40hz, and then allow for pwm tapering back the duty cycle for 60hz operation.

otherwise you might have to have a 100% duty cycle to operate at 40hz if it worked at all, that is if you design for 60hz
and then try to slow it down.

there is some discussion regarding pulsing current not being good for batteries, and there are those on both sides of that arguement
one side likes pulsing for desulfation (which i don't ascribe to) the other wants nice clean low ripple DC (which seems prudent to me)

and there are tons of 60hz battery chargers out there albeit most of the large ones are 3phase and as such low ripple by nature, so maybe
such a system would benefit by a bank of large caps? or a cap/inductor filter? maybe surplus parts?

i have a rather large bank of dc caps from a huge old inverter that i could probably call into duty, but this is something i have no experience in
and would defer to Bruce and other EE types either on this board or elsewhere.

it just seems like a worthy project, at least to me



bob g

[Lloyd]

Bob,

I was doing some reading on the subject, and trying to come up to speed, but atlas I scrapped it because, my lack of being an EE. I have some pdfs, on a system you are describing, also I think I posted a link in another topic where they were using, various of amplifier switching, fets, and mosfet, to rectifie without the big vd.

My pop was a stellar EE, he started at Otis, then to Texas Inst, then to Litton, being he had a family he had to go as a salaried employee...my Mom wouldn't let him risk the start ups. We ended up in Seattle when, Boeing moved us here from Texas. That was the first wave of the tech boom, back then he was a slide rule king, our basement was full of scopes and tubes.

Lloyd

it seems I'm ok at understanding just enough to frustrate myself, and those that take my que.

[Lloyd]

maybe,

We could get halfcrazy to get the boys at mid-nite solar to build such a deal, since an mmpt controller is probable ill suite to create a very effecient gen/controler for bat charging. Especially since the bats are the heart of an efficient off grid.

I think the perfect off grid in a lot of situations is, an array of PV, that can cover some nominal loads while taking the bat through absorption/float, and then let the gen do the bulk. That's how I have my boat set up which is a semi-grid/off grid.

Lloyd

[BruceM]

SCR's can certainly be used for power regulation; at one time they were one of the only options for high voltage and power switching, besides BJT transistors.

SCR's or Triacs (which are just two SCR's) are the control element in many AC motor controls and virtually all dimmers.  Since they can't be switched off (they only go off when current goes to zero), they can't do high speed switching, and for DC, would need large inductors and capacitors for filtering.  You do still see SCR's used as solid state relays for AC in a charger or power supply.

So for a DC output,  it is generally much easier and cheaper to just control it via PWM of a MOSFET, as the MPPT units and AC chargers now all do.  In our case, if using a massive inductor for a filter, then SCR's could be used, if desired.  It's outside my design experience to use them for such an application.

The other method- of going back and regulating the ST with feedback from the battery charger DC (voltage of battery, temperature,  and charge current) is a nice one that we've discussed before.  I do have the schematic elements for this (temperature compensation circuit, AVR, current sense, even transfered onto a single schematic page.  But I have no need for it myself and don't have time to work on it just for fun.

There is a lot to be gained by using the MPPT charge controller that is already there for PV regulation (at least in a 48V system) for the AC battery charger.  It already knows how to do temperature compensation and 3 stage charging, and equalization.

The only bugger we hit is that it would be nice to not have to reset the max battery charge current to limit the AC charger current draw.  Reducing the voltage of the transformer ouput via a switch on an AVR (drop the feedback voltage) or a harmonic dropping resistor might be one way to solve the problem.  Alternately a current limiting chopper circuit would do the trick but we are trying to avoid any custom circuit work.

If Midnite solar was to add a secondary battery charge current limit which could be triggered by an input from an AC charger's DC output, then their MPPT unit would solve the problem in a nice way, and regulate for both AC charger and PV, (or wind).  If the unit was capable of tolerating "lumpy" DC (rectified AC with no filtering), that would be even better;  then all it takes is a transformer and rectifier to have a fancy AC charger.  If the unit had a "charge complete" output for shutting down the AC charger, and an "auto gen start" output signal...it could handle a very sophisticated, fully automated setup.

[Halfcrazy]

I will try feeding my 250 volt unit with power through a rectifier and give an update. when I get 5 minutes.

two points come to mind here. one is the transformers in the previous thread where rated for 30 amps and he had 80 volts so that would turn into almost 60 amps at 48 volts unless I am missing something here I haven't followed that thread real close as I plan to rectify 120 vac then go through a classic.

Second point I will have to ask our engineer how hard it is to take a signal and trigger a lower output current we do have 2 aux relays on this thing and one can be configured for an input.

[BruceM]

Thanks Halfcrazy.  The Midnite solar unit may be the new king of MPPT charge controllers. 
You're right about the buck conversion power efficiency to 48 volts on Marcus's MX80, though it's only 40 amps of charge at a bulk termination voltage of 60V (15V per 12V battery).  An aux input to the Midnight which enables a max wattage input (from an AC charger) would allow the controller to get the most out of the AC charger during bulk charging.  But any implementation of an aux current limit would put it way ahead of the other products.  The MX60/80 has no inputs at all besides the temperature probe.

Llyod-  Perhaps this was a typo; "since an mmpt controller is probable ill suite to create a very effecient gen/controler for bat charging. Especially since the bats are the heart of an efficient off grid."  If you meant that MPPT controllers are ill suited, then you're ill informed.

You should read the specifications and manual on the Flexmax MX80 and the upcoming Midnite product (or Tristar MPPT); they do a very capable job of battery management.  There is no reason they can't be used as the battery management end of an AC charger.  They make a fixed voltage, non-temperature compensated charger look like the barbaric piece of battery abusing  crap that it is. 

In many locations and seasons, PV systems do the entire charging cycle without a problem.  It's only you guys in coastal Washington who don't see the sun very often that think it's only for absorption stage charging. 

[Lloyd]

Hi Bruce,

It's true that PV can do the bulk in a lot of areas...but it takes a fairly large amount of panels, especially in areas that have sun challenges. The NW is a huge off grid market with all of the vacation cabins on islands, and back woods not close to the grid.

The set up for most is to; PV, gen, conservation, and using small bat banks (400-720amphr). The need for well  pumps, and some other 220 v loads...really hits the banks, so a gen is the answer, just bc of the shear costs of a PV sys that could do it all, with our sun challenges.

I guess I am missing something about the mmpt's. If using as a gen controller, what's the most amps it can bulk charge at?

On a 12 volt sys, with a dc gen and a 250 amp alt, you can get close to 220 amps during bulk, make the hot water, run it twice a day, once in the morning, and maybe 1 in the eve. for 1hr each run and burn a little less than 2 litters of fuel. Let the small PV run absorbtion and float. After amortization of gen, and computing fuel costs, I think that's a pretty efficient sys. Bruce, would an mmpt controller do this on an st? If so that would be a great little system.

Thanks for your help Bruce.

Lloyd

[mobile_bob]

i just finished an article on this topic from sandia labs a few weeks ago,and will find it again
and link to it when i can.

the concern of that article was solar panel battery systems often  lose capacity of the batteries
because there is not enough sun hours per day in most area's , most of the year, unless
the system is very well engineered, which acccording to the article most are not.

folks like to live an ongrid lifestyle offgrid and to do so requires a mountain of batteries,
just like horse,  big horse need lots of oats

so you find you have to have a huge panel investment, and then find you need more than one controller
to handle the amount of power, and it gets very expensive real fast, unless of course you engineer well

get it wrong and it gets out of hand in a hurry.

doing bulk charge with a genset makes sense to me, as it is the most efficient part of the charging cycle
and if you are recovering heat for domestic uses, then the overall efficiency can be quite good in my opinion.

the problem as i see it is it is a moving target and no one answer will fit every application

where we need 4, obviously 2+2 will get you there, or 3+1, or 1+1+2, or 1+1+1+1 will get you there as well
and none are the same as 2+2 or each other for that matter.

that makes recommendations very problematic, especially if the user refuses to educate himself beforehand.

personally i think every one of the options discussed on the forum is viable, under certain conditions, and there
is probably one of the options that fits better than all other for each individual application.

it is nice to have a forum populated with folks willing to try and test each option and sort out the pro's and con's
maybe what won't work for one guy, might be the perfect solution for another?

bob g