modified st controller

[BruceM]

Jung's "op -amp cook book"  is only so-so for a beginner with op amps, but is a good reference that you should certainly have.

Radio shack had an old "engineers notebook" series that was good for beginners.  But I'm sure there's some new texts that are decent on B&N or Amazon.

By the way, Bob, your diagram shows the excitation for the ST head coming from the 48V battery bank. That would be fine for a 48V alternator, but its too low voltage for an ST-5, even as a no-surge battery charger.  Much higher than faceplate excitation voltage must be applied, just to compensate for the engine exhaust and compression cycles. (I started with just above faceplate excitation voltage on my ST-3; boy was that pathetic; then I measured and found 400V peak excitation from the harmonic on a surge load, and increased excitation voltage via swapping transformers until  I got comparable motor start performance.)

Excitation can be "lumpy" DC, so in this case  (60V mod) it could come from a step up transformer with the output bridge rectified. Both my AVR designs use 120VAC, rectified, for excitation which is good for ST-3s or ST-5s. (Flicker reduction is better on the ST-3, as the field coil has less induction and 120VDC excitation voltage is able to affect a change in voltage faster.)  

Another design issue:  You can't regulate voltage on one pair of windings separately from the other- so if you want 120VAC on one set of windings (series) and 60VAC (parallel) on the other pair, you can use a version of my basic AVR (fixed voltage regulator) but then you'll have to use something like the PWM Tristar-60 to do battery charge regulation.   That is not a bad solution, as buck converters are often over 90% efficient. This would even work with the harmonic winding, for those who can stand such barbarism.  

One catch- a modest amount of capacitance (100uF of motor run capacitor- nothing else will handle the ripple current without getting very hot) before an inductive filter may  be needed to boost DC voltage for cyclic charging.  In cold weather, the cyclic peak for a 48V bank is 62.4 volts, with a lot of amps, and resulting wire and battery losses above that. The voltage may have to be raised even further to accomodate the PWM Tristar-60; some measurement of charge efficiency and changing the DC voltage would get you to an optimum configuration.
 
If the entire output of the ST head is dedicated to charging, then a smart charge controller/AVR could directly regulate battery charging by regulating the ST output directly to the desired bulk charge current or adsorption voltage.  This would be the ultimate in efficiency and you could slam the whole output of the unit into the battery bank during bulk charging, if the batteries can take it.  

[mobile_bob]

Bruce:

"Jung's "op -amp cook book"  is only so-so for a beginner with op amps, but is a good reference that you should certainly have.
Radio shack had an old "engineers notebook" series that was good for beginners.  But I'm sure there's some new texts that are decent on B&N or Amazon."

i will do some looking around, i have one of the old engineers notebooks, somewhere around here...

"By the way, Bob, your diagram shows the excitation for the ST head coming from the 48V battery bank. That would be fine for a 48V alternator, but its too low voltage for an ST-5, even as a no-surge battery charger.  Much higher than faceplate excitation voltage must be applied, just to compensate for the engine exhaust and compression cycles. (I started with just above faceplate excitation voltage on my ST-3; boy was that pathetic; then I measured and found 400V peak excitation from the harmonic on a surge load, and increased excitation voltage via swapping transformers until  I got comparable motor start performance.)"

i knew when i sketched the concept drawing there was an issue with excitation voltage, what i should have done was perhaps inserted another component group, such as transformer and separate rectifier to provide for excitation power, hadn't given that very much thought because quite frankly i have access to about any reasonable level of excitation current on my machine, (i forget often this is not the case with everyone)
also i neglected to consider the requirements of the st5, my st7.5 needs 56volts @ ~2.6amps at full load iirc. and i guess i quickly assumed that if the batteries are being charged during bulk at 57.6 vdc that i would have enough to provide excitation. but i agree with you that is a flaw in the sketch that needs another option.

"Another design issue:  You can't regulate voltage on one pair of windings separately from the other- so if you want 120VAC on one set of windings (series) and 60VAC (parallel) on the other pair, you can use a version of my basic AVR (fixed voltage regulator) but then you'll have to use something like the PWM Tristar-60 to do battery charge regulation.   That is not a bad solution, as buck converters are often over 90% efficient. This would even work with the barbaric harmonic winding, for those who can stand such barbarism. "

i am aware that one cannot take 120ac off of half the machine while loading the other half for battery charging, unless
he was to regulate for 120operation, and take the other half at full 120 , rectify and feed it to a buck converter for charging
then the regulator maintains the 120 vac side of things, while the buck converter regulates whats left over for stable battery charging.

it might be though that under certain charge conditions, that even this scheme might provide sufficient voltage stability for some loads
for instance perhaps 20amps off the 60vac coils is not too much load as to mess up the 120vac side so much so that it could not be used
for a few hundred watts of AC power just fine. if that is even possible it would have to be determined in testing and some form of control
would have to closely monitor that so that it doesn't whack out and drop voltage too far.  it could be done with another feed back channel
from the 120vac side, wherein if the microcontroller saw that the 120vac side was starting to get too low it would then cut back the charge rate
so things could come back into range.  yes it gets more complicated than what you describe, call me "rube" 

"If the entire output of the ST head is dedicated to charging, then a smart charge controller/AVR could directly regulate battery charging by regulating the ST output directly to the desired bulk charge current or adsorption voltage.  This would be the ultimate in efficiency and you could slam the whole output of the unit into the battery bank during bulk charging, if the batteries can take it."

i would agree that this final option has the highest chance of success, the split system might be made to work with some complicated control
schemes, but we aren't going to know how complex that task might be for sure until it is tried.

the thing is, i can see little reason to need a split alternator system, most folks if they have the need for battery charging will have an inverter
and if they have an inverter it will almost always have a higher rating in kwatts than the generator has, so
it would seem most probable that one would use the total capacity of the st head to do the charging and just use the inverter to provide for
AC power all the time, whether offline or during charging.

i have to keep reminding myself that just because something can be done is no reason it should be done, so by extension
just because some might be able to be done, may not be reason to try and see if it can be done, save for educational purposes.

i guess until we see Brett's conversion results on his st5 it is going to be problematic forecasting what might have to be done in order
to use it most effectively and under what circumstances could it be used in a split system situation (if only as an academic excersize).

Thanks Bruce for your well reasoned response, it is always appreciated

btw, your getting to be an excellent instructor, keep this up and we will make you an honorary Professor around here!

(or is that "ornery") 

just kidding 

bob g

[mkdutchman]

Bruce:

op amps are something i look to as links that tie together a system, and not something i gave due attention to
i guess.

would you have a favorite book on op amps that you would recommend?

bob g

I'm not bruce, but FWIW several electronics books landed in my lap this a couple days ago......my favorite so far is this one, Practical Electronics for Inventors, by Paul Scherz. Many electronic basics and theories, including op-amps are very clearly explained and illustrated. And an online source I've used a lot in the past is this one

http://www.allaboutcircuits.com/

their op-amp section explains opamp theory very well IMO, http://www.allaboutcircuits.com/vol_3/chpt_8/1.html

I still like the book better, it's so much easier to curl up on the couch with a good book than to hunch over the puter screen.......

[BruceM]

Bob,
Definitely Ornery, not Honorary.  I didn't mean to lecture, but wanted to pass on my experience with the ST face plate excitation being an average for resistive loads, at best, and not indicative of what you need for decent regulation in any real circumstances.  I wasted a lot of time on that one.

I'm getting near the end of checkout of my Battery Bank Charge Controller board, and ran into a mystery. The 16 channel analog input multiplexor works fine on the 10 battery shunt monitor signals ( simulated on solderless breadboard now) ONLY if I have a voltmeter on the input or output signal.  Yet there are 10K pull ups on the mux battery shunt inputs on the PCB, and the voltage is changed via opto-isolator pull downs... AND the 4 channel current sensor analog signals going through the same mux worked just fine.  

I give up for today- something odd going on, and I just can't see it.

[mobile_bob]

Bruce:

don't ever be concerned about "lecturing"  i appreciate that sort of thing anywhere i can get it!

and you are very articulate in your presentations, whatever issues we might have in communications is
likely my shortcoming in using the proper verbiage.

i get into trouble a lot over wording or phrasing issues.

my favorite was when i was called an idiot on another forum for calling a small shop engine lift a "cherry picker"

this by a guy who is a used machine tool dealer, and who unlike me hardly has ever used one.

he was like a dog on a bone over that one, i guess folks might have been confused and persuaded to use a bucket lift
truck to lift a lathe into position, anyway that was his problem with me using the term.

anyway, i appreciate learning all i can from you, and really appreciate your patience with me.

thanks

bob g

[EBI-WPO]

Where you are from (or where you grew up) has a lot to do with phraseology. And it can be guaranteed to be a problem with someone from somewhere else, unless you don't do or say anything!

Terry

[dubbleUJay]

What makes it even worst is when your 1st language isn't English !!!
Believe me!

[WGB]

Thanks for the links mkdutchman!
I've been trying to follow you guys, most of the time I'm left in the dust.
I do know switching, starting, machine, HVAC controls, etc., pretty well.
Very soft on PLC type controls.
I do have Mitsubishi software to program one of their cheaper 200 to 300 dollar controller series. I think I could get into doing PLC controls.
I reprogramed an industrial paint line with one.
I have Bill Rogers book, and looking for any info to get up to speed.
Most of what I read, I'm doing good then turn a page and I'm lost, like half the information isn't there?
I have noticed some of my last stupid questions haven't even received many replies. Sorry.
Thanks for the help Bruce!

[BruceM]

I encourage all guys who have had some PLC experience to order one of the US- Picaxe solderless breadboards from PHAnderson.com.  The 18X us a nice starter chip to mess with, the 40X2 is the top of the line.  ($34 and $37).  That is the sum total of your development tool investment.

Play around with Picaxe basic, and you will have gained a new tool that is robust, reliable, cheap and incredibly flexible.  Your fears of micros and programming will melt away quickly, as these chips make it so easy.  You'll find outstanding support on the revolution education forum,  http://www.picaxeforum.co.uk/ 

[Jedon]

I'm trying to figure out if Picaxe etc are like what I have been doing with custom boards, from a programming perspective.
I feed the serial I/O from a board into a device server that converts it to ethernet, then I write C# to read the bits coming in async and decypher the stream from docs that say stuff like bits 4-7 are sensor one values, then the timestamp then battery voltage, then CRC etc. I also go the other direction, forming command packets to send out to the device.
So is this pretty much what programming Picaxe is like? Or is it more like assembler where I need to play with registers and stuff?

[BruceM]

The Pixaxe programming language is a (simple) version of Basic.  No assembler required. 

Their free programming editor lets you step through your code, watching your variables and pins change.  I haven't ever seen an easier product to use.