DC to DC

[b1bone]

I have a fairly hefty supply salvaged from a fairly hefty server. It's output is rated at 750 watts (56v @ 13 amps). This year I will start using a123 cells. I currently have both the 1100s and 2300s (Merry Christmas). I will be making packs of different cells counts. I plan on charging at the field. A/C is not a problem. I would like to take advantage of most of the 56 volts that I can. I would prefer not to play the resistor game and build DC/DC and make the output varaible. It looks like I can make something using some LM338s but I am not sure they can handle the 56v. Does anyone know if they can? If not would anyone have a pointer to a schematic that would help.

[Dan Baldwin]

LM338 are 5 amp linear regulators that are rated at 32 volts of differential so basically, no, they will not work with your 56 volt supply. Since they are linear regulators, they would get extremely hot it you were able to use them at their full rating of 5 amps if you had very much voltage drop at all.

Dan

[ductapemaster]

If you really want to use that power supply, you would need a switching regulator as opposed to a linear (LM338) regulator. They have very high efficiency and generate little heat. The only problem is that to get a lot of amperage out of one, you need to spend some money. Making one that steps down 56 volts at 13 amps to a more reasonable level (say 12 volts) will have to deal with a lot of current, thus making the parts bigger and the cost higher.

If I were you I would sell it on ebay or something and buy yourself a real 12V (or whatever voltage you need) power supply.

Good luck!

-Dan

[b1bone]

Thanks Dan and Dan. 12 volts would be on the very low side Dan. I have a 10s a123 that I plan on using if I can figure out how to charge it. Considered Charles zip charger but the stock pack will fit my application perfectly as it is so I don't want to seperate the cells. Building something from a schematic is not a problem. I just don't know what's out there. Thanks again guys. Time for another plan.

Oh yeah, DB, thanks for taking the time you do building and posting your schematics. I have build a few of them. Thanks again.

[vintage1]

You CAN go linear if you don't mind building something massive, with a lot of power transistors and fins, and maybe a fan..

The brute force method.

You worst case dissipation would be at 0v and 13A output..that 750W.

Probably at least 15 power transistors. On a fan blown heatsink. And thats just for the output. You would need a few more to drive them.

[simingx]

Modifying a SMPSU for different output voltages involves, at minimum:
- Rewinding of the main transformer
- Beefing up output rectifier for the increased current
- Tweaking voltage feedback circuit to get desired voltage
- Modifying overvoltage/overcurrent circuits to operate correctly at new voltage/current
among other safety-related issues...

[b1bone]

V, thanks, I enjoy building these kinds of things when I can but it sounds like it would be more trouble than it's worth.

simingx, I actually thought about opening it up and taking a look inside but I have no experience with switchers. If it was linear I know just enough to be dangerous.

Oh well, maybe it's time to pick that Mastech I've been thinking about.

Now, how many a123s does it take to make 56 volts. :rolleyes:

[dleroi]

Thanks again guys. Time for another plan.



I wouldn't give up so fast. There may be a simpler way of reducing the output voltage.

What do you know about the 56V power supply? Do you have a schematic?

Does it have a 110/220 switch? Do you know for sure that it's a switching PS? How about a picture of the thing?

- Don

[simingx]

Do remember though, that 750W at (say) 13 volts, is 56 amps (conveniently, just the other way round)... you'll need really beefy components on the secondary side to handle those currents.
The amount spent on power semiconductors and high ripple current filter capacitors (I've seen values in the region of 10,000uF and up) may be more than the cost of a new PSU... Unless you like tinkering, in which case money is no object :D


Regulating the output down with a linear regulator is a really bad idea... you'll still be only getting 13 amps out of it, plus the poor regulator will have to get rid of (56-13) x 13 = 560 watts of heat :eek:

Think a halogen lamp....

Another (much safer!) way to do it would be to add something like this: http://sound.westhost.com/project89.htm on the output, but connect the transformer backwards to get a step-down effect. The voltages (12 to 60 V) are conveniently close that you'll probably be able to use the turns ratio suggested, directly. It probably won't even require any feedback... :)

Obviously the ETD39 core will not be enough for anywhere near 700W, but for experimentation it'll be good enough. Even better, if you build one (good for say, 300W?) and manage to make it work, just build another and get 2 isolated supplies from 1...
The only problem I see here is how to get the 12V supply for the switching IC.

[Dan Baldwin]

I'm sure that you could put together something to charge 10S A123's from your 56 volt power supply for less than the price of a Mastech. You might be able to put several 500 watt halogen bulbs in parallel to limit current, but I'm not sure what you'd use to cut off the charge. The DB terminator won't do it. It would be helpful to know what kind of power supply you have.


Now, how many a123s does it take to make 56 volts. :rolleyes:


15 cells would be 54 volts at full charge. :D

Dan

[Chippie]

The only problem I see here is how to get the 12V supply for the switching IC.


Just use a simple series reg with a 12v zener off the primary supply line...I'm sure the current demand by the ic wouldnt be that high.. :)

[AndyKunz]

We make 54V DCUPSs with A123's at work (telcom), using 16S strings.

Andy

[b1bone]

Everbody is talking like 500 watts of wasted energy is a bad thing?? Think about it, if I'm charging at home I could charge my batteries and heat my basement at the same time and it would do wonders for my frozen hands at the field on these cold winter days. ;) OK, everybody wanted pictures. Here are some of the outside. I opened it up and have some from the inside also but won't post unless someone wants to see them. I did take note of some of the logic I saw though. 2 things that got my attention were a mc68hc705j2 ucontroller and in particular a smp220sri. Can't find a data sheet for it though. Could only find this "SINGLE-OUTPUT VOLTAGE-MODE SMPS CIRCUIT".

[b1bone]

Do remember though, that 750W at (say) 13 volts, is 56 amps (conveniently, just the other way round)... you'll need really beefy components on the secondary side to handle those currents.
The amount spent on power semiconductors and high ripple current filter capacitors (I've seen values in the region of 10,000uF and up) may be more than the cost of a new PSU... Unless you like tinkering, in which case money is no object :D


Regulating the output down with a linear regulator is a really bad idea... you'll still be only getting 13 amps out of it, plus the poor regulator will have to get rid of (56-13) x 13 = 560 watts of heat :eek:

Think a halogen lamp....

Another (much safer!) way to do it would be to add something like this: http://sound.westhost.com/project89.htm on the output, but connect the transformer backwards to get a step-down effect. The voltages (12 to 60 V) are conveniently close that you'll probably be able to use the turns ratio suggested, directly. It probably won't even require any feedback... :)

Obviously the ETD39 core will not be enough for anywhere near 700W, but for experimentation it'll be good enough. Even better, if you build one (good for say, 300W?) and manage to make it work, just build another and get 2 isolated supplies from 1...
The only problem I see here is how to get the 12V supply for the switching IC.

This doesn't look bad at all. I might already have some fets that would work. I wind my own motors and just placed and order for more wire. I have all of the even guages from 14 to 30. No problem there. The circuit is simply enough. Chippie's idea for getting the 12 volts should work.

[simingx]

Too bad they look like they're pretty big... I would otherwise buy them off you for powering amplifiers.. :)

[Dan Baldwin]

The power supply says that it is adjustable from 48 to 58 volts. If you set it at 48 volts, it makes things a bit simpler, and it's possible that you can change the feedback path to adjust the voltage even lower.

Dan

[dleroi]

Can we see the inside?