1000watt (24v 40A) powersupply help / General Science and Electronics / Forums

Forums

4hv.org :: Forums :: General Science and Electronics

« Previous topic | Next topic »

1000watt (24v 40A) powersupply help

previous 1 2 3 next

Move Thread

LAN_403

El_Roberto

Fri Nov 28 2008, 07:44PM

Registered Member #1774 Joined: Wed Oct 22 2008, 02:51AM
Location:
Posts: 135

... Im an idiot, since I can get those for about $3 each couldnt I just get like 15 of them and when I need more power I could connect them in parralell? because theres 15 they should be able to handle the 40amps (Im thinking 15 for overkill current handling ability so I wont destroy it). That way I could use 15 different circuits at once as well at differnt voltages. Also I could only get say ten then use a transistor on each output. Someone please correct me if im wrong but it seems like a good idea to me.

EDIT: Actually thinking about it, unless the voltages for all the regulators are exactly the same then the one with higher voltage would draw higher current and would trip its current limiting causing the others to receive higher current witch could trip some of theirs which would end up stopping the whole chain! Would it work if i used the same voltage selecting variable resistor for all ten? or would I still have the same problem because of the small differences in the voltage regulator chips?

Proud Mary

Fri Nov 28 2008, 09:36PM

Registered Member #543 Joined: Tue Feb 20 2007, 04:26PM
Location: UK
Posts: 4992

Roberto, it is not a simple matter to connect three-terminal regulators in parallel as these regulators tend not to share current equally.

However, it is possible at the expense of additional complication, see:
Add Kelvin Sensing and Parallel Capability to 3-Terminal Regulators

It is more usual to use LM317T or LM338 together with one or more (generally inexpensive) pass power transistors.

You can't have too much heavy-duty heat sinking with these circuits, which trade inefficiency for simplicity and economy of construction.

El_Roberto

Sat Nov 29 2008, 02:53AM

Registered Member #1774 Joined: Wed Oct 22 2008, 02:51AM
Location:
Posts: 135

So basically Im back to the problem of there being too much heat generated! Which way of the two you mentioned do you think would work best? I found a circuit online where a seperate smaller transformer was used to power the LM317T and then the large beefy transformer is used to power the 2n3055 transistors, would this work better? I allready have a couple of 2N3055's and about 5 or 6 LM317T's. I have quite a few 120mm computer fans I could use to cool it too. Else I might just give up on making it into one high current powersupply and make it into a few seperate 5-10Amp or so powersupplys in one unit using the simple lm317 and 2n3055 method

Mattski

Sat Nov 29 2008, 04:25AM

Registered Member #1792 Joined: Fri Oct 31 2008, 08:12PM
Location: University of California
Posts: 527

In the application notes section of the datasheet for the part it will typically include a circuit demonstrating the pass through concept with a power resistor. BJTs aren't terrible at sharing current if their bases are all the same voltage, but I'm not sure exactly how good or bad they are. If you have a bunch you might be able to manually pick a well-matched set, and you can also use feedback in the form of an emitter resister to aid in current sharing.

Some quick googling found this reference on BJT paralleling.

Also, do you have a plan to use that much current at the moment? You could always design it for a smaller power level now and upgrade it later when you need to.

Proud Mary

Sat Nov 29 2008, 07:34AM

Registered Member #543 Joined: Tue Feb 20 2007, 04:26PM
Location: UK
Posts: 4992

Rob, if we say for the sake of discussion that your pass transistor has a gain of 10, then a single LM317T (that can output 1.5A with a good heat sink) will be able to control up to 15A.

Don't forget that for a bridge rectifier with a capacitor input filter, AC current is (empirically) about 1.8 times the DC current, so your 35A transformer will not be able to produce more than 20A DC without even considering all the other circuit losses.

Now that you are trying to regulate no more than 20A, your problems have become more manageable!

Good heat sinking can often be achieved by bolting components to the metal case, using mica (or occasionally ceramic) spacing washers together with a thin smear of heat sink grease where the part needs to be electrically isolated. Take care that it REALLY IS insulated from the case before switching on power for the first time! Mica spacers can be fiddly to use!

El_Roberto

Sat Nov 29 2008, 10:29AM

Registered Member #1774 Joined: Wed Oct 22 2008, 02:51AM
Location:
Posts: 135

Well I have most of that stuff allready as my dad used to be really into electroncs, if I have time tommorow I will make a scale version using the parts I allready have (should be good for 15amps or so with what I have (I have two transistors), I will get 5 or so more just for overkill sake to make sure it can handle the heat, and they are really cheap). I also found a fairly large heatsink that (conveniently) has two predrilled area's to fit my 2N3055's. I will use a fan to help as well. My load will be a pile of (25watt, I hope, they look fairly heavy duty and have 25 written on them...) 2.2ohm resistors. Ill post back sometime tommorow when Im done.

Edit: Thats all assuming I can find an electronics store open on a sunday to get a bridge rectifier...(I just broke one of the pins off mine and I cant solder whats left)

Proud Mary

Sat Nov 29 2008, 02:33PM

Registered Member #543 Joined: Tue Feb 20 2007, 04:26PM
Location: UK
Posts: 4992

Good luck with that, lad! Tell us how it goes.

rp181

Sat Nov 29 2008, 04:41PM

Registered Member #1062 Joined: Tue Oct 16 2007, 02:01AM
Location:
Posts: 1529

If the pins broke, try sanding the stub. That allowed me to solder that.

El_Roberto

Sun Nov 30 2008, 03:22AM

Registered Member #1774 Joined: Wed Oct 22 2008, 02:51AM
Location:
Posts: 135

ITS ALIVE!!!! I brought a 50Amp bridge rectifier and put it all together, the rectified voltage is 32v but when using the regulator I can only go up to about 26v, I guess I need to change my variable resistor. With a 2.2ohm load at 26v It draws about 11Amps and the only thing that heats up is my resistors (only rated at 50watt). Once ive worked out the voltage problem Ill order some more parts and fix it all up (Its currently all over my floor). I definately need more transistors because at 12v 1amp it heats up quite fast. What do you guys think of 4 or 5 of these;

Proud Mary

Sun Nov 30 2008, 07:37AM

Registered Member #543 Joined: Tue Feb 20 2007, 04:26PM
Location: UK
Posts: 4992

Attach a small load, and measure the voltage drop across your pass transistor for various settings of the regulator resistor, Robert.

previous 1 2 3 next

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob