Constant current dummy load - home etched, single sided Moderators: adafruit_support_bill, adafruit

[george graves]

I threw this together. Single sided, home etched. Based on Dave's plans:

http://www.eevblog.com/2010/08/01/eevbl ... y-testing/

I used ExpressPCB software, just cause it's so much quicker to quickly whip up a board for home etching(eagle can be a pain at times - footprints and all). Then I usually print it off as a PDF, and tweak the pads in illustrator for home etching (very much needed to make bigger pads, and such for the toner transfer method)

http://dl.dropbox.com/u/2295566/BLOG/dummy%20load%20v1/dummy%20load.pdf

http://dl.dropbox.com/u/2295566/BLOG/dummy%20load%20v1/ExpressSCH.pdf

The only change I really made was running the opamp at 8 volts (7808) for a bit more range. It seems happy at 1 amp. Two amps and things start to really heat up. ;D I'm planning on making a second version down the road. Something arduino controlled with DACs, ADC, multiple mosfets, maybe a fan and a LCD. But this will do for now. All I really need is 500ma - 1000ma.

[john444]

Very nice!

The photos look as if you did not wire the current-sense resistors in a 4-terminal connection.
(I am at work so I could not open the links)
You can avoid several potential problems with 2-terminal resistors.

CC loads can be really handy.

What did you do to have such clean boards?
I have not been very satisfied with the results when I used the toner transfer method.

[adafruit_support_mike]

That's some quality geek banned right there..

Good boards, good build, good photography. It's a thing of functional beauty.

[johngineer]

Very nice, George! I like your R_sense resistor bank -- reminds me of one of those vertical parking deck elevators! :)

[sklarm]

Nice job.

Constant current power supplies are helpful devices. I released a open source hardware kit last month called Da Pimp which works off the same principle. It is used for battery testing, charging and desulfation. It would also be good for power long strings of ultra bright LEDs. I'm surprised there are not more projects like these. Mine puts out a constant .5A when given 120VAC (US) or 1A when 220VAC is used. This is a capacitive charger design. Safety third.

http://screwdecaf.cx/dapimp.html

Da Pimp with Case

dapimpwithcase.jpg (46.06 KiB) Viewed 18342 times

[freaklabs]

Hi George.
The project looks excellent. I am now inspired to build a dummy load too for an AC energy monitoring project I'm doing. Thanks for the post!
Akiba

[skylen]

Nice work. BTW, did you leave the extra op amps floating or did you hook them up as followers?

[IsotopeJ]

Hi George,

You may have already figured this out, but the limiting factor in this circuit isn't the opamp supply voltage. The range is more dependent on the sense resistor. For instance, with a 1 ohm 5 Watt resistor, you're going to dissipate I^2 watts. You'll exceed 5 Watts at 2.24 Amps. Any higher than that you'll burn up your resistors.

Now, before I move on, one consideration is that the OpAmp must be able to supply enough bias voltage to the mosfet base to reach saturation. An amp like the LM324 can go from 0 to Vsupply - 1.5 volts, so you need to look at the datasheet for your mosfet to get the saturation voltage and make sure to drive your opamp at Vsaturation + 1.5 minimum.

For the sake of argument, let's say you're still using a 1 ohm resistor and expect up to 6.5 volts across it, giving you 6.5 amps. At that current your sense resistor will need to dissipate over 42 watts. Yikes. The way to reduce this is to use a smaller value for your sense resistor. In fact, I built this circuit using 40 half-watt 10 ohm resistors in parallel for 0.25 ohms. This way I am able to sink currents up to 8.9 amps. I also added a toggle switch that bypasses 30 of these resistors to leave me with a 1 ohm sense resistor. This way I can switch between a "high" range (about 8ma to 8.9 Amps) and a "low" range of about 2ma to 2.2 Amps.

[bearmos]

Constant current can be extremely useful! I needed one that was fairly beefy for doing some electroplating. I went the TO220 route for the resistor and FET and added some forced air cooling. Components are slightly warm to the touch running at 8 Amps.

Here's a shot of the board

constant Current board

constCurrentCCA.jpg (93.09 KiB) Viewed 13896 times

...with the ridiculous heat sink (salvaged from a Pentium 2 I think):

constant current with heat sink

constCurrentAssembly.jpg (84.24 KiB) Viewed 13896 times

The page below discusses some of the tradeoff's listed in the posts above for component selection (super simple schematic is there as well):
http://twilightrobotics.com/prototyping/constantcurrentsource

[simonclark]

I threw this together. Single sided, home etched. Based on Dave's plans:

http://www.eevblog.com/2010/08/01/eevblog-102-diy-constant-current-dummy-load-for-power-supply-and-battery-testing/

I used ExpressPCB software, just cause it's so much quicker to quickly whip up a board for home etching(eagle can be a pain at times - footprints and all). Then I usually print it off as a PDF, and tweak the pads in illustrator for home etching (very much needed to make bigger pads, and such for the toner transfer method)

http://dl.dropbox.com/u/2295566/BLOG/dummy%20load%20v1/dummy%20load.pdf

http://dl.dropbox.com/u/2295566/BLOG/dummy%20load%20v1/ExpressSCH.pdf

The only change I really made was running the opamp at 8 volts (7808) for a bit more range. It seems happy at 1 amp. Two amps and things start to really heat up. ;D I'm planning on making a second version down the road. Something arduino controlled with DACs, ADC, multiple mosfets, maybe a fan and a LCD. But this will do for now. All I really need is 500ma - 1000ma.

Nice board that you created . nice photography also .each and every wire visible properly .