Electronic Loads

Thread Starter

Wendy

Joined Mar 24, 2008
23,415
I've had several occasions where this piece of test equipment would come in handy. I was think of using this schematic, but it has several issues that I can see.



One of them is a prepackaged Darlington has an unwanted resistor B-E. So I've come up with another idea.



It isn't pretty, but I think it will do the job. I used a Sziklai Pair to reduce the BE voltage drop, and the total transistor will have really high gain (somewhere in the neighborhood of 1e6). The total transistor will probably drop around 1.2V C-E, but I think I can live with that.

It should vary between 0-10A, and the modulation input will let me input a square wave to make the current switch between two values, which will let me see how fast the regulator under test can correct.

I'm interested in feedback on this concept, though I really don't want to scrap the whole idea and design to someone elses vision on how it should be done.
 
Last edited:

k7elp60

Joined Nov 4, 2008
562
Bill,
I'm not sure if I understand what you are trying to achieve. The attached circuit is a DC load bank that I built about 10 years ago that I use as an adjustable load current load for batteries and powersupplies. The current can be varied from near 0 to a little over 10A. R8 controls the current for the load.
Sorry for the poor schematic, circuit was built about 15 years ago and has been used to discharge batteries at a constant current and test power supplies. Some components like the filter capacitors are higher than needed, but at the time used was was on hand. The DPM monitors load voltage and load current. The same circuit has been tested with a IRFZ44 n channel mosfet with excellent results. The npn darlington is mounted on a big heat sink.
 

Attachments

Last edited:

Thread Starter

Wendy

Joined Mar 24, 2008
23,415
Basically what the title says, an electronic load. They don't have to be complex, and I'm planning on external metering.

I did notice one brain fart though, the range on the resistor needs to be 1V, not 10. I'll be back with a revised schematic.
 

Ron H

Joined Apr 14, 2005
7,063
I would probably build a current sink using a MOSFET as the controlled load, with a sense resistor and an op amp in a feedback loop, similar to this one. The advantage of using a MOSFET is that the dropout voltage is Iload *(Rds+Rsense), which can be a lot lower than you can get with BJTs.
 

Bychon

Joined Mar 12, 2010
469
I have an article I cut out of a popular US magazine which elaborates on the circuit in post 4, including a digital panel meter and all the part numbers. What it amounts to is an adjustable current dump. Is it OK to scan it and post it here? Do you care to see it?
 

SgtWookie

Joined Jul 17, 2007
22,230
Bill,
I simulated it, and it does seem to work - at least the static portion. However, the input base current will introduce an error that you just wouldn't have if you used a MOSFET instead.

I didn't bother to simulate the dynamic portion (using a square wave).

You'd be so much better off using a MOSFET and an opamp for controlling the loop. I know you didn't want to hear that, but it's my best suggestion.
 

Thread Starter

Wendy

Joined Mar 24, 2008
23,415
Know of any TO3 case style MOSFETs? Heatsinking is part of my consideration. I plan on being able to overdrive this design and using simple manual feedback (such as the fans) and common sense while using it. TO3 case style is probably the best for basic heatsinks.

The base current would be very low, less than 0.001% of the collector current. My best resistor is 1% with 4 leads, but it is only 5W, so I also have a 3% that is 25W. I plan on using external volt meters where ever needed.

My current thought (currently unfinished) looks something like this...



I'm also going to change the current sense design to add switches and add 1Ω and 10Ω in addition to the 0.1Ω to allow for ranges.

I'm not against using op amps, but it would have to be fast (TLO72?) and also have to be able to drive a MOSFET totally off. One way to do this is a +9VDC / -3VDC power supply. This would probably mean I would have to revive the separate 6VDC batteries for the fans.

Would -3VDC on a gate damage a MOSFET? I don't think it will, but I don't have the experience with them to judge.
 

Attachments

Last edited:

jpanhalt

Joined Jan 18, 2008
11,087
I guess not. Good night.
I'd like to see it. It is related to several of the posts in this thread and would be of interest for general discussion. Such a device might be a lot better than the 50W resistor in a bucket of water I used for my last project.

As for posting a scanned image from the article, not the whole article, that is done all the time and seems to be allowed under the fair use doctrine. Be sure to include a proper citation to the source.

John
 

SgtWookie

Joined Jul 17, 2007
22,230
Know of any TO3 case style MOSFETs?
They used to be pretty common in TO-3, but they've gotten to be so efficient that you can seldom find them anymore. Electronic Goldmine happens to have one or two in stock. Digikey also has some in stock, with a 300W rating.

However, they also have some in what's called a "TO-3P" package; it's like a large TO-220, and some of those are rated for in excess of 500W.

But unless you're using forced water cooling, it'll be hard to get rid of that much heat from a small area. Such a heat sink wouldn't be difficult or expensive to make.

Heatsinking is part of my consideration. I plan on being able to overdrive this design and using simple manual feedback (such as the fans) and common sense while using it. TO3 case style is probably the best for basic heat sinks.
That depends. Junction to case is pretty good with the TO-3. However, IXFY's (or whatever their name is) seems to have some interesting specs with their TO-3P.

The base current would be very low, less than 0.001% of the collector current. My best resistor is 1% with 4 leads, but it is only 5W, so I also have a 3% that is 25W. I plan on using external volt meters where ever needed.
I bought a 1% 1 Ohm 50W resistor a number of years ago. Don't know why I never thought to put four leads on it, but that's a heck of a good idea.

My current thought (currently unfinished) looks something like this...



I'm also going to change the current sense design to add switches and add 1Ω and 10Ω in addition to the 0.1Ω to allow for ranges.
How about adding a comparator circuit to fold it back rather than blowing the fuse?

I'm not against using op amps, but it would have to be fast (TLO72?) and also have to be able to drive a MOSFET totally off. One way to do this is a +9VDC / -3VDC power supply. This would probably mean I would have to revive the separate 6VDC batteries for the fans.
Well, you'll be dissipating a good amount of heat, so you will need more air flow than a fan powered from a 9v battery will give you. Why not just use a salvaged ATX supply for the thing? It'll have all of the voltages you could want. Plenty efficient, too.

Would -3VDC on a gate damage a MOSFET? I don't think it will, but I don't have the experience with them to judge.
They generally have a range of Vgs=+/-20v. Some are less. You can always clamp the gate to the source; one diode'll do it.
 
Last edited:

Thread Starter

Wendy

Joined Mar 24, 2008
23,415
The fuse in this case, along with the diode, is if I do something stupid and connect the power supply backwards. The diode won't need heatsinked, it just needs to take the surge long enough to blow the fuse. I've got a Schottky dual 16A unit that ought to work nicely. I went with fast blow types.

Basically I want this to be brute simple. I have 99% of the parts already, though I'm not sure I can fit this in the case I have.

I hadn't thought about using a MOSFET in a linear mode. Usually this is something to avoid.
 

Thread Starter

Wendy

Joined Mar 24, 2008
23,415
OK, I've come up with an arrangement for the current measurement. I added some test selects for fudge factor measurements. The exact resistances don't make any difference to the circuit, only to the measurement of the current.



I may go with the MOSFET version shown, and parallel around 4 MOSFETs to spread the power dissipation around. Since I think the MOSFET has to go to some really low values I don't see any way around using a +12V power supply (allowing the op amp to reach +10VDC on the output).
 

Attachments

Thread Starter

Wendy

Joined Mar 24, 2008
23,415
OK, (hopefully) final attempt. I tried to design using the other concepts, but this one is so simple my tag line came into play. I really need this chunk of test equipment.



Any LED whose Vf is 1V greater than the voltage drop of CR1 and CR2 will work. Red LEDs are a little iffy, but some will work. With blue or white there is no question.

The circuit is rather passive. The LED is the only powered part (excluding the fans), if it is bright the circuit is probably working well.
 

Attachments

Last edited:

Ron H

Joined Apr 14, 2005
7,063
Looks good, but keep in mind that 2N3055 minimum beta is 5 at Ic=10A. According to the beta graph, typical beta is about 12-15. Your 2N2907 is gonna run out of breath and get really hot at that current level. You might need parallel 2N3055s.
Of course, I have no idea if you really plan to run the load at the full 10A.
 

Ron H

Joined Apr 14, 2005
7,063
Bill, I don't know if there would be any advantage, but you could connect the emitter of Q1 to the emitter of Q3, instead of to the base. The drop between the top of the sense resistor and the wiper of the pot is then only the Vbe of Q1.
You would then need only one diode in series with the pot.
 

Thread Starter

Wendy

Joined Mar 24, 2008
23,415
Don't think I've see that configuration before. I'll keep it in mind.

I'll look for a power transistor for Q2, the point about heating is a good one. Looks like I'm going to need a bigger case, toggle switches rated for 10A aren't too small either. If this gets as hot as I suspect worst case I may need more fans too.

Thanks for the input.
 
Last edited:
Top