who here wants a free 'intelligent' constant current load ??

Thread Starter

chimera

Joined Oct 21, 2010
122
Hello everyone!!

I had been designing a power supply (switching/linear) and needed to test its load regulation. I thought of using power resistors..but eh too big..not really elegant.

So I came up with a solution: its a variable constant current 'dummy' load. I designed a quick pcb for it and now I have several copies of it.

IF ANYONE WHO HAS/WANTS TO BUILD A POWER SUPPLY SHOULD HAVE THIS IN THE TOOL BOX.

If interested, pm me using my username here.

ADMINS: Give this thread a day before moving it to the FLEA MARKET sub forum. Not everyone goes there and I feel its important that ppl should know that they have an alternative other than power resistors for dummy loads.

*****Just trying to give back to the community*****
 

wayneh

Joined Sep 9, 2010
16,400
When you say "constant" current, what does that mean in the context of an AC signal, say 50-1000Hz? We know how a resistive load behaves, and of course it's not constant current. It's a constant average current. Does your circuit behave like a sort of variable resistor, self adjusting to give a constant average current while ignoring short term pulsing?
 

Thread Starter

chimera

Joined Oct 21, 2010
122
When you say "constant" current, what does that mean in the context of an AC signal, say 50-1000Hz? We know how a resistive load behaves, and of course it's not constant current. It's a constant average current. Does your circuit behave like a sort of variable resistor, self adjusting to give a constant average current while ignoring short term pulsing?
the circuit uses an op amp to set the voltage at the base of a N-channel Mosfet. The mosfet is sourcing a 1 ohm (10X 10ohm 1 watt resistors ) load. The output of the op amp feeds the gate of the mosfet. The mosfet is powered by what ever converter you want to test (like is load regulation). There is a voltage reference present in the cirucit for the op amp. Ofcourse, there will the need for a heat sink.

ENJOY
 

wayneh

Joined Sep 9, 2010
16,400
the circuit uses an op amp to set the voltage at the base of a N-channel Mosfet. The mosfet is sourcing a 1 ohm (10X 10ohm 1 watt resistors ) load. The output of the op amp feeds the gate of the mosfet. The mosfet is powered by what ever converter you want to test (like is load regulation). There is a voltage reference present in the cirucit for the op amp.
That's a traditional and standard approach. How is that "nothing like" what was in that thread?

I've been thinking about this topic for a while. See this thread, for instance.
 

wayneh

Joined Sep 9, 2010
16,400
I see you've offered your prototype for sale on E-bay. Caveat emptor.

If you won't answer questions about your circuit here, why would anyone here want to buy it?

For example, you've said your circuit is powered by the device under test. What if the power supply is 3.3v? Unless you've used the right op-amp and an unusual MOSFET, it won't work very well. A normal MOSFET won't turn on at that low a gate voltage.

On the other end of things, your ad claims 2.5A at 20v, or 50W of heat to be dissipated in those 10 little resistors. I could be wrong, but they don't look like 5W resistors. Even if they are, driving them to their maximum power without heat sinking will turn them into fuses and even risk a fire.
 

Thread Starter

chimera

Joined Oct 21, 2010
122
When I read something about an op-amp and ten resistors I immediately thought of EEVBlog and Dave Jones....

http://www.eevblog.com/2010/08/01/e...my-load-for-power-supply-and-battery-testing/
the 10 ohm approach for a resulting value of 1ohm for voltage detection, current interpolation, ADC, DACs is a standard approach. As a matter of fact, I remember i came across this approach a long time ago during one my labs at my uni during undergrad

Dave is a fantastic asset to engineers everywhere! As a matter of fact, I often ask him advice on his forum for some work related designs
 
Last edited:

Thread Starter

chimera

Joined Oct 21, 2010
122
I see you've offered your prototype for sale on E-bay. Caveat emptor.

If you won't answer questions about your circuit here, why would anyone here want to buy it?

For example, you've said your circuit is powered by the device under test. What if the power supply is 3.3v? Unless you've used the right op-amp and an unusual MOSFET, it won't work very well. A normal MOSFET won't turn on at that low a gate voltage.

On the other end of things, your ad claims 2.5A at 20v, or 50W of heat to be dissipated in those 10 little resistors. I could be wrong, but they don't look like 5W resistors. Even if they are, driving them to their maximum power without heat sinking will turn them into fuses and even risk a fire.
Sorry for the late reply.. i had been busy at work.. I tell u ..i love designing stuff but i cant stand corporate bulshit!

Anyways..

Yup, but if you remember ece 101 courses, 10 ohms times 10 in parallel..gives u 1 ohm and ... the current (let assume 2.5Amps, which is whats its rated at) gets divided and equally (approx) flows thru the resistors. So each resistor will see a max of 0.625 watts at full power (assuming u mount a big enough heat sink on the transistor). The resistors I have used have a power rating of 1 watt and like i mentioned in my posting on ebay.. every component is from known manufacturers...so the resistors used are by Vishay Dale.. top notch stuff

Also, 20V is the absolute maximum input for the variable load. For instance, you can provide 20V from a power supply that you want to test. U can then change the current drawn by varying a voltage divider. This will help you determine the load regulation, drop out voltage of ur power supply under test or the DC to DC converter you have designed.

KEY TO THIS DESIGN: a precise voltage reference (provided you thermal shield it sufficiently using good pcb layout design)

In the case if it is a 3.3V power supply, which you want to test, you will only be able to go up to 350mA of current. I tested it before typing it out (and I tested in several pieces..not just one..this ensures good deviation). But then again, usually power supplies or if you are designing a DC DC converter, tend to be 5V and upwards..3.3V is usually in the linear regulator territory and it usually wont be included in high power applications..

I have to say.. U NEED A BIGGER, much BIGGER HEAT SINK for 20V. The one provided with it will cause the junction temp to go over 110C. I have tested this in the lab and by calculations.

About the mosfet, its a low gate voltage mosfet..RFP12N10 by FairChild.. so I knew I can use it :p

BTW.. i remember i posted here some time ago that if any one wanted this for free... they should have let me know.. and I am still very much up for it..I like helping peeps of AAC and EEV forums. No one asked.

If still interested, PM me.

That's a traditional and standard approach. How is that "nothing like" what was in that thread?

I've been thinking about this topic for a while. See this thread, for instance.
Yes you are right. It is quite a similar design. I had not seen it as I did not search for one. I ended up simply creating one and had the intention of having it available for fellow engineers and hobbyists to conveniently have at hand.

Thank you for providing that link.
 
Last edited:
Top