Constant current load for testing power supplies (guidelines)

[budm]

Re: Constant current load for testing power supplies (guidelines)

Sch PDF, provided by Analog Devices.

[gabiz_ro]

Re: Constant current load for testing power supplies (guidelines)

budm
This was interesting but what it lacks is what mosfet or transistors was used
Do you have any clue what was used

http://www.linear.com/solutions/1845

You ask about that?

Is on photo of schematic
Q1 Q2 2N3904
Q3 Q4 FQA140N10

[eccerr0r]

Re: Constant current load for testing power supplies (guidelines)

Selfish answer: no, because I don't see any real use for it on my bench... The way I see it, a PSU needs to be able to deal with constantly varying loads, sometimes even in the kilohertz range. Since I don't need to accurately measure the limits of a PSU - just loading it down with light bulbs is sufficient to get a rough estimate of its capabilities.

BTW, that linear.com sample idea for design, I would assume that:

1. They used 4x470mΩ because you can now get 4x the wattage compared to one individual resistor of the size. Likely the smaller resistors are cheaper.
2. Those transistors are probably not critical. Probably something like IRF520 will do.

[Dannyx]

Re: Constant current load for testing power supplies (guidelines)

Has anyone else been messing around with this project ?

[Dannyx]

Re: Constant current load for testing power supplies (guidelines)

Also, stupid question, but why so many resistors in series there ? 470mOhms times 4

[sam_sam_sam]

Re: Constant current load for testing power supplies (guidelines)

budm
This was interesting but what it lacks is what mosfet or transistors was used
Do you have any clue what was used

[budm]

Re: Constant current load for testing power supplies (guidelines)

http://www.linear.com/solutions/1845

[Dannyx]

Re: Constant current load for testing power supplies (guidelines)

Yeah that would be interesting. That was the original point of this post: for people to come up with their own ideas and designs revolving around the main topic, since I'm sure there are various ways of achieving it, so perhaps in the end we'll combine all those ideas and end up with something that's both functional and revolutionary

[Th3_uN1Qu3]

Re: Constant current load for testing power supplies (guidelines)

I have a MOSFET load schematic i designed a few years back but never got around to building, i can share if you wish. It has only been simulated but it is such a basic circuit that it is guaranteed to work. It's based on the standard opamp + voltage reference + current shunt topology.

[Dannyx]

Re: Constant current load for testing power supplies (guidelines)

I added the load this time. The resistor divider you mentioned should go across RLoad then ?

[Dannyx]

Re: Constant current load for testing power supplies (guidelines)

I shall attempt a re-draw to see what turns out

[momaka]

Re: Constant current load for testing power supplies (guidelines)

I drew a schematic but stopped at the following part

So from what I understand, the pot's wiper goes to the NI input (what you call the op-amp's reference), but what about the Inverting input ? I assume that goes to the anode of the zener diode (hence the shunt's "output"). Nifty

Still not quite there yet.

Inverting input goes to a resistor divider network connected across the load. The resistor divider network basically "measures" the voltage across the load and the output of that network goes to the inverting op-amp pin for feedback. Meanwhile, the Zener-pot circuit sets the "reference" voltage on the non-inverting input of the op-amp. Essentially, the op-amp compares the "feedbacK" from the resistor divider network to the "reference" voltage. If the feedback voltage is low, the op-amp turns On the transistor to increase the voltage going to the load. And if the feedback voltage is high, the op-amp turns Off the transistor to reduce the voltage.

Also, your circuit is missing the load. R1 should NOT be your load. R1 is what we will use later for implementing the CC circuit. The load will be located between R1 and your transistor's Emitter/Source (depending on if you use MOSFET or BJT... but if you're going to use an op-amp to drive that circuit, you're better off using a MOSFET, unless you can find a good Darlington BJT).

I am still watching this post to see what you come with I would like like to build something like this as well

I'll see if I can build a circuit like that in the next few weeks, given that we might have some cold weather and I might not be able to do much outside.

[sam_sam_sam]

Re: Constant current load for testing power supplies (guidelines)

I am still watching this post to see what you come with I would like like to build something like this as well

[Dannyx]

Re: Constant current load for testing power supplies (guidelines)

I drew a schematic but stopped at the following part

Then use a pot connected between the Zener's anode and cathode and pot output connected to the op-amp reference input so that you can adjust the voltage to your DUT.

So from what I understand, the pot's wiper goes to the NI input (what you call the op-amp's reference), but what about the Inverting input ? I assume that goes to the anode of the zener diode (hence the shunt's "output"). Nifty

[momaka]

Re: Constant current load for testing power supplies (guidelines)

I originally conceived the reference by putting a pot between GND and VCC (let's say 5v for argument's sake). It would also work with a zener diode for a fixed reference, but that's beside the point (since I'm trying to put what you suggested into perspective). The pot method works, but I have to "manually" adjust it if the resistance changes, because that reference would stay the same regardless of the load, so what you're suggesting is using the feedback from the shunt as the ref ?

Use a Zener to get a stable regulated voltage. Connect the Zener's anode to the upper end of your shunt resistor instead of ground (that way, your Zener's reference voltage will vary with voltage across the shunt due to load.

Then use a pot connected between the Zener's anode and cathode and pot output connected to the op-amp reference input so that you can adjust the voltage to your DUT.

[Dannyx]

Re: Constant current load for testing power supplies (guidelines)

That's because you are probably using ground as the reference point for your CV. You must use the high-side of the current shunt resistor of your CC as the ground reference for your CV, if you want the voltage across the DUT to be correct.

I originally conceived the reference by putting a pot between GND and VCC (let's say 5v for argument's sake). It would also work with a zener diode for a fixed reference, but that's beside the point (since I'm trying to put what you suggested into perspective). The pot method works, but I have to "manually" adjust it if the resistance changes, because that reference would stay the same regardless of the load, so what you're suggesting is using the feedback from the shunt as the ref ? Sorry if I didn't get it - I'm a bit rusty right now having left electronics projects behind for a brief period

[momaka]

Re: Constant current load for testing power supplies (guidelines)

It all makes sense in my head but building the thing is a whole different story I know how to build a CV supply (op-amp and feedback loop) at its most basic, but its voltage would have to be adjusted up/down based on the current draw. If the resistance drifts up/down (which is what gave me the idea of using the lighbulb as a load, albeit a variable one), the supply would also need to compensate with more/less voltage from the supply under test via the pass element...that's how I see it at least.

That's because you are probably using ground as the reference point for your CV. You must use the high-side of the current shunt resistor of your CC as the ground reference for your CV, if you want the voltage across the DUT to be correct.

[Dannyx]

Re: Constant current load for testing power supplies (guidelines)

It all makes sense in my head but building the thing is a whole different story I know how to build a CV supply (op-amp and feedback loop) at its most basic, but its voltage would have to be adjusted up/down based on the current draw. If the resistance drifts up/down (which is what gave me the idea of using the lighbulb as a load, albeit a variable one), the supply would also need to compensate with more/less voltage from the supply under test via the pass element...that's how I see it at least.

[momaka]

Re: Constant current load for testing power supplies (guidelines)

A CV AND CC supply IMO....at the end of the day, isn't a CV supply something thar maintains a constant voltage across a load regardless of the load itself ? I'd need to couple that with a CC section so the current doesn't drift down as resistance goes up....pretty complex

Well, you will be controlling the current electronically by measuring the voltage across the (known) resistance of your current shunt (i.e. the Nichrome wire) and adjusting the voltage accordingly with an active device (BJT, MOSFET, etc). Meanwhile, the CV section can be done by a simple linear regulator with a pass-transistor to allow you to draw heavier currents - at least that's how I would build the thing (I've actually been thinking about it a lot lately).

I assume it heavily depends on the length and cross-section and stuff like that. I've got a few hundred meters of nichrome on my hands to play around with

Yes, exactly.
More heat in less surface area = hotter Nichrome wire. So you will likely want a few long parallel strands to keep the surface area high and the resistance low. Think of it as building a low-resistance, high-power resistor out of many high-resistance, low-power resistors.

[Dannyx]

Re: Constant current load for testing power supplies (guidelines)

I recall some type of nichrome changed around 1.5% or so every hundred degrees Celsius ...

I assume it heavily depends on the length and cross-section and stuff like that. I've got a few hundred meters of nichrome on my hands to play around with