Simulating load on LED drivers.

[budm]

Re: Simulating load on LED drivers.

How would you connect the circuit provide by TOM to the LED driver board?
The Cathode return of the LED is not connected directly to GND in the LED power supply board, it is connected to the PWM control and current sensing resistor to GND.

[lookimback]

Re: Simulating load on LED drivers.

I thought I'd be able to float the ground so the ground points in the schematic would be the cathode return. But, I'm guessing that I would have a problem with the logic ground not being at the same potential. But, if I measured the resistance from the cathode return to ground, and set the logic ground to the same potential, would that solve it?

[budm]

Re: Simulating load on LED drivers.

The way the LED driver circuits is set up is to have the PWM Controlled and sensing resistor on the sink side of the LED, Booted Voltage is fed directly to the Anode of the LED string.

[tom66]

Re: Simulating load on LED drivers.

That is why I suggested the circuit should be powered from a battery.
The circuit would respond quick enough to 100Hz PWM, possibly the feed forward cap around the op-amp would need tweaking but it should work ok as-is.

[lookimback]

Re: Simulating load on LED drivers.

So, how would I connect that circuit to the LED driver output?

[tom66]

Re: Simulating load on LED drivers.

Between the + and - of the LED driver output, where the LED string would normally go.

[lookimback]

Re: Simulating load on LED drivers.

So, the battery current output to + and ground to - ?

[tom66]

Re: Simulating load on LED drivers.

Across the MOSFET between the drain and other end of the resistor.
You will want to modify that circuit of course.
Why don't you test it on the bench first. You should draw a schematic first.

[lookimback]

Re: Simulating load on LED drivers.

You will want to modify that circuit of course.
Why don't you test it on the bench first. You should draw a schematic first.

Yes, I'll have to figure out how much power this could potentially see and choose MOSFETs accordingly. Also a really big heat sink. When I actually get time to start this, which hopefully will be next month, I'll draw up a schematic post it. Definitely will bench test it. Another question: I'm assuming ISET is where I'd put a potentiometer to set it. Would the other side of that pot go to the other side of resistors R8/R9(between them and R10)?

[tom66]

Re: Simulating load on LED drivers.

Yes, I'll have to figure out how much power this could potentially see and choose MOSFETs accordingly. Also a really big heat sink. When I actually get time to start this, which hopefully will be next month, I'll draw up a schematic post it. Definitely will bench test it. Another question: I'm assuming ISET is where I'd put a potentiometer to set it. Would the other side of that pot go to the other side of resistors R8/R9(between them and R10)?

ISET is a voltage input. So the pot would go across a voltage reference ideally. If you just use the battery as the reference then you may find the current changes as the battery slowly runs flat. One possibility is you could use a 5V regulator or similar as a reference source. The other thing you will need to do is make sure the current scale is set correctly. That circuit is designed for a fairly high discharge current of up to 1.5 amps. You probably only want a fifth of that, so you could look at changing some of the resistors. Finally, the circuit itself can actually be simplified by removing one of the op-amps, just have a Google for current sink circuits, it will be a little less precise, but it will work fine for your application.

[lookimback]

Re: Simulating load on LED drivers.

ISET is a voltage input. So the pot would go across a voltage reference ideally. If you just use the battery as the reference then you may find the current changes as the battery slowly runs flat. One possibility is you could use a 5V regulator or similar as a reference source. The other thing you will need to do is make sure the current scale is set correctly. That circuit is designed for a fairly high discharge current of up to 1.5 amps. You probably only want a fifth of that, so you could look at changing some of the resistors. Finally, the circuit itself can actually be simplified by removing one of the op-amps, just have a Google for current sink circuits, it will be a little less precise, but it will work fine for your application.

I'll add an LDO for the reference. I had saw a simple one with one op-amp before you posted this one. Would I still need the parallel MOSFETs if I'm sinking that much less current?

[tom66]

Re: Simulating load on LED drivers.

The parallel MOSFETs are used in this application to dissipate as much power as possible in a small area.

You can use one big heatsunk MOSFET or hundreds of SMD MOSFETs, it depends on what you want to do, but soldering hundreds of SMD MOSFETs doesn't sound fun to me .

[lookimback]

Re: Simulating load on LED drivers.

The parallel MOSFETs are used in this application to dissipate as much power as possible in a small area.

You can use one big heatsunk MOSFET or hundreds of SMD MOSFETs, it depends on what you want to do, but soldering hundreds of SMD MOSFETs doesn't sound fun to me .

Be great if I needed practice, but I think I've mastered the technique already and want to have time for other projects and family as well. I think I'll go with the one big MOSFET.