PWM LED Driver thoughts

[ajp8868]

Hi all

I'm currently fixing a DJ light where the LED doesn't come on. I've checked the led itself and the voltage is getting to it, however the control circuit isn't working by the looks of it.

It's a constant current type circuit and it uses a VAS1210 to control a MOSFET.

My query is this:

The datasheet for this VAS1210 states either pwm control or voltage up to 1.2v, however when I've measured the adj pin, it appears it's being applied with ~5v when the led is (supposed to be) fully on and it steps down slowly from there when the LED is dimmed via control. I'm fairly certain the datasheet says the adj pin can handle up to 6v, but is 5v too high or could I actually be reading a PWM signal as 5v?

Hope that makes sense, sorry if it doesn't!

[Buriedcode]

I'm not clear on what exactly is going on - when yous say:

I've checked the led itself and the voltage is getting to it, however the control circuit isn't working by the looks of it.

If the LED die - or LED string/array - is getting voltage, and it is sufficient for the LED(s), then they have failed, rather than the driver.

My query is this:

The datasheet for this VAS1210 states either pwm control or voltage up to 1.2v, however when I've measured the adj pin, it appears it's being applied with ~5v when the led is (supposed to be) fully on and it steps down slowly from there when the LED is dimmed via control. I'm fairly certain the datasheet says the adj pin can handle up to 6v, but is 5v too high or could I actually be reading a PWM signal as 5v?

Hope that makes sense, sorry if it doesn't!

If you're measuring with a multimeter, then it is likely you are measuring PWM - multimeters react/update slowly and can average PWM to DC.  Most meters with a frequency counter with duty % which is handy for spotting this.  A few photos of the PCB - and the LED's - so we can see how its put together, or how many LEDs are in a string, what type etc.. woudl help a lot.   

If there is say, two white LEDs in series, and you're measuring ~4V across them, then they won't light, but if its say 4 in series (4*3.4V min = 13.6V) and you measure >14V, but the LED's won't light, then one oor all of the LEDs in that string have failed.  We can't know without more info

[ajp8868]

I don't think the LED has failed, because if I apply voltage to the MOSFET gate then the led turns on. The MOSFET is on the negative side of the LED.

Ah ok I will test it in duty cycle mode to see if it makes more sense. I'll also get some pictures tonight of the board to upload EDIT: Picture attached

The thing is, even with the ADJ pin showing at 5v, the output to drive the MOSFET gate is showing at less than 1v - so I'm assuming this might not be enough to drive the MOSFET? It's model is TPCA8016

Thanks for your help!

[Buriedcode]

I don't think the LED has failed, because if I apply voltage to the MOSFET gate then the led turns on. The MOSFET is on the negative side of the LED.

Ah ok I will test it in duty cycle mode to see if it makes more sense. I'll also get some pictures tonight of the board to upload EDIT: Picture attached

The thing is, even with the ADJ pin showing at 5v, the output to drive the MOSFET gate is showing at less than 1v - so I'm assuming this might not be enough to drive the MOSFET? It's model is TPCA8016

Thanks for your help!

Ahh gotcha, yes, 1V isn't enough for an N channel MOSFET gate, even the low logic level ones, so I think you're right - the controller chip is faulty. 

As a side note - its always nice when manufacturers print the part values on the board - makes replacing any anonymous SMD caps much easier!

[ajp8868]

Ah brill, fingers crossed that's the problem! I'm waiting on the chips arriving so I'll update once replaced.

Yes Chauvet (or at least the manufacturer that makes their lights) are quite good for that. Has helped me many times so far.

Thanks again!

[ajp8868]

It was the led driver, hooray!

The exact same fault has happened in 4 lights. Seems to be quite a bad failure rate.

[Buriedcode]

I wonder if its the MOSFET driver failing, or somehow theres overvoltage on VIN?

For the same fault in multiple units, thats either a bad batch or devices, or - I'm geussing - something about the components they've chosen which make that failure more common.

But, I guess it doesn't really matter, glad you got it working dude 

[ajp8868]

I was thinking the same, though these units have had a hard life in a club ceiling for around 7 years so maybe that chip is prone to overheating.

Hey ho, at least it's an easy fix if it happens again!

[themadhippy]

though these units have had a hard life in a club ceiling for around 7 years so maybe that chip is prone to overheating.

not just heat thats your enemy in such locations,smoke fluid gets everywere bringing dust and grime with it,few years build up of that dont help.At least cracked oil aint as common these days.