hi,


I've got a "DONE" ADS-42W900-LEP constant current LED psu outputting 900mA. I want to lower the current. The PSU does not seem to have a dedicated driver IC. What component should I look for? (update: more pcb circuit tracing, detected a driver ic. https://forum.allaboutcircuits.com/...nt-current-led-psu.196058/page-2#post-1851033)

The board does have a 22kohm resistor on the low V side but I'm not sure if it controls the current value.

Any ideas?

Thanks!

 

The current is likely set by a small value resistor in series with one of the outputs. If you can find that resistor, changing it might work. I would not try to increase the current at all. Perhaps you could decrease it.

 

The current is likely set by a small value resistor in series with one of the outputs. If you can find that resistor, changing it might work. I would not try to increase the current at all. Perhaps you could decrease it.

I only want to reduce the current.

I've traced the output low V side of the pcb, there is no resistor in series. So I reckon the high V side does the current adjustment by PWMing the mosfet? High-side sensing as in this article https://www.ledsmagazine.com/leds-s...-efficiency-without-a-sense-resistor-magazine ?

 

I removed some of the white goo from the pcb uncovering a 6 pin ic, 0H-A26, one of its pins is connected through a 10Kohm resistor to the gate of the svf10n6 mosfet. I could not find a datasheet for the 0H-A26. Not sure how continue. Any ideas?

 

there is no resistor in series.

It might be that arched length of wire at the bottom right corner of the pic. If so, unsoldering one end and adding more wire of a similar gauge in series should decrease the regulated current. That said, I would have expected any current shunt to be nearer to the power MOSFET.

 

PWM does not control current, it controls duty cycle, which controls power.

A constant current supply adjusts its output voltage to control the current. There pretty much has to be a current sense resistor. That loop of wire does look suspiciously like it is being be used as a resistor.

 

The wire by the big cap is a jumper, j1.

The low V side of the pcb has only 3 connections to the high V side. Two wires from transformer (red and black) and a capacitor or some kind of safety element (blue by the mosfet radiator). Would this be enough to send current readings to the high V side?

And what about 0H-A26 , could it be one of those ic's which do high side sensing, therby eleiminiating the need for a sensing resistor on the low V side?

 

Would this be enough to send current readings to the high V side?

The current is likely regulated on the low voltage side.

 

here's a schematic of the low voltage side:

cy4 is the blue component between the high and low V sides

 

If that is correct, the current regulation is done by magic.

 

Simulation in LTspice shows that the Rds(on) of a MOSFET doesn't vary much with temperature and drops a voltage proportional to drain current. Since a transformer primary current varies with secondary load current it would therefore be possible to use that voltage drop to sense the load current, with Rds(on) acting as a sense resistor in place of the usual one on the secondary side. I'm guessing that's what is happening in this LED driver.

 

... it would therefore be possible to use that voltage drop to sense the load current, with Rds(on) acting as a sense resistor in place of the usual one on the secondary side. I'm guessing that's what is happening in this LED driver.

could this psu be a Isolated Flyback SMPS? Like the one on the attached screenshot, using IRS2982S chip

https://www.infineon.com/dgdl/Infineon-IRS2982S-DataSheet-v13_01-EN.pdf

 

The SMPS above uses a current sense resistor connected between the MOSFET source and ground. So not the same as I was envisioning in post #11.

 

Hello there,

Yes a sense resistor from source to ground would work both as an overcurrent and regulation current sense. If your circuit has that, that's the way it works then. Transformers are sometimes sampled in the primary side so it would make sense. That way the circuit could detect a problem too.

If no other way and you do not want to lower the current too much, maybe you can shunt some of the current to ground with a resistor. I am not sure how much voltage are you getting on the output though that could be a big waste of energy.

On one of the units I worked on in the distant past we sensed the current in the collector or drain instead of the emitter or source.

If you can draw the entire schematic we can figure out exactly how this works without knowing what that 6 pin IC does.
I am betting it is either an opto coupler or a special type of switching regulator which works with just one transistor just like this appears to do.

What is the input voltage and what is the output voltage?

 

What is the input voltage and what is the output voltage?

Input 220V, output 30-60V, depends on how many leds are connected (in series)

I'm gonna have to poke around the pcb, it will be pain in the ass 'cause it's all so small and crammed into such a small footprint

---


Perhaps it would just be easier to use a pwm module to regulate the duty cycle of the LED psu output thereby dimming the leds...

 

Perhaps it would just be easier to use a pwm module to regulate the duty cycle of the LED psu output thereby dimming the leds...

That might wreak havoc in its current regulation.

 

I wonder if the MOSFET is one something like this ? It wouldn't need the usual low-Ohm current-sense resistor.

 

Input 220V, output 30-60V, depends on how many leds are connected (in series)

I'm gonna have to poke around the pcb, it will be pain in the ass 'cause it's all so small and crammed into such a small footprint

---


Perhaps it would just be easier to use a pwm module to regulate the duty cycle of the LED psu output thereby dimming the leds...

So how much current do the LEDs draw, did you mention that yet?
I ask because if they are 20ma LEDs it would be easy to shunt some of the current to ground with a resistor, and the resistor would not have to dissipate any more than 2 watts for any voltage level.

You might be able to attach your own constant current circuit to the output. It would probably go to 60v for low current settings, but your external circuit would be limiting the current not the original circuit.

Can you find a new device that will do this for you? Maybe that's easier you don't have to do anything else but invest some cash.

 

So how much current do the LEDs draw, did you mention that yet?

not, sure did not measure that yet

the smd leds are either 1W or 3W, wired in series, 15 pieces

 

Can you find a new device that will do this for you? Maybe that's easier you don't have to do anything else but invest some cash.

well I've got 12 of these so....