Stupid AMC7135's

I have a 105c board here with 8x7135 but it only pulls 2.44A when it should be pulling 2.8A. Is there any way to determine which 7135 is not working? Thanks :slight_smile:

Could be your meter or the battery your using.

  • What battery are you using, & is it fully charged?
  • If you are using a DMM, the stock leads will not be sufficient for measuring with.

Battery is Pana 18650B fully charged. I have eliminated battery and multimeter as causes because I have another driver the same and it measures 2.78 with the same battery and meter…

Stock teeny wires from driver to emitter? Using at least 22Ga wires?

They are the same wires as on the other driver that measures 2.78A

Depending on where you got the NCRB, the protection circuit could also be junk.

Its unprotected :stuck_out_tongue:

What emitter? An XP-E or rebel wont pull the full draw…

If it’s not the meter then I think his conclusion might be the correct one, stupid 7135’s. There’s no way to test them on the board. One of the Oshpark projects is a board for testing them before they get mounted but otherwise you would have to desolder and pry up each output pin in succession until you found the bad chip. I would instead do a very close visual inspection and pray for a bad solder joint on one.

Well that sucks… Just had a thought though! What about shorting out the 2 outside pins on the 7135 to bypass it? Should the light dim a little bit with each one that is bypassed? And if one isnt working nothing will change when shorted out?.. Maybe? :stuck_out_tongue:

They are all in parallel. What you do to one affects each and every other one. That’s the problem, isolating the behavior.

Short of removing the chips one by one you could cut the pwm traces one by one with an exacto knife as that would be easy to bridge afterwards.

Whoops, knew I was missing something :frowning:

Might just desolder them one by one because I need the practise anyway :slight_smile:

Or cut the pwm traces if that is easier… What are the pwm
traces?

If you have extra chips then after finding the culprit you could take a stab at stacking a few.

Yeah I will try that :slight_smile: . I got a pack of 10 7135 350mA ones from fasttech but they have a different number on the top of them compared to the existing chips on the board… does that matter or will they still work?

They will work just fine, same component, different mfg’s.

What LED is this running?

Great, thanks! A xm-l2 u3 2a on copper noctigon