Tim Biller has his amplifier back - and it is now working (I hope!). So what was the fault? Well if you go back and refer, I was convinced it was a connection problem of some type somewhere very near the TPA3112 chip - but we just couldnt find it (a common problem! - quite a few things are written off in repair shops for this very reason). So we changed the TPA3112 chip, which was a total PITA which did improve some things, but didnt fix the problem. Eventually we found it. Because of how Class D amplifiers work (search for TPA3112 or Class D is you want to know how) they require a passive low pass filter on the output - and one capable of handling the full range of speaker voltages and currents. This typically consists of a couple of inductors and capacitors. And on this board it is a push-pull output, so the two output pins are connected to chokes. The board was laid out for leaded, radially-mounted chokes (inductors) on a 6mm pitch. The factory appears to have run out of 6mm inductors so they decided to use 5mm pitch inductors instead. So they set their pick n place machine to ram those into the holes instead. Either that or maybe a large mallet. This can work, but it forms a very sharp bend in the lead - stretching the outside of the kink. Worse - this is completely hidden underneath the chokes. With Vibration (remember - this is an amp and speaker) the sharp bend can break. And so it was. WORSE STILL - when you turn the board upside down and start poking around with your meter to check for connectivity - you put a tiny bit of pressure on the choke and thats just enough to push the broken leads together so your meter lies to you and tells you the connection is just fine. Grrrr.
Posted on: Wed, 03 Sep 2014 08:21:38 +0000
Trending Topics
Recently Viewed Topics
© 2015