Hi jean-paul, indeed this is good advice. I think 90% of my broken stuff is due measuring/testing on a "live" unit.
Agreed and I already made similar changes long before I broke this amp with my stray probe.
On the one vs both channel getting new transistors I intend to test if its worth doing but first I want to make sure the one channel is performing as it should with new parts.
That said the repaired channel is up and running (been playing for 48 hours) using original drivers and predrivers taken from another amp. This was a temporary test to confirm the amp would work before I ordered new output transistors. I tested the amp taking out the damaged power transistor (thanks for the tip on this!) and run it with only 2 sets of the original output transistors and it worked.
Anyways, I have listened to it and I agree with anatech any differences are inaudible. However doing a 19/20khz IMD test indicates the repaired channel is performing slightly better (around 3dB) but I won't now for sure until I change the other channel as well.
That said the repaired channel is up and running (been playing for 48 hours) using original drivers and predrivers taken from another amp. This was a temporary test to confirm the amp would work before I ordered new output transistors. I tested the amp taking out the damaged power transistor (thanks for the tip on this!) and run it with only 2 sets of the original output transistors and it worked.
Anyways, I have listened to it and I agree with anatech any differences are inaudible. However doing a 19/20khz IMD test indicates the repaired channel is performing slightly better (around 3dB) but I won't now for sure until I change the other channel as well.
Simply having devices that match each other will reduce distortion. In manufacturing they don't have the luxury of time, the most expensive part of building or repairing anything. They just install devices without testing or grading. When I was training to be a tech, they had to measure and grade transistors since there was such a wide variance in performance. Reliability could depend on it. Today it is mostly distortion performance as modern devices don't vary nearly as much.
One prime area for improving performance is the voltage amp section, specifically the differential pair match. There are many other areas you can improve things as well.
I often see distortion reduction after repair. That is because of some attention to detail, not just that the parts are new.
One prime area for improving performance is the voltage amp section, specifically the differential pair match. There are many other areas you can improve things as well.
I often see distortion reduction after repair. That is because of some attention to detail, not just that the parts are new.
It is an elementary procedural fault. Quite easy to correct, just flip the switch "I will never again move test clips while the device is energized". Yeah the one with the green LED. When working with higher voltages and power it is kind of mandatory/life saving. Suppress the idea that test pins of DMMs exist to be moved in energized electronic devices with very short clearance between parts pins. Implement the mantra: "my test probes are short circuits".
We are human and make human mistakes because we think to be faster, gain time, are in a hurry etc. This very often leads to neglect, forgotten temporary stuff, 99% finished things, defects, electrical shocks, defective devices ... besides the always existing risks of clips that fall off, devices that have unexpected quirks. Risks add up.
Unfortunately I witnessed the consequences of such mistakes on a larger scale. It is just not worth it, take your time and work safely. That having said, it is spectacular to see such things when no one was hurt 🙂
Last edited:
Hi Jean-Paul,
Absolutely true! I once had a tech that blew amplifiers up while setting bias current. Finally I had to let him go because he refused to do thing safely.
Often you have to resort to connecting wires to test points you want to measure and bring them out, then assemble the unit so you can test it. Time consuming, but safe. Poor design costs in maintenance, therefore the end user. Manufacturers often no longer consider this.
Absolutely true! I once had a tech that blew amplifiers up while setting bias current. Finally I had to let him go because he refused to do thing safely.
Often you have to resort to connecting wires to test points you want to measure and bring them out, then assemble the unit so you can test it. Time consuming, but safe. Poor design costs in maintenance, therefore the end user. Manufacturers often no longer consider this.
