First post on a 15 year old puzzle. Spent too many hours troubleshooting this 15 years ago without any luck finding root cause.
Years ago receiver was fine then little sound is coming from the left ch of my Harmon Kardon HK 930 receiver. Happens on all inputs and isolated the problem to the first transistor (TR501) in left ch of the preamp section. No gain at the collector output seen on scope with 1 KHz test tone. Right ch works normally. After a bunch of part substitutions, measuring signals, putting old parts back left ch TR501 starts to work again. No idea why.
Same left ch problem comes back recently. Take the TR501 off the pcb and it measures perfectly on a transistor checker. Replaced with new transistor, no change. Remove and measured TR501 and it is fine, therefore old and new TR is not damaged. Measured the resistors on left and right ch and they are within 2-3% of each other. Removed signal path electrolytic caps. Their values and ESR measure fine. Replaced them with new caps, no change. Measured the DC voltages around TR501 and next gain stage transistor TR501. Bias and other DC voltages are almost identical. Removed TR502 thinking TR502 base is dragging TR501 collector down, no change. Measured the resistance of tracks between parts. Continuity is fine. I redesigned +18V DC circuit to reduce ripple, noise and less drain on 35V rectified DS power supply. It's much better now.
What type of intermittent defect could this be cold solder joint, broken track?
Years ago receiver was fine then little sound is coming from the left ch of my Harmon Kardon HK 930 receiver. Happens on all inputs and isolated the problem to the first transistor (TR501) in left ch of the preamp section. No gain at the collector output seen on scope with 1 KHz test tone. Right ch works normally. After a bunch of part substitutions, measuring signals, putting old parts back left ch TR501 starts to work again. No idea why.
Same left ch problem comes back recently. Take the TR501 off the pcb and it measures perfectly on a transistor checker. Replaced with new transistor, no change. Remove and measured TR501 and it is fine, therefore old and new TR is not damaged. Measured the resistors on left and right ch and they are within 2-3% of each other. Removed signal path electrolytic caps. Their values and ESR measure fine. Replaced them with new caps, no change. Measured the DC voltages around TR501 and next gain stage transistor TR501. Bias and other DC voltages are almost identical. Removed TR502 thinking TR502 base is dragging TR501 collector down, no change. Measured the resistance of tracks between parts. Continuity is fine. I redesigned +18V DC circuit to reduce ripple, noise and less drain on 35V rectified DS power supply. It's much better now.
What type of intermittent defect could this be cold solder joint, broken track?
This is an amplifier with negative feedback in two loops. One is for DC from emitter of T2 (I'll omit 50x for easyness) and base of Q1. The other is in AC from T2 collector to T1 emitter. As the global gain is high (two common emitter stages) the signal level at T1 collector must be low. This is because it is amplifying the difference of the incoming signal and the output and this vectorial sum is very low under normal conditions. So I don't see any issue per se.
Moreover: as in the second stage the emitter is grounded via C553(?), the input resistance of Q2 is very low (Hie) thus loads severely Q1's output.
Moreover: as in the second stage the emitter is grounded via C553(?), the input resistance of Q2 is very low (Hie) thus loads severely Q1's output.
Last edited:
Completely agree with Osvaldo.
Swap TR1 with TR2 (both 2SC644). Problem may be caused by weak (original) TR2. Then replace.
C303 and C304 can measure fine, but can have problems too, they are within the gain loops. Only 6.3V and 16V specs.
Solder joints and tracks are always to suspect, especially near heat sources.
Two times R508 on the pcb (red arrow), 1.5k and 5.6K, the 5.6k is on the collector of TR2 thus R510 (compare R560).
Swap TR1 with TR2 (both 2SC644). Problem may be caused by weak (original) TR2. Then replace.
C303 and C304 can measure fine, but can have problems too, they are within the gain loops. Only 6.3V and 16V specs.
Solder joints and tracks are always to suspect, especially near heat sources.
Two times R508 on the pcb (red arrow), 1.5k and 5.6K, the 5.6k is on the collector of TR2 thus R510 (compare R560).
Thank you both for the input. Agreed the schematic and board layout has errors. I was trying to breadboard the section with two 2N3904 going from the layout and didn't notice R508 appears twice. Swapping the transistors is a good idea. I remembered removing Q2 many years ago to see if gain returns but no luck. Not sure if I had a scope back then but it's useful to see what's happening.
Need to be careful because all the soldering/desoldering makes the pads fragile. Luckily this old pcb has thick copper traces.
I will report my findings.
Need to be careful because all the soldering/desoldering makes the pads fragile. Luckily this old pcb has thick copper traces.
I will report my findings.
To avoid destroy the pads to solder, use the lowest solder iron temp compatible with solder junction to soften, and the less time posibble.
No argument with any of the above posts.
As Osvaldo noted, the amp has two feedback paths. The DC bias (R504 & R507) is clearly working since the pair biases correctly. Health of the AC path (R505 & R506) is not so clear. These two resistors set gain of (1+ 15K/820) = ~19.3 I suggest applying sine wave to input and tracing with a scope.
Good luck!
As Osvaldo noted, the amp has two feedback paths. The DC bias (R504 & R507) is clearly working since the pair biases correctly. Health of the AC path (R505 & R506) is not so clear. These two resistors set gain of (1+ 15K/820) = ~19.3 I suggest applying sine wave to input and tracing with a scope.
Good luck!