Amplifier sensitive to resistor type

[Diver300]

I've been repairing a Crest Audio CA9.

Crest have very usefully posted the schematic here:-

https://www.crestaudio.com/media/schematics/ca_schematic_set1.pdf

In the main amplifier, Q22 had blown, resulting in several of the resistors going up in smoke. I changed all of them and Q21 and Q22.

It seemed to be working, until I put it all back together, when R139 burned out. I realised that the tab of Q22 was shorting to the heatsink, which only mattered when the heatsink was earthed.

Anyhow I changed R139 but the amplifier was oscillating at MHz frequencies, only at the positive peak of each cycle.

Eventually, I fixed it by changing R146 and R147 from wirewound to carbon film resistors.

(https://uk.farnell.com/1903870 to https://uk.farnell.com/9338039)

It seems that something changed when R139 blew, but I have no idea what. It also seems that the inductance of R146/R147 is very important. I was surprised how critical it was.

 

[JimB]

Your post inspired me to have a dig in a box of old resistors and do some measurements.

The results are as per the attached picture.

The resistor at the bottom of the picture is a carbon film resistor, the rest are all wire wound types which are well known for being inductive.

JimB

 

[Diver300]

Thanks for that. There is certainly a huge difference between the 3rd (50 Ω, 1.9 µH) and the 5th (39 Ω, 0.04 µH).

I think that there must have been a similar difference on the two 10 Ω resistors that I was using.