17-01-2016, 04:17 PM
Just been playing with another one of these. The usual LF rumbling noise...
Twiddle RV3, just in case. No difference.
Change TR1, which makes it about 5dB quieter. But still too noisy...
Change TR2, which makes no difference. Often the way, but it's nicer to have a pair of BC108s rather than a BC108 and a BC148.
Tried R1, but it hardly made any difference. Refit the original. Likewise for R4. But a replacement R3 made things about 3dB quieter...
Noise is now -45dBqs (unweighted, quasi-peak responding rectifier). Better, but noise still audible through loudspeaker. Should be about 10dB less noise, so something must be going on...
As before, the noise is affected by the position of the bass control, so it must be related to TR1. We've already checked R1, R3 and R4, so...
And this time it was indeed R5. Replacing it brought the noise down to -55dBqs, which is pretty much as quiet as I've seen from these. One day I must do a blanket replacement of all resistors with film types (along with replacing all transistors) to see what the original design was capable of.
This time, the noisy resistors were standard axial 10% carbon composition types. The resistors that were OK were the "box" types that caused me grief earlier on; neither type is more or less reliable in my experience. Of course, both were out of spec in resistance terms (6k4 and 1k12).
Still, these sets are worth the effort
Twiddle RV3, just in case. No difference.
Change TR1, which makes it about 5dB quieter. But still too noisy...
Change TR2, which makes no difference. Often the way, but it's nicer to have a pair of BC108s rather than a BC108 and a BC148.
Tried R1, but it hardly made any difference. Refit the original. Likewise for R4. But a replacement R3 made things about 3dB quieter...
Noise is now -45dBqs (unweighted, quasi-peak responding rectifier). Better, but noise still audible through loudspeaker. Should be about 10dB less noise, so something must be going on...
As before, the noise is affected by the position of the bass control, so it must be related to TR1. We've already checked R1, R3 and R4, so...
(25-04-2012, 11:50 PM)Mark Hennessy Wrote: ...and it's probably not R5.
And this time it was indeed R5. Replacing it brought the noise down to -55dBqs, which is pretty much as quiet as I've seen from these. One day I must do a blanket replacement of all resistors with film types (along with replacing all transistors) to see what the original design was capable of.
This time, the noisy resistors were standard axial 10% carbon composition types. The resistors that were OK were the "box" types that caused me grief earlier on; neither type is more or less reliable in my experience. Of course, both were out of spec in resistance terms (6k4 and 1k12).
Still, these sets are worth the effort
