Fair point, Nigel. The only reason I did not suggest that is that I've never had a problem with a small capacitance with large degen resistors in the input diff pair. At 470 ohms 100pF gives you a unity gain frequency of only 3.3MHz rather than 10MHz, which is more usual.
But worth trying all the same.
Especially if the wiring is not the best.
I'm not sure if oldjack still has the basic configuration or changed things over to Patrick's options. Did you change the caps, Jack?
But worth trying all the same.
Especially if the wiring is not the best.
I'm not sure if oldjack still has the basic configuration or changed things over to Patrick's options. Did you change the caps, Jack?
Sometimes the designer sets C1 to be as little as possible to get best bandwidth. Then someone else comes along building it from a different batch or newer batch of transistors and it oscillates.
Patrick, the grounds from the power supply caps are all taken to a single screw connector.
Each of the signal grounds are taken separately to a star point connector.
The two ground connectors are then taken separately to a common earth connection on the power supply.
There is no chassis or case, this is only a free standing pcb with integral heatsinks.
I did some more readings, with the signal generator connected but not switched on the input reads 5mv and the output (connected to 8R resistive) reads -46mv.
With nothing connected at the input, q1b is 115mv, output is 65mv.
q2b is 117mv (from feedback line).
Each of the signal grounds are taken separately to a star point connector.
The two ground connectors are then taken separately to a common earth connection on the power supply.
There is no chassis or case, this is only a free standing pcb with integral heatsinks.
I did some more readings, with the signal generator connected but not switched on the input reads 5mv and the output (connected to 8R resistive) reads -46mv.
With nothing connected at the input, q1b is 115mv, output is 65mv.
q2b is 117mv (from feedback line).
Is it your heatsinks fixed to a metal plate below the amp board, if there is a metal plate, please connect a ground wire to that metal plate for the amp's grounding.
You tested your amp output is ok, temp is ok, then your amp is ok to work. The small output DC offset below 100mv is acceptable.
Next time you build an amp with matched input transistors the output DC offset will be better.
You tested your amp output is ok, temp is ok, then your amp is ok to work. The small output DC offset below 100mv is acceptable.
Next time you build an amp with matched input transistors the output DC offset will be better.
Have managed to fit variable resistors r7 and R8 to second channel.
Adjusted reading between q10e and Q11e to 20mv.
With no input, reading at input 50mv, output 0.5mv, 8r resistive load attached.
Fitted pot to input and connected signal generator, input reading 0mv, output 50mv.
Connected cd player, input reading 0mv, output -50mv.
From above readings not sure what is happening as the pattern is different from the other channel.
Ran both channels with no input for 30min, heatsinks Luke warm.
Ran both channels from signal generator for 30min with +/-15 volt sine output, heatsinks Luke warm. Sine waves look clean.
I think I will now redesign the pcb and incorporate the recommended mods.
How close should q21 and q2 be matched.
Thanks everybody for the input, without it there is no way I could have sorted any of the problems out.
Adjusted reading between q10e and Q11e to 20mv.
With no input, reading at input 50mv, output 0.5mv, 8r resistive load attached.
Fitted pot to input and connected signal generator, input reading 0mv, output 50mv.
Connected cd player, input reading 0mv, output -50mv.
From above readings not sure what is happening as the pattern is different from the other channel.
Ran both channels with no input for 30min, heatsinks Luke warm.
Ran both channels from signal generator for 30min with +/-15 volt sine output, heatsinks Luke warm. Sine waves look clean.
I think I will now redesign the pcb and incorporate the recommended mods.
How close should q21 and q2 be matched.
Thanks everybody for the input, without it there is no way I could have sorted any of the problems out.
you can just using a multimeter with small signal transistor hfe testing function to test a few tansistors and select a pair of same/almost same high hfe reading transistors is ok for the matching.
This is normal in amp build and no need to concern.
You should concern both amp channel output wave form and level is the same when they inputted same signal level.
If both channel output wave form and level is ok, then everything is ok.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.