Quad 303 Voltage Query

[Craig Sawyers]

Everything is down to too much current flowing through Q102, Q107 and R116, R117.

From the measurements at the emitter of TR102 at 0.7V, 7ma must flow through R114 (100 ohms). That is confirmed by the voltage at the junction of R116/R117/C106. This the OP measured to be 34.2V, which means that 67 - 34.2V is across R116 (4k7) giving a current through this resistor of 7mA.

For that chain to work properly half that current needs to flow - 3.5mA. That sets the base of TR102 at 33V. The Vbe multiplier TR107 then sets the base of TR104 at 33-1.7 = 31.3

Which are all about what they ought to be.

So the base of TR102 is not being biassed properly. The emitter should be at 0.35V to give 3.5mA collector current, and the base should be about 1V. If the base voltage is correct at 1V, MR100/101 should not be conducting.

I suspect that those diodes are there to prevent an input transient getting through and giving an unsafe bias of the output power transistors. Alas both designers (Peter Baxandall of the output triples and bias system, and Mike Albinson of the input section) are long pushing up daisies. So we cannot ask them.

The next thing to check are the value of R104, R130 and R112. You might have to lift one end of these three to make an accurate resistance measurement.

Craig

[PJL]

There seem to be multiple circuit diagrams and I was working off the one without TR107. I am now working from 'Quad 303-33 Service Supplement' although it still does not reference a Q102 or Q107.

The output (pin 5) is about 3V when it should be 33V and this results in TR100 base and emitter being low, TR101 being cut-off and consequently TR102 being fully on. It's not immediately obvious why this would cause the top triplet to run low current though.

The service manual setup requires RV100 to be adjusted to get 33.5V on pin 5 so this could be a good place to start. Does iRV100 do anything to the voltage on pin 5?

[GlennJobson]

OK, first anomaly...

R104, R130 and R112

I have lifted one end of each on each driver, they are marked the same per board, so each one 'should' be the correct value according to my list.

Reading are

Left Channel - R104 40.6K (incorrect?) R112 153K R130 39K

Right Channel - R104 22K (correct?) R112 154K R130 39K

[Craig Sawyers]

That the 22k is 40.6k explains why the collector voltage of TR100 is roughly half what it ought to be.

Replace it with the correct value of 22k and see what happens. There may be other problems, but that would be a good start.

Craig

[GlennJobson]

OK, some success.

I replaced R104 with a new 22K, also R112 as it tested fine but looked 'cracked'.

I now can test all point on the left channel and the voltages are identical, within 0.1V of the same points on the right channel.

Connecting the speakers I have very good audio from the left and not quite so loud from the right. I have no pre-amp connected, connected directly to the inputs on each board.

Thinking about all of the results I have found on the board given the advice from this thread I can't find a difference on the channel boards now, both tests exactly the same as far as I can test.

What should I look for next.

Thanks again...

[Craig Sawyers]

So what you are saying is that the repaired L channel is louder (whatever that means) than the R channel?

Anyway, since high value resistors go higher (or that seems to be the case) the high resistance value gain setting resistor is R113, which should be 82k on both channels.

Check the measured value on L and R channels.

Craig

[nutteronthebus]

early transistor amps can suffer from epoxy rot ( PNP type more than NPN ) see https://www.vintage-radio.net/forum/...ight=epoxy+rot

[Craig Sawyers]

Well yes, epoxy rot is certainly a thing.

But the OP says now the repaired L channel is working having replaced R104, DC voltages from both channels are now near identical (within 0.1V). If the bias conditions are identical it strongly suggests that there is no epoxy rot.

Since the DC voltages are OK, the AC gain is set by resistors - R113 (82k), R111 (2k2), R108 (22k) and R101 (22k). Any of those could be off, so they should all be checked, starting with the 82k R113, then R108 and R101, and finally R111. On both channels.

Craig

[GlennJobson]

OK, found some time to remove the resistors and test...

R113 L 83.3K R 82.7K
R111 L 2K2 R 2K2
R108 L 22.1K R 22.1K (same value, different 'look' - different manufacturer?)
R101 L 22.1K R 22.1K (same value, different 'look' - different manufacturer?)

The boards are the original board and nobody has removed any of these components before me by the look of the solder joints.

[Craig Sawyers]

OK - so the AC gain determining resistors are correct. The slight difference in value of R113 amounts to a gain difference of 0.06dB - which is insignificant.

How certain are you that the gain of the two boards is different? Can you quantify that?

Craig

[GlennJobson]

>How certain are you that the gain of the two boards is different? Can you quantify that?

In this instance I am a member of the general public and not an expert. So the comment I made about the L and R channel output may have been misleading.

>So what you are saying is that the repaired L channel is louder (whatever that means) than the R channel?

Yes, and to clarify my use of the word 'louder', the common term that would be used to describe the audible output from one speaker being noticeably different to that of the other.

The test was though perhaps not made in optimal conditions.

I have a test tone from a single sine oscillator of perhaps an unknown 'level', it has an output dial but it is not quantified other than 1-6. This was connected directly to the channel boards, the inputs where the Quad connections arrive from the amplifiers back plate.

I tested each board with a 6ohm speaker connected to the correct output. The repaired board was producing a noticeable louder output.

This week I have been working on the Quad preamp, with this connected using the 4-pin cable there is now no obvious difference.

Additional tinkering has been done on both boards, I have checked all of the solder connections on the board, not sure that was actually necessary but all of this combined with the Quad preamp I think all is well.

Thank you for your patience and all the helpful suggestions.

[Craig Sawyers]

Good - all seems well. So the only problem was R104.

Craig