Non top boost AC-30 in for repair.

[Mark]

Thanks for your replies, I do appreciate that.

PDF64, I have previous subbed in a 16uF filter cap and the previously replaced 8uF cap was out of circuit. I have checked the power supply node with a oscilloscope and the signal isn’t sitting on top of the supply rail.

I also tried putting a plastic MKT 2.2uF cap across the cathode bypass cap to ensure the ESR wasn’t too high on that part. The 22uF cap is new and it was fine.

Phil, the issue at the moment isn’t that I’m in the normal channel which is turned down and signal is getting through. It’s that I’m in the normal channel which is turned down and I can turn up the brilliant channel volume and I can clearly hear the guitar signal. The guitar is more prominently heard when I plug into the brilliant channel and turn it to zero and then turn up the normal channel. Not a lot is heard till I reach twelve o’clock on the normal volume pot.

Here are the three volume pot measurements.

Vib/trem measures 637K. When the pot is at zero, I measured 63 ohms between wiper and earth.

Normal pot measures 627K. When the pot is at zero, I measured 19.4 ohms.

Brilliant pot measures 887K. When the pot is at zero, I measured 31.5 ohms. The brilliant pot is measuring quite high but I don’t think it is problematic.

[pompeiisneaks]

hmm correct me if this seems off, but since they share a triode doesn't that mean that the signal can go through the commonly coupled cathodes and out the other side of that triode?

I'd think that's expected behavior unless you decoupled the two cathodes of the 12AX7

~Phil

[Mark]

I'd think that's expected behavior unless you decoupled the two cathodes of the 12AX7

The cathodes are shared though they are bypassed by a 22uF cap. That should shunt the signal on the cathode to earth, I thought the ESR of the electro might be an issue so I bypassed the 22uF cap with a 2.2uF MKT plastic cap and there was no difference. I have never experienced anything like this with amps with a shared bypassed cathode. I have played a vintage 5E3 Deluxe and mucked around with the volume controls, as you do, but I found the unused volume control effected the timbre somewhat, but I never had an increase in volume.

I still have the normal channel plate and cathode disconnected and I found there was no bleed through but there was a slight change in timbre when rotating the normal channel volume pot.

I will try splitting the cathodes and see if it makes any difference.

Thanks for your input Phil, I need new ideas and to go back and examine stuff I might have brushed over.

[pdf64]

Apologies for harping on about trying a different ECC83 in there, but have you?

[Mark]

No problem at all PDF64, I was in the process of doing this myself. I did tested all the valves with an Orange Valve tested that a friend owns and they all came up good. Though, I’m unaware at what voltage the valves are being tested and what tests are taking place.

https://orangeamps.com/products/accesso ... ve-tester/



I think I have swapped 12AX7’s and had the same result. Mind you, I have been caught out on that one before when servicing an SVT. I swapped 12DW7 valves and got the same results. What I didn’t know is, these 12DW7’s had the same bizarre fault in them. When I put new 12DW7’s in the amp magically worked.

I’ll let you know tomorrow how it goes.

[Mark]

Hi PDF64, I tried swapping valves and it made no difference whatsoever. I will try removing the wiper connections and try to find out how the signal is getting into the other triode.

[pdf64]

Given time and vibration, the rather thick solid core wires typically used for early 60s AC30s can be prone to developing failing solder joints.
It's a total guess, but I wonder if instead of the regular signal path, the bleed through is getting to the LTP via the 0V common? eg a slight differential of relative circuit common points, perhaps due to a problem solder joint on the 0V return side

[Mark]

Thanks for your continued interest PDF64.

It is possible that 5mVpp is making it through 0V. Though I have moved the earth location of the 47K resistor around.

Another weird thing is, if I short the normal channel grid to earth (whilst playing through the brilliant channel), the signal is amplified even more. I have mentioned this before and I still don’t understand why it’s happening.

As shown below the amplitude is 32mVpp. While the amplitude without the input earthed is 14mVpp. This is obviously a good clue, though I don’t gather the importance of it yet.

9F3A6D92-6454-46CE-BE8D-002397FA325E.jpeg

1E71C44B-B43A-42CB-9CA2-D0B809BA9F9B.jpeg

This is the other thing that I don’t understand, how is the signal larger with the cathode disconnected from the circuit?

I will earth the grid and see if that has any effect. Waiting for enlightenment.

1C968035-F2DA-4B32-AC41-4C22D0AA80B3.jpeg

[Mark]

I mounted all the components on the valve socket and the was no difference to the fault symptoms. Hmm, what to do next?

Whilst shopping I thought there is nothing on the circuits, it is all suspended off the socket. The filter cap is different, the only things the same are the power supply node, the earth point and the valve socket.

I will try a different power supply node and a different earth point, and finally a different socket, though it sounds like madness to replace the socket for this sort of fault.

I'll report findings

(1) Changing the power supply node from correct node resulted in a signal with less distortion and more treble content bleeding through. I thought the volume was a bit louder too.

(2) I mounted the earth directly on the negative lead of the filter cap. Absolutely no difference at all.

F4E0B402-62E4-4961-9524-C1E3BB7D1B50.jpeg

5AA5BED4-6213-4113-9BE3-4104740504C4.jpeg

[Mark]

Here is a quick demonstration of the fault.

[Alnico]

I found if I plugged into one channel and turned up the other channel. Nothing happens till I get to 12:00 on the pot, then there was an increase in volume, the brilliant channel sounded more like the normal channel. Plugging into the normal channel and turned up the brilliant channel, the volume did increase but not to the same degree as the other way around.

You got me curious, so I pulled out my '63. It behaves exactly as you described above. Four input Marshalls do this too.

[pompeiisneaks]

I started thinking about this again and stared at the schematic. I threw a dud idea about the cathode and you were right, the cathode bypass cap should shunt signal to ground BUT then I looked again and they share anodes as well, the only difference here is that the signal coming out of one triode would have to go through two 220k or 440k resistance to get to the other side of the anode and bleed into the other channel. Not sure if that's it, but it seems to be probable to me.

The data that Alnico just gave seems to confirm it's expected behavior for these amps means there's nothing to see here, move along

On my AC30/4 that I built you can definitely hear the trem thumping even at 0 volume if you've got it dialed up to max.

~Phil

[Mark]

All input to this issue are good, even the dud suggestions.

I still am not convinced this isn't a fault, I had Aaron who is on this forum check his non top boost AC-30 treble model and his amp doesn't behave the same way as the one on the bench. I have asked a few other AC-30 owners what their amp does and I'm waiting for the verdict. the amp even sounds better to my ears (for what that's worth) when there is no bleed through, there seems to be more clarity.

[Mark]

I have still been plugging away with this amp and getting nowhere. I thought I’d look at the noise the amp has. It’s not terribly loud, but it is noticeable.

I expected the hum to be in the preamp stage, but it is in the P.I. stage. The filters caps have been replaced (sounded much better), I have even mounted resistors directly to the valve socket in order to establish if there is a conducting board. No change occurred.

This is what I am seeing, the main rail voltage for the amp is 354VDC.

The node for the P.I. stage is 315VDC.

The area being measured is the P.I. stage which consists of resistors mounted directly on the valve socket with no input caps or output caps present. The P.I. is isolated apart from the power supply.

The plate voltage for the brilliant/normal channel side is 239VDC with 100mVpp at 50hz present.

The plate voltage for the tremolo channel is 242VDC with 43mVpp at 50hz.

The grid voltage on the brilliant/normal channel side is 39.3VDC with 6mVpp at 50hz.

The grid voltage on the tremolo channel is 39.7mVpp with 6mVpp at 50hz.

The shared cathode voltage is 61.9VDC with 66mVpp at 50hz.

The voltage on the top of the 47K tail resistor is 61VDC with 61mVpp at 50hz.

Any thoughts?

[pdf64]

Does the hum stop if the LTP phase splitter valve is removed?