Help with homebrew amp

[paggerman]

Hi all, thanks for you help in the past; I hope that I can count on your collective wisdom again.

I've built a push pull amp using a chassis and PT from a radiogram, 6N2P triodes and 6P14P output valves. The pre-amp is built to the specs of Merlin's switchable cascade/parallel pre amp, a James tone stack, cathodyne PI and regular PP power stage following exactly the power amp of my first build (apart from valve type). When I pick strings in parallel mode it sounds really rich and full. In cascade mode it sounds thinner, but no problems.

I have two problems. Problem 1 - If I strum the sound breaks up and fizzles out and soon comes back when I stop strumming. Problem 2 is that when I flick the switch from parallel to cascade I lose ALL sound for around 45 seconds, then it slowly comes back - a bit like when warming up from power on. Switching cascade to parallel is OK. With no signal, it's really quiet - no 50Hz hum at all.

To address problem 1, I have taken the approach that this is 'blocking distortion', but I don't know as I can't find a sound sample to compare. I have reduced the values of the post PI coupling caps to 22nF from 47nF and have added a 750k resistor to the grid of the PI. This has marginally improved the problem, but it is still severe. Because of this problem I haven't even tried to turn the volume up beyond what would be '2' if there were any markings.

For problem 2 - I have scratched my head a lot, followed the circuit diagram through and come up with nothing to change.

I have measured voltages and have shown them on the attached diagram. If anyone would care to have a look and let me know if you see anything suspicious, I would really appreciate it.

Very many thanks.

[loudthud]

Why such a small resistor feeding the HT3 node? I think you need a much larger resistor there to slow down changes in B+. Have you tried a more conventional arrangement with HT3 and HT4 reversed?

The input impedance of the PI is very high due to the boot-strap affect around the 470K. Try reducing the 22nF cap on the input of the PI to 1000pF or even lower until you start to get a bass roll off.

[tubeswell]

If I strum the sound breaks up and fizzles out and soon comes back when I stop strumming. Problem 2 is that when I flick the switch from parallel to cascade I lose ALL sound for around 45 seconds, then it slowly comes back - a bit like when warming up from power on. Switching cascade to parallel is OK. With no signal, it's really quiet - no 50Hz hum at all.

Cool looking schematic. Have you metered the voltages on the filter caps when it does this?

[martin manning]

Is your power supply really wired as shown in your schematic? Specifically, the cathodes of both rectifiers should be connected to the top of the 220k resistor.

[paggerman]

good call

Is your power supply really wired as shown in your schematic? Specifically, the cathodes of both rectifiers should be connected to the top of the 220k resistor.

Thanks Martin - well spotted. No, it's not really like the schematic (and never was!).

I haven't had much time in the last week or so to get a fix of solder fumes - hopefully tonight.

I will be rearranging HT3 and HT4, putting a much bigger resistance between HT2 and HT3 nodes (3 x 100k 1w in parallel) and a 1000pf in front of the PI. I will update the schematic and report back.

Many thanks

All changes listed above made this evening - no improvement. I think my next move will be to reduce the value of the anode resistors of the triodes - I'll try them one at time, starting at the input stage. Unless anyone has another suggestion? I'm going to concentrate on the fizzling out when strumming because then I can use it, and work on the switch problem later.

Cheers

[paggerman]

Just an update on where I am at with this build. I dismantled the switchable cascade / parallel input and rebuilt it as per a diagram I picked up for the Marshall 18Watt. Works beautifully on the 'HI' input, but I just get a big farting noise on the 'LO' input (some guitar sound is barely audible). I haven't even had a look underneath to debug the dud input. I really like the parallel input triode sound. And the whole thing has a lot less buzz than any of my other builds. I'll try the switchable input again on my next build - I'm getting a 'don't fix something that ain't broke' feeling about this one.

Update: got the DVM out, checked continuity and found that the connection from the 'LO' input to earth was dry. A quick dab of solder, plugged in and at low volume and all seems well. I'll give it a workout tomorrow and update the schematic to reflect the changes. Cheers all!

[Bruce / Mission Amps]

Re draw the schematic to what the amp actually is... and where is the preamp volume control in this thing?
Naturally, I don't think your 500K pot, wired as a variable resistor, is the volume control, possibly some kind of sloping midrange but,.... are you thinking your distortion boosting master volume pot, ahead of the cathodyne driver, is the "volume control"?
These modified Baxandall tone-vol circuits confuse me sometimes. ha ha

[paggerman]

Resolution of problem and updated schematic

Happy New Year to everyone. Thanks for all your help and guidance on this. I finally got the root problem - I had used some solid wire I picked up from somewhere and tried to use it jumper between two terminals on an input jack. The wire was clearly unsuitable for purpose as the solder joint was very dry and wouldn't solder in. I was at risk of melting the jack before soldering the wire. So I replaced it with 2cm of offcut wire and it now works brilliantly. I really like the tone.
Here's the schematic of the finished product - enjoy (if you're into that sort of thing)!

Red15 schematic.pdf