The grounding scheme is very strange. All signal grounds, including power supply, should go to a single point, and should float from the chassis. This point should ideally be at the lowest signal potential of the circuit, that is, the input. AC ground goes to the chassis for safety. Remember that power supply decoupling caps for a given stage are part of the AC signal return for that stage. Separate them and you've created a much larger AC loop for the circuit. For example, the 3.3uF decoupling caps for the 6N1P should be grounded to the same point as the grid resistors and cathode resistors for that tube.
If you need shielding from the top plate or chassis, you can link the chassis to the signal ground at one point, but it is better to leave it floating, IMO.
In your case, I think the problem is one of proper grounding of the input grid. Your source is acting as a ground reference for the 6N1P grids, and when you unplug it they are left floating somehow and causing the 6N1P to oscillate.
If you need shielding from the top plate or chassis, you can link the chassis to the signal ground at one point, but it is better to leave it floating, IMO.
In your case, I think the problem is one of proper grounding of the input grid. Your source is acting as a ground reference for the 6N1P grids, and when you unplug it they are left floating somehow and causing the 6N1P to oscillate.
Again, if your grid-to-ground path is lengthy, you might be picking up RF, EMI or other interference when the source is disconnected. The grid-leak resistor, cathode resistor and all other ground paths for that tube should be tied closely together, then referenced to the rest of the circuit ground.
The star ground for circuit and PS grounds both tie to the IEC’s earth ground. Again, it’s in a wood chassis so a chassis ground like would be used on a metal enclosure isn’t an option.
It could be the circuit ground tied to IEC that's a problem. Also, if there's not metal plane at all, your transformers may not be saftey grounded. I've used wood and Delrin material and I tie all the transformer shells to IEC ground and let the circuit ground float. Just a suggestion.
We're only assuming that the symptom is motorboating (VLF oscillation) but it sounds like it could be squeeging (VHF oscillation in repetetive bursts at an audio rate) instead. "Star" grounds encourage this with their long multiple ground loops. They're problematical in a stereo amplifier, at best.
I'd like to second grovergardner's concerns about PE/safety ground wiring. Something as inherently dangerous as a wooden chassis requires extra attention to PE concerns. You might want to investigate the bridge diode // 10R method of isolating signal ground from PE (for low voltage) with failure protection at > 1.4 Volt +/.
All good fortune,
Chris
I'd like to second grovergardner's concerns about PE/safety ground wiring. Something as inherently dangerous as a wooden chassis requires extra attention to PE concerns. You might want to investigate the bridge diode // 10R method of isolating signal ground from PE (for low voltage) with failure protection at > 1.4 Volt +/.
All good fortune,
Chris
Thanks. I haven’t had much time today to work on this but will try to get a photo uploaded tomorrow.
I previously had this amp running in a metal enclosure with a circuit board and didn’t experience this issue so the first thing that comes to my mind is where the differences are between the circuit board and point to point wiring. Everything on the circuit board shared the same ground plane which was also tied to IEC so it never crossed my mind that there could be a problem there.
I previously had this amp running in a metal enclosure with a circuit board and didn’t experience this issue so the first thing that comes to my mind is where the differences are between the circuit board and point to point wiring. Everything on the circuit board shared the same ground plane which was also tied to IEC so it never crossed my mind that there could be a problem there.
It is low frequency feedback
So in general a tighter board
layout might hide issues.
Compared to a larger point to point spacing.
Input wires can be very small actually.
And shielding goes to star ground not case.
Often use tiny tiny headphone wires with shielding
for tube inputs. And keep signal grounds
away from safety ground.
Its just a zillion feet of green wire attached to neutral buss.
110/120 is half tap on the transformer.
So " ground" is just neutral on the buss.
So as usual expect noise or other nonsense
dealing with grounding myths and chassis ground safety.
1k grid stop can be enough
but with the volume pots removed
depending on rotation they would make it a higher value.
Could increase them from 1k to any standard value upward.
5.6k , 8.2k , 10k possibly up to 20k
depending on miller capacitance.
Assume same old 150p to 200p
Even if high around 200p
that grid stop can be as high as 22 to 27k
and not worry about bandwidth
So in general a tighter board
layout might hide issues.
Compared to a larger point to point spacing.
Input wires can be very small actually.
And shielding goes to star ground not case.
Often use tiny tiny headphone wires with shielding
for tube inputs. And keep signal grounds
away from safety ground.
Its just a zillion feet of green wire attached to neutral buss.
110/120 is half tap on the transformer.
So " ground" is just neutral on the buss.
So as usual expect noise or other nonsense
dealing with grounding myths and chassis ground safety.
1k grid stop can be enough
but with the volume pots removed
depending on rotation they would make it a higher value.
Could increase them from 1k to any standard value upward.
5.6k , 8.2k , 10k possibly up to 20k
depending on miller capacitance.
Assume same old 150p to 200p
Even if high around 200p
that grid stop can be as high as 22 to 27k
and not worry about bandwidth
Following up on some previous posts, without the input tube in I measure 100k resistance between the input (+) and ground.
For interest’s sake I put a 2 prong adapter on the power cord to lift the whole ground. With that in place the right channel still motorboats but the left channel did not. I reinserted the left RCA, removed it again, and it MB’d too. Then re-inserted left RCA, removed it a third time and no MB. So I can’t explain what is happening but it is likely ground related.
I don’t have time today but this week sometime I will tie the transformer to earth ground and remove the rest of the audio circuit from the earth at the IEC connection.
Unless someone has a better suggestion after seeing the photo.
For interest’s sake I put a 2 prong adapter on the power cord to lift the whole ground. With that in place the right channel still motorboats but the left channel did not. I reinserted the left RCA, removed it again, and it MB’d too. Then re-inserted left RCA, removed it a third time and no MB. So I can’t explain what is happening but it is likely ground related.
I don’t have time today but this week sometime I will tie the transformer to earth ground and remove the rest of the audio circuit from the earth at the IEC connection.
Unless someone has a better suggestion after seeing the photo.