Been working through a couple of these units, and can’t seem to get it stable at high volumes.
All voltages seem good despite it being converted to solidstate rectifier (pi filter added to get voltages back to 20v high of published).
The tertiary feedback tap measures, and seems to be doing its do.
It oscillates with unit volume max, and preamp feeding it (10k outZ) through both a 15095 bridging octal module, or 15335 bridging 1:1.
I have it biased at around 60% diss.
Seems to get to a point where it just lights up.
The snubber across the power tube plates is new.
Thinking about adding a snubber across the phase inverter feeds, but dont want to shave highs. Any insight most welcome.
All voltages seem good despite it being converted to solidstate rectifier (pi filter added to get voltages back to 20v high of published).
The tertiary feedback tap measures, and seems to be doing its do.
It oscillates with unit volume max, and preamp feeding it (10k outZ) through both a 15095 bridging octal module, or 15335 bridging 1:1.
I have it biased at around 60% diss.
Seems to get to a point where it just lights up.
The snubber across the power tube plates is new.
Thinking about adding a snubber across the phase inverter feeds, but dont want to shave highs. Any insight most welcome.
I have seen the Altec 1570 do the same when you turn the volume all the way up.
Likely a stopper resistor in the input stage will fix you.
When the volume is not at full setting the pot acts as a stopper.
When you turn the volume all the way up you do not have a stopper.
Try 10K as close to the input tube pin 7 as you can get it.
Likely a stopper resistor in the input stage will fix you.
When the volume is not at full setting the pot acts as a stopper.
When you turn the volume all the way up you do not have a stopper.
Try 10K as close to the input tube pin 7 as you can get it.
The preamp Zout at 10k also seems excessively high (if it is indeed 10k Zout which is very unusual),
that will cause all kind of sensitivities to cap loading and coupling between in and out.
Jan
that will cause all kind of sensitivities to cap loading and coupling between in and out.
Jan
Have you measured the 'oscillation'? Is it a low or high frequency? Does it happen straight away, or 'build up' over a second or so when input vol pot is rotated to max?
"I have it biased at around 60% diss. Seems to get to a point where it just lights up." Is that comment about the 5u4's arcing ?
"The snubber across the power tube plates is new." Are you referring to C12 across the output tube plates? If so, then that cap needs a really high AC voltage rating (not just a high dc voltage rating).
"Thinking about adding a snubber across the phase inverter feeds, but dont want to shave highs." I think you are referring to a step filter across R4, or the rather quirky coupling between the cathodyne PI and the driver stage?
At the moment, I wouldn't start changing the feedback compensation part values, or adding step networks, other than making sure you have a nominally resistive load always connected (ie. a speaker with zobel network to maintain a somewhat constant impedance. over frequency). Modifying the feedback is a somewhat specialist topic.
The amp is very similar in topology to a Williamson amp, so a lot its assessment also relates to your amp, although I am intrigued about the coupling setup to the driver stage.
"I have it biased at around 60% diss. Seems to get to a point where it just lights up." Is that comment about the 5u4's arcing ?
"The snubber across the power tube plates is new." Are you referring to C12 across the output tube plates? If so, then that cap needs a really high AC voltage rating (not just a high dc voltage rating).
"Thinking about adding a snubber across the phase inverter feeds, but dont want to shave highs." I think you are referring to a step filter across R4, or the rather quirky coupling between the cathodyne PI and the driver stage?
At the moment, I wouldn't start changing the feedback compensation part values, or adding step networks, other than making sure you have a nominally resistive load always connected (ie. a speaker with zobel network to maintain a somewhat constant impedance. over frequency). Modifying the feedback is a somewhat specialist topic.
The amp is very similar in topology to a Williamson amp, so a lot its assessment also relates to your amp, although I am intrigued about the coupling setup to the driver stage.
All excellent ideas, and they even crossed my mind too!
RE: plate to plate cap ac handling.
I didn’t think about high AC swings, despite the part living where AC is highest. Thanks
will update this evening whence back in there.
The unit has previously been converted to SS rectifier. Then a large pi CRC filter off of reservoir, to get all B+ drops nearish spec voltage.
got around 475 on plates, -45v on grids ”currently“.
RE: plate to plate cap ac handling.
I didn’t think about high AC swings, despite the part living where AC is highest. Thanks
will update this evening whence back in there.
The unit has previously been converted to SS rectifier. Then a large pi CRC filter off of reservoir, to get all B+ drops nearish spec voltage.
got around 475 on plates, -45v on grids ”currently“.
Clean the volume control with elecro-wash or deoxit f5.: A cleaner that has a light oil in it.
I would also place a 1M resistor from the tube grid to dc ground to stabilize the volume control since the bias current is dependent on the quality of the control's wiper.
If not fixed yet:
Lift R11 and C8 out of circuit. if it hums or oscillate, check the transformer for a center tap open to ground on the heater.
Change transformer due to shorted windings.
Otherwise:
If it doesn't oscillate after lifting the negative feedback loop reversed the connection on the output transformer on the feedback loop.
Also, there should be a grid stopper in the first tube where the feedback loop is returned. A valid resistor size is the same value as the feedback loop: (2200 ohms).
Last edited:
The input network comes strapped around a filter. The 100k to ground on that input puts leak at max to 70kish. Pot has been replaced. 1m was already in there when I received it.
oscillated quicker in pot rotation. Went to 250kA for pot and it improved. Despite 6cg7 detailing bias R value as 2m2. I found that any leak value over 100k increased the potential for oscillation.
RE: tertiary NFB
Seems as though the color coding for the OT matches the wiring of most units I’ve seen pictures of online. I didn’t check phase of NFB, but it seems correct. Will inspect.
The 10k source is from an old preamp unit that states 10k outz.
oscillated quicker in pot rotation. Went to 250kA for pot and it improved. Despite 6cg7 detailing bias R value as 2m2. I found that any leak value over 100k increased the potential for oscillation.
RE: tertiary NFB
Seems as though the color coding for the OT matches the wiring of most units I’ve seen pictures of online. I didn’t check phase of NFB, but it seems correct. Will inspect.
The 10k source is from an old preamp unit that states 10k outz.
Parallel tubes can oscillate at VHF frequencies, where the output wiring forms a tuned circuit. Any changes in wire length or location in this area could induce oscillations to begin. Many amplifiers of this design added plate stopper resistors (50-100 Ohms) or inductors, or both in parallel.
Goldstache, when you get time it would be good to try and identify the frequency of oscillation, as that could easily alter the focus on any remedies.
One aspect to keep in mind is that the amp would not have been unstable in new condition, which implies the 'cause' of instability is more likely to be in a degraded part or subsequent change in parts and/or cabling. That's not to say that worthwhile modifications based on 60 years of collective experience (since amp manufacture) shouldn't be seriously considered, but rather that they may not align with your initial issue.
Even the act of changing the input pot could have altered the screened wiring run, or grounding, and change the level of parasitic feedback to the input grid. Also have you swapped the 6CG7's around, and/or used other 6CG7, to see if the oscillation conditions remain the same or vary ?
One aspect to keep in mind is that the amp would not have been unstable in new condition, which implies the 'cause' of instability is more likely to be in a degraded part or subsequent change in parts and/or cabling. That's not to say that worthwhile modifications based on 60 years of collective experience (since amp manufacture) shouldn't be seriously considered, but rather that they may not align with your initial issue.
Even the act of changing the input pot could have altered the screened wiring run, or grounding, and change the level of parasitic feedback to the input grid. Also have you swapped the 6CG7's around, and/or used other 6CG7, to see if the oscillation conditions remain the same or vary ?
Thanks so much for all the replies.
haven‘t revisited yet. But will update as I go.
I didn’t look at osc frequency on scope. But it’s above 14k, my personal threshold. lol
Monitoring various valves in output section, bias is very solid and then at some point input swamps and it lights up. ( current ramps at periodic intervals, and a perceived thump is audible whilst.
the only mods done on this are the ones it came in with.
recapped. (Screen supply cap added)
SS rectifier and CRC to knock noise down and return to somewhat stated voltages.
panel mounted TRS input jack near to input terminals. Jack is grounded through chassis connection there.
a small bit of unshielded wire from TRS to Input iron.
haven‘t revisited yet. But will update as I go.
I didn’t look at osc frequency on scope. But it’s above 14k, my personal threshold. lol
Monitoring various valves in output section, bias is very solid and then at some point input swamps and it lights up. ( current ramps at periodic intervals, and a perceived thump is audible whilst.
the only mods done on this are the ones it came in with.
recapped. (Screen supply cap added)
SS rectifier and CRC to knock noise down and return to somewhat stated voltages.
panel mounted TRS input jack near to input terminals. Jack is grounded through chassis connection there.
a small bit of unshielded wire from TRS to Input iron.
Adding a screen supply cap might alter the output stage behavior in this case. Might be worth trying without the screen supply cap.
Welp I rewired the input/volume/grid v1.
added 10k series grid R.
still acts funny
Not funny like before where it would seemingly start to oscillate.
everything seems peachy until I get to about 110vac on variac. Start to hear a little fizzle crackle from speaker (8ohm 15” 100w)
Its almost like maybe damage has been done before and possibly there’s an insulation issue in one of the iron chunks.
I still didn’t measure freq of onset condition (oscillation) as it’s not thumping anymore, just some yuck when nearing full wall.
The presentation of this particular failure is kinda scary, as current jumps around when it happens (all tubes), so I don’t leave it on long enough to get an AC measurement of possible parasitic.
all windings meter typical R measurements and nothing shorts to case
I’m gutting the SS rectifier and returning it to tube to see if there’s a primary insulation issue and it’s moot. Or the secondary getting pretty high and causing havoc.
will update, many thanks.
added 10k series grid R.
still acts funny
Not funny like before where it would seemingly start to oscillate.
everything seems peachy until I get to about 110vac on variac. Start to hear a little fizzle crackle from speaker (8ohm 15” 100w)
Its almost like maybe damage has been done before and possibly there’s an insulation issue in one of the iron chunks.
I still didn’t measure freq of onset condition (oscillation) as it’s not thumping anymore, just some yuck when nearing full wall.
The presentation of this particular failure is kinda scary, as current jumps around when it happens (all tubes), so I don’t leave it on long enough to get an AC measurement of possible parasitic.
all windings meter typical R measurements and nothing shorts to case
I’m gutting the SS rectifier and returning it to tube to see if there’s a primary insulation issue and it’s moot. Or the secondary getting pretty high and causing havoc.
will update, many thanks.
Ok, wired it back to stock (tube recto) schematic, except for the reservoir is 100uF instead of 80uF.
aside from that heres what’s altered.
-10k series grid R right on V1 input pin (7)
- 220pf in series with 10k across V1 plate R (R4)
no more audible crackle at full 120vac wall.
Does still start to audibly thump when AC input gets above a certain threshold. Sorry don’t have a vac input number, but it’s well below the .9vac spec on schemo.
all tubes have been swapped and swapped. Don’t think it’s a valve issue.
Input level seems to just light up osc in power supply.
No signal applied, unit volume at zero (grounded), idles just fine.
Presently I have it biased kinda cold 25ma per KT77/470v on plates.
aside from that heres what’s altered.
-10k series grid R right on V1 input pin (7)
- 220pf in series with 10k across V1 plate R (R4)
no more audible crackle at full 120vac wall.
Does still start to audibly thump when AC input gets above a certain threshold. Sorry don’t have a vac input number, but it’s well below the .9vac spec on schemo.
all tubes have been swapped and swapped. Don’t think it’s a valve issue.
Input level seems to just light up osc in power supply.
No signal applied, unit volume at zero (grounded), idles just fine.
Presently I have it biased kinda cold 25ma per KT77/470v on plates.
Did you remove the screen bypass cap? Just so it is really stock configuration?
After that my guess would be some drifting part either coupling cap or resistor.
After that my guess would be some drifting part either coupling cap or resistor.
The screen cap is gone.
Stock 3 filter wiring (100/50/50uF)
All caps are swapped. Only original part in preamp/bias/power is the combo of series resistors that make up cathode of V2 and they measure good?
totally stable idle and when signal is around 250mv it starts thumping.
Even at moderate listening levels the current at power tubes barely winces on transient peaks.
Then I get input level to a place where it starts thumping. And bias swings around dramatically.
Stock 3 filter wiring (100/50/50uF)
All caps are swapped. Only original part in preamp/bias/power is the combo of series resistors that make up cathode of V2 and they measure good?
totally stable idle and when signal is around 250mv it starts thumping.
Even at moderate listening levels the current at power tubes barely winces on transient peaks.
Then I get input level to a place where it starts thumping. And bias swings around dramatically.
Just a thought... make sure that C3 doesn't have any issues.
Most feedback designs are optimized with the forward gain maximized and then 1/β sets the overall gain. C3 ensures the stability of the feedback loop by bypassing AC around the 1800Ω cathode bias resistor. If C3 had any high frequency issues then you might see some of the bias resistance at AC. This would reduce β and lead to instability.
Most feedback designs are optimized with the forward gain maximized and then 1/β sets the overall gain. C3 ensures the stability of the feedback loop by bypassing AC around the 1800Ω cathode bias resistor. If C3 had any high frequency issues then you might see some of the bias resistance at AC. This would reduce β and lead to instability.