Hello everyone.
I built the amp using attached schematic.
The problem is that it oscillates. Frequency of oscillation varies from about 1kHz to 20kHz. I do not hear all frequencies but I can see them with an oscilloscope. I can stop the oscillation and change its frequency and intensity by moving of anode wires of output tubes. However even when I find such place where the amp is not oscillate it is so unstable that I don’t like such solution (it is enough to move even a little part of either wire for a couple of millimeters to start the oscillation). It is very interesting that changing the places of the anode wires (changing the phase of the OT) does not change the situation significantly (it only changes the frequency of the oscillation). In all tests Gain was at least in the middle (usually at maximum).
When I put signal input wire to ground oscillation stops.
DepositFiles
(Is there any better place for files?)
The sound of the amp is cool (when I find the place for anode wires at small level of volume).
What shell I check further?
I built the amp using attached schematic.
The problem is that it oscillates. Frequency of oscillation varies from about 1kHz to 20kHz. I do not hear all frequencies but I can see them with an oscilloscope. I can stop the oscillation and change its frequency and intensity by moving of anode wires of output tubes. However even when I find such place where the amp is not oscillate it is so unstable that I don’t like such solution (it is enough to move even a little part of either wire for a couple of millimeters to start the oscillation). It is very interesting that changing the places of the anode wires (changing the phase of the OT) does not change the situation significantly (it only changes the frequency of the oscillation). In all tests Gain was at least in the middle (usually at maximum).
When I put signal input wire to ground oscillation stops.
DepositFiles
(Is there any better place for files?)
The sound of the amp is cool (when I find the place for anode wires at small level of volume).
What shell I check further?
Does it oscillate with the volumes turned right down?.
Often it's down to weither a mistake in construction, or poor layout.
With valve amps you can at least remove valves to help isolate the problem.
Often it's down to weither a mistake in construction, or poor layout.
With valve amps you can at least remove valves to help isolate the problem.
No
What do you mean under a mistake in construction? Schematic mistake?
What should I check primarily if it is poor layout?
If I put away any tube - oscillation stops. Without output tubes there is also no oscillation.
No, a wiring error.
Post a picture of it's construction - but it's essential to keep inputs well away from outputs.
If oscillation stops when you remove any, then it must be feedback from the output to the input - does it do it with no input plugged in (I'm assuming you've used the shorting jack socket on the input?.
But as V1 and V3 are on different channels, only unplugging one of them should stop it, the other (as it's not connected) shouldn't make any difference.
I will a little later. (But the input is far from output).
If I put out the cable from the input (in this case the input is connected to the ground) - the oscillation stops. If I leave the cable in (the length of the cable is 4 meters) but put the signal wire to ground on not connected end of the cable at maximum volume level small oscillation is still detectible.
The problem appears only on distortion channel (V1,V2). I disconnected everything which has no deal with the distortion channel (the character of oscillation did not changed).
Try adding some low value caps along the wires: a .1uF 600V across supply rail and chassis (or GND if chassis is floating), and see if oscillation frequency varies in frequency and or amplitude. Also add a "snubber" in the output stage. A snubber is a R-C series circuit in parallel to the output transformer, try for start, a resistor of about 1.3 times the load impedance (say, 6500ohm for a 5K output transformer), and a .01uF 2KV or more, and re-chech amplitude/frequency of the oscillation. If oscillation disappears, fix it in the circuit. Use the lower cap and the greater resistor that dampens the oscillation.
Caution: resistor not inductive (Means not wired wound), and be careful with high voltages in the circuit, and the resistor may become hot!!!
Caution: resistor not inductive (Means not wired wound), and be careful with high voltages in the circuit, and the resistor may become hot!!!
The first thing to do is get the phase right. The output at the speaker should be the same phase as the input at Loop_out. The phase inverter circuit take care of making the feedback subtract from the input. If the phase is incorrect, reverse the PS1 and PS2 lines between the phase inverter and the power tubes.
There are no grid stoppers on the output tubes. Install a 10K resistor at the power tube socket in series with the pin 5 wires.
There is a mistake the way the schematic is drawn. It shows a virtual ground near F3 and F5 in addition to the ground on pin 4 of the DC supply for the preamp tubes. Remove the 100 ohm resistors. If you have not implemented the DC heater supply, leave the 100 ohm resistors in.
There are no grid stoppers on the output tubes. Install a 10K resistor at the power tube socket in series with the pin 5 wires.
There is a mistake the way the schematic is drawn. It shows a virtual ground near F3 and F5 in addition to the ground on pin 4 of the DC supply for the preamp tubes. Remove the 100 ohm resistors. If you have not implemented the DC heater supply, leave the 100 ohm resistors in.
I tried this - nothing changed.
Adding the snubber changes the frequency of oscillation but doesn't eliminate it.
I changed the phase but I can't define which phase is right because the oscillation remains in any case. When I change the phase - the frequency of oscillation changes and to make the amp oscillation more quiet I have to move anode wires of the output tubes in other place (lead dressing?).
They are already there (forgot to mention) - the picture is the same with and without them.
I removed those resistors - oscillation still remains.
I think that it is important:the oscillation varies hugely when I move anode wires of the output tubes (in any phase).
You said you had an oscilloscope. Use that to make sure the input signal (at the phase inverter) and the speaker signal are in phase. You will need to use a second channel for triggering or the scopes external trigger.
What value are the grid stoppers? Marshall uses 5.6K which is usually high enough but not always, that's why I suggested 10K.
They shouldn't have anything to do with the oscillation, maybe hum but the oscillation needs to be fixed first.
That's what makes me think of the grid stopper issue. Try to route those wires away for the grids and phase inverter circuits. Twist them together when possible.
- Status
- Not open for further replies.