Sometimes too much negative feedback can make an amp unstable, and thus susceptible to self-oscillation.
An example would be lowering (or raising on the low end) the frequency of the dominant pole, thus reducing the amount of feedback at frequency extremes. That reduces the effectiveness of the feedback, increasing high or low frequency distortion, and perhaps changing the distortion profile for the worse. The result just might not be as “good” as you initially expected.
LCR meter can tell you nothing about wrong wiring of ground connections inside of the amp.
Yes I agree @ Wb. I'm not suspecting cap issues all are major brand from digikey. I have had a little bit of a breakthrough last night was able to knock down the hum quite a bit with some wiring experimenting. So this design is using some individual boards from one of the members and it has separate front end/driver boards with each having gnd connections. The only time I get the hum is when I connect a stereo cable to input. The design utilizes a commom PS and common star gnd. When both channels are connected it defines a loop from drawing out the schematic. My next test will to be reconfigure gnd wiring to have only 1 wire going back to the star. Anyway I'll keep posted. This is a good learning experience and I will draw it all out. Be a good example for some of the newbies I'm sure.
Hello,
I've builted the Elektor Claus Byrith 4-30 (EL34 PP) amplifier and Im facing a wired oscilation, also with the NFB circuit disconnected.
(all the tests and scope pics: NFB off)
The problem start to appears with low output power level, 2w i.e.
The frequency of the blob is 63khz.
If I remove the power tubes there is no oscillation in the other stages (pre and LTP phase inverter) both ok:
Some pictures of the circuit, i've already tried to change the ground layout and bring the decoupling caps closer to the ef86 and 12ax7.
One funny thing that I noticed is: If I touch the grid pin of one EL34 with the multimer probe the oscillation disapears (low output levels).
I've builted the Elektor Claus Byrith 4-30 (EL34 PP) amplifier and Im facing a wired oscilation, also with the NFB circuit disconnected.
(all the tests and scope pics: NFB off)
The problem start to appears with low output power level, 2w i.e.
The frequency of the blob is 63khz.
If I remove the power tubes there is no oscillation in the other stages (pre and LTP phase inverter) both ok:
Some pictures of the circuit, i've already tried to change the ground layout and bring the decoupling caps closer to the ef86 and 12ax7.
One funny thing that I noticed is: If I touch the grid pin of one EL34 with the multimer probe the oscillation disapears (low output levels).
Check the bias resistor 390K and the grid stopper 2.2K. Also the screen resistors 1K . There are might be bad resistors or bad solder joints. You can use a wood stick to knock on the components to find out if there is any bad solder joints.
It's hard to see how the 1k screen stoppers are wired. The EL34 anode wire also seems to pass close to a PI stage wire and bias wiring. The squegger style oscillation could also be due to capacitance coupling from an EL34 anode wire to preceding stage high-impedance wiring - keep those anode wires as far apart from other wiring as practical - or move other wiring further away, or on the other side of ground wiring.
Unrelated to that, it looks like your main 0V wire from power supply zone doesn't go directly to the 10R cathodes TP0, which is the high current audio signal loop.
Unrelated to that, it looks like your main 0V wire from power supply zone doesn't go directly to the 10R cathodes TP0, which is the high current audio signal loop.
You could also try putting a 47pF, 50pF, or maybe a 120pF cap rated at 500V or greater between pins 1 and 6 of V2, also try putting a 47pF in parallel with R2 (1M), you might need a higher cap value though.
As the amp also oscillates with global NFB disconnected, I'm wondering if the screen taps are correctly wired.
Plate on one side of the primary and screen on the other would result in positive feedback.
Plate on one side of the primary and screen on the other would result in positive feedback.
Hello, in the next week I will try to make some changes in the layout. I'll keep you guys updated.
I have got a single ended 6550 tube amp that oscillates violently when in Ultralinear mode, but is stable in triode mode. no negative feedback, cathode bias. Is it any trick to make it stable?
Triode UL switch. 100ohm resistor screen to anode when i triode mode. Not sure about the transformer values. Oscillates at 4 ohm tap and 8.2ohm load in UL mode
Not sure how to check that, in works fine in triode mode. In UL I kan hear the tubes vibrate and transformer "scream" , can see the oscillation in the oscilloscope, cannot recall exact frequency but the whole ting protest in UL mode, but are OK in triode
Last edited:
If the OT is connected 'upside down' it would work fine in T-mode but may have excessive feedback in UL-mode. That might be an issue.
Jan
Jan
Last edited:
Ok, arbitrarily number the primary leads 1, 2, and 3, according to their colors.
Measure the following resistances:
1-2
2-3
1-3
Measure the following resistances:
1-2
2-3
1-3