Recently I designed and built an integrated amplifier with a circlotron output stage. I noticed since the begin a noise that unavoidably is always there.
Even if i pull out the tubes that noise appears, so i did try to trace it behind until i reach the mains transformer. As long as one specific of the two floating secondaries of the circlotron is connected the noise appears. If I change the floating secondaries of place with each other, I also have that noise, which changes place if i disconnect one of them, so I think the problem comes from one of the secondaries. It isn't a noise like 50/60Hz hum, but I don't know how to describe it.
Can you suggest solutions for this problem? The transformer is a custom made one for my project.
Thanks.
Even if i pull out the tubes that noise appears, so i did try to trace it behind until i reach the mains transformer. As long as one specific of the two floating secondaries of the circlotron is connected the noise appears. If I change the floating secondaries of place with each other, I also have that noise, which changes place if i disconnect one of them, so I think the problem comes from one of the secondaries. It isn't a noise like 50/60Hz hum, but I don't know how to describe it.
Can you suggest solutions for this problem? The transformer is a custom made one for my project.
Thanks.
Are you using bridge, or full wave rectification on each secondaries? If so, good. If your using half wave rectification, you can try reversing the wires on one of the secondaries and it may help, but you are genrally asking for trouble.
Is it a mechanical buzz? When used in a configuration where the noise occurs, does the waveform coming from your transformer (measured where the transformer secondary connects to the rectifier) look more like a square wave than a sine wave?
I've noticed most tube amp transformers will buzz when using solid state rectifiers working into a large capacitance with no inductors or resistors placed beforehand. The higher the current demand on the transformer, the more it will buzz as the peaks of the secondary waveform spew their guts trying to keep the capacitor charged. You end up with a waveform that looks more like a square wave. The harmonics of this square wave like waveform become mechanically audible from the transformer.
Mechanical damping (potting the transformer, foam padding in the endbells, ect) can help, designing a power supply for lower repeatative peak current can help and using a higher current transformer can help.
Is it a mechanical buzz? When used in a configuration where the noise occurs, does the waveform coming from your transformer (measured where the transformer secondary connects to the rectifier) look more like a square wave than a sine wave?
I've noticed most tube amp transformers will buzz when using solid state rectifiers working into a large capacitance with no inductors or resistors placed beforehand. The higher the current demand on the transformer, the more it will buzz as the peaks of the secondary waveform spew their guts trying to keep the capacitor charged. You end up with a waveform that looks more like a square wave. The harmonics of this square wave like waveform become mechanically audible from the transformer.
Mechanical damping (potting the transformer, foam padding in the endbells, ect) can help, designing a power supply for lower repeatative peak current can help and using a higher current transformer can help.
It's not the toroidal transformer that is buzzing. The noise is heard on the speaker, even without the tubes connected. I'm convinced the issue is related to the parasitic capacitance between the two secondaries of the circlotron. It is probably creating a ground loop that is making this noise...
The noise is proabably corona effects on the di-electric....
Very common with power transformers of this size...
Corona can occur at these mains voltages...
The primary and secondary are typically out of phase at the di-electric interface..
So you would have 220V on the primary finish and lets say 350V on start of the first layer of secondary..
Since these are out of phase you have a potential of the peak difference not RMS...
This would be 805V peak difference across the di-electric gradient...
With the neutral and hot flipped you would reduce this voltage gradient to 495 volts...may or may not be low enough to solve the problem...
When the mylar tape is tensioned to wrap around the primary windings it sometimes stretches too thin...plus most if not all MYLAR tapes used overseas mostly in china are not from 3M but rather a bogus copy than has micro porous pockets on the sticky side...this creates the ideal conditions for corona crackling...you can confirm this by putting an AM radio next to the transformer and seeing if this crackle comes through the radio...
cerrem
Very common with power transformers of this size...
Corona can occur at these mains voltages...
The primary and secondary are typically out of phase at the di-electric interface..
So you would have 220V on the primary finish and lets say 350V on start of the first layer of secondary..
Since these are out of phase you have a potential of the peak difference not RMS...
This would be 805V peak difference across the di-electric gradient...
With the neutral and hot flipped you would reduce this voltage gradient to 495 volts...may or may not be low enough to solve the problem...
When the mylar tape is tensioned to wrap around the primary windings it sometimes stretches too thin...plus most if not all MYLAR tapes used overseas mostly in china are not from 3M but rather a bogus copy than has micro porous pockets on the sticky side...this creates the ideal conditions for corona crackling...you can confirm this by putting an AM radio next to the transformer and seeing if this crackle comes through the radio...
cerrem
Last edited:
There is a good chance you have the the magnetic feild from the power transformer leaking into the core of the audio output transformer. The test would be t move them far apartt say maybe 3 r 4 feet) and see if that fixes the problem. If distance fixes the problem, you can't build a two foot long chassis so try rotating transformers, yu might find that one of then at 45 or 90 degree rotation makes it quiet.
With the tubes pulled there is not many other paths other then magnetic coupling
The amplifier is NOT an OTL design.
I'm having multiple problems after analyzing a square wave at an 8 Ohms pure resistive load at the oscilloscope.
It is clear that the signal has a 100Hz noise in it, although it is not hum for sure, since hum appears like a DC wave with ripple (i think i'm right about this).
Putting an oscilloscope probe over the two circlotron supplies (after the filter caps), they are clearly different. I made a picture in paint to illustrate this problem: http://feupload.fe.up.pt/get/5rKk8LrkNeJat9r
Another thing is that i'm having an oscillation with a frequency of 125kHz. One thing I did not do... Putting grid stop resistors in the output valves. Do you think it will solve this problem?
I'm having multiple problems after analyzing a square wave at an 8 Ohms pure resistive load at the oscilloscope.
It is clear that the signal has a 100Hz noise in it, although it is not hum for sure, since hum appears like a DC wave with ripple (i think i'm right about this).
Putting an oscilloscope probe over the two circlotron supplies (after the filter caps), they are clearly different. I made a picture in paint to illustrate this problem: http://feupload.fe.up.pt/get/5rKk8LrkNeJat9r
Another thing is that i'm having an oscillation with a frequency of 125kHz. One thing I did not do... Putting grid stop resistors in the output valves. Do you think it will solve this problem?
Last edited:
OK. I soldered grid resistors on the output and also at the preamp valve grids.
That obvious ghost oscillation in the square wave with 230kHz disappeared but instead of it, there is a ghost signal which is side by side with the square wave at the 8ohm load... Any ideas?
That is textbook perfect picture of power supply ripple. It is unavoidable but you can control the amplitude. If it's only a few millivolts you are OK. If it is about a volt then it''s big. You picture lacks a vertical scale.
Here goes a picture of the signal on the 8ohm load without global negative feedback. As you can see, there's the signal, and a ghost one and I already tried another oscilloscope and probes... Channel 1 is the input that is a 40mV peak to peak signal with a frequency of 10kHz:
An externally hosted image should be here but it was not working when we last tested it.
Last edited:
The ripple you can see there is normal, but that blur around the wave when the filter is discharging is really strange and I think it's the main cause of the oscillations in my amplifier.
I tried it already with a good analog oscilloscope and the two signals were there... I don't think the problem has something to do with the scope.
It's only a mono amplifier.
Yep, this signal is gone when disconnecting the power supply. I already had noticed that. Any extra suggestions?
And yes I hear noise at the speakers when I disconnect all tubes.
Thank you all very much for your time...
Last edited:
If you pull tubes one at a time from the input to the output does the noise change?
IE is the noise from the pre or power?
Dont do this if its DC coupled!
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.