I don't know if placebo effect or not, but my amplifier sounds better if I insert a parallel LR filter in the anode path, just before the usual plate resistor.
This is called RF anode choke or "parasitic choke". Usually I read in radio schematics anything above 1mH should work to purge possible RF noise, with the knee out of the audio band but before MF band.
I put 15K res // 1.5mH inductor, but it was a quite random choice.
What do you think? And which values are ok?
This is called RF anode choke or "parasitic choke". Usually I read in radio schematics anything above 1mH should work to purge possible RF noise, with the knee out of the audio band but before MF band.
I put 15K res // 1.5mH inductor, but it was a quite random choice.
What do you think? And which values are ok?
You have effectively placed 15K in parallel with your plate resistor in the audio band. THis will significantly reduce the gain of the ECC83 and alter its operating point, both of which will likely change the sound. If you want it to reduce RF interference it needs to be in series with the plate resistor.
Cheers
IAn
Cheers
IAn
Yes it is in serie with 150K plate resitor.
Sorry I was not clear: first I have the LR cell (15K//1.5mH) and then, in series, there is the plate resistor (150K)
Which values are reccomended for the LR cell?
Sorry I was not clear: first I have the LR cell (15K//1.5mH) and then, in series, there is the plate resistor (150K)
Which values are reccomended for the LR cell?
To be honest I have never found an rf choke necessary. If RF really is a problem then it is probably better to stop it getting in in the first place at the input connector.
Cheers
Ian
Exactly what is happening in your case is hard to tell without some wide range frequency response testing. Every real inductor has a Self Resonant Frequency due to its built in stray capacitance. This causes its apparent inductance to increase as the operating frequency approaches the SRF. At the SRF the inductor acts like a near open circuit (parallel resonance). Above the resonant frequency the inductor begins to act more like a capacitor than an inductor.
A quick scan of the Digikey web site reveals typical SRF's of 1.5 mH inductors range from 500 KHz to 3.4 MHz.
You have inserted a parallel resonant tank circuit in your amplifier's plate circuit. If the amplifier stage is a triode, the Miller capacitance may be high enough to create a text book oscillator. Yes, an active device can be an amplifier and an oscillator at the same time, often without obvious indications. This could shift the operating parameters enough to affect the sound.
I usually go over my amplifier designs with an RF spectrum analyzer, but I was an RF engineer by day, so I have these kinds of things in my lab. If you have a scope try bringing the probe near the glass envelope of the tube or the case of the inductor. Ideally you should see nothing except some line frequency noise. An AM radio often makes a good RF detector.
https://www.everythingrf.com/community/what-is-self-resonant-frequency
A quick scan of the Digikey web site reveals typical SRF's of 1.5 mH inductors range from 500 KHz to 3.4 MHz.
You have inserted a parallel resonant tank circuit in your amplifier's plate circuit. If the amplifier stage is a triode, the Miller capacitance may be high enough to create a text book oscillator. Yes, an active device can be an amplifier and an oscillator at the same time, often without obvious indications. This could shift the operating parameters enough to affect the sound.
I usually go over my amplifier designs with an RF spectrum analyzer, but I was an RF engineer by day, so I have these kinds of things in my lab. If you have a scope try bringing the probe near the glass envelope of the tube or the case of the inductor. Ideally you should see nothing except some line frequency noise. An AM radio often makes a good RF detector.
https://www.everythingrf.com/community/what-is-self-resonant-frequency
You shouldn’t put 150k plate resistor in parallel with any choke, before calculating the new resistance value which would affect 12AX7 voltage gain and distortion.
A general 1mH RF choke rated 100-300mA might only have 10 ohm DCR. Paralleling it with 150K makes total resistance close to 10 ohm and you would lose almost all gain of 12AX7.
You can try, instead, put them in series, where the 12AX7 plate -> RF choke -> 150K plate resistor -> B+
This will create a LPF high frequency roll off around 28Mhz which we human being would never hear anyway, but the rolling off effect somehow has “smoothing” tone in the high note so you might feel the high frequency response sounds smoother.
A general 1mH RF choke rated 100-300mA might only have 10 ohm DCR. Paralleling it with 150K makes total resistance close to 10 ohm and you would lose almost all gain of 12AX7.
You can try, instead, put them in series, where the 12AX7 plate -> RF choke -> 150K plate resistor -> B+
This will create a LPF high frequency roll off around 28Mhz which we human being would never hear anyway, but the rolling off effect somehow has “smoothing” tone in the high note so you might feel the high frequency response sounds smoother.