Erratum; not EF184 but should have mentioned 12GN7 which has a sim transconductance. It is a flighty tube.
I had the same problem with a pentode, CCS don t run correctly that move the voltage in horizontal normally go to the maximum B+ with risk for the tube. My decision was to change by a triode.
Maybe the pentode can be regulated by gyrator but I didn't try
Maybe the pentode can be regulated by gyrator but I didn't try
A pentode will give the same problem: you can't set one current to two different values at the same time. Something has to give. For a pentode or a heavily degenerated triode it is anode voltage which gives, by going to max or min values allowed by the rest of the circuit.
Well, I replaced the CCS with a resistor. I adapted the D3A to a matching operating point to allow for a 7K7 plate resistor and implemented a second FET for channel separation. Works well and sounds good.
Actually the mu of the D3A appears to be about the same as with the CCS which was a pleasant surprise to me.. 🙂
Actually the mu of the D3A appears to be about the same as with the CCS which was a pleasant surprise to me.. 🙂
Your oversight is the D3A; it is a flighty high gm tube; I got exactly the same oscillatory results using a lower gm 12BY7. (sensitive to loading). What is this telling one ? The SS CCS has an ft bandwidth equal or higher than the tube which has much parasitic wiring; and the SS CCS has a complex arrays of transistors/fets, ready to create an excellent oscillator from such. With such a wild ft; an early mobile phone is all one needs to create AM in an audio circuit.
If one is desperate to use a CCS with high gm stuff; then experiment using simple low ft, 2 transistor discretes and accept small caps across various parts. However, many BC transistor types have a gain product ft to 100's MHz. despite 'taming' the circuit, I found as you have; no difference in sound performance, so the technical effort doesn't warrant it.
richy
This is not an HF parasitic, but a DC mismatch caused by the anode current and cathode current of a triode being set to different values - or at least trying to set them to different values.
I regularly use DN2540 cascode CCS as high gm tubes (trioded D3a, E280F, E810F, C3g) anode load without any problem. In most cases - in fact always 🙂 - in the cathode I use LEDs.
Automatically put carbon comp resistors to FETs gate, tubes grids, and trioded tubes G2, G3 too. One case out of ten is required to use ferrit beads and/or anode stoppers.
I avoid to use these tubes as DC coupled/DC biased circuit from previous stage, but regularly DC coupled its to power tubes.
Automatically put carbon comp resistors to FETs gate, tubes grids, and trioded tubes G2, G3 too. One case out of ten is required to use ferrit beads and/or anode stoppers.
I avoid to use these tubes as DC coupled/DC biased circuit from previous stage, but regularly DC coupled its to power tubes.
Where did you get this design from?
[I like the idea of using pentodes just on the fact we don't have to deal with the issue of miller capacitance but is that an actual stand alone phono stage or a correction amplifier as you call it? I see the series zeners adding up to 327 volts or thereabouts but how much variation are you getting on the plate voltage where the ccs is plating loading that tube?
[I like the idea of using pentodes just on the fact we don't have to deal with the issue of miller capacitance but is that an actual stand alone phono stage or a correction amplifier as you call it? I see the series zeners adding up to 327 volts or thereabouts but how much variation are you getting on the plate voltage where the ccs is plating loading that tube?
- Status
- Not open for further replies.