rjm 2 tube phonostage

[hacknet]

i was studying the circuit above and i was thinking of replacing the 2 30k plate load resistors to pnp ring of two constant current sources.

i remember reading before that the output impedance of the stage is a part of the riaa calaculations but i did a simulation on LTcad and it seems the riaa curve doesnt deviate when replaced with an ideal current source.

do i have to tweak the riaa values if i were to replace the Rp with a ccs ?

thanks a million guys!!

[soeren]

I can not directly answer your last question, although I think that it will not have a big influence. Could'nt you just try it and see with at test record if it works? And adjust if there are problems

I have some other comments:

First you should get rid of the cathode capacitors C5 and C6 as they have a very detoriating influence on the audio quality. If you are just cutting them off, then you will loose some gain and introduce a little amount of NFB but very local and benign, and it will have positive influence on the subjective audio quality. If you have gain enough, that is the simplest solution, if not you should either introduce at third stage or obtain the negative grid voltage in some other way, there are many solutions from batteries to a separate PS.

Second there should be a much better separation between stage 1 and 2, in your circuit they are 'talking' not to say 'shouting' to each other in opposite phase via the resistor R12 which is far too small. This is very audible. Most people think that the PS should basically protect the circuit from the ripple from the AC, but it is equally important to protect the individualt stages from each other.

My solution to the problems would be to increase the B voltage dramatically, say 300V and then increase R4 and R9 accordingly to obtain the right current and replace R12 with a series of RC couplings where there should be good high frequency capacitors in parallel with the electrolytics. You may not forget that there is a tremendous amount of treble in the first stage, and that is why a RIAA in my mind is the most difficult amplifier to design. You see that this approach also gives you something near to a constant current source.

If you have access to a scope, then it is very instructive to see what is going on around R12 or its replacement, it can be quite chocking. High voltages, take care of both yourself and the scope!

The quality of the signal capacitors is extremely important, and that is why you have to get rid of C5 and C6 as there to my knowledge does not exist a realistic capacitor of 100µF.

Take care of the high tensions,
Søren

[hacknet]

hey! i was thinking of doing just that..

i was thinking of getting rid of the cathode capacitors and adding a 5687 to compensate for the gain lost and to bring the gain up alittle.

im also thinking of individually regulated stages and channels. this should be fun.

[SY]

Removing the cathode degeneration will upset the RIAA, so be prepared to trim. It will also raise the source resistance of the second triode.

[hacknet]

yeah, thats what i was thinking of.. in that case, replacing the plate resistor with a ccs will change the zout even more, i think there is alot of riaa tweaking ahead of me..

[rjm]

This circuit isn't very original, but the values I chose were chosen for a reason.

If you use black gates for C5 and C6 there is no adverse affect on the sound, and you have the benefit of higher gain and low output impedance. Plus the RIAA values I derived from the simulation will be correct...

R12 was chosen so that the ripple into the first stage is cancelled by the ripple introduced into the second. Its also a reasonably low value such that the power supply impedance into the first stage at low frequencies is low. I'd rather keep a low power supply impedance throughout to reduce crosstalk or intermodulation between stages, rather than trying to brickwall each stage using huge, slow RC networks.

Replacing the plate resistors with CCS of one sort or another is a reasonable upgrade path except you'll have to increase R5 to compensate for the lower output impedance of V1.

I'd be the first to admit that this phono stage isn't really high performance. I enjoyed how it sounded very much though. Why not give it a shot and see how you like it first before embarking on extensive modification of the original.

/rjm

[hacknet]

yeah.. thats a good idea. its a fun project to keep me occupied.!