His Master's Noise: A Thoroughly Modern Tube Phono Preamp

For boards, contact Jack at www.tech-diy.com . I'm rebuilding the project on one of the boards he designed and I suspect it will look a lot better than my current point-to-point version. (Disclaimer: no commercial interest in board sales, but I did get a free board)

I'm finishing up an article on a tone control line amp- once that's done, I'll write up my MM amp. Until then, you can turn this into a very good MM preamp if your cartridge is tolerant of capacitance (e.g., Grado): just eliminate the input transformer and change the first stage grid resistor from 6k8 to 47k. If your cartridge needs a low input C (e.g., the Audio-Technica 150MLX that I use), my upcoming MM-only design will take care of that.
 
Thanks. Drew. No, the MM version will look quite a bit different. I made the mistake of changing jobs and moving to Chicago, so that delayed things a bit. I managed, though, to get some of the basic topology worked out and breadboarded before I relocated, so once my lab is unpacked and set up, things should move pretty quickly (famous last words).
 
Hey Sy,

I know how that feels, I moved 3 years ago and still trying to sort a listening post!

This is the cart I'm using, the music maker Classic LE:
Variable-reluctance stereo phono cartridge with line-contact stylus and silver coils.
Output: 4.0 mV output
Loading: 47K ohms (not capacitance sensitive.)
Tracking Force: 1.6 grams.
VTA/SRA Alignment: Front face of cartridge perpendicular to record surface (viewed from the side).

Any thoughts?

Thanks,

Drew.
 
Hey Sy,

Having an idea to double the stages and run it balanced then perhaps into a circlotron 6SN7 stage and not need the output cap.

Capacitance of course it is irrelevant, it won't do anything to phase......not. Not that I completely understand all this yet but I'm sure it come to me.

Thanks Sy.
 
Hi SY, just completed your phono amp using DIY-tech boards! Was able to pack all together in single chassis including the power transformers; with shorted inputs it is perfectly quiet; a bit of hum when connected to the cartridge; have to try twisting the wires from input RCA to the board.

Really impressed with the sound, thanks for making this project public!
As well, thanks to Jack for making the PC boards!

I am using SAE 1000E HOMC cartridge, no input transformer, just 47K resistor on input.
SAE spec is: 2.5mV output, Load impedance: 47K (non-critical), Load capacitance: non critical; it was apparently made by Coral.

Two things I found strange:
1) From 4 D3a tubes, with CCS adjusted at 20mA (there is 19.8V drop across 1K plate resistor) I got huge plate voltage variations between the tubes: 140V, 163V, 179V and 199V! Did you maybe get simmilar experience? Should I buy some more D3a’s to be able to find two that are more similar?

2) My 160B+ is spot on, at 160.5V, but second B+ instead of 260V is 210V; it is all the same on both channels; my unregulated HV is at 339V. Do you think I should tweak the 260V regulator to get closer to 260V or 210 should be fine?
 
Thanks, will get my B+ up, closer to 260, as you sugest. I checked all resistors in 260V regulator and they are alll correct; really strange I am getting only 210, especially since 160 regulator is spot on.
Would you mind sugesting resistor values to adjust to get up from 210; I guess R25 and R26 would set the output voltage, but please, if you could help a bit with right way to calculate actual values.
 
You might want to look closely at the error amp transistor and the voltage reference. Could you measure the voltages at EBC of both transistors on the oddball channel?

The circuit functions as a basic feedback amp. The reference diode is the noninverting input, the two resistors from output to ground form the feedback voltage divider. I don't remember their numbers offhand, but let's say that the upper one is R1 and the lower one is R2. Vout(R2/(R1 + R2)) = Vref + 0.7V (the Vbe drop of the error amp).
 
OK, thanks, so it makes sense:

On 160V regulator I got 160V(8.2K/(160K+8.2K))=Vref+.7V; Vref comes to 7.1V; very close to nominal 6.9V for LM329.

On 160V regulator I got now 210V(8.2K/(220K+8.2K))=Vref+.7V; Vref comes to 6.84V; very close to nominal 6.9V for LM329.

In order to get 260 I could adjust R2 to 6.6K to get 260V(6.6K/(220K+6.6K))=Vref+.7V; V ref comes to 6.87V; very close to nominal 6.9V for LM329.

I will parallel my 8.2K with 51K I got here; this should do the trick!