MC Input Stage

[Onvinyl]

Hi,
in the urge to build something useful with parts I already have in my bin, I sketched a phono input stage tonight.
I build the attached circuit, input stage is biased with 6mA (IDSS 7.5), the FC-Fets are around 11mA, the most juicy I have.
I can't do distortion measurements, a square looks fine up to 70kHz, gain is plenty.

The working conditions for the FET's need refinement (wish I had some more V-grade parts)

PSU Voltage seems a bit low as well.

The input offset is around 4mV, which is a bit high for my liking, so no listening tests so far (I even havent a second riaa stage handy...)

Should be no problem to change it for non-inv input, however.

My reasoning was:
-Push-Pull enhances linearity
-No Diffamp helps noise
-Since MC-Carts are low on voltage and relativly juicy on current, Folded Cascode might present mostly it's sonic advantages here.
- haven't seen this topo for a phonopre here.

So, what do you think, is this a worthwhile approach?

Rüdiger

[SY]

Won't R7 and R8 cause a loss of S/N since they're essentially in series between the cartridge and the FET inputs?

[Onvinyl]

Originally posted by SY
Won't R7 and R8 cause a loss of S/N since they're essentially in series between the cartridge and the FET inputs?

Hm, yes, that's true. I need them to bias the FET's, maybe I should look for lower IDSS in the first stage.
But then, 4mV Offset with Zout of 50R gives 0.08mA, which seems too high for a poor MC-Cart, doesn't it?


Rüdiger

[oshifis]

I built the same topology with (medium) power transistors that are constructed with many parallel sections, good for low noise. I used a capacitor in the position of R5 in your figure. That capacitor made a huge difference in sound, depending on the type and value. Worth playing with it.

[analog_sa]

Originally posted by Onvinyl
- haven't seen this topo for a phonopre here.


Rüdiger

If you apply the input to the gates it's pretty much identical to a Vendetta. And it solves the offset issue without having to match hundreds of fets.

[Onvinyl]

As far as I know, the vendetta is differentialvendetta

And yes, if I can't make it run, I have to use the gates. Since I have good results in the phonnoclone with the inv. input, and the famous hiraga le prepre is using emitter input, I thought this might work better (at least with a middle to low impedance Cart with flat Frequency over impedance plot. Just wish there were such plots...)
Rüdiger

[Onvinyl]

Originally posted by oshifis
I built the same topology with (medium) power transistors that are constructed with many parallel sections, good for low noise. I used a capacitor in the position of R5 in your figure. That capacitor made a huge difference in sound, depending on the type and value. Worth playing with it.

Yes, I probably will parallel some FET's in the front end later.

Do you mean you *replace* R5 with a cap? I can't see how the thing could be working then?

Rüdiger

[analog_sa]

Originally posted by Onvinyl
As far as I know, the vendetta is differential

Wrong.

[Onvinyl]

Hm, ok, at least I know now it will work cool!

Thanks,
Rüdiger

[Onvinyl]

is there a way to vary the models of the FET's in LTSpice?

I get substantial differences in sim and reality looking at DC-operating points.

Rüdiger