MPP

[MiiB]

please put a drawing up.. i don't have the schematic here....so I have no reference to the R19

 

[hesener]

here you go. I am a little bit concerned with these inconsistencies, was under the impression that this R19 is 15k, and thats what I put in my paradise, and its working very well with it.

Let's take this schematic as the base to adress any changes going forward please
thanks

 

[hesener]

ohoh....... I reviewed my files, and here is a schematic from jan 19, that implements a few more changes that had been discussed in this trail
R19=15k
220Ohm "Neumann" resistor
but, R47 / R51 = 22 Ohm

hope that clarifies it.... I don't think I will be going through 50pages of postings to find out, lets take the working boards to measure and finetune......

 

[hesener]

... this time with the file....

 

[Joachim Gerhard]

I am blind tonight. Where is R19 ?

 

[Joachim Gerhard]

I found it. It drains the current in the current mirror of the buffer. R19 is 15kOhm in my schematic.

 

[brianco]

Joachim....

Your message box is full again!

Brian

 

[sampler]

What i would really like to see is your design in parallel symmetric ( if better or necessary ) with a balanced, differential input stage. J-Fet input or not. If we only use N-Channels this is not a component sourcing problem anyway. We can call that the Starless-Balanced and i would 100% support that design.

Believe me, I would like it too, I had to spend couple weekends improving my tt wiring, shielding and cabling to get rid of all that mains stuff down to -80dB, and with balanced that would be no problem at all. But then again, with balanced we have two opposing currents that can't be mirrored directly or it will cancel out, and I really love all this transimpedance mirrors feeding riaa network stuff. This concept of starless and following phonos got me intrigued in first place. So this has to go back to drawing board from scratch, or if someone has something to propose regarding balanced front end with 2 in phase currents output (preferably no LTP) I'm all yours. Meanwhile, more bjt sorting... more meditation

 

[Joachim Gerhard]

Here is a simple idea. The output would be differential of cause so we can not simply sum it for unbalanced. It would cancel. For unbalanced we need a high CMR INA like the Thats 1200. The Fets at the input should be cascoded. This shows only the principle.

 

[Joachim Gerhard]

From base T7 to emitter T5 there is a connection missing, sorry.

 

[Audiofanatic]

WOW Mr G.

Here is a simple idea. The output would be differential of cause so we can not simply sum it for unbalanced. It would cancel. For unbalanced we need a high CMR INA like the Thats 1200. The Fets at the input should be cascoded. This shows only the principle.

Cute !

Audiofanatic

 

[Joachim Gerhard]

Yes, it is very simple. I do not know if it would work exactly like this and we have to check offset issues but it is a start.

 

[Audiofanatic]

Current mirror

I ones did something like this before with the front as is but the output with onlt T11 and T14 and not a RIAA but a simple preamp.

I must look for it, I have it somewhere lying around.

I love current mirrors too


The output had CCS on each transistor leg (emitter ).


Offset was not a problem due to a cap !


Audiofanatic

 

[Joachim Gerhard]

I solved the problem with the balanced input and Samplers Mirror-Buffer. It is again so stupid simple that i had to smile. The minus pole of the cartridge simple goes into the sources. The differential INA at the output is now obsolete.
The node between the sources can even have a DC shift against ground because there will not flow current into cartridge. The gates of the Fet isolate that current, so this node can be trimmed for minimum distortion. With the resistor values given ( 5 and 2 Ohm ) and the N-Channels running on 7mA each plus the P-Channels running on 8mA each the offset null and the distortion null fall aprox. in one place.

 

[Joachim Gerhard]

I would of cause cascode the Fets. That can be arranged as a driven cascode too.

 

[Joachim Gerhard]

Here is the input gain cell.

 

[RCruz]

Now you are surfing again... cool

 

[Jacques Merde]

The node between the sources can even have a DC shift against ground because there will not flow current into cartridge. The gates of the Fet isolate that current...

If the node between the FET sources is not at ground potential, how do you prevent current flow through R12?

 

[Audiofanatic]

Perfect !

I have to admit it's a different animal !
There is something about it that I don't quite seem to understand, but I'll be back !
I like it.



Audiofanatic

 

[Joachim Gerhard]

Jacques, you are right. R12 should be high impedance or the source node trimmed.