MPP

[Benedetto]

The 1N829 sadly costs a litte more.

 

[MiiB]

I agree....will post it tomorrow...then we have a chance to sleep on it

Joachim will you do a perfbofeard..or is it tooo complex for that..??

According to the nice sweedish work i belive theres are more options to tha zener...maybe 3 of theese could do a good figure if run at 5 mA

BCX55/C3V9 (0.5W 3.9V):
#1 @ 1mA: 1.7 1.8 1.7 1.6 1.6 uV
#1 @ 5mA: 1.1 1.1 1.1 1.2 1.1 uV
#1 @ 20mA: 0.94 9.76 0.84 0.79 0.77 uV
#2@1mA: 1.6 1.6 1.6 1.5 1.6 uV
#2@5mA: 1.2 1.1 1.1 1.1 1.1 uV
#2 @ 20mA: 0.97 0.78 0.86 0.76 0.77 uV

 

[sampler]

Guys, don't loose your sleep over that reference. I think even LM329 decoupled for low FC would do the job. At least it has well documented noise over BW in datasheet, so 5uV wideband is realistic with right decoupling.

 

[Joachim Gerhard]

I can build that, no problem. The PCB can be designed parallel . When i find problems i can report.

 

[FdW]

This is what I have. There seems to be 163MHz 1mV, but with the 'naked' circuit on the table I do think it is environmental. It is less than 1mV, right on the limmit of the scope.

 

[MiiB]

Wow....amzing...

 

[Onvinyl]

Frans, what is the div/V?

Rüdiger

 

[FdW]

Frans, what is the div/V?

Rüdiger

2mV / Div

 

[hesener]

I can build that, no problem. The PCB can be designed parallel . When i find problems i can report.

cool - this is progressing really fast. have time monday night to start working....

 

[MiiB]

This is where I'am Now..Freeze...

Think that PCB and Perfboard can comence...

Wrong designation on two helper transistors, all should be BC327-337

By the way, that buffer looks awesome...don't know what to call it.. driven diamond cascode..??

michael

 

[RCruz]

why R15 set to 5k ?

 

[Joachim Gerhard]

R15, R16 should be the same. Magic distortion cancelation ?

 

[MiiB]

my mistake should be 2.2K

 

[MiiB]

Updated... Think most of the ooops's are taken out...

michael

 

[FdW]

R12/13 100u not 1, that is 1F?

 

[MiiB]

hmmm that is what happens when you simulate deep in the circuit, you have to set the caps high, or completely remove them, Will set them back to right value later today....but where it say 1 its supposed to be 10mF

 

[FdW]

There is still a small improvement to be made. R17/R29 can be split in two (2 x 220) and from the center a capacitor could be connected to ground (470/1000uF). I know the power supply is very good, and this is not ‘needed’, but still…

 

[FdW]

And have a look at C3/4/8/9

 

[MiiB]

Will add that...and post new schematic later toady.... to be clear I suppose you mean R27 and R 29.. True that takes even more noise out, and reduces the resistor current noise at the same time.. good thinking..! (again)

The C-3-4-8-9 is supposed to be 10mF and 3.3 Uf for decoupling, though the decoupling is not needed on a high impedance current mirror, I can still hear it, and if at Mundorf tinfoil cap is used it add's a pleasant calm warmth to the sound. ( just one of those things you can hear, but not explain)

 

[FdW]

Ooops... 27/29 yes