Best low noise regulator?

[Salas]

Originally posted by Marinos
Thanks,
I think that tomorrow I'll be listening to them.

Yes, listen and tell me first impressions Vs batteries you used to have. They will sound better after 48hours of use, I know from experience. Also by then you will be able to see if the Zeners changed a little, and also able to tell if the OPAs will appreciate the matching of Zeners.

[Marinos]

Hi salas,

Listening impressions are very good. For me the main sonic advantage of the batteries was the homogeneous presentation they have. The sound with the shunts has the same homogeneity, plus much more detail, plus explosive dynamics in a musical manner. Enough with magazine like words.... A very worthwhile experiment. Shunts stay..

As you can imagine, for the time being, my pre is a rats nest. In order to close the project I would like your opinion on these :
- Is splitting the PSU for each channel (dual mono) worth the effort ?
- Each channel uses 2 OPA627s. Should I experiment with higher current ?

Thank you salas.

[Salas]

I am happy that you like the particular shunts. There is much voicing put in to them besides tech considerations. As you see, batteries are not the holy grail once believed to be.

-Yes, dual mono is worthy. You will notice increased spatial differentiation and more concrete overall sound.

-No, 90mA running CCS current is many times what 2 OPA627s can ever dream of asking for.

Did you ever try them forced class A output stage via resistor from neg to out by the way? I had devised an OPA627-37 line pre in 1991 (had to make an expensive special order for the OPAs when I eyed them in their then fresh instrumentation Difet brochure), and I remember that forcing class A in their output was worthwhile.

*The shunts are going to improve their subjective qualities after enough play time.

**They are going to sound a bit smoother when the BJT's Idss collector load is gonna be 2SK170BL, I saw that Spencer has already mailed them to you.

***Its nice to know that the 2N5459s did the job nicely too.

[Marinos]

A question I had from the beginning was if you left leds without a capacitor bypass intentionally. But it's never late for questions like this ..

Thank you for the info about Spencer .... (I didn't notice it) and last but not least thank you again for your support.

[Salas]

You are welcome.

Yes I left it out intentionally, because with actively loaded BJT driver, the performance is adequate enough so not to be tempted to add bootstrap via cap that can introduce some side effects too, since no elcap is perfect. We run the LEDs with enough current on individual CCS so to be very low noise anyway. A simpler simplistic shunt then.

Its nice to know that the shunts show their benefits in high PSRR op amp applications too.

Happy Easter, enjoy your new sound.

[Marinos]

Happy Easter to you too.

[Marinos]

Hi salas,

I'm planning the dual mono version of shunts and I need your opinion on these:

- Because mounting 4 shunts on the existing chassis is not practical, I would like to use some board level heatsinks (see photo) that I already have. These are for TO220 packages, 35mm wide, 38 mm tall, Rth 11 C/W aprox. What do you think ? Is it adequate for Q4 (Q1 will be left bare) ?

- Fanos restocked GRs so I can use GRs for Q5 and Q6 as well. Should I use them or BLs are a better option for that positions ?

- If I'm interpreting things right there are 4 parts on the schematic. The current setting (R1,Leds,Q1,Q5,R4), the voltage setting (D1,C2,Q6,R5), the regulator bias (Q2,Q3,R3), the current regulator (Q4,R6,C4). The 2 last parts should be in close proximity (layout wise) ?

[Salas]

Hello,

1. For your current and voltage they seem more than OK for Q4. Q1 can be left bare in your situation, but a silly small sink like those for 7812 or 317 that just sit on air, supported by the semiconductor screw only, would make you feel better for long term reliability, given the Greek Summer coming with some heatwaves to boot.

2. BL for Q5, high Idss ones, around 10mA and over. Get a few GRs though so to stock Q6 for possible future shunts like those, because GR has the lowest pinch off under Vbe and that helps the under Zener CCS performance.

3. It better be compact as a whole.

[Marinos]

It seems that we exchanged easter greetings too early...

Can you comment on the brakedown of the schematic ? This will help me a lot..

What about a version of your shunts for lamp roasting ?

[Salas]

LEDs, R1, and Q1's Vgs determine the constant current. Q1 creates it. Q5 locks the current through the LEDs for even better voltage reference regarding steadiness and noise. At near 10mA they are doing better according to LED master SY.

Zener is the Vref for Vout. Q6 locks it current wise to its benefit.

Q2 actively loads Q3 for good gm*LoadImpedance regulation loop.

Q4 does the current duty. C4 terminates closely.

I am just putting together a 16A one for the lamb motor. Beer closed circuit sink cooling featured. I am having problems determining the fluid's quality. I am leaning towards Kaiser at the moment.