Geller Labs SVR, Lars Walenius black edition with 587LQ

[HighVoltage]

These are 5 Geller Labs SVR boards as a special Black Lars Walenius Edition,
and gold plated PCB by our late friend LARS.
 
I was lucky to get these boards and was more lucky to get original LQ ceramic chips.
4 x 587LQ ceramic chips for 10V output
1 x 586LQ ceramic chip for 5V output

Now I need to get some good cases and then will take some long term measurements.
Then the next step will be to measure the TC and apply some NTC to make these boards
in to SVR-T boards with temperature compensation, that was developed by LARS for
Geller Labs.

The original Gellerlabs website is still up but Joe Geller is no longer selling these parts:
http://www.gellerlabs.com/Voltage%20References.htm



 

[Andreas]

Hello,

why do you use sockets?
in my experience they only make trouble (at least after some years).

you should populate the 100nF capacitor at the output which is in the BOM but not on the PCB-layout
(but there are enough holes at the output left).

with best regards

Andreas

[HighVoltage]

Hello Andreas

Sockets, because I was not sure if the 587LQ would work properly and I might had to exchange a few.
After a few days running good, I am planning to remove the sockets

100nF on the output is a great idea.
Thank you, I will add them.

[Andreas]

Hello,

not my idea. It was a addition on the web site (optional) after Lars found out that it is necessary to reduce EMI errors.

with best regards

Andreas

[HighVoltage]

Thanks Andreas, I will implement the suggestion.

Now, about two hours later I get these results:

One SDEV is higher than the rest
1: 2.8uV
2: 4.6 uV
3: 2.5 uV
4: 1.1 uV
5: 3 uV

But what surprises me are the spikes every 30min or so to be seen on the 34470A and the 34461A,
both at 100 NPLC. I don't think I have outside disturbances. But then, the boards are not shielded
and no capacitor on the output. But still very strange to happen in such intervals.
 

[Andreas]

Hello,

It seems to be some common cause effect.
-> either the power supply
-> or some mains disturbance which is coupled over the power supply or one of the instruments.

I would check with battery supply and/or the output voltage of the power supply.
The question is how good is the PSRR to achieve 1 ppm jumps?

with best regards

Andreas

[HighVoltage]

almost a day and 16.000 readings later:

StandardDiviation:
1: 3.16 uV
2: 13.6 uV
3: 2.88 uV
4: 4.7 uV
5: 6 uV

It seems the #2 has a problem with the 13.6 uV SDEV
But in general they seem ok, with around 3 uV SDEV without shielding and box around them.

I will power them now with a battery only and see is the spike will disappear.