There is high feedback in the sense of the error amplifier (Q2 Q3 cascode) automatically reacting at the output node.
Why I asked, I wanted to use your PSU for high precision master clocks in DAC, HTPC and etc. In this case is importan general noise and phase noise near 1Hz and lower.
For example, PVD-Audio for their USB PCIe card stoped on variant with low-noise stabilizers without loop feedback by voltage.
How you think, is your PSU suitable for MCLK with very low level of phase noise, especially on 1Hz ?
For example, PVD-Audio for their USB PCIe card stoped on variant with low-noise stabilizers without loop feedback by voltage.
How you think, is your PSU suitable for MCLK with very low level of phase noise, especially on 1Hz ?
Yes if you put 1000uF C2. See third picture in post#1 noise sim chart. Grey bottom curve. Only 4nVrtHz @ 1Hz rise.
You listen and decide. Noisewise when the 1.3 has 1000uF C2 they are comparable. But they work on different principles and open loop gains.
Very interesting will be results of such comparison.
On ebay sells PSU for clocks from Sallarz
R-Uhr High Precision Low Jitter Reference Clock, ultra low noise Regulator | eBay
It is based on voltage reference
who knows is it good good schematics or not ....
Last edited:
I recommend these.
Kingbright WP113IDT
the vf does not wander much between samples.
I sell them.
J113 can be used within range for current production. Salas has posted this prior
J113 can be used within range for current production. Salas has posted this prior
Thanks Teabag. I got my hands on a batch of PF5102 and problem solved. I will measure them for Idss. Is it correct you need Idss lower than 7,5mA if you will use bibs @ 5V. Otherwise you are fine with any PF5102?
Hi Salas
Run into a problem with one of my neg regs. It outputs roughly 8 v less than it should compared to a good neg reg.
Voltages across R5 and R6 on the bad one measure roughly half that of voltages measured on the good one. Voltages across all the other resistors measure about the same between the bad and good reg. Also the LEDs appear slightly dimmer compared to a good one. Please advice.
Thanks. Nash
Run into a problem with one of my neg regs. It outputs roughly 8 v less than it should compared to a good neg reg.
Voltages across R5 and R6 on the bad one measure roughly half that of voltages measured on the good one. Voltages across all the other resistors measure about the same between the bad and good reg. Also the LEDs appear slightly dimmer compared to a good one. Please advice.
Thanks. Nash
I've replaced the above 12v trafo with a 15v. One extreme to another! This one gives 17v ac and so 22v dc rectified into the Ubib reg. Therefore I have almost 10v drop over the system. Low current though and the jeatsinks are 30c after an hour....powering 2 x AD1862 and opamps. No music signal running through them though.
Hopefully this is a suitable setup now and better than the precious 4v drop.
Q2 was bad.
On a separate note is it safe to use acetone to clean the flux residue on these boards?
Thanks for your quick help.
nash
Yes indeed, ten Volt drop significantly lessens M1's parasitic capacitance vs four Volt across. Makes a better performing CCS out of it. As long as dissipation isn't forbidding the higher voltage transformer luxury.
