Salas, is it just me, or the schematic of the positive regulator has a mistake? Connection of J2 and Q4 seems quite suspicious to me 😎
Its good now, I have seen it some hours ago and restored it. It was just a glitch in the presentation drawing, the boards are correct.
Thank you, Salas! Later today I plan on reworking some of my BiB1.1 boards to V1.3 and compare it with the "original".
Especially for Soekris which asks for around +/- 10V rails, the regulator is in a good range for its M2 output Mosfets to have substantial voltage across for lower parasitic capacitance and there is good voltage space for all its active components. Thus the response and phase margin are at a more optimum state.
As it squeezes down to 5V applications all those voltages go down to just decent enough. That's why I noted better put less than 7.5mA IDSS J3 if for 5V output use. Not for it to have demanding for voltage space Vgs (off). From 6V upwards setting not even that suggestion matters. I have checked on the bench that all the parameters from its most squeezed 5V setting up to 40V are adequately stable and that is performing well through such a remarkably extensive setting range though.
Salas, what is the reason of having two LEDs? With my BiB regulators I ended up using only a single zener diode*. Sonically I liked it better than LED + pot.
_______
*) I stay away form 8V2 ones!
_______
*) I stay away form 8V2 ones!
Needs two LEDS drop to keep the Q3's base above its collector level when at bottom range Vout. Non temperature compensated Zeners are substantial pink noise generators.
Here is a quick and dirty assembly on one of my "standard" BiB boards. A combination of red and green LEDs and no pot gives around 4.9V.
This will supply the USB side of JLSounds XMOS transport for my AD1865 DAC. It has absolutely the same type of components as the "standard" BiB I use in there. This is to ensure proper listening tests.
After some burn-in, listening sessions are in order
This will supply the USB side of JLSounds XMOS transport for my AD1865 DAC. It has absolutely the same type of components as the "standard" BiB I use in there. This is to ensure proper listening tests.
After some burn-in, listening sessions are in order
Attachments
Precarious experiment to mod another circuit & board towards the 1.3. It has high OLG, cascode, and CFP. Those are sensitive works. It needs some things be done exact and some parts to be exact. I see four wire mode in your pic. On Kelvin wiring it will be a pig to be stable if ever. If this experiment will be easily stable on Kelvin it means it behaves much differently. But good luck.
You should at least verify its basically stable without any AC gremlins on the rail by using the scope. Post#1 has a 1us/2mV 20MHZ limited picture I measured at the Ultrabib's +/- rails that can serve for your reference.
George, nice. This shunt reg does not need high spare current to play well as you saw. But calculate load pk with your lowest impedance headphones for your DCG3's both channels to make sure that your common PSU's current limit setting won't clip at worst case demand. There is a guidance post I made about its consumption details named "Few words about power" #79 in the DCG3's thread: Salas DCG3 preamp (line & headphone) - Page 8 - diyAudio
As I understand, I need to up the current on the positive rail a bit for my 62 Ohm cans. But again only this is for the worst case scenario, as I really do not push them very hard. Not hard to do though, just solder another resistor from the top.
I did. No gremlins or other mythical creatures on the scope when powering the XMOS board. Funny thing was I got oscillation "packages" when loaded with dummy load consisting of 100R || 100uF PanaFC 😕
Hmm, thank you, Salas. Probably the 3D rework of existing BiB pcb is playing tricks on me. What is C3 you have on board?