Not changing for M1, but only for M2. So IRF630 M2 test only in your positive example.
Simulation shows a little greater (4 times more than IRL530) output impedance (120uOhm up to 10kHz, 365uOhm at 100kHz).
BTW with IRL3803 the Rout below 10uOhm up to 10kHz, and 60uOhm at 100kHz.
p.s. +5V, 0.5A
Last edited:
I am expecting higher with 630 too, that is why I have the 530 as standard, although my model shows using the 530 & 630 closer than 4x but with more phase margin. Layout and cables shall mostly swamp such μΩ dif worries though. Sims ah... alright, but we also have Lemon who is the real food testing crocodile around here 
Hi,
I'm going to use ultraBIB for a DAC with on board regulators and I can't decide about the proper implementation: bypass on board regs and supply the DAC directly from BIB or use it as pre-regulator? What would you suggest?
I was also thinking of making the extra positive board from the group buy as a linear PSU for my Raspberry Pi3. Does these mosfets IRF9640/640 sufficient to give 5.1V 2A? Do I need to change anything else apart from the mosfets as big heat sinks I already have good ones, so I am covered in that part.
Also I have couple of spare positive boards from BiB 1.1 version, so can I use these boards with the above mosfets for 5.1V @ 2A?
Thanks
Those MOSFETS for M1&M2 and a 0.27R 5W R1 should make and handle 2.2A CC in the 1.3. What else to change depends on load dynamics. If the Pi3 allows for high wasted (spare) current at times like 0.5A on M2 then R9 should be higher value 1k. If it allows for 0.75A wasted at times then 1.5k. Heat and phase margin in super shunts go worst during high ratio of pk/average load consumption cycles by computers. Because a super shunt (old name of CCS biased shunt regs) must burn itself what's not put out to the load. About using the 1.1 for that I haven't analyzed but possibly.
No bad feelings here, but I always preferred BiB over the Ref-D in both analogue and digital circuits. These are my subjective ¢5.
Simulation wise IRF9640/IRF640 are better choice for the shunt transistor. I didn't go for a practical comparison, tho
The priority was to first make a plug-in upgrade to the many years popular BiB 1.1 so to can advance the already built systems around it and skip its increasingly faked (or increasingly rare and expensive when real) Toshiba JFETS issue. Also to much simplify the pre-configuration. But I don't exclude I will design any additional physical formats for the 1.3 successor if it will nowadays prove as much popular as the 1.1 was.