Thanks for the Ultra BiB Salas.
I have build one for a diy AD1865 dac.
Actually 4
3x +5v and one -5v
Sounds is very good.
On the scope under load it showed some RF thing on the output,nothing big i could hear.
I tamed it with a little coil of 220uH in series with the output,and to be sure i put a ferrite beat on every gate of the big Fet,s.
Everything is running cool not overheating.
The scope pictures are from before and after.
I have build one for a diy AD1865 dac.
Actually 4
3x +5v and one -5v
Sounds is very good.
On the scope under load it showed some RF thing on the output,nothing big i could hear.
I tamed it with a little coil of 220uH in series with the output,and to be sure i put a ferrite beat on every gate of the big Fet,s.
Everything is running cool not overheating.
The scope pictures are from before and after.
Attachments
Hmm the layout is fairly critical for this circuit and there is no ground plane in your utilization so in the end you did not do too bad with measures against RFI 
P.S.
How much spare current you allow? The CCSmA-LoadmA difference I mean i.e. shunt current spare mA burning on the output sink.
P.S.
How much spare current you allow? The CCSmA-LoadmA difference I mean i.e. shunt current spare mA burning on the output sink.
Salas, do you know what the max gauge wire the 1.3 UltraBibs will accept via direct soldering?
To calculate whether heatsinks will be adequate, I've used the below example assuming 15vout @ 300 CC powered by a transformer with a 15v secondary.
Is the only way to know voltage drop across M1 to measure it or can this be determined ahead of time given transformer specs? Also, let's assume Celcius above ambient is 3.9 degrees for M1 and M2. Assuming they will be connected to the same heatsink, how do you calculate the temperature rise of the heatsink - do you simpluy add them?
I've used the below heatsink as an example (.67 c/w) but given 300 mA at 15v this seems like overkill if temperature rise is only 8 degrees above ambient.
40mm Heatsinks – diyAudio Store
M2
Watts dissipated for M2 will be Vout*Shunt current i.e. 18V*0.125A=2.25W
Vout 15
CC 0.3
Watt 4.5
C/W Heatsink 0.67
Temp 3.015
inefficiency X 1.3
C above ambient 3.9195
M1
M1 the dissipation comes from (35-18V)*CC
Secondary v 15
V across C1 ?
C above ambient #VALUE!
To calculate whether heatsinks will be adequate, I've used the below example assuming 15vout @ 300 CC powered by a transformer with a 15v secondary.
Is the only way to know voltage drop across M1 to measure it or can this be determined ahead of time given transformer specs? Also, let's assume Celcius above ambient is 3.9 degrees for M1 and M2. Assuming they will be connected to the same heatsink, how do you calculate the temperature rise of the heatsink - do you simpluy add them?
I've used the below heatsink as an example (.67 c/w) but given 300 mA at 15v this seems like overkill if temperature rise is only 8 degrees above ambient.
40mm Heatsinks – diyAudio Store
M2
Watts dissipated for M2 will be Vout*Shunt current i.e. 18V*0.125A=2.25W
Vout 15
CC 0.3
Watt 4.5
C/W Heatsink 0.67
Temp 3.015
inefficiency X 1.3
C above ambient 3.9195
M1
M1 the dissipation comes from (35-18V)*CC
Secondary v 15
V across C1 ?
C above ambient #VALUE!
Correct remarks, the layout of this reg has some points of caution you could stumble on when spinning your home board, I went through a couple of prototypes to optimize the GB board too.
122mA is good enough, but at what output voltage? Because dissipation in Watt is Vout*SpareCurrent. Small voltage I assume because the sinks in your photos are small.
@stew
M1 dissipation is VDSM1*CCS. To calculate M1 & M2 heat effect on a common sink simply add their dissipation to be in the ballpark. Also see: Calculating heat sinks
M1 dissipation is VDSM1*CCS. To calculate M1 & M2 heat effect on a common sink simply add their dissipation to be in the ballpark. Also see: Calculating heat sinks
I wonder why is possible that reservoir capacitor quality affects output of shunt regulator. I would expect that above decent C1 quality, there should be no audible or measurable difference. This capacitor is loaded with CCS. There should be almost no load variation. Does it’s quality (leakage, ESR..) affect how clean is constant current? Or, does output load variation, despite shunt, modulate constant current to some extent?
Most likely, there are other reasons I don’t have clue about.
I have to build two +-15V power supplies. UBIB boards are on the way from USA. Hopefully they will arrive despite this Corona turmoil.
Now is time to pick a good components. I can follow capacitor remarks from first pages (Nichicon Gold Tune etc.) but I like to understand why am I doing something.
Now is time to pick a good components. I can follow capacitor remarks from first pages (Nichicon Gold Tune etc.) but I like to understand why am I doing something.
