L-Adapter

Maybe an insulation issue? Does it work if you will lift the PCB off the chassis? Don't worry about developing heat without sinks when unloaded.

yes that did it. It now measures perfect when everything is detached from the chassis. Now I have to find the cause. Q2 has a isolation pad. the rectifying diodes did not get and isolation pas as a I assumed that these do not need isolation

thx
 
diyAudio Chief Moderator
Joined 2002
Paid Member
Or a solder joint stresses when the build gets fastened down? Rework the bridge diodes and Q2 joints when assembled to the chassis so they may reposition pins a little in the pads holes.
To self relieve possible mechanical tensions. It's usually recommended in power amps and PSU builds to last solder chassis or sink loaded parts pins. After fixed in place for that reason.
 
Maybe an insulation issue? Does it work if you will lift the PCB off the chassis? Don't worry about developing heat without sinks when unloaded.

Or a solder joint stresses when the build gets fastened down? Rework the bridge diodes and Q2 joints when assembled to the chassis so they may reposition pins a little in the pads holes.
To self relieve possible mechanical tensions. It's usually recommended in power amps and PSU builds to last solder chassis or sink loaded parts pins. After fixed in place for that reason.

That last thing is exactly what I did. First populated the pcb, then mounted all chassis mounted parts. Fixed the pcb's on the standoffs and as a last step I soldered the rectifier diodes and Q2 to avoid solder joint stresses when tensioning the screws for the chassis mounted parts.
 
I used a plastic shoulder washer on Q2 and mounted everything again on the chassis, checking voltages after every connection. Now all is fine. I think it was the srew for Q2 just touching the edge of the metal.

I am using it now on the RME ADI2. There is a clear difference. Tighter bass, more details and above all more rest. Let's see what it does in a few days.

In this version I used 2 Mundorf 10000uf/40V caps which i already had. I also have 6 pieces of Vishay/BC-components 15000uf/40V leftovers from a project.
Could i use these to build the next PSU? That will be a 19V version for a Intel NUC as Roon end point
 
I have done a prototype board and fired it up. Seems to work but with less voltage reduction. I have used some leds i have at home and they seems not to work out any good. Any suggestions for replacements? As you can see i need common single leds. I tried to measure them for voltage drop at 5mA but did not come out good in reality.
 

Attachments

  • Ladapter.JPG
    Ladapter.JPG
    241.7 KB · Views: 266
I bought some Cisco Meraki ms220 8 port switches very cheap. I want to experiment with a LPSU upgrade and maybe later a OCXO clock. I will loose the original SMPS, also the one for the POE. I only need max .5A for the switch and 1 A for the clock. Would the L-adapter be a good choice or rather go to UltraBIB’s?

Regards,

Harold
 
diyAudio Chief Moderator
Joined 2002
Paid Member
You should add together the peak DC current consumption of that system's components first, and compare the transformer's current spec for both secondaries wired in parallel. Making sure it's sufficient. Keep in mind that due to full bridge rectification, the DC current spec is about 60% of the transformer's secondary side AC current spec.
 
diyAudio Chief Moderator
Joined 2002
Paid Member
A 50VA 9V toroidal example you can find full specs for, is Antek AS-0509. Other good makes should be of comparable spec.

Computers run on low average consumption compared to their peak spec when not pushed in peripheral configuration and heavy tasks. It's mainly about booting correctly in audio use. It's that phase when they pull most.

In such mini computer + dac systems 50VA is usually sufficient but always check the parameters of your own configuration and possible modes of use.
 
diyAudio Chief Moderator
Joined 2002
Paid Member
L-Adapter can handle up to 100W output if with very good external sinking. Q2 dissipation in Watts is the voltage across its collector to emitter (Vce) multiplied by the output current it passes. Vce can be easily probed between the pins of D11 protection diode that spans Q2. Multimeter's red probe to cathode pin, black probe to anode pin.