USSPA Build thread

Thanks to Fab to share pair of these preamp boards and even in this pandemic I could receive the boards in a months time. I had the matched original Toshiba 2sk2013/313 from the USSA-5 build as well as matched pairs of 2SJ74/170BL grade with 10.1/10.2mA IDSS. So decided to build the preamp to pair and got all the parts in 1% tolerance limits per the BoM and the excellent manual by Fab.

I have decided to power it using Salas BiB psu boards dual mono hence built one set initially to test keeping in mind the current requirements of the preamp so used 1R+0.33R (in series) current limiting resistor. The psu with a single amp is holding on well with no heat on the mosfets which is good. Used an rcore rated 0-15 1A*2 dual secondary and 2 independent for dual mono.

Yet to hear the preamp will do that over the weekend hopefully when I can make the psu ready and power it up in pure dual mono configuration. Right now the bias and offset is pretty stable even after an hour the bias stays close to 50mV and the offset around 0.3-0.5mV which is just awesome and I guess thats because of the closely matched transistors. I also did not see any heat on the heat sinks at all pretty cool compared to the ACP+ :)

Here are the construction pictures for reference.





 
Sure Do as of now the bias/offset is perfect stable without any movement even after an hour with couple of adjustments. Will do the second board and hear the amp in this test cabinet for now but need to wait for another rcore for dual mono as I goofed up with the ratings. The one above being tested is rated 0-15v 1A x 2 dual secondary but the other one is rated 0-15v 0.5A x 2 only and not sure even with dual mono and per Fab's manual the preamp consumes min of 1A and the rcore may not be sufficient to be used for one channel.

Thanks Harry, will keep my progress posted here and see how the listening goes.
 
Hi Manniraj
The board only consumes <100ma but the more you are away from maximum in a psu preamp the best it is.;).. but 500ma should be enough.

Fab

Thanks fab then that seems to be well timed. I think the confusion is because of the manual which states:

4 Power supply requirements (for 2 channels)
Well regulated Voltage: +/- 12v to +/- 15vdc (with heatsink shown). Use very good PSU here
since PSRR of CFA topology is not that good.
Current: >= 1A to be on the safe side.


Hence I got the 2A rated RCORE for single channel board :) But per your post above I guess even my 1A rated RCORE is also sufficient per channel.

Thanks
 
Hello Fab,

I finished up both the amp boards and using single Salas Ultra BiB board shorting the ground of both the polarity powered. I could set the bias to 50mV and the offset being around 0.00-0.5mV on both the channels. Ran for 2-3 hrs and readjusted the bias/offset. Now to listening I connected a Alps blue velvet 10k volume pot and wired initially using single core unshielded cables from the input to the pot and then to the amp to the output RCA jacks. All RCA jacks are isolated using the supplied bushes and I have tried both grounding the ground of psu via a Thermistor in series to the chassis as well as without any.
There is a white kind of noise which I can hear as I tested using my iPhone HPs as well as my Senneheiser HD598. I am unable to get rid of this noise even with multiple different cabling/wiring, keeping the Rcore outside the chassis, keeping the volume pot outside the chassis. The first few pics below is a small cabinet which I tried and it was a tight fit so I went back to my longer cabinet and the issue is the same.

When I connect this preamp to my USSA5 the noise is even more and increases with volume pot to some extent and gets thinner when the pot is at the maximum. Also another observation I see is that when the amp boards are powered and biased I see the voltage on the psu dropped to around +/-14.56vdc. But without load its giving perfect +/-15vdc and dropping is around 500mV.

i have also isolated all the 4 mounting points of the amp boards from the chassis as you can see in one of the pic by using a longer nylon spacer and keep the other 3 legs over the chassis. Used Canare interconnect cable for signal wiring which I usually use in all my preamps with zero hum/noise issues. This is the first time I am getting this kind of noise in the preamp section. Hence confused a bit as the power supply is very clean and use the same with my ACP+ which works absolutely hum free in similar kind of small cabinet dual mono. I feel this isn't ground loop but some other issue as it seems like white noise or something else unable to find the route cause.

Any suggestions to get rid of this noise which sometimes gives headache to me when listening over HPs. Also when listening to music the noise is there but because of the music its not audible enough but during some passages I could feel it is there :)






Thanks
 
Last edited:
One more update I changed the transformer to a 0-18v 1A *2 and I still get the noise and its a bit loud on my Troels + USSA5. I have also changed the pot to a Alps 20k and different sorts of grounding but not sure whats causing the issue.
Tried and changed the bias point to 60mV and the offset was close to 0 and no change to the noise.

My last suspect is the Salas clone psu but I have used these in my ACP+ without any noise issues.

Thanks
 
Last edited:

fab

Member
Joined 2004
Paid Member
Hi manniraj
You reported “... I see the voltage on the psu dropped to around +/-14.56vdc. But without load its giving perfect +/-15vdc and dropping is around 500mV....”

This is definitely wrong. There is current leaking into a wrong path somewhere. 50ma should have no voltage drop at all on a good psu.
Verify also R18 value if not open. You can short it also since it is to prevent ground loop.

“.... and changed the bias point to 60mV...”
60mV/ 4.75 ohms = 13ma but bias should be in the order of 50ma (or more).


Try only one channel powered and connected at a time and notice if different behaviour.

I had noticed your cabinet had already mounting post present, just make sure they do not touch the USSPA pcbs.

I sure have no noise issue and nobody reported such issues so continue checking for faults or bad components or damaged cables...

Fab
 

fab

Member
Joined 2004
Paid Member
Thanks manniraj
One possible mistake is to have the leads of the thermistor touching the output transistors heatsink (or mounting screw) on the USSPA board.

Make sure to increase your bias current because your output transistors at 13ma are still in the transition part of the curve before going linear so under biased ...that could explain at least part of your problem.

Is your PSU a shunt or series regulator? If shunt one then normally require to adjust it for proper output feed current.

Keep us posted on your investigation.

Fab
 
Last edited:
Thanks manniraj
One possible mistake is to have the leads of the thermistor touching the output transistors heatsink (or mounting screw) on the USSPA board.

Make sure to increase your bias current because your output transistors at 13ma are still in the transition part of the curve before going linear so under biased ...that could explain at least part of your problem.

Is your PSU a shunt or series regulator? If shunt one then normally require to adjust it for proper output feed current.

Keep us posted on your investigation.

Fab

One update Fab, regarding the noise issue its the Salas psu which is giving this I guess as I isolated to that one after replacing and checking every other part. I replaced the Salas with my old Denoiser psu and viola no noise and absolutely dead silent. You are right about the bias point as I got my calculation wrong, so now I have set it to 238mV (which is around 50mA current) and now I see the heat sinks are a bit warm after 30-45 mins of playback but not so hot. I am using your suggested 1.5" sized heat sink.

Salas ultra bib so its a shunt regulated psu and I used the current limiting resistor (R1) on Salas with 1.3R 3W considering Salas suggested formula "R1=600/CCmA" keeping CCmA as around 500mA. So I guess that is enough current for USSPA both the channels running on a single PSU but I need to debug what else might be the issue like something leaking on the psu side which is going into the preamp and making the noise amplified. But I would like to use the Salas as I use it in all my builds right from Soekris, ACP+, Korg B1 :)

Regarding the thermistor touching the heat sink I followed your manual and put heat shrink tubes on the legs all the way to the end so its only the thermistor which touches the heat sinks and not its legs. I guess that is how it is supposed to be done. Let me know if other wise.

Thanks again.
 
Last edited:
Glad that you have identified the board responsible for the source of the noise issue.:cheers:
I think shunt regulator current does not need to be more than double the load current. The USSPA mono consumes <60ma.

Fab

Yes Fab, so keeping 2 channels of the preamp makes it 120mA and making it double around 240mA is what I should target if using a single psu for both the boards. My plan is to make it dual mono down the line once I can make it either Salas only (real estate is the problem with 4 psu boards) or VRDN much smaller in side and i can stack it one over the other. But even better is the low noise LT30xx regulator based psu but those are expensive.

thanks will keep you posted.