I am not so sure what is contentious, but current is going through the load and returns to the power through the ground. In real circuits, the problem is in the return path. This is why Dario finds that connecting PGND to safety earth no so desirable. I would be interested in other views on the cause of Dario's experience. My view is something needs to be done in the power supply. The layout seems not too bad as it is. I would recommend again, a different snubber. But since I have a totally different power supply design, I guess this thread will have different considerations.
Last edited:
I have to agree, and to a large extent, this has been accomplished in the various MyRef layouts, i.e. they're on the same PCB, but well isolated.
One minor quibble I have left is the speaker return current trace. At the moment, it goes to PGND, the main high-current star ground. I think there's a reasonable case to return it to the junction of C1, C2 instead.
follow the signal current from either capacitor through the amplifier out to the speaker terminal. Now mirror that route with the return signal cable.
You will find that part of the route already has a return cable there.
You only need to insert the return cable where the route does not exist.
This completely removes the need to take speaker return all the way back to the node between the main smoothing caps.
That's why your version of myref is done as is.
Last edited:
one channel of my newly built rev e kit is being problematic... 
in a normal start up the relay clicks on faster than the other channel and dc
reads -40v(!)
without the op amps this channel doesn't even start but the relay clicks and the led
comes on momentarily when i power it
off.
in both situations( with and without the op amp) the temperature for the 3886 registers much higher than the 'healthy' channel.
i had this same problem with my older kits which was the reason why i scraped the project n started this one. i did carry over all the caps
from the previous project, however; which makes me wonder if one of the caps have gone faulty...
please help, gents. thank you in advance
in a normal start up the relay clicks on faster than the other channel and dc
reads -40v(!)
without the op amps this channel doesn't even start but the relay clicks and the led
comes on momentarily when i power it
off.
in both situations( with and without the op amp) the temperature for the 3886 registers much higher than the 'healthy' channel.
i had this same problem with my older kits which was the reason why i scraped the project n started this one. i did carry over all the caps
from the previous project, however; which makes me wonder if one of the caps have gone faulty...
please help, gents. thank you in advance
The relay implements speaker isolation at start up with a delay.
at shutdown with a near instant relay release
Dc detect and isolate. This also prevents speaker connection if the amp sends DC to the DC detect at start up both before and after the start up delay.
If the relay clicks over and you are seeing -40Vdc at the speaker terminal then there are at least two defects.
1.) the amplifier is sending DC to the speaker.
2.) the DC detect and isolate is not working.
Get the DC detect and relay trigger circuit working first.
at shutdown with a near instant relay release
Dc detect and isolate. This also prevents speaker connection if the amp sends DC to the DC detect at start up both before and after the start up delay.
If the relay clicks over and you are seeing -40Vdc at the speaker terminal then there are at least two defects.
1.) the amplifier is sending DC to the speaker.
2.) the DC detect and isolate is not working.
Get the DC detect and relay trigger circuit working first.
to further complicate the matter...
the healthy side now reads 100mv dc, and all ive done since the initial reading was swap around the op amps(actually lf01 modules). the bad side
still reads constant 40mv.
i dared hooking them up to the speakers and as expected,
it hums. the healthy channel, even with the higher dc
reading, is silent as night.
the healthy side now reads 100mv dc, and all ive done since the initial reading was swap around the op amps(actually lf01 modules). the bad side
still reads constant 40mv.
i dared hooking them up to the speakers and as expected,
it hums. the healthy channel, even with the higher dc
reading, is silent as night.
thank you for your help, andrew, but I'm not knowledgeable enough to follow through without a step by step instruction
i only have a basic multimeter for ac,dc, and ohm. so i guess the cap test will have to wait until i get a better tester.
the both channels work, btw, and the sound is as good as anything i heard, but there's a prominent hiss on the bad channel.
so to highlight the symptoms of the bad channel once more-
relay doesn't click without the 318
3886 is warmer \
40mv dc
hisses (hums louder without anything connected)
sporadically pops, cackles before the music starts
*premature start is no longer an issue. the relay for both of them start at about same time now, though it clicks off slower than the healthy channel on the power down.
ugh, i just don't know anymore... bad solder joint, maybe? I reflowed them and checked them thoroughly during the assembly, though..
i only have a basic multimeter for ac,dc, and ohm. so i guess the cap test will have to wait until i get a better tester.
the both channels work, btw, and the sound is as good as anything i heard, but there's a prominent hiss on the bad channel.
so to highlight the symptoms of the bad channel once more-
relay doesn't click without the 318
3886 is warmer \
40mv dc
hisses (hums louder without anything connected)
sporadically pops, cackles before the music starts
*premature start is no longer an issue. the relay for both of them start at about same time now, though it clicks off slower than the healthy channel on the power down.
ugh, i just don't know anymore... bad solder joint, maybe? I reflowed them and checked them thoroughly during the assembly, though..
Last edited:
47pF in C10?
I fear they're too much...
In this moment I'm settled with these values:
C10 33pF
C34 27pF
C32 150pF
It sounds great!
do you have 40Vdc or 40mVdc for the bad channel output offset?
If you have prominent hiss and hums on the bad channel, how can
You found 27pF film cap? I have some PS films caps of those values, but since they are old, the solder does not seem to stick.
I'm probably going to play around with C10 and C34 some more. The values I'm trying now seems to cause the amp to run warmer. Not sure why. But again, I'm modding other things as well.
No, I'm using Silver Micas
Also the original alternate compensation run warmer, a sign of overcompensation?
oh well, since it makes music even with the bad channel, might as well post pics of it
here's when it was young and full of hope.
now burdened with alimony
I'd like to call it the host for obvious reason
since there's not enough real estate up top, the choice was made to install the lf01 module under board, which made the board taller than it is wider- hence the vertical mounting. caps are mundorf sio 2.2uf, mundorf au/ag & supreme (c7 &9), sic safco 220pf, and a russian teflon for the 22uf spot.
wooden casings and new trafos are in the near future.
and the hissing has disappeared since reflowing the solder but the relay issue remains...
as for the sound- well, i don't wanna fill up a whole page
here's when it was young and full of hope.
An externally hosted image should be here but it was not working when we last tested it.
now burdened with alimony
An externally hosted image should be here but it was not working when we last tested it.
I'd like to call it the host for obvious reason
since there's not enough real estate up top, the choice was made to install the lf01 module under board, which made the board taller than it is wider- hence the vertical mounting. caps are mundorf sio 2.2uf, mundorf au/ag & supreme (c7 &9), sic safco 220pf, and a russian teflon for the 22uf spot.
wooden casings and new trafos are in the near future.
and the hissing has disappeared since reflowing the solder but the relay issue remains...
as for the sound- well, i don't wanna fill up a whole page
Last edited:
You have taken the Myref to another level.Please post more about your impressions when it settles down 
