But knows very well what beer is! As do I. Well well, perhaps we cross path at some other time.
Back to bias.
Luvrockin: I have reduced ventilation right now. But given +/- 25 degree celsius ambient, my sinks now measure 54,5 degrees on top of the sinks, 50 at the bottom.
My total dissipation is 122,5 watts per sink, counting unloaded rail voltage.
Counting only actual voltage at the rails under full load, the dissipation is 112 watts per channel/sink.
If you wanna stick with that as a max to start with, you can do the math the following way:
(112/4/rail voltage) x 0,5 = voltage drop over each source resistor (not as measured on test point I would think).
For 32v rails under load, that equates to 0,437vDC as read across source resistor. Which equates 0A875 per device, right in Sangrams sweet spot.
I would recommend you consider both 6L6s way and Sangram/Audiosans.
If the latter, I would consider the following:
1: Connect speakers and listen to music with current bias.
2: Repeat this a few days. Watch bias, measure sink temps under different usages.
3: Increase bias in small steps, listening in between, measuring temps. Look for current hogging too, between devices. Measure each source resistor with several dmms to identify this.
4: Does increasing bias make a difference? Is it worth the temp increase? Consider this all the time.
And by God, be careful!
Edit: This is not a path without risk. Your amp did blow twice. And even though you removed the diodes, it is not certain they themselves where the root cause of the blow-ups. They might have been caused by faults elsewhere in the circuit, and I don’t feel the root cause has been found yet. Sticking with 6L6s advice is the absolute safest route, and possibly the best one too.
Regards,
Andy
Back to bias.
Luvrockin: I have reduced ventilation right now. But given +/- 25 degree celsius ambient, my sinks now measure 54,5 degrees on top of the sinks, 50 at the bottom.
My total dissipation is 122,5 watts per sink, counting unloaded rail voltage.
Counting only actual voltage at the rails under full load, the dissipation is 112 watts per channel/sink.
If you wanna stick with that as a max to start with, you can do the math the following way:
(112/4/rail voltage) x 0,5 = voltage drop over each source resistor (not as measured on test point I would think).
For 32v rails under load, that equates to 0,437vDC as read across source resistor. Which equates 0A875 per device, right in Sangrams sweet spot.
I would recommend you consider both 6L6s way and Sangram/Audiosans.
If the latter, I would consider the following:
1: Connect speakers and listen to music with current bias.
2: Repeat this a few days. Watch bias, measure sink temps under different usages.
3: Increase bias in small steps, listening in between, measuring temps. Look for current hogging too, between devices. Measure each source resistor with several dmms to identify this.
4: Does increasing bias make a difference? Is it worth the temp increase? Consider this all the time.
And by God, be careful!
Edit: This is not a path without risk. Your amp did blow twice. And even though you removed the diodes, it is not certain they themselves where the root cause of the blow-ups. They might have been caused by faults elsewhere in the circuit, and I don’t feel the root cause has been found yet. Sticking with 6L6s advice is the absolute safest route, and possibly the best one too.
Regards,
Andy
Last edited:
Hello Guys, thank you all for the info. I’m back at it after an extended break. Andynor, you’re a lucky man only having to worry if your Conona’s are chilled enough to enjoy (assuming you drink you beer cold) rather than if it’s gonna get you sick lol! So I’ve taken the advice from the group and connected up some music. Using an iPad playing Qobuz, some cheapo speakers, cheap power cord and junk cables started to listen. The dc offset is staying pretty stable at about 15.8mv on the right channel and -5/-6 on the left. It sounds very nice all considered. The only thing, I have a slight hum coming out of the right channel. The left is silent. I tried to rotate the transformers slightly but the hum got slightly louder so I moved it back. Any thoughts?
Nice! There is music. Hum is always tricky, and often is caused by a combination of factors. Tranny-turning that helps indicates one liable source.
I suggest overview and detail pics of the amp. Focus also on grounding wires, and how and which wires are twisted together etc. There are several ways to improve this.
Check out hifisonix how-to on grounding, Google it and see.
Be aware, increasing bias will increase ripple on the rails, and as such can increase hum in cases of either too much ripple in relation to PSRR, but also if the issue is common impedance related.
I know. Reducing bias and temps to more moderate levels have removed every drop of hum in my system. But along the way, improving wiring schemes reduced it significantly.
A learning process.
Regards,
Andy
I suggest overview and detail pics of the amp. Focus also on grounding wires, and how and which wires are twisted together etc. There are several ways to improve this.
Check out hifisonix how-to on grounding, Google it and see.
Be aware, increasing bias will increase ripple on the rails, and as such can increase hum in cases of either too much ripple in relation to PSRR, but also if the issue is common impedance related.
I know. Reducing bias and temps to more moderate levels have removed every drop of hum in my system. But along the way, improving wiring schemes reduced it significantly.
A learning process.
Regards,
Andy
First off. Connect a source that is grounded. Ipad's, phones and that sort of sources can give you a lot om hum.
make sure the signalpath from RCA to amp is not to close to "high" voltage wires. Press the ear of the RCA ground connector out and towards the senter and twist the signal wires good(if possible).
Here we go! All eyes on deck. My sinks are running at 51C with ambient @ 23.5C or 74*f and it’s been playing music for about 45 minutes. I don’t know if I can wait to put it in some real speakers. Audiosan, I will connect up to preamp and check in the AM. Just had some pics in the phone.
Attachments
Last edited:
Your PSU gnds are connected 1 cable on each side of the PSU. High currents will be travelling across these, as your gnd planes are connected together just at that exact point. You have in effect created a loop area. Though not nescessarily the main problem, it can contribute to «hum exacerbation», in loss of a better definition.
Recommendation: unsolder, and connect them as close to each other as possible on the euroblocks on ONE side of the PSU halves.
I need to go to bed, so no time to investigate pics further right now. Cheering you on!
Regards,
Andreas
Recommendation: unsolder, and connect them as close to each other as possible on the euroblocks on ONE side of the PSU halves.
I need to go to bed, so no time to investigate pics further right now. Cheering you on!
Regards,
Andreas
Last edited:
Yes. But he has made 4 different gnd connections points. There should ideally be only 2. the way it is connected, there is one loop area per PSU. Not ideal, unless increased chance of hum is the intention.
The boards aren’t very intuitive wrt this, so it is try and succed or fail, but one always succeeds in the end.
I stand my ground: I recommend unsolder the two gnds on the problem channel, group them into euroblocks on one side, then try the other euroblock.
Also, if you have enough wire, you should try turning the tranny even more to fin the null point.
Regards,
Andy
The boards aren’t very intuitive wrt this, so it is try and succed or fail, but one always succeeds in the end.
I stand my ground: I recommend unsolder the two gnds on the problem channel, group them into euroblocks on one side, then try the other euroblock.
Also, if you have enough wire, you should try turning the tranny even more to fin the null point.
Regards,
Andy
Last edited:
Andy, It’s near impossible to see but one green wire on each side goes to the fe boards and the other 2 run down to the same terminal it the grounding block. Also, with the boards jumpered, isn’t the ground all common? Audiosan, I will give that a shot. You’re talking about by the fe boards?
Yes. I guess the easy way is to route the wires from PSU to amp alongside the bottom and leave the signal wires where they are.
Perhaps it is common, but it is not nescessarily the quietest, because:
1: all 4 gnd points are very close to the high ripple currents travelling between the two OSU halves 0 volts. Possible consequence: hum.
2: The gnd points are on different sides of this connection. This creates a loop area. Possible consequence: hum
This could be part of the issue, but does not have to be. But is there room for some improvement? Of course! There always is, isn’t it?
Cheers,
Andy
1: all 4 gnd points are very close to the high ripple currents travelling between the two OSU halves 0 volts. Possible consequence: hum.
2: The gnd points are on different sides of this connection. This creates a loop area. Possible consequence: hum
This could be part of the issue, but does not have to be. But is there room for some improvement? Of course! There always is, isn’t it?
Cheers,
Andy
Last edited: