I used it exactly this way:
http://www.diyaudio.com/forums/showthread.php?postid=611337#post611337
There's no turning back.
http://www.diyaudio.com/forums/showthread.php?postid=611337#post611337
There's no turning back.
Grounds and BrianGT 3886 Board
I've just built one of these up. See here. Being that I replaced my homebrew point to point boards with Brian's, and that these were based on my incomplete understanding of "star" ground, I found it difficult and frustrating to connect my single transformer to the 16 diodes on the boards, and ended up using only 8, as shown on page 15 of the 3875 board manual (after much study, both PS boards appear to be similar). I'm Using 8 because I blew up 4 but that's another story.
Anyway, I actually have every single ground (that I recall) tied to the sole "star" ground point. No hum. I'll post my implementation in a few days, but I may have lucked into a good, though not politically correct, grounding arrangement.
Yes, signal ground, power ground, chassis ground, everything to the same point, including speaker binding post ground. And I only use one of the 3 grounds available on the L/R boards from Brian. Like I'm going to run 6 wires to ground when ther is 0.0 ohms between them. Lucky me.
I've just built one of these up. See here. Being that I replaced my homebrew point to point boards with Brian's, and that these were based on my incomplete understanding of "star" ground, I found it difficult and frustrating to connect my single transformer to the 16 diodes on the boards, and ended up using only 8, as shown on page 15 of the 3875 board manual (after much study, both PS boards appear to be similar). I'm Using 8 because I blew up 4 but that's another story.
Anyway, I actually have every single ground (that I recall) tied to the sole "star" ground point. No hum. I'll post my implementation in a few days, but I may have lucked into a good, though not politically correct, grounding arrangement.
Yes, signal ground, power ground, chassis ground, everything to the same point, including speaker binding post ground. And I only use one of the 3 grounds available on the L/R boards from Brian. Like I'm going to run 6 wires to ground when ther is 0.0 ohms between them. Lucky me.
Have been away for a few days with work, but have just managed to re-wire the ground to the two baords as suggested by Carlos. The noise was reduced very slightly, but it is still there. I have also tried moving the power supply board a little further away from the transformer, but there was zero difference. I have taken a few scope shots of the noise. Perhaps someone can offer some more suggestions. My boards are a little further apart than Carlos's and mounted on separate heat sinks. My next course of action will be to mount both boards close together on one sink.
The first scope shot is of the outputs of both amps with the two inputs tied, scale is 20mV per division, 5ms. Second shot is with the top trace showing the output of the power supply board, scale .5mV. Looks like the noise is definately comong down the power supply lines. Now what to do?
Thanks for the continuing help,
Chris
The first scope shot is of the outputs of both amps with the two inputs tied, scale is 20mV per division, 5ms. Second shot is with the top trace showing the output of the power supply board, scale .5mV. Looks like the noise is definately comong down the power supply lines. Now what to do?
Thanks for the continuing help,
Chris
Attachments
I am having a similar problem when both RCA r connected, but mine being a 7294 and not a 3886.
My contruction has only 1 transformer, 2 rectifiers and 2 cap banks feeding 1 amp each. The ground is shared coz it is coming from the centertap of the T/f.
After reading this and some other threads here, I sketched out the grounding and it looked something like a 'V', with the amps at the 2 top ends and the PSU at the apex.
Hence, whenI put in the preamp, it kinda closed the loop at the top (the preamp has common ground).
So I went home, armed with croco-link and started experimenting.
I
1) I clipped both grounds of the RCA togther n turned on the amp. NO HUM despite the ground forming a loop!!
2) Connect both inputs to the pre with pre 'off'. HUM when amp is powered up.
3) Connect both inputs to pre with pre's powercord completely removed from powerpoint. STILL HUMS!
4) Connect single input but linking the 'HOT' pin by crococlip to the 2nd amp. HUM!!
5) Connect single input to pre & 2nd input crococlip to ground. WORSE! it went 'eeeeeeeee'....
6) changed both cables (just in case it's just as simple as that). STILL HUM!!
7) Soldered a 390R in series witt he input (like the grid stopper resistor in tubes). STILL HUM!!
Now, I'm beginning to doubt it has anything to do with a circulating current when the grounds connected form a loop coz I did (1) and it was dead quiet!
has it got somethng to do with the input?
What to test next??
My contruction has only 1 transformer, 2 rectifiers and 2 cap banks feeding 1 amp each. The ground is shared coz it is coming from the centertap of the T/f.
After reading this and some other threads here, I sketched out the grounding and it looked something like a 'V', with the amps at the 2 top ends and the PSU at the apex.
Hence, whenI put in the preamp, it kinda closed the loop at the top (the preamp has common ground).
So I went home, armed with croco-link and started experimenting.
I
1) I clipped both grounds of the RCA togther n turned on the amp. NO HUM despite the ground forming a loop!!
2) Connect both inputs to the pre with pre 'off'. HUM when amp is powered up.
3) Connect both inputs to pre with pre's powercord completely removed from powerpoint. STILL HUMS!
4) Connect single input but linking the 'HOT' pin by crococlip to the 2nd amp. HUM!!
5) Connect single input to pre & 2nd input crococlip to ground. WORSE! it went 'eeeeeeeee'....
6) changed both cables (just in case it's just as simple as that). STILL HUM!!
7) Soldered a 390R in series witt he input (like the grid stopper resistor in tubes). STILL HUM!!
Now, I'm beginning to doubt it has anything to do with a circulating current when the grounds connected form a loop coz I did (1) and it was dead quiet!
has it got somethng to do with the input?
What to test next??
To be a bit more constructive,
Here goes
I had that same problem a while back. Really crazy and cranky inputs! I ended up on the dead end as most would, chasing grounding problems. In the end I had changed everything in the amp with no effect whatsoever, except the rectifierboard. I even put in another trafo! Guess what, when I put in a new rectifier the amp went totally quiet. With the crazy inputs you have there, I would put in a new rectifierbridge! It's no big deal and worth a try, I recon If it doesn't work, I'll buy you a beer Steen.
BTW I P2P wired the rectifierboard like this.
Attachments
Hmmm, come to think of it, I cant even answer that clearly. I just threw out the darn rectifierboard that made me work a week to get rid of that hum. I didn't bother to change the diodes one by one to see if that was it. Wish I knew, though.
If you mean a sixpack for each, you got a deal
Steen
try this
Hi Chris,
I've just found this post. Have you solved the problem with hum ? I've been experimenting a bit with grounding using one transformer and one/two bridges for 2 channel-GC. Due to the relatively long run betwen bridge and GC boards I've had alwais hum appearing at the moment when connecting both left and right channels. The solution CarlosFM shows didn't work for me (bad interpretation/implementation - or too efficient spkr (?): residual hum always audible at 20-30 cm 96dB spkr).
Next arrangement solved my problem (47 labs use the same in 4717 integrated amp :
If you use 4 separated power supply cables per board (-,pg-,pg+,+) and left/right power cables are joined at the rectifier, you have to do:
1) deconnect rca's (-) (ground) cable from both GC pcb boards
2) create chassis star ground (csg) near to RCAs (to form a kind of equilateral triangle rca - csg - rca)
3) connect RCAs (-) to the csg (join left and right rca ground here)
4) connect csg to the left pcb AT (NEAR) THE POINT where left speaker (-) is connected
5) connect csg to the right pcb AT (NEAR) THE POINT where right speaker (-) is connected. (this is tricky, 1 mm make a difference ! )
you can also try to alter position between transformer and amplifier. With this arrangement there is no more hum, but there is always a slight noise, ear ~ 5cm from speaker.
chrish said:
Hi Chris,
I've just found this post. Have you solved the problem with hum ? I've been experimenting a bit with grounding using one transformer and one/two bridges for 2 channel-GC. Due to the relatively long run betwen bridge and GC boards I've had alwais hum appearing at the moment when connecting both left and right channels. The solution CarlosFM shows didn't work for me (bad interpretation/implementation - or too efficient spkr (?): residual hum always audible at 20-30 cm 96dB spkr).
Next arrangement solved my problem (47 labs use the same in 4717 integrated amp
:If you use 4 separated power supply cables per board (-,pg-,pg+,+) and left/right power cables are joined at the rectifier, you have to do:
1) deconnect rca's (-) (ground) cable from both GC pcb boards
2) create chassis star ground (csg) near to RCAs (to form a kind of equilateral triangle rca - csg - rca)
3) connect RCAs (-) to the csg (join left and right rca ground here)
4) connect csg to the left pcb AT (NEAR) THE POINT where left speaker (-) is connected
5) connect csg to the right pcb AT (NEAR) THE POINT where right speaker (-) is connected. (this is tricky, 1 mm make a difference ! )
you can also try to alter position between transformer and amplifier. With this arrangement there is no more hum, but there is always a slight noise, ear ~ 5cm from speaker.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.