Ever since I had some new boards made for my amp I haven't been able to get it to work right. The signal is static on it but if I take my fingers and touch the driver tubes it gets loud and the static quits. I am touching the glass only. I was pushing on the tubes and noticed it change and I thought there was some internal issues with the circuit board but it affects it without putting any pressure on the tubes. Just touching the glass is enough.
I thought maybe a bad ground and its maybe working because its grounding thru me? The audio ground I mean. I have the rca jacks on the back connected to star ground but I am going to try running them straight to the pot and see if that matters. Any ideas are welcome.
D3a drivers with 300B SE output tubes.
I thought maybe a bad ground and its maybe working because its grounding thru me? The audio ground I mean. I have the rca jacks on the back connected to star ground but I am going to try running them straight to the pot and see if that matters. Any ideas are welcome.
D3a drivers with 300B SE output tubes.
Sounds like a high frequency oscillation or instability. Can you post the schematic and photos of the board?
Last edited:
You are adding stray capacitance, which can cause/cure/change instabilities in a circuit.
Could be layout related. Does it still happen without the 300B?
Could be layout related. Does it still happen without the 300B?
Last edited:
Rewiring ground made no difference. I changed tubes to 2 new ones and its still screwed up but putting my hand by the tube does nothing now. It must be the boards. It has to be.
I will when i get a chance ill post pics. I design pcbs for a living. I am thinking its an internal defect, not a layout issue. I routinly design boards with 16 layers, impedance matching, mixed analog and digital grounding etc. This is a simple 4 layer board. But its possible. I have tried several of the 5 boards I had made and same exact issue with every one. I have already changed it to a 2 layer board, increased clearances, reduced ground pours to a single plane, increased size of holes in a few places and increased keep out areas around all mounting holes. So the new boards just plain have to work. Want me to post gerbers? picture of each layer? Or ???
Also this amp has been working for 2 years. I just got tired of the old chassis because working on anything was a pain in the *** and after shocking myself with 500v 3 times I grew tired of it. 500v hurts! Hell it blew a hole in my skin and made me bleed once. Didn't think it possble to bleed form being shocked but alas it is.
Also this amp has been working for 2 years. I just got tired of the old chassis because working on anything was a pain in the *** and after shocking myself with 500v 3 times I grew tired of it. 500v hurts! Hell it blew a hole in my skin and made me bleed once. Didn't think it possble to bleed form being shocked but alas it is.
In-line ferrite bead is recommended for almost all the pins of D3a
to impede ringing.
It is easy for point-to-point wiring, not sure
how to do this for planar PCB technology ?
to impede ringing.
It is easy for point-to-point wiring, not sure
how to do this for planar PCB technology ?
If the board worked before being installed in the new chassis, it is probably ok.
Check the input and output wiring, and the common connections. Keep the input socket and input wiring
away from other parts of the circuit, especially away from the output transformer leads and the output
terminals. Bring the input common to the supply common separately from the output stage common
and the speaker common.
Last edited:
So the identical schematic has worked before properly, but not on the new board rev.
Could there be a net list error, maybe a capacitor connected to the wrong point or layer, etc?
Do you have a breadboard version to hack on?
Could there be a net list error, maybe a capacitor connected to the wrong point or layer, etc?
Do you have a breadboard version to hack on?
My new boards will be here in a day or two.
I reduced the 4 layers to 2
Increased clearances
Increased keep out zone around all mounting points
Connected only one sides pour to ground ( I usually connect them all and stitch the hell out of it with vias unless it a dig/analog board then I lay it out carefully and only connect the two grounds in one spot using 0 ohm resistors so we can put caps across them if needed.)
Among the things I tried with the current board:
populated 3 of the 5 I have. All were the same.
At one point I had it lying on top of the chassis with no screws in it and it worked for 4 hours. Sounded great. But then it digressed,
swapped driver tubes back and forth
swapper output tubes back and forth.
bought new coupling caps ( mostly because I wanted to_
Pulled the amp 100% apart and rebuilt in two chassis seperating the power supply.
I am quite sick of working on this thing.
I reduced the 4 layers to 2
Increased clearances
Increased keep out zone around all mounting points
Connected only one sides pour to ground ( I usually connect them all and stitch the hell out of it with vias unless it a dig/analog board then I lay it out carefully and only connect the two grounds in one spot using 0 ohm resistors so we can put caps across them if needed.)
Among the things I tried with the current board:
populated 3 of the 5 I have. All were the same.
At one point I had it lying on top of the chassis with no screws in it and it worked for 4 hours. Sounded great. But then it digressed,
swapped driver tubes back and forth
swapper output tubes back and forth.
bought new coupling caps ( mostly because I wanted to_
Pulled the amp 100% apart and rebuilt in two chassis seperating the power supply.
I am quite sick of working on this thing.
If both channels of the board are unstable (with different layouts), the D3A could as mentioned
need ferrite bead(s). This is a very wide bandwidth tube. You could cut the traces close to the
tube and add a jumper with the bead, for each tube terminal. I would start with the screen 2,
then the grid, cathode, plate. Do one bead at a time.
Also I would try local bypassing across each filament, 0.1uF.
Last thing to try is adding local decoupling on the board for the power supplies,
a larger electrolytic and a smaller ceramic, for each channel.
need ferrite bead(s). This is a very wide bandwidth tube. You could cut the traces close to the
tube and add a jumper with the bead, for each tube terminal. I would start with the screen 2,
then the grid, cathode, plate. Do one bead at a time.
Also I would try local bypassing across each filament, 0.1uF.
Last thing to try is adding local decoupling on the board for the power supplies,
a larger electrolytic and a smaller ceramic, for each channel.
Last edited:
Also check your tube sockets for corrosion and poor contact. Check inserting valves has not lifted a PCB track. Check for dry joints which may arc and cause static type noise.
Last edited:
WOO HOO New boards fixed it bigtime...
I have my stereo back ....
breaking in some new Deulend caps. got the cast copper since they were on sale. Also swapped out the resistors on the input to Vishay Precision Z foils.
I have been slumming it with a chinese el84 amp and man what a difference...
Also have a brand new pair of telefunken D3a tubes in it. But not a lick of noise. None...
Thanks to those that attempted to help.
Now I need to finish designing the final chassis for it. going with a couple mono block chassis. Need to buy another power transformers...
Jeff
I have my stereo back ....
breaking in some new Deulend caps. got the cast copper since they were on sale. Also swapped out the resistors on the input to Vishay Precision Z foils.
I have been slumming it with a chinese el84 amp and man what a difference...
Also have a brand new pair of telefunken D3a tubes in it. But not a lick of noise. None...
Thanks to those that attempted to help.
Now I need to finish designing the final chassis for it. going with a couple mono block chassis. Need to buy another power transformers...
Jeff