I just rebuilt a revibe that I failed to get working about a decade ago. Sluckeys revibe pdf was very helpful this time around. On the first power on, I was getting some weird noise from the tremolo section, almost like motorboating, but not quite. It was independent of the LFO speed.
Assuming I had a bad cap, I jumpered in a new .22uf on the vibrato inverter, problem solved. Solder in the new cap, problem returns. If I Jumpered from the new cap to anywhere on the same node, problem goes away. Resolder everything, no change. For some reason I thought to run a 180p cap from the .22 to the plate, across the existing wire, and the thumping completely went away. More experimenting, with the cap connected on one side, and touching the cap body the thumping goes away.
Eventually, I shielded the grid wire on that tube and problem solved. Can anyone explain what was going on with that 180p cap, and why it would have any effect jumpered across a wire?
revibe tremolo parasite
Moderators: pompeiisneaks, Colossal
revibe tremolo parasite
You do not have the required permissions to view the files attached to this post.
Re: revibe tremolo parasite
I worked on an Eico hf-20 last year that had a 47pf cap across a wire on the phono input. I partially assumed it was a schematic revision error at the time. I just can't wrap my head around how a cap does anything while it is shorted.
You do not have the required permissions to view the files attached to this post.
Re: revibe tremolo parasite
Look at this pic...
You do not have the required permissions to view the files attached to this post.
Re: revibe tremolo parasite
Nope, but I can see how it would look that way. I've got the revibe working perfectly now with the shielded cable. I could get the thumping sound to go away if I moved the unshielded grid wire far away from everything else, and it was also working fine with the dodgy cap arrangement on the plate. I'm just confused why that cap worked. Some kind of strange filter?
Re: revibe tremolo parasite
Yep, some weird kind of filter. That's the only thing that explains the situation.
If you could connect an oscilloscope to the plate of the tube, you'd probably see RF on part of the signal swing as you ran signal through it. I say "if you could" because in situations like this where a tiny capacitance is causing the presence/absence of the problem, even the resistance/capacitance of a 10X scope probe can make the problem vanish. There are scope probes that will do this, but they're special beasts, not widely available. But I digress.
A wire is not a zero ohm connection. It is a series RL with a capacitance across it. Ham radio people who built their own equipment used to know this, as I've seen charts of the highest frequency that a cap is still a cap versus lead length. The inductance of the wire can cause RF issues, including parasitic oscillation and tuning in radio stations. What makes this tricky is that often it's the quirks of the exact tube that it interacts with that makes the thing go into and out of oscillation as the tube's internal parameters change - just a bit - as the signal wanders up and down.
"Gimmicks" were a well known ham radio builder's trick. Builders would tune in stages of RF by soldering a length of twisted pair wire between two places, and then snipping off some of the twisted pair for tuning. The prevailing wisdom was that the gimmick added about 1pF per inch.
The way to fix it is to move the "filter" frequency so much that the rest of the amplification loop can't break into Nyquist gain-phase oscillation. That's what adding a cap to ... somewhere ... is doing. It's changing the filter resonance. It's strange to think, but a capacitance across a length of wire is in fact a resonant filter at some frequency. There is already a capacitance there, in the wire itself. Adding a cap across it adds a massive amount more.
That is, either use a sledgehammer (the added cap) or get tricky and change something else. My preference is to make the amplifying device unable to respond at the suspected RF. I would see if I could add either ferrite beads in the signal wires and/or power supply leads, or use RF killer capsresistors on the tube/transistor.
This kind of thing is especially bad with MOSFETs. In some ways, they're like tubes on steroids, complete with 'roid rage. Their input impedance is often higher, and their frequency response is vastly higher.
If you could connect an oscilloscope to the plate of the tube, you'd probably see RF on part of the signal swing as you ran signal through it. I say "if you could" because in situations like this where a tiny capacitance is causing the presence/absence of the problem, even the resistance/capacitance of a 10X scope probe can make the problem vanish. There are scope probes that will do this, but they're special beasts, not widely available. But I digress.
A wire is not a zero ohm connection. It is a series RL with a capacitance across it. Ham radio people who built their own equipment used to know this, as I've seen charts of the highest frequency that a cap is still a cap versus lead length. The inductance of the wire can cause RF issues, including parasitic oscillation and tuning in radio stations. What makes this tricky is that often it's the quirks of the exact tube that it interacts with that makes the thing go into and out of oscillation as the tube's internal parameters change - just a bit - as the signal wanders up and down.
"Gimmicks" were a well known ham radio builder's trick. Builders would tune in stages of RF by soldering a length of twisted pair wire between two places, and then snipping off some of the twisted pair for tuning. The prevailing wisdom was that the gimmick added about 1pF per inch.
The way to fix it is to move the "filter" frequency so much that the rest of the amplification loop can't break into Nyquist gain-phase oscillation. That's what adding a cap to ... somewhere ... is doing. It's changing the filter resonance. It's strange to think, but a capacitance across a length of wire is in fact a resonant filter at some frequency. There is already a capacitance there, in the wire itself. Adding a cap across it adds a massive amount more.
That is, either use a sledgehammer (the added cap) or get tricky and change something else. My preference is to make the amplifying device unable to respond at the suspected RF. I would see if I could add either ferrite beads in the signal wires and/or power supply leads, or use RF killer capsresistors on the tube/transistor.
This kind of thing is especially bad with MOSFETs. In some ways, they're like tubes on steroids, complete with 'roid rage. Their input impedance is often higher, and their frequency response is vastly higher.
"It's not what we don't know that gets us in trouble. It's what we know for sure that just ain't so"
Mark Twain
Mark Twain
Re: revibe tremolo parasite
Thanks R.G., interesting. I replicated the problem with an unshielded wire which, of course, took extra effort to find the one position where it would oscillate instead of the one position where it wouldn't... I tried putting different sized ferrite beads on a few parts of the power supply rail and on the plate load resistor but no change. That chassis is too tight for me to want to experiment any more when I already have a solution but it's nice to have a couple more tools if I run into a problem like that again.
Re: revibe tremolo parasite
I think your cap trick and shielded wire trick are just band aids that are covering up the real problem. V5 deals with a freq range of 3 to 10Hz. I've seen quite a few revibes, especially based on Hoffman's layout that work just fine. I think there is an underlying logical problem in the circuit. Something like poor ground, bad connection, etc. I would keep looking.
Re: revibe tremolo parasite
I played with it a bit more. At this point, the unshielded plate and grid wires need to be parallel and on top of each other to get the issue to happen, so maybe an underlying problem was already solved. The only thumping noise I can get now follows the LFO speed.
I previously swapped v4 and v5 (both JJ ecc83S) to rule out a tube problem and there was no change. This time I tried an EHX and I can no longer get the plate and grid wires to interact at all. I switched between all 3 tubes once more to make sure I wasn't missing anything. Could there be something about the structure of the 83S's that doesn't play nice under these very specific circumstances?
I previously swapped v4 and v5 (both JJ ecc83S) to rule out a tube problem and there was no change. This time I tried an EHX and I can no longer get the plate and grid wires to interact at all. I switched between all 3 tubes once more to make sure I wasn't missing anything. Could there be something about the structure of the 83S's that doesn't play nice under these very specific circumstances?