Wendigo preamp noise

[fossilshark]

As i'm sure you more experienced guys saw coming, this layout I came up with is garbage 

So I botched in a rectifier and some 2200uf caps for the heaters, after doing that the consistent buzz was gone. I then narrowed down the noise to the first 2 gain stages and thought "nice, a little shielded cable should fix this"

I added shielded cable to the long wire going to the output and ziptied it to the rectifier output for some mechanical stability and I threw in some shielded cable on the longer wires in the first 2 stages (only grounding 1 side of the shielding of course) and... well the noise problem is 10x worse.

With no input plugged in, I turn up the volume and gain pots and they all make a horrid screeching noise past a certain point that it did not do before I added shielded cable. I thought shielded cable makes noise go away? How can I post a sound clip for you all?

This preamp is supposed to be used at a gig on the 11th as I have spent way too long on it and I am growing more and more concerned that this layout is just absolutely terrible and I won't be able to get the noise level low enough for a live band situation.

Any advice is greatly appreciated.

[fossilshark]

Small update: did the same thing with the guitar plugged in and guitar volume all the way down, the screeching only happens with both gain channels dimed and the volume almost all the way dimed. There is still a decent amount of noise I am not comfortable with.

[sluckey]

This preamp is supposed to be used at a gig on the 11th

Go to plan B. Take the pressure off and fix this thing at your leisure.

[fossilshark]

I mean its real close to being done its just this one noise mystery now. I've taken way too long on it as it is. I just don't get how adding shielded cable could make MORE noise issues.

[sluckey]

If done properly shielded cable should not introduce more noise. I suspect you may have suffered some collateral damage while installing shielded cable.

[fossilshark]

Perhaps. I beeped out the shielded connections and made sure no signals were shorted to grounds, mabey i caused a faulty ground connection in the preamp?

I really really hate working with this shielded cable it melts so easily when soldering and theres no way to make it look pretty in a layout, might look around for cloth insulated solid core shielded cable.

[sluckey]

I switched to RG-316 a long time ago. It will not melt.

[PRR]

> shielded cable to the long wire going to the output and ziptied it to the rectifier output for some mechanical stability

Possibly the worst thing you could do. Do you understand what a rectifier's output looks/sounds like? No "shield" is perfect. Imagine drinking water in a cloth hose, lashed to a cloth sewer line. Yeah it leaks "less", not zero.

Unlash it, rig it to be semi-safe, listen as you move that signal wire near/far from the rectifier line.

[fossilshark]

Yes i figured that out by moving the lines around, whoops!

So i used a decoupling cap to ground out the signal starting from the input and found the lines going to these rotary switches are causing 90% of the noise. If i shield all those lines, would it matter where i ground the shielding to? Im mostly using bus ground layout except for a few things going directly to ground.

[fossilshark]

I think I'm just going to tap out on this. My choices are spend another 4 hours adding shielded cable that will ultimately make the amp more susceptible to mechanical vibrations or I can just accept that this layout fkn sucks and I need to spin up a PCB for the next iteration, or use a much bigger chassis.

Thank you all for your help, if my friend makes any professional recordings of it I will post them here.

[PRR]

Switching teeny capacitors like that, we usually mount the caps on the switch, not run wires elsewhere.

Yes your scheme is mechanically robust yet easy to replace/modify.

But I think you found one reason why pFd caps should not be on long wires. (I agree that 4" did not seem long at first glance, so thanks for the observation.)

[tdvt]

I switched to RG-316 a long time ago. It will not melt.

Semi-related query;
Looked at the RG-316 linked, what is the "50Ω" designation referring to?

[sluckey]

Semi-related query;
Looked at the RG-316 linked, what is the "50Ω" designation referring to?

50Ω cable is the industry standard for RF circuits such as communications radio to antenna cable. The antenna will have a 50Ω impedance and the radio will have a 50Ω impedance. So naturally you would use a 50Ω cable to connect the two. Doing so will insure maximum power transfer and minimum VSWR.

75Ω cable is the industry standard for video circuits such as your RG-59 or RG-6 cable TV systems. Same reasons as using 50Ω cable for RF circuits.

None of this matters at audio frequencies. You're not interested in transferring power. You're only interested in the excellent shielding properties of the RG-*** cables. Heck, shielded audio patch cables don't even have an impedance designation because it's not necessary. Generally speaking, more expensive audio cables have better shielding and lower loss (resistance/foot).

[tdvt]

Thank you for the explanation.

I have used plenty of 75Ω & remember 30Ω flat TV antenna wire, but wasn't sure how those designations originated.

I know some of my bulk XLR cable has single-digit resistance per 100 meters

[sluckey]

TV flat twin lead is 300Ω.

[tdvt]

TV flat twin lead is 300Ω.

I told you I remembered it....

[PRR]

If you run an infinite length of cable, and look at high-infinite frequency, you will "see" 50 Ohms looking into the cable. If you want to connect a 50 ohm load at >1MHz, the 50 ohm cable is a good fit. Other loads or cable impedances will reflect power back to the source.

And as Sluckey says, this does NOT matter for stage audio.

Even if it did: you have to use perverse cable dimensions to get away from 50/75 ohm cable. There was a 105 Ohm video cable with a hair-fine center conductor. 600 Ohm nominal (on loooong lines) telephone cable is spaced more than a foot. (and actually droops from >900 in bass to 350 in treble).