How to isolate in Protoboards

[LaloFP]

Hi!

My question is How do you isolate when testing cyrcuits on protoboards.

A few weeks ago I was building the Mad Professor Deep Blue Delay, and exploring with mods and changes. Sometimes I got a horrible distortion in the signal, and, without changing anything, it disappeared.

Then I realiced that if I were touching directly to the floor with my feet without shoe, the distortion goes away. Then that trick didnt worked anymore and the distortion depended on where I where on the room 

I filaized the delay, put it in a metal box and never heard that distortion again.

But today.... Im testing a MorningGlory (from https://aionelectronics.com/project/cerulean-bluesbreaker-overdrive/) and I face the same problem. But discovered it when making a mod.

The C4 capacitor is 10n, and I got the distortion, but if I replace C4 with 47n, the distortion is gone.

I recorded that noise for you:
https://soundcloud.com/lalofp/sound-check

First you can hear the distortion (exactly the same as with the Delay), then I switch to the 47n cap and the distortion is gone. Then I switch again to the 10n cap and the distortion is back. The Gain is turned all the way down.

So... how can I isolate the protoboard without putting all the things inside a big metal box?

You cant experiment with that kind of distortions... 

[LaloFP]

Or do you think that that distortion is not comming from outside?

Now I doubt if the distortion is because of something wrong on my circuit.

But in the Delay case, it sometimes came and sometimes didnt, without any (theorically, maybe I moved some element) change in the circuit.

Can it be a weak connection of a component? Maybe the protoboard is inserting that distortion because a component is not strongly connected and makes a false contact... I dont know...

what do you think?

[anotherjim]

It's something I find I have to live with...

The protoboard is a great thing, but it is not an ideal environment for sensitive circuits.

Watch out for gum on component legs from any paper bandolier component carrier,  R's, C's and D's can have that and it easily transfers to the protoboard contacts if you don't clean or shorten their wires.

External interference is a present danger. Where I am, it changes with the time of day and/or the weather. Sometimes the source is closer to home - lights, soldering irons, heaters etc.

Some circuits suffer oscillation (parasitic oscillation) under certain conditions. The oscillations can be very high frequency but cause audible distortion. Sometimes this is absent on the protoboard circuit but occurs when boxed - or it's the other way around and goes away when boxed. The difference is the contact strips in the protoboard which have a large area compared to any circuit board give extra capacitance (stray capacitance) to everything connected. That said, an oscilloscope capable of reading well over 10Mhz is the only help to see if this is happening.

[LaloFP]

thank you jim 

now I will look up components with gum and see if the distortion stops when cleaning them

[LaloFP]

I discovered something:

In the audio I shared with you I was using a 3PDT to switch from the 10n cap to the 47n cap.

In that scenario, when using the 10n cap, there was distortion, and when using the 47n cap, there wasnt.

I also swap the caps to use the same wire of the 3PDT for the 10n and the 47n cap, to discard the option of beign the 3DPT the cause of the distortion, but nothing changed, the 10n cap showed distortion and the 47n not.

BUT! if I replace all the 3PDT thing and use directly and only the 10n cap or the 47n cap, ther is no distortion in none of them.

So, cant be this?: the distortion was induced by the 3PDT or wires used with the 3PDT, but the 47n cap filtered that distortion to ground, and the 10n cap didnt?

[anotherjim]

Switches and pots can pick up noise when their metalwork isn't earthed. That's another "proto" problem. You could try wrapping a bare wire on the switches fixing nut and connecting to the protoboard power ground strip and see if that's a cure.

[LaloFP]

Thank you again Jim!