Are My Traces Causing The RF Interference?

[railhead]

I'm still refining my fuzz clone, and my latest PCB layout revision seems to have introduced major RFI, and no component changes were made. I can run my previous PCB pedal right next to this new one, and the previous one is dead silent. Here's the "silent" version:




Here's the new one:




The main changes I made were to move the pot connections to the top of the PCB to make wiring easier. You'll see, too, that I have a ground going along the top so that I can wire the out jack and volume pot to ground -- and the ground is crossing under the pot out. Could this be causing the interference?

I know I can add a cap to kill the RFI, I'm just curious as to why this PCB layout is causing it.

Thoughts?

[e45tg4t3]

well, afaik you should try to avoid 90° traces like in your old layout... maybe this could make this layout silent....

Benny

[soulsonic]

In the old layout the input lug is much closer to the ground connexions than with the newer one. This is most likely the issue. To take care of it the best, I think at ground plane is in order - or at least draw a "Guard Ring" around the input panda pad. I've done it a couple times and it works really well in most cases for keeping radio problem at bay.
The ground crossing over or under things won't usually cause interference - in most cases, the capacitance it introduces helps filter interference out. It acts like a small cap to ground; that's built into the board.

[Paul Perry (Frostwave)]

I don't think right angle traes are a hassle until you get up to 500 Mhz.
Do you have both layouts operating at the moment?
Could be some other variable..

[railhead]

In the old layout the input lug is much closer to the ground connexions than with the newer one. This is most likely the issue. To take care of it the best, I think at ground plane is in order - or at least draw a "Guard Ring" around the input panda pad. I've done it a couple times and it works really well in most cases for keeping radio problem at bay.
The ground crossing over or under things won't usually cause interference - in most cases, the capacitance it introduces helps filter interference out. It acts like a small cap to ground; that's built into the board.

If I'm understanding you correctly, you're saying it's better to have the input closer to the grounds, right? Like I had in my first layout? If that's what could be causing the RF, then moving that is no biggie.

As for the ground plate, are you saying to maybe make one large enough so that it extends the grounds further tot he right, thus closer to the input? If so, I like that idea better (saves etchant).

I'm not sure what you mean when you mention adding a guard ring. Do you mean just an empty trace between the input and ground?

[darron]

In the old layout the input lug is much closer to the ground connexions than with the newer one. This is most likely the issue. To take care of it the best, I think at ground plane is in order - or at least draw a "Guard Ring" around the input panda pad. I've done it a couple times and it works really well in most cases for keeping radio problem at bay.
The ground crossing over or under things won't usually cause interference - in most cases, the capacitance it introduces helps filter interference out. It acts like a small cap to ground; that's built into the board.

If I'm understanding you correctly, you're saying it's better to have the input closer to the grounds, right? Like I had in my first layout? If that's what could be causing the RF, then moving that is no biggie.

As for the ground plate, are you saying to maybe make one large enough so that it extends the grounds further tot he right, thus closer to the input? If so, I like that idea better (saves etchant).

I'm not sure what you mean when you mention adding a guard ring. Do you mean just an empty trace between the input and ground?

for the ground ring, i assume you mean like all the layouts on tonepad?
http://www.tonepad.com/getFileInfo.asp?id=82 for example. i'm curious to know what benefit this adds also, and any interesting reads that anybody might know of giving layout tips for audio.

[soulsonic]

That Tonepad layout doesn't have an input guard ring, though it does have the input situated very closely to the ground traces.
Here is an app note that shows guard rings and give an explanation: http://www.tranzistoare.ro/datasheets/480/49178_DS.pdf
Look on page 7 of the PDF. It shows this practice in relationship to panda opamps, but I think it can work just as well with most other things.
Of course, if you do a ground plane where it has alot of ground surface area surrounding the input pad, then you've accomplished the same thing as near as I can tell.

[darron]

That Tonepad layout doesn't have an input guard ring, though it does have the input situated very closely to the ground traces.
Here is an app note that shows guard rings and give an explanation: http://www.tranzistoare.ro/datasheets/480/49178_DS.pdf
Look on page 7 of the PDF. It shows this practice in relationship to panda opamps, but I think it can work just as well with most other things.
Of course, if you do a ground plane where it has alot of ground surface area surrounding the input pad, then you've accomplished the same thing as near as I can tell.

that's very interesting soulsonic. i'll have a look into this business. thanks (:

[railhead]

Well, I have an interesting (somewhat) update. First off, I went ahead and etched a new PCB loaded with ground planes, put together a new fuzz, thought all was well -- then a started hearing the radio again.

Ugh.

BUT, I took more time to ascertain what was really going on, and I noticed that the RFI got louder as I turned down the volume on my Strat, which, in short, led me to suspect my Start was picking up the RFI. So, I grabbed my Les Paul, plugged it in, and I don't hear any radio noise at all. After all was said and done, the culprit was -- apparently -- my Strat and not my pedal. This has taught me a few things:

1. I should be more confident in my layouts, assuming I've taken the time to make sure things are as they should be.
2. If something's not sounding right, it never hurts to try the pedal using a different guitar/amp.
3. You may as well load the PCB with ground planes since it can't (really) hurt, plus it saves on etchant.

Thanks for all the help, though!

[darron]

Well, I have an interesting (somewhat) update. First off, I went ahead and etched a new PCB loaded with ground planes, put together a new fuzz, thought all was well -- then a started hearing the radio again.

Ugh.

BUT, I took more time to ascertain what was really going on, and I noticed that the RFI got louder as I turned down the volume on my Strat, which, in short, led me to suspect my Start was picking up the RFI. So, I grabbed my Les Paul, plugged it in, and I don't hear any radio noise at all. After all was said and done, the culprit was -- apparently -- my Strat and not my pedal. This has taught me a few things:

1. I should be more confident in my layouts, assuming I've taken the time to make sure things are as they should be.
2. If something's not sounding right, it never hurts to try the pedal using a different guitar/amp.
3. You may as well load the PCB with ground planes since it can't (really) hurt, plus it saves on etchant.

Thanks for all the help, though!

hmmm... i've got an idea for you to try. it might not work but it won't take long to find out. my first build was a fuzz face and i used the negative ground version. it worked really cool, but once you turned the gain up on the effect or the volume down on the guitar it would start making noise, oscillating and motorboating. it wasn't as bad or even went away on humbuckers though.

try putting a 1000uf or so capacitor on the power. this might not do the trick, but if you have a massive capacitor lying around then you might get lucky (: