Tips for Using the Cliff EMI/RFI Washer

[Big Monk]

I’m in the process of making about 50 I/O boards right now and I thought it might be useful to show people how I’ve adapted the Tayda PCB isolated jacks for use on my board and with the Cliff ground washer.

If my count is right, I have about 300 or so units out in the wild with the Cliff washer.

I have it flawlessly integrated now but a few problems caused issues early on:

1.) It’s really meant to mate with Cliff jacks

2.) Jacks were not sitting flat onto the board with it on

3.) Modifications to the washer often caused more issues than helped.

Here’s what I do:

1.) You need to notch, score and cutout the plastic “lip” underneath the jack:





2.) Bend the plastic washer seating tab on the EMI/RFI washer:




3.) Install the original plastic washer PLUS an additional one (the Tayda jacks only come with one so I order a few hundred extras every couple batches). Fit the EMI/RFI washer so the tooth part slides just over the exposed female part of the jack:



4.) Notice that the PCB pins for the ground washer are slightly elevated. Gently bend the tab in toward the jack body:



The jack should now sit flush with the PCB:

 

[Ginsly]

I only use metal panel-mount jacks, so I’ve never heard of these… Are they essentially to ground a plastic jack to the enclosure, or is there something else going on?

 

[Big Monk]

I only use metal panel-mount jacks, so I’ve never heard of these… Are they essentially to ground a plastic jack to the enclosure, or is there something else going on?

Within the context of grounding best practice, you only want to ground the enclosure at a single point.

By isolating both jacks and making the ground connection at the input through the ground washer, you essentially make enclosure ground there.

 

[PedalBuilder]

Within the context of grounding best practice, you only want to ground the enclosure at a single point.

By isolating both jacks and making the ground connection at the input through the ground washer, you essentially make enclosure ground there.

Have you noticed any difference in noise levels by grounding at a single point vs grounding at both jacks?

 

[Big Monk]

Have you noticed any difference in noise levels by grounding at a single point vs grounding at both jacks?

If I am being totally honest? Yes, but not an order of magnitude type reduction.

I've gone through a few phases where I first isolated, with a washer, the metal output jack and used shielded wire, etc. I then went to isolated jacks with a board mounted ground spring and shielded wire.

This is the last iteration that I have tried and compared standard hobbyist "I don't care" grounding, it is slightly less noisy overall. Slightly being the key word.

The fact that it is so easily implemented into my builds is the driver for doing it and NOT the slight noise reduction. Best practice is always best when it doesn't require tons of effort.

 

[Ginsly]

Within the context of grounding best practice, you only want to ground the enclosure at a single point

But your 3pdt and pots still make contact with the enclosure, right? Wouldn’t the enclosure be grounded in many spots?

 

[Big Monk]

But your 3pdt and pots still make contact with the enclosure, right? Wouldn’t the enclosure be grounded in many spots?

There is no continuity between pot lugs/switch lugs and the cases. At least there shouldn't be. This is the mechanical vs. signal ground topic in practice.

In a roundabout way, the cases DO connect to ground through the enclosure but the EMI/RFI ground washer at the input effectively serves as the single point of contact, as it makes the shortest and most direct connection from the enclosure to ground, theoretically speaking.

Like I said, in practice we are talking about "inside baseball" here. If this was a more involved or difficult process, it wouldn't be worth it.

 

[lowpitch]

Any particular reason to ground to the input rather than the output jack? My thinking: input will be more sensitive to noise interference through parasitic capacitance, especially in high gain effects.

 

[Big Monk]

Any particular reason to ground to the input rather than the output jack? My thinking: input will be more sensitive to noise interference through parasitic capacitance, especially in high gain effects.

I don;t have a good reason other than R.G. Keen told me that was the best place.

 

[lowpitch]

I don;t have a good reason other than R.G. Keen told me that was the best place.

View attachment 93717

Ah, I may have been thinking about this backwards. If I'm understanding correctly, the idea is to provide a fast and direct path to shield ground for any noise picked up along the way from guitar to pedal for example, so it doesn't have to travel through any ground wires in close proximity to the main circuit within the enclosure.

 

[Big Monk]

Ah, I may have been thinking about this backwards. If I'm understanding correctly, the idea is to provide a fast and direct path to shield ground for any noise picked up along the way from guitar to pedal for example, so it doesn't have to travel through any ground wires in close proximity to the main circuit within the enclosure.

That's how I always thought of it.

An immediate ground path to shunt nasties.

 

[Brett]

Thanks for sharing! I'd asked about this a while back in another thread and hoped to have a follow-up from you:

Is that alright? I can post measurements here tomorrow.

I'd still be interested if you're willing to share.

In a roundabout way, the cases DO connect to ground through the enclosure

I seem to recall Chuck D. Bones mentioning that a reliable ground connection to potentiometer case is generally considered good practice to reduce unwanted noise in high-gain circuits. While the case on a standard 16mm Alpha pot doesn't share an electrical connection to its lugs, grounding the case can help shield/contain stray EMI.

 

[Big Monk]

I'd still be interested if you're willing to share.

Absolutely. I don;t have the design files for my I/O board on my thumb drive at work but I can post them sometime later tonight.