What's up with EHX putting caps on the jacks?

[j_flanders]

I plugged straight into my amp today and noticed how much treble I'm apparently losing through my pedal board.
All the pedals should be true bypass, or are at least using 3PDT switches.

I started measuring each patch cable and each pedal to see what resistance or capacitance they added.
As far as resistance goes they all measured open, so no hard wire bypass with hanging inputs.

But on both my big box Holy grail and Deluxe Memory Man I measured a rather high capacitance in bypass.
Opening up my Deluxe Memory Man I see a 470pF cap soldered across the output jack.
I didn't open my HG but remember seeing a similar thing last time I had it open, though I think the caps were on the pcb there, not straight on the jacks.

On some of the nano's that I have traced, I noticed a small cap around 100pF on both input and output, allways in circuit.
I don't think I have seen a factory schematic with them added but surely have seen them on schematics where people traced an EHX pedal. (See example below).

Why would they do that? Why not move the cap after/before the bypass switch to keep them out of the signal chain when bypassing the pedal.
If you have 5 of these pedals on your board it quickly adds up...
Examples:

Or this one:

[antonis]

Those 100pF caps act as RF rectification prevention..
(especially those with series ferrite bead..)

[j_flanders]

Those 100pF caps act as RF rectification prevention..
(especially those with series ferrite bead..)

I kinda knew that, but the question is why they're also there/active in 'true' bypass state/mode?
Why don't they put the ones at the input after the switch and the ones at the output before the switch?

[r080]

I suppose putting it right on the jack shuts it down right away before it gets anywhere. Those designs might have also been from a time that many people would have had more than one buffered, and might not have noticed it.

[anotherjim]

Interesting. I think it's possible there was a product engineer enacting RF emission regulations in a totalitarian fashion. Filtering every connection to the outside world is how to do it (the DC jack too). The tone suck from many such pedals on top of the cable capacitance wasn't considered.

[j_flanders]

Those designs might have also been from a time that many people would have had more than one buffered, and might not have noticed it.

Although there are some older (big box) pedal where there's a resistor + cap at the output, the 100 or 120pF cap + ferrite bead is something that started with their newer (nano + XO) pedals.

enacting RF emission regulations in a totalitarian fashion.

So far, that's the only reasonable explanation.
With the cables connecting to the pcb first (instead of straight from jacks to footswitch) and seeing a 3PDT inside it's very much not apparent at first sight that all these are not really true bypass in a manner which most consider true bypass to be.

(the DC jack too).

Ow, I didn't even notice they did it there as well.

[Phoenix]

EHX/New Sensor reached a $450kUSD settlement with the US FCC in 2014 for code violations.

The addition of cap/ferrite to their products was most certainly because of this case. It's a brute-force approach to prevent conducted emissions that would have essentially no R&D cost and require little re-engineering of existing product lines or re-work of inventory, and "tone" would have been entirely secondary to compliance with regs in the most rapid and cost-effective manner.

I've not reverse-engineered any more recent EHX products, they may or may not still feature this issue. With the costs of a good EMC compliance engineer, the potential for failing of the quite expensive compliance testing, re-spin of designs if so, or equipment for in-house pre-compliance testing, it may still make economic sense for them to stick with the brute-force approach if that's working for them. It's easy to argue that if you've got a big true-bypass pedalboard it's adviseable to be using a buffer somewhere earlier in the chain anyway.

[j_flanders]

EHX/New Sensor reached a $450kUSD settlement with the US FCC in 2014 for code violations.

The addition of cap/ferrite to their products was most certainly because of this case. It's a brute-force approach to prevent conducted emissions that would have essentially no R&D cost and require little re-engineering of existing product lines or re-work of inventory, and "tone" would have been entirely secondary to compliance with regs in the most rapid and cost-effective manner.

I've not reverse-engineered any more recent EHX products, they may or may not still feature this issue. With the costs of a good EMC compliance engineer, the potential for failing of the quite expensive compliance testing, re-spin of designs if so, or equipment for in-house pre-compliance testing, it may still make economic sense for them to stick with the brute-force approach if that's working for them. It's easy to argue that if you've got a big true-bypass pedalboard it's adviseable to be using a buffer somewhere earlier in the chain anyway.

Thanks! Great post and also a good read in the links.

[Mark Hammer]

I suspect it is also an adaptation to the omnipresence of switching power supplies.  Yes, it does nothing for the cleanliness of the power, but heterodyning produced by the interaction of a switching power supply and digital effect circuitry has a way of leaking out the audio jacks.

[j_flanders]

I suspect it is also an adaptation to the omnipresence of switching power supplies.

In the manual of most of those nano and xo pedals, EHX explictly states to only use the adapter supplied with the pedal. And those are never of the switching type.

Use only the 96DC-200mA adapter the [insert pedal name] comes supplied with.
Do not use any other DC adapters. Using other DC adapters, even those made by Electro-Harmonix, could cause harm to the unit, the adapter or you.

[Mark Hammer]

They ALL say to use their power supplies, because they don't want to deal with repairs or returns resulting from people using completely inappropriate supplies.

[anotherjim]

Note the order of the L-C filters in the EHX schemes. They are all facing outward. To stop RF getting out, not getting in.

There are of course a huge array of components to support EMC regulations. The "Pi" filters are very common. A suitable filter goes in EVERY outside connection, even a footswitch jack.
https://www.mouser.co.uk/datasheet/2/281/ENFD0006-1140683.pdf.

[Ben N]

Re. the New Sensor FCC consent decree, HG is a digital device, but DMM?

[anotherjim]

BBD clock is a square wave. Even if the clock frequency is only just ultrasonic, the harmonics are RF.

[j_flanders]

but DMM?

DMM surely is (according to their definition of digital) , but what about the examples I posted: LPB-1 and Germanium OD.
They're not too different from a Big Muff when compared to the Holy Grail, DMM and Small Clone, and the text in one of the links, which explains what is considered as a digital device, says that the BMP is not digital:

http://www.effectsbay.com/2014/05/fcc-regulations-for-pedals/

Taking the Electro-Harmonix (EHX) brand as an example, the Big Muff π distortion pedal is not considered a Class B digital device because it does not generate any signals above 9kHz nor does it use any digital techniques, though it can process signals in excess of 9kHz that are present at its input.

The EHX Small Clone analog chorus pedal on the other hand is considered to be a Class B digital device. It uses a logic IC to output a digital clock signal ranging from 20kHz to 80kHz. The Small Clone exceeds the digital device minimum of 9kHz by close to a factor of 10.

[anotherjim]

Emissions and analogue audio is a can of worms. You could argue that clipping can produce RF harmonics, or that mixed signal in a non-linear circuit can produce higher intermodulation sum frequencies. You might also argue (I won't!) that the likes of EHX are selling a mixture of compatible devices (stompboxes) that may comprise a "system" (pedalboard) and if one element can produce RFI then all of their products should suppress it.

[amz-fx]

The LPB is a simple transistor booster and should not be generating any frequencies that will be a problem. However, if they felt a need for this filtering, they could have used these same components by putting them on the pcb side of the switch where they would not impact the audio signal when the circuit is bypassed.

http://www.muzique.com/news/modding-the-eh-lpb-1-nano/

regards, Jack