Tube PHONO preamp interference - RFI, EMI, bad grounding?

@lewm This has to be corrected though. As I pointed out earlier, if the chassis isn't directly grounded it will act as an antenna rather than a shield.

Back in the 70's and 80's in South Dakota where TV stations were rare, farmers used to park an old tractor on a hill near their house and then point their TV antenna at that rather than in the direction of the TV station. They were using the tractor to collect RF energy which is then re-radiated.

We don't yet know if this sound is air borne, but if it is it will be important for the chassis to be grounded. Its also important for electrical shock hazard safety.

The EAR has rave reviews, and it just seems improbable that Tim de Paravicini with his reputation would make this kind of mistake.

Tim passed away (RIP) about 2 years ago. The simple explanation is this unit was built incorrectly.

By my way of thinking, that actually makes the most sense if you are going to float the audio grounds; you want the upstream gear to be grounded at the same potential as that of the on board audio circuit. It seems that you want the ground lug to be at AC or earth ground, which is certainly not "wrong" either.

@lewm You want the ground post at chassis potential, not the audio ground. The grounding system is supposed to be ignored by the audio.

So the ground post is simply bolted to the chassis, along with the center pin of the IEC connector. The 33 Ohm resistor is then placed from the audio circuit ground to the chassis.

This sounds like an easy fix! It may not fix the RFi, but it won't hurt and it will make the unit safe and in compliance with EU directives.

I have measured the impedance between the ground of the RCA input/output and the chassis. You’re right - it’s 0.01 ohms.

Seems like there’s a poor grounding on the inlet. I wrote to EAR if that’s on purpose or not.

Thanks!

What can I do to find out what causes this 33 ohm resistance?

@filipm Grounding in audio products is poorly understood, especially in the context of meeting safety requirements! In the US its a bit of the Wild West; some know what they are doing and others don’t, often using techniques from the 1950s that are no longer considered kosher in the age of grounded outlets.

That 33 Ohm resistance isn’t a bad thing- its just that its in the wrong place. How it should have worked is that the connection between the IEC and the chassis should have read zero Ohms and 33 Ohms between the RCA grounds and chassis. In this way the chassis is held at ground by the power cord- and noise collected by the chassis is grounded. The audio circuit is thus floating at ground potential owing to the 33 Ohm resistor but shielded from the outside world. In this way ground loops are avoided even though the power cord is grounded.

Right now the chassis is less effective as a shield it can act as a bit of an antenna instead. You don’t get ground loops though, because there is a 33 Ohm resistor between the AC power cord ground and the audio ground. But in practice, this approach is noisier insofar as RFI and the like are concerned.

IMO the preamp needs to be modified- ’fixed’ or whatever you want to call it, but not so much to fix the RFI issue as it is to fix the potential shock hazard should something go wrong. You really want that chassis grounded so if something like a power switch were shorted to the chassis (for example if the unit were dropped and then plugged in), the fuse would instantly blow, protecting the user. Right now what would happen in that scenario is the 33 Ohm resistor would likely fail, leaving the chassis live, causing damage to equipment downstream and also being a possibility of a lethal shock hazard.

This is of course based on what you’ve told me so far.

I have measured the impedance between the ground plug on the inlet and the grounding pin of the chassis (or any screw) - it’s 33.2 ohms.

REF150 - the impedance between the ground on the inlet and any screw on the chassis - 0.01 ohms (0.00-0.02)

@filipm 

IF the first statement is correct and about the EAR (and succeeding posts suggest this is so) THEN the EAR is incorrectly grounded.

So we need to double check:

The ground pin of the IEC connection should tie directly to the chassis of the EAR. To not do so puts the unit at odds with EU directives for voltage safety.

The 33 Ohm resistance would then occur between the chassis (or ground pin of the IEC connection since they are the same) and the ground of the RCA connectors.

So here's another measurement to take. What is the resistance between the RCA connectors and the chassis? If zero Ohms or very nearly so, I think we found the problem despite remonstrations from EAR.

@filipm I think you have already determined that the ground pin of the IEC connector is tied to chassis. Is there an impedance between the chassis and the ground of the input signal? It may be that a component in the grounding scheme of the phono section has failed. This could cause noise to be injected into the phono section.

Regarding the power transformer noise in the rest of your equipment, some of that might be reduced by using a DC Blocker, provided the transformers in question are toroidal.

Its a common mistake to assume that physical transformer noise is getting amplified when you have problems like this. IME that is extremely rare; IMO you have two different problems.

If tips from phono stage manufs including Jud @ Joule-Electra(RIP), Kevin@K&K and everyone here couldn't resolve it, it was time to move on. 

@sbank If that was a JE phono section, the input gain stage was so starved for current that it easily rectifies RF energy and so will be prone to RFI problems. I think Jud didn't want to increase the current to a more linear region as gain was going to be sacrificed.

@filipm OK. That's 99 44/100% the EAR is at fault. What happens when it is not running off of the battery?

I would make sure the EAR is grounded via the power cord.

@filipm 

That it makes more noise with the inputs shorted is telling.

Of course you are living in a high RFI zone; that is why everything works fine elsewhere. Could you clarify this comment:

About the ARC - I have tried the same only with a Bryston B60 integrated, the noise has always been there.

Was the ARC in the system with the Bryston and is that with or without the EAR?

@filipm 

Although I agree with most of what @theaudioamp has suggested, I'm yet to be convinced this is actually the fault of the phono section, even though the phono section is clearly able to amplify the noise.

First off the 12AX7s have nothing to do with this. They are not somehow inherently sensitive to RFI. How their grid circuits are set up is an entirely different matter...

The length and exposure of the 20 foot XLR cables is a clue.

I think that what is happening is there is noise being picked up and injected into the ground of the ARC (which plays the same sound at a much lower level...). Since the ARC does not support the balanced standard (even though otherwise is balanced) a noise is manifest.

To sort out what is going on first short the inputs of the phono section to be sure its OK. Removing the top cover and no additional noise points to the idea that the phono section really isn't the culprit. But if shorting the phono inputs shuts it up then the phono section needs mods to protect it from demodulating RF (demodulation is what causes nearly all RFI problems). I've seen this sort of thing caused by poor tonearm cables too, but if the phono section is adequately protected from RFI its usually no concern.

If shorting the inputs has no effect I would be trying to use different interconnect cables. A lot of 'high end audio' manufacturers have a poor grasp of how balanced line is supposed to work so I would be trying a cheap microphone cable like you can get at Guitar Center since such a cable will be wired correctly and will also have an effective shield. Note that this cable is between the preamp and amp; the idea here is that noise is being injected elsewhere from the phono section but is able to be amplified by the phono section; the noise is in the grounds of the preamp and is being passed to the phono section.