Valve amplifier picking up EMI/RFI

[Gabe001]

Just wondering if I could pick your brains

I noticed that my armstrong 220 clone (el84x4 and ecc83x3) was making a random faint clicking/popping noise through the speakers (both of them). I'm not sure whether this was an issue from the start or whether this is new, but I only noticed it a week ago. It's not very obvious. It affects both channels. Swapping el84s with those in Mullard 5-10 made no difference.

Disconnecting the preamp made no difference. Switching speakers made no difference, although it's more obvious in my higher sensitivity mission speakers

So I fired up my Mullard 5-10 and the problem still persists. In fact it's worse on the 5-10.

This is a relief as it's not a problem with the amplifier per se. Of course it could still be the power tubes, but it's unlikely that all 6 of them are bad.

I think my amps are picking up RFI/EMI, environmental or via mains.

Using my little experience from restoring radios, I tried a mains filter cap between mains L and N and it didn't help. However, a 22nf cap between mains neutral and chassis (which is grounded to mains earth), greatly attenuated this random noise. I'm not quite happy with this solution due, just in case the amp is plugged into an incorrectly wired socket. Rob Robinette calls this a death cap.

Lots of possible culprits in my room ; Power line ethernet adapters, WiFi router, TV with WiFi, z wave home automation devices and the usual appliances.

Has anyone got any experience with this, or any ideas?. A quick Google reveals that the belkin gold series surge protector with interference protection has helped someone in the past, who localised the problem to the power line ethernet adapter (homeplug). Another person who had a similar issue localised the source to the sky box.

Thanks
Gabriel

[Radio Wrangler]

Any new appliances in the house. How often are the clicks. Do they vary with time of day?

It could also be due to the failure of a suppressor in a boiler, fridge or freezer etc.

Try unplugging everything else.

David

[murphyv310]

Try winding the mains cable around a ferrite ring or use the clip on ferrites on the mains cable. Sounds like a mains bourne interference issue.
Why on earth are you plugging into an incorrectly wired socket? If you know about it then fix it!

[Gabe001]

Originally Posted by murphyv310

Try winding the mains cable around a ferrite ring or use the clip on ferrites on the mains cable. Sounds like a mains bourne interference issue.
Why on earth are you plugging into an incorrectly wired socket? If you know about it then fix it!

I'm not!. My sockets are fine. I can't vouch for other people's though

[Gabe001]

Originally Posted by Radio Wrangler

Any new appliances in the house. How often are the clicks. Do they vary with time of day?

It could also be due to the failure of a suppressor in a boiler, fridge or freezer etc.

Try unplugging everything else.

David

I'd say every 15-20 seconds. I don't think they vary significantly during the day. Sounds like I've got my work cut out unplugging everything. Was hoping to avoid this

Murphy310: would these ferrites be ok? (See attached)

[Roger Ramjet]

Hi Gabriel,

Given the improvement after you fitted a capacitor between the incoming mains Neutral & Earth then it would seem to be a mains borne issue, so out of the possible culprits you list I, would suspect the Power Line adapters because they piggy back signals direct onto the house wiring.

Rog

[murphyv310]

Originally Posted by Gabe001

I'd say every 15-20 seconds. I don't think they vary significantly during the day. Sounds like I've got my work cut out unplugging everything. Was hoping to avoid this

Murphy310: would these ferrites be ok? (See attached)

Hi
Yes these ferrites should help.
David is correct, try disconnecting items around the house and switching off the mains to the boiler etc. To be fair it must be fairly powerful to effect a valve amp. Ferrite beads onto the EL84 grid coupling caps can also help. I get breakthrough on my Mullard 3-3 when transmitting on 2 meters with 50 watts but it's only slight.

[G6Tanuki]

Speaker cables can be a big issue, specially in amps that use negative feedback, it can provide a nasty reverse path for noise to get back into the low level audio stages. Think of the speaker leads as an end fed wire antenna...

Winding the speaker leads round a ferrite rod or looping them through a ferrite toroidal core was the first step when in the past I had issues with hi-fi stuff picking up our mobile radio transmitters.

[lesmw0sec]

Regarding clicks, I have a bedside rechargeable phone which can be picked up causing clicks on AM, but only when it is fully charged.

[Gabe001]

Originally Posted by G6Tanuki

Speaker cables can be a big issue, specially in amps that use negative feedback, it can provide a nasty reverse path for noise to get back into the low level audio stages. Think of the speaker leads as an end fed wire antenna...

Winding the speaker leads round a ferrite rod or looping them through a ferrite toroidal core was the first step when in the past I had issues with hi-fi stuff picking up our mobile radio transmitters.

Can I use the ferrite beads shown in the attachment above here too?

[Radio Wrangler]

Don't know about the ferrite beads in the attachment. There are lots of different ferrites with rather widely varying properties. It's a bit like the possibly apocryphal story of the man who went into a shop and asked to buy "A resistor".

Ian White wrote a good article on suppression ferrite selection for his RADCOM "In Practice column. Have a look for his website: http://www.ifwtech.co.uk/g3sek/
and you'll see the size of his suggested core in his hand in a photo near the bottom.

These are not cheap cores.

Note that they are chosen for good loss in the HF bands from about 1MHz upwards, and require the cable to do several passes through the hole. They work at RF. I use them.

BUT your interference could be baseband clicks and transients on the mains with much lower frequency components. Most ferrite cores are going to lose their effectiveness down there.

15-20 seconds spacing sounds like a thermostat or maybe an intermittent SMPS keeping something topped up.

Valve amps are pretty good at ignoring RF muck. Without any evidence or having seen the problem myself, I still lean towards the switching clicks on the mains rather than RF. The capacitor mention seems to fit this. I think it needs something fairly powerful to upset a valve amp.

With an RFI fix (or clicks) you rarely get as much improvement as you'd like in suppressing one place. Ideally you want to do some fixing at both the source and at the victim.

If it is a switching click, it could well be a contact getting zapped periodically without its suppressor capacitor.

Until you've found the blighter, it's mostly guesswork.

David

[Diabolical Artificer]

Scope the input and HT rail after the rectifier before the reservoir/smoothing cap if your scope can take it, that and have a look at the mains though just hook onto your heater winding to see if you can see any spikes. Another area to check is if any of the valves have a dodgy internal connection.

I have some big ferrites rings and some the same as in that attachment I can send you to try if you want em Gabriel, drop me an email later

Andy..

[Gabe001]

Not sure if this will work, recorded with my phone next to speaker. You can here 3 clicks/pops at around 8-10 sec. Ignore the noise at the end, that's just me handling the phone.

Andy the problem is in 2 different amps. It seems to be the more sensitive amps that are worse affected. The Olson doesn't have any audible noises but then the input sensitivity of this is much less than the Armstrong/mullard

A 22nf cap between rectified HT and chassis (mains earth) didn't help. Maybe a caps from each heater winding to ground would if that's possible, the scope on the heaters is a good idea I think.

The only new recent installation in the house is a smart meter. I hope that's not the problem. The powerline adapters seem ok - they can be noisy but aren't the problem. I've started unplugging appliances without success so far, but I haven't done the boiler/cooler/fridge/freezer yet.

I'll email you re ferrite bead to try if you've got some spare

[Gabe001]

David, thanks for the link and your help

I see now why you said that I need to find the culprit. The ferrites are tailored to offending frequency.

Can I confirm that I'd need one with a significantly wider inside diameter than the cable, so that I can wind it through several times?

Also, does a ferrite core work against mains borne interference, or is it for airborne signal shielding only?

G

[60136 Alcazar]

Originally Posted by Gabe001

However, a 22nf cap between mains neutral and chassis (which is grounded to mains earth), greatly attenuated this random noise.

In the old days there was a standard circuit for reducing mains-borne interference, consisting of capacitors from live to earth, neutral to earth and live to neutral. The live to neutral was on one side of an RF coil in each of the live and neutral feeds, whilst the other two were on the other side of the coils. These days X and Y rated capacitors would need to be fitted. That said the first line of defence back then was establishing if the interference was being generated inside or outside your home - if outside you would fit the above circuit; if inside you would track down the culprit and suppress at source!

[Gabe001]

Originally Posted by 60136 Alcazar

In the old days there was a standard circuit for reducing mains-borne interference, consisting of capacitors from live to earth, neutral to earth and live to neutral. The live to neutral was on one side of an RF coil in each of the live and neutral feeds, whilst the other two were on the other side of the coils. These days X and Y rated capacitors would need to be fitted. That said the first line of defence back then was establishing if the interference was being generated inside or outside your home - if outside you would fit the above circuit; if inside you would track down the culprit and suppress at source!

Yes thanks John, my attempt at mitigating this problem came from the radios I worked on, some of which have these filtering caps in some flavour or another.

I have concerns about a live to earth cap in an exposed chassis though. Apart from the safety of it (maybe I'm worrying a bit too much, given that the chassis is earthed) the current leakage of approx 3ma (depending on the cap size) may trigger the circuit breaker (I think)

G

[Radio Wrangler]

Originally Posted by Gabe001

I see now why you said that I need to find the culprit. The ferrites are tailored to offending frequency.

Can I confirm that I'd need one with a significantly wider inside diameter than the cable, so that I can wind it through several times?

Also, does a ferrite core work against mains borne interference, or is it for airborne signal shielding only?

G

For lower frequency suppression, It's usual to have a large ring-shaped core to get the cable through the hole a few/several times. Inductance increases as the square of the number of turns.

For each type of ferrite material there are bands of frequency where it has different behaviours. At the lowest frequencies, it makes inductance, but not enough to be useful for any purpose, so the core might as well not be there. Next up in frequency is a range where it makes useful inductance with low losses. As we go higher in frequency, the losses get bigger. You can no longer make high-Q inductors, but there are still some uses and the losses are still OK for making acceptably efficient transformers. As we go higher still in frequency the losses become dominant. The ferrite isn't junk, though. Up here it becomes a valuable suppression material. It absorbs RF interference, turning it to heat - which is a lot better than just reflecting it.

If your interference is RF, ferrite cores are effective. If it's 'fridge clicks' then you may get better results from capacitors.

Oh, and most of the cores mentioned so far seem to be 'split cores' in plastic cases. They are very convenient for snapping around a cable but it's harder getting more than one turn in. The real sting is in their price. To make a split core, they first make an unsplit core, then use a diamond saw to cut it in two. The cut faces then need polishing to allow them to be mated with minimum air voids... more diamonds... is this beginning to sound expensive? Save money by using cores without splits. They work a little better too. Use split cores for cables where you can't remove a connector to do the threading.

You can get the same issue with two different amps, and it's worse with more sensitive amps.... I'm beginning to suspect an earth loop problem.

David

[G6Tanuki]

If it is mains-borne interference, you can get a range of line-filters designed for the job; here I've got a couple made by Roxburgh that are built into a back-to-back IEC kettle-lead-style plug/socket/box which makes insertion in the mains-lead quite easy. Hopefully they're made to an appropriate standard so you don't need to worry about leakage currents, whether to use X or Y class capacitors etc.

[wd40addict]

In my case power line ethernet adapters caused something like what you're describing and were ejected from the house. We subsequently had a smartmeter fitted and I've not noticed any issues with it.

[Gabe001]

Originally Posted by G6Tanuki

If it is mains-borne interference, you can get a range of line-filters designed for the job; here I've got a couple made by Roxburgh that are built into a back-to-back IEC kettle-lead-style plug/socket/box which makes insertion in the mains-lead quite easy. Hopefully they're made to an appropriate standard so you don't need to worry about leakage currents, whether to use X or Y class capacitors etc.

Is this it? A bit more pricey than I expected