Freezer causes audible clicks on audio amplifier, but only when on same ring

[Obelix27]

Hi - my 1 year old table top freezer causes audible clicks on my (good quality) audio equipment every 15-30 minutes as it turns on/off. What is particularly surprising to me is that the clicks only happen when the freezer and audio are on the same mains ring. I have monitored the output of the audio equipment on a storage scope and the clicks pretty much disappear if say the freezer is on the kitchen mains ring and the audio on the living room ring.
Most of the wiring in my flat, and certainly the kitchen and most/all the living room, were rewired 7 years ago. And a new Wylex fuse box installed.
I don't understand why putting them on different mains rings "fixes" the problem. Even if I connect them both to the kitchen mains ring, which was definitely completely renewed 7 years ago, then I see the clicks .
Any ideas what is going on?

 

[James]

What is the make and model of the freezer?
Also are you moving it physically further away when you plug it in the kitchen socket?

 

[Obelix27]

It is a small Russell Hobbs RHTTFZ1B table top freezer.

I haven't physically moved the freezer or amplifier. I am just using two extension leads for all the experiments, and only when they are plugged into the same ring do I get the problem. Odd as obviously the rings meet at the same place!

 

[Lucien Nunes]

The clicks are caused by the inductive switching transient as the thermostat contacts open or close. Interference can be both conducted along the mains lead and radiated from the freezer and cabling. If the freezer and audio system are on the same circuit, the coupling between them for conducted wideband interference may be quite high. When they are on separate circuits, to get from one to the other the interference must pass through the consumer unit. The impedance of the MCBs in series with each circuit, along with the low impedance looking back towards the supply, will significantly attenuate the interference possibly to below audible level.

Which is the culprit? So far it is hard to say. Good audio equipment should be fairly tolerant of average mains-borne interference. I design professional audio gear and you can happily start and stop any kind of freezer / welder / monster hydraulic press / other aggressive-sounding machine on the same circuit as one of my gadgets without any clicks or pops. But even top quality domestic listening gear is not always provided with good electrical defences against the outside world. OTOH, some domestic appliances are unduly prone to radiating hash, making you wonder what suppression they added to the sample units to get them to pass EMC tests.

Try putting a VDR-type surge suppressor at the freezer end of its cable, and/or a capacitive filter at the HiFi end, or some other permutation of your choosing.

 

[Obelix27]

The clicks are caused by the inductive switching transient as the thermostat contacts open or close. Interference can be both conducted along the mains lead and radiated from the freezer and cabling. If the freezer and audio system are on the same circuit, the coupling between them for conducted wideband interference may be quite high. When they are on separate circuits, to get from one to the other the interference must pass through the consumer unit. The impedance of the MCBs in series with each circuit, along with the low impedance looking back towards the supply, will significantly attenuate the interference possibly to below audible level.

Which is the culprit? So far it is hard to say. Good audio equipment should be fairly tolerant of average mains-borne interference. I design professional audio gear and you can happily start and stop any kind of freezer / welder / monster hydraulic press / other aggressive-sounding machine on the same circuit as one of my gadgets without any clicks or pops. But even top quality domestic listening gear is not always provided with good electrical defences against the outside world. OTOH, some domestic appliances are unduly prone to radiating hash, making you wonder what suppression they added to the sample units to get them to pass EMC tests.

Try putting a VDR-type surge suppressor at the freezer end of its cable, and/or a capacitive filter at the HiFi end, or some other permutation of your choosing.

Thanks very much for the detailed response. The audio equipment is Quad 909, Quad 707 and Quad 405 power amplifiers which are good quality British made amps, that weigh about 12kg each so have a serious power supply inside!

The MCB/fusebox presumably has essentially zero impedance. But unless the freezer interference is very high frequency (I guess radio frequency) I would have thought that the freezer's noise must be the same over my entire home. I don't see how the interference can go "one way" along the mains cable. Or is it a result of the small but non zero resistance of the mains cables?

FWIW Russell Hobbs say "When the internal temperature of the fridge drops below the required level, a capacitor will activate which starts up the compressor. When this happens the fridge has a greater electrical draw than during it's ambient cycle". I really struggle with this not being a euphemism for "terrible design"!! How much can a small freezer compressor draw, and the problem occurs equally when the compressor turns off.

Any idea about a mains filter / suppressor? Taking to account the freezer cost £100 I don't want to spend a lot. The freezer is still under guaranty but I guess the chances of getting a refund are about zero. Unless I can somehow challenge them on the EMC tests...

 

[DPG]

I used to have a Quad 405-2 amplifier. Very nice it was too. I built a separate power supply for it external to the amp and with nice big audio grade capacitors. I'd like to say it made a big difference but unfortunately my hearing is not sensitive enough to detect such differences.

 

[Lucien Nunes]

The Quads are excellent amplifiers, but neither the 405 nor 606 (I'm not familiar with the innards of the 909) have any specific shielding between the mains wiring and the signal, nor input filtering on the connector. Therefore I would expect them to be at least moderately susceptible to nasty noises on the mains.

Are you sure it's the power amplifiers picking the noise up, not the source device / preamp / interconnections? If you try to localise the point it is getting in by disconnecting cables, don't forget to short-circuit any active but unused inputs. E.g. if you unplug the control unit from the power amp, the power amp inputs will be unduly sensitive to noise until shorted. I can't recall the input sensitivity or impedance but I use a Quad 520 for my workbench monitoring system and it is very ticklish with the inputs open.

 

[Darkwood]

@Lucien Nunes

Is there anything you don't ruddy know lol.. excellent responses and deffo not my field.

 

[Obelix27]

The Quads are excellent amplifiers, but neither the 405 nor 606 (I'm not familiar with the innards of the 909) have any specific shielding between the mains wiring and the signal, nor input filtering on the connector. Therefore I would expect them to be at least moderately susceptible to nasty noises on the mains.

Are you sure it's the power amplifiers picking the noise up, not the source device / preamp / interconnections? If you try to localise the point it is getting in by disconnecting cables, don't forget to short-circuit any active but unused inputs. E.g. if you unplug the control unit from the power amp, the power amp inputs will be unduly sensitive to noise until shorted. I can't recall the input sensitivity or impedance but I use a Quad 520 for my workbench monitoring system and it is very ticklish with the inputs open.

I have unplugged the Quad amp from the preamp and everything and shorted the inputs and I still get the spikes. I did some more tests yesterday (unfortunately I forgot to short the Quad's inputs for all the tests) where I plugged the Quad into the same socket as the Russell Hobb and had it physically close. Then I turned the freezer off, and then tried it with the Quad next to my Bosch freezer (and the Russell Hobb off). The results are pretty surprising! The Russell Hobb causes a lot of spikes (overload of measuring equipment is 0dB), whereas the Bosch freezer is nice and quiet. And of course there is a background of 50Hz noise.
Going across circuits reduces the spikes a lot.
How can I get Russell Hobb to take this thing back??? I know little about EMC tests but surely it can't pass them?!

 

[Lucien Nunes]

deffo not my field

I have an unfair advantage here because professional analaogue audio happens to be one of my regular fields of work. I often get calls to make gadgets or hack and customise other products. Because the stuff I make travels the workd and must work without fail under whatever conditions it encounters, I have to be very confident about sensitivity to interference, earth loops, bad power etc.

The real challenge is that sometimes the design cycle is very short. I have made PCBs while the courier is outside waiting to take away the finished unit. So I sometimes have to be very conservative in design because it absolutely must work perfectly first time. There is no time to debug anything and the penalty for failure is severe.

E.g. I made a little control box today in between other jobs (it wasn't what I was supposed to be doing.) 8 opamps on a piece of stripboard, four pots and two switches on the top. soldering one bit while designing the next.. Wired it up, put the lid on, tested it and handed it over. I don't know where it will be tomorrow but it won't click, thump or buzz and I would bet you £10k that it will still be working in 20 years (no warranty on the pots!)

Rambling... time to go home.