7

My fridge is sending bad clacks to my loudspeakers when it turns on/off, I fear for the tweeters. I'd add a snubber/suppressor to the compressor but my notions about RC filters from school are 15 years old and I have no idea how to dimension it.

It's an old european fridge, so mains are 220V 50Hz, and the compressor mentions two other values In=0.8 and Icc=6.3, which I suppose are nominal and startup current. Do I need to compute the motor inductance? What condensator/resistor values to pick? Is an RC snubber even the correct fix?

How to estimate the RC values for snubbing an A/C motor?


Trying another approach. I've found a some usual R/C values for suppressors. For instance this one is 47ohm / 0.1µF, giving 33kHz of cutoff frequency... but the datasheet says up to 62Hz? Why is there 3 orders of magnitude difference ?


EDIT: in the end I added a capacitor to the fridge and that resolved the problems. That was quite long ago so I don't remember the exact value, but if anyone wants it just bump me…

stevenvh
  • 145,145
  • 21
  • 455
  • 667
  • Argh. I just noticed this question is 1 1/2 years old *after* writing a answer. I should have checked, but we've been getting a lot of old questions dredged up from the deep past lately. – Olin Lathrop Aug 04 '12 at 14:58
  • @Olin - Sorry, my fault. I'm editing/cleaning up tags, and these had to be done manually, because the same tag was used in two different meanings. That's why they bumped the list. Nevertheless, thanks for answering, it's always useful. – stevenvh Aug 04 '12 at 15:20

3 Answers3

3

You may want to consider using an NTC inrush limiter in series with the compressor instead of a snubber. Make sure you choose one well-sized for the steady-state current of the appliance.

You may also wish to try putting an appropriately-sized metal-oxide varistor across the mains to clamp any badness generated by the compressor.

A safety-recognized X-capacitor across the mains may also help suppress noise.

There are also off-the-shelf EMI filters that could help with the noise if you prefer a pre-fab solution. I would guess that the motors in a washing machine wouldn't be so different from those in a fridge from an EMI standpoint.

Adam Lawrence
  • 32,921
  • 3
  • 58
  • 110
  • About the thermistor, I doubt it would help with the spike when the motor shuts down? Also, your varistor/capacitor links are exchanged :) – Damien Pollet Jan 10 '11 at 17:05
  • The varistor looks like a protection for exceptional surges, like lightning? I get a spike each time the fridge starts or stops… – Damien Pollet Jan 10 '11 at 17:17
  • The NTC will limit inrush current when the compressor starts, yes. Once it's "hot" it drops to a low resistance and as you stated won't likely help much on turn-off. – Adam Lawrence Jan 10 '11 at 20:47
2

The fridge may be dumping noise on your power line, but the real problem is the amplifier driving your speakers. Either the amp itself has bad line filtering, or your setup allows noise to get into its audio inputs.

Disconnect the audio input from the amp and see if the fridge can still cause clicks in the speakers. If so, then the amp has a crappy power supply circuit and needs help in the form of a line filter.

If the noise goes away when the input to the amp has been disconnected, then the noise is already on the audio signal by the time it gets to the amp. There are two likely causes to this, bad shielding and bad grounding. If you are using regular off the shelf shielded audio cable and it is properly connected at each end, then shielding is probably not the cause. The most likely cause is then a ground loop. Is the amp plugged into the same outlet as whatever equipment is producing the line level audio signals going into it? If not, then you have a ground loop. The fridge is probably causing a significant ground bounce when it turns on and off, and this is getting into your audio signal due to the different ground points. Plug all the audio equipment into the same outlet strip plugged into a single outlet in the house. Adding a line filter to that wouldn't be a bad idea.

You can try to reduce the noise the fridge is making, but that's just one noise source of many that the audio system needs to be immune to anyway. Crap on the power line happens, regularly. The audio system should be set up to deal with it properly.

Olin Lathrop
  • 310,974
  • 36
  • 428
  • 915
  • It would be more useful to disconnect the amp *and short it's inputs*. Leaving the amp with floating inputs may simply exchange one noise source for another. – Connor Wolf Aug 05 '12 at 06:18
  • @Fake: Good point. – Olin Lathrop Aug 05 '12 at 12:27
  • As I said in my comment on @tyblu's answer, the clacks happen when the sound source is my laptop, not when it's my desk computer. The laptop powerbrick does not have a ground connection, the way I use it. I can't imagine that recent Genelec monitors would be badly built electrically, and my experience matches your suggestion that the noise is transmitted through the source. In the end I added a capacitor to the fridge and that fixed the problem; I did have a longer cord with ground connection for the laptop's power brick, but I don't remember if I tried that… – Damien Pollet Aug 05 '12 at 18:00
  • 1
    See? It never hurts to post a good answer :-) – stevenvh Aug 05 '12 at 18:48
2

Your fridge may have poor grounding. I've run into more than a few old ones that had a floating ground that intermittently 'closed' to Earth, one of which gave me my first close encounter with mains power (and later used to shock people at parties, but that is another tale). Measure the compressor chassis voltage versus mains Earth/ground (or see if there's only 2 prongs in the plug). Of course ya can't just short the chassis to ground straight away if this is the issue -- see if there's a constant current path, first. I'm not sure why there would be, but it's possible! If there is, I'd look into it some more: why is the chassis energized? Is some insulation worn through? Oh yeah - don't get killed or anything.

The other way to go about this is to protect the speakers directly. This can be done with clamping diodes or other TVS parts.

tyblu
  • 8,167
  • 6
  • 40
  • 70
  • There are green/yellow wires and the plug should have 3 prongs, but I have to get back behind it to measure anything. As for the speakers, they are near-field monitors each with their internal amp/alimentation so I'm not sure what I can do to them; the clack most probably comes from the signal wire because they only happen when they are plugged to my laptop; the desktop computer must have a better ground connexion than the laptop's power brick. – Damien Pollet Jan 11 '11 at 08:30