Trying to find a vibration or sound sensing switch / circuit

[chrisv2]

I am building an intercooler for my home air compressor. This is basically a heat exchanger with a fan blowing over it, to help drop the temp of hot air as it exits the compressor pump head. I am also integrating a water trap, to pull the condensate out before the air continues on to the main tank of the compressor.

This unit needs to be turned "on" when the compressor is running, then turn "off" when the compressor stops. I could hardwire it into the control box of the compressor but I'd rather not do that, as the compressor is 220V and this intercooler will be 120V (plus I just don't want to mess around with the compressor wiring).

I was thinking that if there was some sort of simple sensor/switch I could adhere to the side of the compressor, to detect either the vibration of the compressor or the sound -- and then turn my intercooler on, and likewise in the absence of this condition turn the intercooler off -- this might work?

Needs to be simple, inexpensive and reliable. Do you know of anything out there already or is this easy enough to build?

 

[Arouse1973]

Hello
Yes indeed, there are literally hundreds of vibration sensors out there with varying sensitivities. Where are you based, what electronic components suppliers do you have where you are?
Adam

 

[chrisv2]

USA. I was searching more and couldn't find anything that was "simple" enough. I did find a very simple "shake switch", which is basically a conductor positioned inside a coil of "springy" wire, and will make contact as the switch is jounced around:

http://www.amazon.com/gp/product/B0..._m=ATVPDKIKX0DER&pf_rd_r=0PSG8S536HJHCZE5RV4A

If I used something like this, with a transistor to increase the current capacity (enough to power the coil of a relay), and I added a large enough capacitor to smooth out the "jitter" of the switch bouncing around -- do you think it would work to keep the relay energized while the compressor is running? I was doing some simple calculations and it seems like a 3300uf capacitor ought to be plenty to keep a relay with a coil resistance of 100 ohms energized for around 500 milliseconds, which should be more than enough headroom to cover the bouncing of the switch. When my compressor turns off, all goes quiet and the relay should de-energize.

Does this make sense or am I off on a dumb path?

 

[Arouse1973]

I think you make perfect sense. Let me see if I can find some others for you.
Adam

 

[chrisv2]

awesome, thanks! I am trying to keep this ultra-simple. My idea needs: shake switch, transistor, resistors, capacitor, transformer, rectifier, relay. Already a high part count...ugh...

 

[Arouse1973]

Ok so yeah I found the same switch, I think it might work. Just putting a simple circuit together for you.
Adam

 

[Arouse1973]

This is as about as simple as it gets. U1 is your vibration sensor. We might need to modify the circuit depending on how long the switch stays closed for in between vibrations. You will need to choose a transistor that can drive your relay and also the diode needed will depend on the relay also. We may need to change the transistor to a Darlington if greater gain is required or we could use a MOSFET. It's a start for you anyway.
Adam

 

[chrisv2]

Very cool, thank you Adam! I was thinking that a relay like this might work: http://www.digikey.com/product-detail/en/T9AS5D12-12/PB306-ND/254523

I'd like to use a nice-sized 12" fan to blow over the heat exchanger. I should probably just plan on this entire thing being 12VDC that way I can get a DC fan and also run your circuit off the same power supply.

My compressor is pretty big, it's a 7.5HP electric motor and 80 gallon tank. You know when it is running. So, when it is running that little shake switch should constantly be banging around making contact. Is C2 meant to smooth out the jitter from the switch?

 

[Tha fios agaibh]

Problem is, most USA ac circuits don't have a neutral ran to it. you can't derive 120v from a 3 wire branch circuit without using the equipment ground as the return. This is a code violation and not safe. It'd be smarter to use a 240/120v stepdown transformer on the compressor contactor, or just use a 240v motor for your intercooler.

 

[chrisv2]

That's exactly why I used 6/3-WG copper when I ran the circuit for the compressor ;-)

I'm thinking of using a 12VDC automotive fan - a typical 12" or 14" radiator fan. I can get one for under $30. The same spec 120V or 240V fan is suddenly a $200 item.

 

[Tha fios agaibh]

Good deal!

 

[Tha fios agaibh]

Isn't it almost as expensive to add a 12v power supply?

Clever circuit by Adam, but why not just use the contactor or an aux contact to power your circuit?

 

[chrisv2]

I thought of that but I don't want to hard-wire into the compressor control unit. Also I have another compressor and was thinking I might want to move the intercooler from one to the other. Plus I wanted to see if there is a way to do the "sense vibration and keep relay engaged" trick

 

[Arouse1973]

Very cool, thank you Adam! I was thinking that a relay like this might work: http://www.digikey.com/product-detail/en/T9AS5D12-12/PB306-ND/254523

I'd like to use a nice-sized 12" fan to blow over the heat exchanger. I should probably just plan on this entire thing being 12VDC that way I can get a DC fan and also run your circuit off the same power supply.

My compressor is pretty big, it's a 7.5HP electric motor and 80 gallon tank. You know when it is running. So, when it is running that little shake switch should constantly be banging around making contact. Is C2 meant to smooth out the jitter from the switch?

Yes C2 is for supplying enough current to the base of the transistor when the switch break momentarily between vibes. The transistor I was thinking of using is a bit overkill as far as current handling goes, I didn't know what relay you were going to use. But now I do I can use a lower power transistor. I may need to add a small base resistor now to limit inrush current when the vibration sensor switches off. I'll build something and let you know.
Adam

 

[Arouse1973]

Here is a much simpler circuit using a MOSFET. The MOSFET is a bit over kill but should work Ok. Feel free to look for another one if you want. The circuit switches on in about 100 ms and off in about 50 ms. To increase the turn on and turn off time increase the value of C1.
Adam

 

[chrisv2]

Very cool - thank you Adam. I think the MOSFET is perfect and will give a wide tolerance for the choice of relay. I just have a couple of questions:

- the resistors are ok to be 1/4 watt?
- no harm in going with a larger value for C2 (say 2200uf) - right? I just want to make sure it stays on if the shake switch is less consistently "on" than I imagine it will be.

Elegant design, I think it's great!

 

[Arouse1973]

Hi Chris

Resistors are Ok to be standard metal film 1/4 W. The problem with going too high is that it will take longer to come on in the first place. If the switch is not coming on as constantly as you would like it then it may struggle to come on in the first place.

You could change the value of R5 to increase the turn off delay.

Let me change the value of R5 and see what happens to the timings

Adam

 

[Arouse1973]

Sorry Chris I gave you the wrong timings. It's about 0.5 s with a 22 K. If you change this to 47 K it's about a second. But the downside is it will take the relay a little longer to turn off. The MOSFET will heat up a little bit more, but for that length of time it will be fine.
Adam

 

[chrisv2]

I'm going to have to observe my compressor when it runs but I have to imagine that a 500ms "stay alive" time is going to be plenty.

 

[Arouse1973]

Yeah it would be good to give it a go. We can always tweak the circuit if we need to. Sometimes the simple circuits are the best, sometimes not. No harm in starting out simple though.
Adam