PDA

View Full Version : Snubbing mech relay EMI from inductive load- Need advice on cap to use on the AC side



damage31
01-18-2012, 11:58 PM
Question: I am having issues with my stamp resetting when the relay it controls opens up. The relay controls an inductive water pump (toms aqualifter) load that is 18watts @ 120v AC.

I have found a site (http://www.controlanything.com/Relay/Device/DME1P56KF) that explains via a video pretty much exactly what I am seeing happen. Their solution is to put a cap on the AC line in parallel with the load. There is a schematic of what I would do on the above link. I am specifically thinking about using: a .56uF 400VDC / 200VAC Polyester Film Capacitor.


I am using a radioshack relay as seen here (http://www.radioshack.com/product/index.jsp?productId=2062481)


My question is: Is this (http://search.digikey.com/us/en/products/B32933A3564K/495-4359-ND/2415252)digikey part basically the same spec as the .56uF 400VDC / 200VAC Polyester Film Capacitor? I would assume so. Other than a slightly higher AC rating, it appears to fit the bill.

2nd Question: is this (http://search.digikey.com/us/en/products/ECQ-E4564KF/EF4564-ND/56594)part able to be used for emi supression? It doesnt call it out on the digikey web page like the other one above,but it is quite a bit cheaper. I'm not sure what the difference is between the two, but must have something to do with the EMI capability i would guess. This one here (the chearper one) appears to be closer in spec to the controlanything cap, but doesnt call out EMI.


I am open to suggestions...

PJ Allen
01-19-2012, 12:19 AM
Are you using a flyback diode across the relay coil?

bill190
01-19-2012, 03:17 AM
Google the words...
relay flyback diode

bsnut
01-19-2012, 06:51 AM
I am going to answer your questions.

My question is: Is this digikey part basically the same spec as the .56uF 400VDC / 200VAC Polyester Film Capacitor? I would assume so. Other than a slightly higher AC rating, it appears to fit the bill.
That cap will work for what you are trying to do. Remember that the voltage of the cap needs be above the operating voltage for what you are trying to do.

Are you using a flyback diode across the relay coil?
If you don't have flyback diode across the relay coil, you will need one. This diode across the load (relay coil or solenoid) protects the transistor from spikes when the load is turn off.

damage31
01-19-2012, 01:02 PM
thanks all- yes I am using a diode across the relay coil. I forget the exact part number, but it's a 1N400x series... something i had in my parts bin.
I am also using a 2n2222 to actually fire the relay coil. I can actuate the relay all day long with no AC load and not have any stamp issues.


My problem is specifically when the relay turns off- AND the AC load is connected...
my stamp will occasionally reset. (similar to what is shown in the video from my first post)

I figure that there must be a EMI spike from opening the relay contacts that is causing my stamp to reset. (but i dont have access to a oscilloscope to confirm the back emi, I just assume it is there)


Thus far in my attempt to fix this, I have sprinkled: .1uf caps around my stamp and the 4 other IC's power pins, as well as added a couple of 10uf, as well as a 4700uf just before my 7805 vin ...This has helped reduce the frequency of the stamps resets quite a bit, but not 100%.


I am running 3 LTC1257's, 1 DS1302, and the 1 5v relay + the stamp from a 7805 with a heat sink. The 7805 is cool to the touch. My circuits power consumption is very low, i forget the exact spec i measured a few months ago, but it was near or just over 90-100ma with the relay on if i remember right.

bill190
01-19-2012, 01:20 PM
Where is the relay physically located?

Near or on the same pcb as the Stamp? If yes, might try moving the relay away from the Stamp and see what happens. Then the AC wires will be away from the stamp.

damage31
01-19-2012, 06:49 PM
I'm limited at the moment- currently they share space in a project box, but theres ~1.5 inches between the relay and the stamp.

I think ill try the cap, and if that doesnt work, ill have to break the relay out into a different project box.

But I suspect that even a diff project box wont help till i can control the relay emi...

Mickster
01-19-2012, 06:58 PM
Is it out of the question to use a "zero-volt-switching" Solid State Relay?

Maybe a much smaller version of this:

http://www.omega.com/pptst/SSRL240_660.html (http://www.omega.com/pptst/SSRL240_660.html)

Mickster

damage31
01-19-2012, 07:28 PM
I did some reading and the best I can tell is that highly inductive loads are not recommended for 0 cross SSR because the current lags the voltage, so the reverse breakdown is still an issue, and could destroy the SSR.

0 cross is a good choice for resistive loads

Beau Schwabe (Parallax)
01-19-2012, 08:03 PM
Years ago I had a similar issue controlling a 240VAC industrial vacuum system.... BS2 controlled a DPDT 30Amp relay.

I found that the reset line to the BS2 was sensitive to the equipment noise, especially since I had about a 30 foot umbilical link acting as an antenna between the BS2 and the PC.

While a capacitor works well for inductive loads that are AC, and a diode works well for inductive loads that are DC, my solution was to 'harden" the reset-ability of the BS2 by tying a 1k resistor from the ATN (pin 3) of the BS2 to Vss (pin 4).

Mickster
01-19-2012, 08:04 PM
I did some reading and the best I can tell is that highly inductive loads are not recommended for 0 cross SSR because the current lags the voltage, so the reverse breakdown is still an issue, and could destroy the SSR.

0 cross is a good choice for resistive loads

I don't know where you got that from but.....

http://www.opto22.com/site/pr_details.aspx?cid=4&item=OAC5


M (http://www.opto22.com/site/pr_details.aspx?cid=4&item=OAC5)ickster

damage31
01-19-2012, 08:24 PM
that opto22 SSR is different than the couple I saw on digikey- i forget off hand who's SSR i was looking at, but i think it might have been a crydom

that opto22 one looks interesting: "Turn-off Time 1/2 cycle maximum-zero current"

I'll have a peek into it so see if i can get more info.

Mickster
01-19-2012, 08:32 PM
They've been around for decades, actually. We switched to the Potter Brumfield version because they were significantly lower in cost but they didn't have the OPTO-22's retaining screw. We drove solenoid powered hydraulic valves and never had an issue. We have since switched to DC.

Mickster