arc suppressor for relays

You ordered parts without looking at the data sheets? Or do you need help in interpreting the data?

Here's a typical TVS data sheet:

Take, for example, the P6KE220CA.

No more than 500nA at 188V At 1mA nominally 220V, but no less than 209V or more than 231V

At 1.85A (surely beyond your required clamping current) no more than

328V. At 10.3A, no more than 388V.

It's a bipolar device, so the ratings apply for either polarity of applied voltage.

I'd put it across the solenoid coil.

You could also try one of these, which you may have parts for in your junk box:

Those built-up ones are stupid expensive, you can just try a 630V

0.1uF film cap series with a 1/2W 100R resistor for < 1/10 the cost.

As an aside (which doesn't actually answer your original question), were you aware that relays need to be mounted so that the gap between the contacts is vertical if you want the longest life? This is one of these useful facts that was known by designers in the 1920s (or even earlier) and appeared in various textbooks on telephony, but now seems to have been forgotten.

The reason is that it reduces the chance of dirt falling on the contact surfaces and helps to dislodge any loose debris resulting from contact wear.

I have come across an industrial diswasher that was forever burning out small power relays simply because they were wrongly mounted.

--
~ Adrian Tuddenham ~ 
(Remove the ".invalid"s and add ".co.uk" to reply) 
www.poppyrecords.co.uk

I did look at the data sheets, they cover a wide range of devices, and have very little descriptive info on what the columns of numbers actually mean. I think I could have ordered a slightly higher voltage part and been sure it would be fine. I ordered the P4KE200CA. But, looking a bit closer, they show a range of min and max clamping voltage that look like they will be OK, but right on the edge. For peace of mind, I may just install one and then order some 220 or 240 V ones later to retrofit the rest of the unit.

Not a lot of space to put this on the board, I've got maybe 3mm on the back of the board and can put it between the pins of the relay. I don't think the caps would fit, but the 400W TVS should.

Thanks much,

Jon

These relays are semi-sealed (totally enclosed by plastic, but probably not hermetic (although the boards are probably washed after assembly, so they may be darn close to hermetic) so there is no external dirt to get in them. It is now burning the contacts severely, and you can hear the solenoid valves rattling and buzzing as the contacts arc. So, this is not a little speck of dust, but massively burned-away contacts. I expect on post-mortem, the contacts will be totally vaporized.

Jon

Can you use a solid state relay? Maybe off the board?

Another "designed by monkeys" consumer product.

Get some DC controlled SSR's (solid state relays), that can operate from 120 DC Or AC, diode them from the coils supply, place a cap on the control terminal of the SSR. The M1, M2 terminals will go across the contacts. What this does is the SSR first comes on before the contacts close, there by removing the load. when the contacts close, the SSR will not conduct and thus it won't be doing any work.

The cap on the control terminal allows for a slight time off delay, this allows for the relay to open and the SSR will be conducting for a very short time afterwards..

Actually, I think there are add on products that already do this.. Not sure.

Jamie.

There are 8 of these. If I can find a drop-in SSR that matches the pinout, I sure will do this the next time it needs to be opened up. Off the board won't work, there really is limited space in the "console" that sits above the the washer top/rear. If it was only one relay, it would have already gotten hacked as you suggest.

Yeah, they must sell a LOT of these $275 controller boards.

Jon

No room for double relays. I'm not sure your double relay scheme will fix the problem, anyway, it occurs on shut-off! The SSR would shut off first, then the mechanical relay would drop out later, and still arc (although the off SSR might have an internal snubber or avalanche mode that would reduce the arc.) First, there are EIGHT of these in a tiny space, so adding 8 more relays will never work. But, why use TWO relays when one good one will do it?

A cap big enough to delay the turnoff of the SSR could damage the relay driver chip, and would also keep the magnetic relay on.

Jon

I guess the relay switches 120 volts AC, right? What drives the relay? Is it

5 volts, other?

Is there an optoisolated triac arrangement available that could mount to the solenoids and wire to the board directly?

Can you observe the relay contacts to see if your snubber idea is effective? You might try something like an X rated 0.068 uF and 10 ohms in series across the contacts. You want to suppress the high voltage from the solenoid coil. Almost any amount will help.

What is the make of the washer? I would want to avoid it for sure.

Here is an AC opto coup[led triac that can switch 0.1 A/ 400 volts continuos. Google the datasheet from TI for some additional ideas to increase the current if needed.

tm

"Jon Elson"

** The relays are arcing because of the back emf from the solenoid coils at the instant of opening.

The time honoured fix is a "snubber " across the relay contacts.

100nF ( class X2) and 100 ohms in series is a good start.

... Phil

Keep in mind that SSRs virtually always fail 'on' = flood. Mechanical relays generally fail off (though I've seen them stick on very, very rare occasion and a very badly made relay).

Best regards, Spehro Pefhany

--
"it's the network..."                          "The Journey is the reward" 
speff@interlog.com             Info for manufacturers: http://www.trexon.com 
Embedded software/hardware/analog  Info for designers:  http://www.speff.com

Not sure about MOC3023 specifically but as I recall the MOCs are made specifically for driving real TRIACs, and are rated for that service ONLY.

The heavily inductive load (if this is indeed the failure mode -- remember, _it's EDMing right through entire blob-O'-metal contacts_) might not even allow the TRIAC to turn off, ever. The solution in either case (thrysitor SSD or mechanical) is an arc suppressor or snubber, as mentioned elsewhere.

The reason MOS SSDs were suggested is because they turn off so slowly, absorbing switching energy and damping transients. They have to be rated for enough peak thermal dissipation, which unfortunately isn't normally part of the datasheet. One would suppose the manufacturer rates them appropriately for the load current, both continuous, turning on, and turning off.

RC snubbers may work better than TVSs or MOVs because the latter allow any voltage up to breakdown. A spark will start on the (mechanical) contacts as soon as they open, growing in voltage roughly proportional to gap length. If the arc voltage never hits TVS/MOV breakdown, it simply won't do anything. MOVs may have some advantage because they have fairly high capacitance.

Now, a capacitor helps greatly because it limits turn-off slew rate; unfortunately, a cap across the contacts is a similarly bad idea at turn-on. A series resistor yields an acceptable compromise between turn-off peak voltage and turn-on peak current. For the solenoids in a washer I'd be very surprised if 0.1uF / 100 ohm didn't fix things completely. (A sudden transient would have to carry 2A to reach the same peak voltage of 200V that a TVS might begin conducting at -- more than enough, since solenoids can't carry more peak current than the steady state peak current, which is fractional ampere.)

Tim

--
Deep Friar: a very philosophical monk. 
Website: http://webpages.charter.net/dawill/tmoranwms

I once welded a 60A contactor. I guess they don't like switching into fault current on a residential 50A, 240V circuit (probably ~2kA peak). ;o)

Tim

--
Deep Friar: a very philosophical monk. 
Website: http://webpages.charter.net/dawill/tmoranwms

You didn't absorb the post.. The diode prevents the other side from back draining. SSR have very little current at the control terminals, they are so low infact, that we have to put burden resistors on them in many cases to prevent cross talk from other control lines in the race way from triggering them on.

THe scheme does work because we do it.. It has increased the life contacts driving heavy loads greatly.

We use lots of 24 volt logic and in area's where we have a potential of this happening, we place a small SSR across the contacts with a diode and cap on the control brick, sometimes we need to have a drain R there because it stays on too long. Which you do not want to happen, i'll burn out the SSR if it can't handle the full load for prolong periods.

btw, they make softstart contactors with this topology in them.

Jamie

In another response I suggested a snubber and I think the OP asked about that originally. We do not know any details about the solenoid current, etc. so yes, reviewing the datasheet for the MOC is in order. It does have example designs for controlling a secondary triac. For fail safe, maybe put two in series, who knows? It appears space limitations are forcing the use of too small a relay. Using a 0.1uf with a fuseable resistor would be my first choice plus an annual relay inspection :).

The real crime is they didn't even reserve the space for basic protection like this. I put at least some effort into ESD / overvoltage / overcurrent protections on all connections entering and leaving my PCBs.

If nothing else, one could hack into the solenoid wires and place the snubber there. It's low frequency energy, so it's not as big a deal if it's *at* the switch or not.

Tim

--
Deep Friar: a very philosophical monk. 
Website: http://webpages.charter.net/dawill/tmoranwms

Sears kenmore Oasis or Whirlpool Cabrio or Calypso. Probably a dozen more brands using the same control. All of these have a big knob in the center of the panel to select the cycle. The later ones have LEDs around the knob to verify the cycle selection, I think the old ones just had a pointer on the knob.

Jon

The problem here is I have practically no space to put a classic snubber. The R is small, but the cap will be fairly big. A TVS device will fit for sure. I think I'm going to have to retrofit one SSR for the failed relay just to keep washing clothes until the replacement relays come in. That will give me some time to research the situation, and do some experiments with some spare solenoid valves to determine the best snubber R/C if I can find components that fit, or work with the TVS devices to be sure they will work reliably.

Thanks,

Jon