Shop OBEX P1 Docs P2 Docs Learn Events
Relay contacts welding together....is there a simple solution? — Parallax Forums

Relay contacts welding together....is there a simple solution?

MicksterMickster Posts: 2,694
edited 2014-04-22 18:06 in General Discussion
These are name-brand relays (Phoenix Contact) and rated at 2A switching of an inductive 24v DC load (solenoid coil). We are right at 1.8A and switching at approximately every 5 seconds. Going through these relays at an alarming rate but they are a very small package that I can't simply replace due to limited space. The load solenoids do have snubber (free-wheel) diodes, BTW. Is there something I can do to alleviate the contact arcing?
«1

Comments

  • idbruceidbruce Posts: 6,197
    edited 2014-04-16 09:45
    Mickster

    Is the 1.8A constant current or max current. I am assuming that it is constant current, and that your solenoids might be drawing more current when first applied. I could be wrong, but that would be my guess. Measure the max current draw through the entire cycle. If you ever exceed 2A, I would say that is where your problem lies.

    Just guessing of course.
  • Duane C. JohnsonDuane C. Johnson Posts: 955
    edited 2014-04-16 09:59
    Hi Mickster;
    Mickster wrote: »
    These are name-brand relays (Phoenix Contact) and rated at 2A switching of an inductive 24v DC load (solenoid coil). We are right at 1.8A and switching at approximately every 5 seconds. Going through these relays at an alarming rate but they are a very small package that I can't simply replace due to limited space. The load solenoids do have snubber (free-wheel) diodes, BTW. Is there something I can do to alleviate the contact arcing?
    Show us the EXACT part number of the Relays, Solenoids, and the driver circuitry.
    Also the timing and rise/fall times of the input signals.

    In general, I HATE RELAYS, especially when made to operate rapidly.
    That being said, good relays in properly designed circuits should have good lifetimes.

    Give us the specks and we may find the problem.

    Duane J
  • sidecar-racersidecar-racer Posts: 82
    edited 2014-04-16 10:01
    Mickster,
    Generally contact arcing is caused by the inductive kickback when the points open (Note: the inductive kickback goes both positive and negative so a diode won't get it all). On old-style ignition systems (points) if the capacitor in parallel with the points failed or even was the wrong(?) value, contact pitting a failure would occur.
    Rick
  • LoopyBytelooseLoopyByteloose Posts: 12,537
    edited 2014-04-16 10:06
    Mickster wrote: »
    These are name-brand relays (Phoenix Contact) and rated at 2A switching of an inductive 24v DC load (solenoid coil). We are right at 1.8A and switching at approximately every 5 seconds. Going through these relays at an alarming rate but they are a very small package that I can't simply replace due to limited space. The load solenoids do have snubber (free-wheel) diodes, BTW. Is there something I can do to alleviate the contact arcing?

    You may not like thiis answer, but the 'derating' factor for inductive loads which is normally applied to limit the current capacity of the relay to 40% of the advertised rating (motors derate down to 20%!).

    http://cp.literature.agilent.com/litweb/pdf/5988-6917EN.pdf

    In other words, for 1.8 amps of inductive load, you need a 4.5 amp or better rated relay. Read the derating document attached, that is the guideline.
    There are other things to do enhance the longevity of the points, but those really should come into consideration after you apply a derating according to the type of load.

    I generally don't bother with anything under a 10amp rated relay.. there are several different derating factors, but inductive loads are severe.

    And DC use is usually more aggressive since metal migration only in one direction from point-to-point. So you often see a DC rating of 24VDC or 120VAC due to this behavior.
  • Beau SchwabeBeau Schwabe Posts: 6,566
    edited 2014-04-16 10:13
    For similar reasons a capacitor is placed across motor brushes, placing an appropriate capacitor across the relay contacts will do wonder to prolong the life of the contacts.... BTW also done for reed switches.

    But as Duane requested ... a schematic and part number would help us out here in determining a proper solution.


    Contact Protection and Arc Suppression Methods for Mechanical Relays:
    http://www.industrologic.com/mechrela.htm
  • LawsonLawson Posts: 870
    edited 2014-04-16 12:00
    That switching rate also sounds like an application better suited to a PV opto-isolator + FET or a solid state relay.

    Marty
  • MicksterMickster Posts: 2,694
    edited 2014-04-16 14:47
    Many thanks to all.

    I do normally use SSRs but went with this device due to space constraints.
  • LoopyBytelooseLoopyByteloose Posts: 12,537
    edited 2014-04-16 21:16
    Well, if you need small space and tolerant of hostile environment...automotive relays seem a better way to go. I am wary of any relay that tries to be too tiny.

    The one below is rated at 24VDC and 20 amps and I suspect you would be much better off with it. You can get sockets with pig-tailed wires as well.

    I cannot seem to locate anything the exact same size that goes above the 6amps you already have.

    http://www.amazon.com/24-VOLT-RELAY-20-AMP/dp/B004XCGCN0
  • TubularTubular Posts: 4,703
    edited 2014-04-16 23:39
    Maybe this SSR from Panasonic could help. I don't think the pins would be exactly the same but its external dimensions are the same
    http://media.digikey.com/pdf/Data%20Sheets/Panasonic%20Electric%20Works%20PDFs/AQ-E%20Relays.pdf

    But as Beau says a decent cap across the contacts should solve it
  • Peter JakackiPeter Jakacki Posts: 10,193
    edited 2014-04-17 00:47
    I normally just drive these kinds of loads with very small mosfets, small like dual mosfets in an 8-pin SOIC, they have no problem whatsoever in handling this type of load and they are very reliable. Does it need to be isolated or can they have a common ground between the load and the logic?
    And ditto for what Duane said, especially the I HATE RELAYS :)
  • CuriousOneCuriousOne Posts: 931
    edited 2014-04-17 01:55
    When engineering faced such problems in early 20th century, when there's need to drive large motors/etc and sparks were nasty, they've introduced "rotary" relays - it consisted of rotary, sliding contacts with variable resistance, so first touch was made thru large resistance part, so no sparks, and then it rotated and slided into low resistance part.
  • Peter JakackiPeter Jakacki Posts: 10,193
    edited 2014-04-17 02:22
    CuriousOne wrote: »
    When engineering faced such problems in early 20th century, when there's need to drive large motors/etc and sparks were nasty, they've introduced "rotary" relays - it consisted of rotary, sliding contacts with variable resistance, so first touch was made thru large resistance part, so no sparks, and then it rotated and slided into low resistance part.

    Oh great, but I think you are 100 years too late.

    BTW, the sparks are not generated when the contact makes but when the contact breaks. Normally if a relay is used then a capacitor is uses across the contacts to snub the spark but why use a relay.
  • MicksterMickster Posts: 2,694
    edited 2014-04-17 03:07
    I normally just drive these kinds of loads with very small mosfets, small like dual mosfets in an 8-pin SOIC, they have no problem whatsoever in handling this type of load and they are very reliable. Does it need to be isolated or can they have a common ground between the load and the logic?
    And ditto for what Duane said, especially the I HATE RELAYS :)
    1. It's a retrofit package that I supply for similar-but-different machinery. We can never be sure of the machine being equipped with AC, DC or a mixture of both solenoids
    2. My customers are typically 24/7 automotive. Maintenance guys can always work with relays and they like parts to be easily locatable/replaceable/by-passable.
    I forwarded Beau's info and they seem to like this idea. I will use a cleaner solution in the future though.

    Many thanks, once again!
  • Erik FriesenErik Friesen Posts: 1,071
    edited 2014-04-17 18:10
    Use a tvs diode to contain the field collapse. You could also add some series resistance, would probably help, and still pull in the solenoid.
  • MicksterMickster Posts: 2,694
    edited 2014-04-17 18:26
    As stated in the OP, this is already provided...this is not a back-emf issue.
  • Erik FriesenErik Friesen Posts: 1,071
    edited 2014-04-17 18:28
    I see that now, sorry. Have you scoped this? Can you current scope this as well?
  • CuriousOneCuriousOne Posts: 931
    edited 2014-04-17 22:01
    As the rotary element is circle shaped, contact breaking is made again on high resistance point, so no sparks there.

    Surplus sales sites still may have them in inventory...
  • LoopyBytelooseLoopyByteloose Posts: 12,537
    edited 2014-04-18 05:51
    @Mickster
    Let us know how Beau's solution plays out. I have long thought that adding a resistor and a capacitor might alway be a MUST include part of correct relay points protection. You happen to be in a situation where you are able to confirm or deny that this is enough.
  • RforbesRforbes Posts: 281
    edited 2014-04-18 11:35
    @Mickster-

    Have you checked this out from Phoenix Contact? Maybe this SSR would work for your application- https://www.phoenixcontact.com/online/portal/us?uri=pxc-oc-itemdetail:pid=2966595&library=usen&pcck=P-16-06-01-02&tab=1
  • Mark_TMark_T Posts: 1,981
    edited 2014-04-18 12:21
    It seems the arc power is too high for the relay - 2A at 24V is 48W, that's a lot of arc power available
    and I suspect the free-wheel diode isn't getting much of a look-in.

    Suggest trying a snubber or TVS, not free-wheel diode, on the relay _winding_, so it springs open faster.
  • MicksterMickster Posts: 2,694
    edited 2014-04-18 13:24
    Rforbes wrote: »
    @Mickster-

    Have you checked this out from Phoenix Contact? Maybe this SSR would work for your application- https://www.phoenixcontact.com/online/portal/us?uri=pxc-oc-itemdetail:pid=2966595&library=usen&pcck=P-16-06-01-02&tab=1


    Ah, yes. That could be a last resort. I would need dozens of them and they are ~$40 each!
  • T ChapT Chap Posts: 4,223
    edited 2014-04-18 16:13
    Can you provide a dual purpose solution using a 2 mosfets: one mode configured to switch AC, or jumper to a mode to switch DC.
  • bill190bill190 Posts: 769
    edited 2014-04-20 09:18
    Search google.com for the following including quotes...

    "relay contact life"

    You will find a very good paper on this in one of the first search results.
  • kwinnkwinn Posts: 8,697
    edited 2014-04-20 11:51
    A bridge rectifier and diode like this should work. Add a mosfet, optoisolator and current limiting resistor between the + output of the bridge rectifier and suppression diode and you have a solid state AC/DC relay coil driver.
    521 x 168 - 14K
  • MicksterMickster Posts: 2,694
    edited 2014-04-20 14:42
    Thanks for the suggestions, guys....the rectifier idea has me thinking for the future.

    Problem is that I can't just cobble new circuitry together as my name will be mud as soon as maintenance has to replace something (customer is big auto). Beau's suppressor idea is favorite right now because it is "reasonable". I even sourced pre-packaged R-C network modules that are intended for this.

    Naturally, my customer is several thousands of miles from me so I need to wait for him to add the components.....this individual gets more pleasure from complaining than resolving...I'm sure you have experienced similar :-)

    I shall update this thread as and when I receive feedback.
  • jmgjmg Posts: 15,173
    edited 2014-04-20 17:59
    Mickster wrote: »
    We are right at 1.8A and switching at approximately every 5 seconds.
    ... BTW. Is there something I can do to alleviate the contact arcing?
    ...

    We can never be sure of the machine being equipped with AC, DC or a mixture of both solenoids

    The heading says "Relay contacts welding together", but the text asks about arcing ?

    What sort of cycle life times and failure-cycle-counts are you seeing ?

    Contact release is going to be more of a problem as the inductor acts like a constant current source.

    If your target can be AC or DC loads, then that's rather harder to cover. - An R-C solution, may still draw enough power to not release a AC solenoid.

    Relays at close to rated loads have fairly low life cycle numbers, and 5s cycle time, is going to clock up a shipload of cycles.

    I think you will need to bite the bullet, and change to a AC SSR to have a big impact on failure rates.
  • jmgjmg Posts: 15,173
    edited 2014-04-20 18:12
    Mickster wrote: »
    I do normally use SSRs but went with this device due to space constraints.

    Cute, but not especially cheap ?

    In a broadly similar form factor, it looks like IXYS have a number of AC/DC SSR choices

    One example
    CPC1909J RELAY 60VAC/DC 6.5A ISOPLUS264 500+ $5.03
  • MicksterMickster Posts: 2,694
    edited 2014-04-21 08:15
    @jmg

    The machine is currently operating @ 140 hrs/week and I guess we have been losing a relay on an average of 1000 hrs. We have used these relays for years on AC solenoids and failures have been few and far between. The problem is with the DC inductive load.
    Funny you should mention the IXYS devices as I recently downloaded datasheets of some of their other SSRs....I could not find a price somehow but that part number worked when I Googled (don't know what I was searching for before). $5 is Digikey's single-piece price which is very reasonable.

    Problem remains though that these big plants prefer garden variety off-the-shelf components to minimize downtime....but they don't want to stock spares...aaarrrggghh!
  • LoopyBytelooseLoopyByteloose Posts: 12,537
    edited 2014-04-21 10:46
    Hmmm... the DC loads seem to be wearing out the points much faster simply because metal migrates only in one direction of the relay points with DC. With AC, it shuttles back and forth.. thus lasts longer.

    Every auto used to change points and a capacitor rigorously for a tune-up. So adding the resistor and capacitor to increase the longevity is likely to help out. After all those were DC setups.

    I just wonder what all the telephone company relays did to last. I do know they had people running around in sub-stations changing relays all day, but could it really have been that much? And then the issue of relays controlling elevations seemed to work quite well for a long time.

    It would be optimal to get something rated for higher amps and higher DC volts both.
    It seems like some real testing with SSRs would be worthwhile. $40 might seem cheap if it really lasts 10x longer.
  • jmgjmg Posts: 15,173
    edited 2014-04-21 13:28
    Mickster wrote: »
    The machine is currently operating @ 140 hrs/week and I guess we have been losing a relay on an average of 1000 hrs. We have used these relays for years on AC solenoids and failures have been few and far between. The problem is with the DC inductive load.
    1000 hrs at 5 seconds, is 720000, which I would call quite good (ie above expected), pushing along close to rated load.

    AC loads will tend to self-quench, as the current passes thru 0 often, whilst DC works like a current source on release.

    A couple of life-cycle curves, (usually spec'd with the easier resistive loads)
    http://www.mechanical-relay.com/images/dy03a_2.gif
    https://www.ia.omron.com/Images/e5cn_e5cn-u_sp_413-106673.jpg


    Mickster wrote: »
    Funny you should mention the IXYS devices as I recently downloaded datasheets of some of their other SSRs....I could not find a price somehow but that part number worked when I Googled (don't know what I was searching for before). $5 is Digikey's single-piece price which is very reasonable.

    Problem remains though that these big plants prefer garden variety off-the-shelf components to minimize downtime....but they don't want to stock spares...aaarrrggghh!

    Hehe, Of course, with a SSR they will not need spares :)
    - anyway you can easily include a couple with each product, just to make admin happy. ;)

    It will not take long to get to the 1000 hours they expect to see failures at.

    It sounds like you have these packed together ? - if that is the case, choose a SSR that is somewhat over-rated, like the link I gave, which will be ~ 324mW at 1.8A max.
    SSRs work better cooler, and everything is more reliable cooler.
Sign In or Register to comment.