4
\$\begingroup\$

The relays that I am using are JS1-12V-F.

The relays are being used on a testing apparatus to automatically test PCB components for accuracy.

When powered off for a long time (> 3 minutes), the resistance between Common and the Normally Closed leg is ~0 Ω (0.3 Ω). When powered on, the resistance between the Common and Normally Closed leg is ~infinity Ω (O.L.). If the power is kept on for 120 seconds (2 minutes) and then it is powered off, the resistance 1 second later will be > 10 Ω (as high as 60 Ω I've seen). The resistance will decay over a few seconds to maybe a couple minutes.

If the power is kept on for only a few seconds, then the resistance might only be as high as 3 Ω, but when there are 4 relays in series, the component being measured can be much higher than it truly is. For instance, an 82 Ω resistor will read 94 Ω and be considered bad. Or a 70 Ω bad resistor will read 84 Ω and be considered good.

Is anyone familiar with this phenomenon? Google has not been very helpful. I searched "relay temporarily has resistance when closing" and similar searches, but they weren't very helpful

\$\endgroup\$
9
  • 2
    \$\begingroup\$ Are you testing the relay alone, not connected to any circuit? \$\endgroup\$
    – Eugene Sh.
    Commented Aug 9, 2022 at 18:35
  • 1
    \$\begingroup\$ How much current flows through the relay contacts when they're closed? Does your relay have a Wetting Current spec? \$\endgroup\$
    – brhans
    Commented Aug 9, 2022 at 18:36
  • \$\begingroup\$ @Eugene Sh. The purpose of the relays is to switch between different first party resistors. But I have been specifically testing the relays (which are on a 3rd party PCB) without the first party resistors \$\endgroup\$
    – Zeno
    Commented Aug 9, 2022 at 18:42
  • \$\begingroup\$ Normally-closed contacts are generally less reliable than normally-open, since the return spring operation can be quite leisurely, in comparison to the slamming action when the NO contacts close, which helps to punch through any contaminants. \$\endgroup\$
    – jayben
    Commented Aug 9, 2022 at 18:54
  • \$\begingroup\$ @brhans I believe that the DMM that I am using sends 1mA when measuring the 82 Ohm resistor. I could not find "wetting current" or "fretting" in the data sheet. I do not think that it was even taken into account when it was designed. \$\endgroup\$
    – Zeno
    Commented Aug 9, 2022 at 18:55

2 Answers 2

9
\$\begingroup\$

The JS1-12V-F just isn't good enough for low current applications such as testing PCB components. You need a relay with a specified operating contact current that can be as low as 100 μA (or even lower). For instance the JS1-12V-F has a minimum switching current of 100 mA at 5 volts i.e. no good for sensitive or low current circuits: -

enter image description here

You shouldn't skimp on test equipment relays because it's more trouble than it's worth. The best relays are hermetically sealed but, you will probably get away with an OMRON G5 relay: -

enter image description here

\$\endgroup\$
4
  • 1
    \$\begingroup\$ Thank you for the info. Maybe I will be allowed to make one with good components. \$\endgroup\$
    – Zeno
    Commented Aug 9, 2022 at 19:04
  • \$\begingroup\$ I cant find the minimum switching current for this relay is it "failure rate"? components.omron.com/us-en/datasheet_pdf/J117-E1.pdf \$\endgroup\$
    – Zeno
    Commented Aug 9, 2022 at 19:31
  • \$\begingroup\$ @Zeno that's because it ain't suitable. Don't skimp. If you need a high current relay that works at 10 uA and 16 amps then use two relays. The spec indicates that contact resistance is measured with 1 amp flowing (top of page 3 and note 1) and that is a massive indicator that it isn't suitable for microamp switching. I did mention G5 relays and not G2 relays. \$\endgroup\$
    – Andy aka
    Commented Aug 9, 2022 at 19:34
  • 1
    \$\begingroup\$ Gotcha, I, not trying to skimp. It's not coming out of my pocket. I'm just trying to find the right ones that are active and in good enough supply \$\endgroup\$
    – Zeno
    Commented Aug 9, 2022 at 19:49
2
\$\begingroup\$

Could you use an opto-MOS photocoupler? Here is one for about a dollar that has less than 100 mOhms ON resistance and can handle 2A at 40V. However, it does not have a NC contact. You can get an AQW612 with one form A and one form B, but they are about $6, and have about 1 Ohm ON resistance. But as the ON resistance is known and consistent, you can just subtract it from the reading.

\$\endgroup\$
1
  • \$\begingroup\$ Off-state leakage current doesn't appear to be specified for item 1. Item 2 is hard to get to the data sheet. I also suspect that parasitic capacitance of both items might be a show-stopper for many in-circuit tests. \$\endgroup\$
    – Andy aka
    Commented Aug 10, 2022 at 7:38

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.