Shop OBEX P1 Docs P2 Docs Learn Events
Minimum current and voltage specification for relay contacts? — Parallax Forums

Minimum current and voltage specification for relay contacts?

MJHanaganMJHanagan Posts: 189
edited 2012-11-19 15:47 in Propeller 1
I need to switch some samples from a high voltage/high current AC circuit to a low voltage/low current DC circuit. My initial thought was to use a DPDT relay for each sample (NC contacts going to the AC circuit and the NO contacts going to the DC circuit). When I looked at a few datasheets for relays I noticed a specification for "mimimum switching capacity" (see for example the specification section on page 3 in the attached datasheet below). These minimum values seem to be on the order of 5V and 50-100mA. The low voltage DC circuit I have will use only ~3-4V and =<10mA of current.

Does anyone know how this "minimum switching capacity" specification may impact how I am tying to used the relay?

Panasonic JW Relay datasheet: Panasonic JW Relays.pdf

Comments

  • RaymanRayman Posts: 14,669
    edited 2012-11-18 16:16
    If you want to control a relay with the prop, you probably want something like a ULN2803 inbetween it and the Prop...
  • MJHanaganMJHanagan Posts: 189
    edited 2012-11-18 16:37
    Rayman wrote: »
    If you want to control a relay with the prop, you probably want something like a ULN2803 inbetween it and the Prop...

    Indeed, I plan on powering the coil of the relay using some intermediate device like a darlington array (with a flyback diode). The specification in question is related to the contact rating, not the coil.
  • jmgjmg Posts: 15,173
    edited 2012-11-18 17:12
    MJHanagan wrote: »
    I need to switch some samples from a high voltage/high current AC circuit to a low voltage/low current DC circuit.

    ? switch some samples ?
    Do you mean like a flying capacitor, only done using a relay ?
    MJHanagan wrote: »
    Does anyone know how this "minimum switching capacity" specification may impact how I am tying to used the relay?

    You can bump into an issue around wetting currents, with very low current values.
    Above a certain current/voltage, the small arcing energy works to clean any film/oxides off the contacts.

    http://en.wikipedia.org/wiki/Switch#Wetting_current

    For this reason, signal relays often use Gold contacts - but then you need to avoid currents high enough to damage the gold.
  • Dr_AculaDr_Acula Posts: 5,484
    edited 2012-11-18 17:35
    Tricky design challenge. Relay contacts optimised for both low voltage DC and high voltage AC. Any way of buffering the AC, eg with an optocoupler? Or another relay? Or a small transformer?

    Any more information about the circuit as there might be other solutions.
  • MJHanaganMJHanagan Posts: 189
    edited 2012-11-18 18:12
    Dr_Acula wrote: »
    Tricky design challenge. Relay contacts optimised for both low voltage DC and high voltage AC. Any way of buffering the AC, eg with an optocoupler? Or another relay? Or a small transformer?

    Any more information about the circuit as there might be other solutions.

    The AC circuit sends 2-5 Amps of current at a voltage of 120-240V (it heats up the DUT). The DC circuit measures the DUT's resistance after it cools back down to room temperature. I use the Prop to control the TRIAC driver to deliver a specified amount of power to each of the six DUTs and measure the AC current in each DUT.

    Sounds like instead of a single DPDT relay you are suggesingt one DPST power relay for the AC circuit and another DPST signal relay for the DC circuit? This might work as long as the signal relay could handle the 250 VAC on the unswitched COM terminal. Is that asking a bit much from a signal relay?
  • jmgjmg Posts: 15,173
    edited 2012-11-18 19:07
    MJHanagan wrote: »
    The AC circuit sends 2-5 Amps of current at a voltage of 120-240V (it heats up the DUT). The DC circuit measures the DUT's resistance after it cools back down to room temperature. I use the Prop to control the TRIAC driver to deliver a specified amount of power to each of the six DUTs and measure the AC current in each DUT.

    Sounds like instead of a single DPDT relay you are suggesingt one DPST power relay for the AC circuit and another DPST signal relay for the DC circuit? This might work as long as the signal relay could handle the 250 VAC on the unswitched COM terminal. Is that asking a bit much from a signal relay?

    ...plus you have the lack of interlock in two relays, whilst a SPDT mechanically cannot have Mains on your measure side.

    Might be easier to just ensure your measure current is over 10mA, and use a std relay - those values suggest it would be possible ?
  • Duane C. JohnsonDuane C. Johnson Posts: 955
    edited 2012-11-18 19:41
    Hi MJ;
    MJHanagan wrote: »
    Does anyone know how this "minimum switching capacity" specification may impact how I am tying to used the relay?
    Ditto to jmg's wetting current comment.

    Mercury wetted reed relays can do this.
    While these are significantly more expensive than conventional dry contact relays they are quite desirable for making "laboratory quality" measurements. Generally they can switch relatively high currents and exhibit significantly lower contact resistance. They have no "Wetting Current" problems.

    An example of a Mercury wetted reed relay, however this only has a 50 VA contact.
    But I didn't look very hard.
    American Relays, Inc.

    And the mercury plunger or mercury displacement relays. Much higher current.

    Duane J
  • MJHanaganMJHanagan Posts: 189
    edited 2012-11-19 10:27
    jmg wrote: »
    ...plus you have the lack of interlock in two relays, whilst a SPDT mechanically cannot have Mains on your measure side.

    Might be easier to just ensure your measure current is over 10mA, and use a std relay - those values suggest it would be possible ?

    The samples have a very strong PTCR factor so when measuring the resistance I need to use as little a current as possible to minimize Joule heating. I was hoping to make several measurements over a nominal 0.1 second time period to enhance the accuracy of the measurement (I am trying to accurately resolve differences in the sample's resistance over long periods of time of less than 1%). Even at a 10 mA current I can see the resistance increasing over time using a good quality electrometer. If I cannot find a suitable solution to the 10 mA relay contact problem I may need to opt for some faster measurements at 100 mA in order to minimize the self-heating. The single DPDTrelay certainly makes things simpler and safer compared to two separate relays and the nasty outcome if the Mains and the DC relays are both accidentially activated!
  • MJHanaganMJHanagan Posts: 189
    edited 2012-11-19 10:33
    Hi MJ;

    Ditto to jmg's wetting current comment.

    Mercury wetted reed relays can do this.
    While these are significantly more expensive than conventional dry contact relays they are quite desirable for making "laboratory quality" measurements. Generally they can switch relatively high currents and exhibit significantly lower contact resistance. They have no "Wetting Current" problems.

    An example of a Mercury wetted reed relay, however this only has a 50 VA contact.
    But I didn't look very hard.
    American Relays, Inc.

    And the mercury plunger or mercury displacement relays. Much higher current.

    Duane J

    The "Mains" relay could be a common DPST relay and the DC could be one of these DPST reed relays. Having the third set of contacts on the Mains relay might come in handy as a safety feature by interlocking the coil current for the DC relay. The DC relay couldn't be energized unless the Mains relay is deenergized (or vice versa).
  • LoopyBytelooseLoopyByteloose Posts: 12,537
    edited 2012-11-19 11:22
    Dr_Acula wrote: »
    Tricky design challenge. Relay contacts optimised for both low voltage DC and high voltage AC. Any way of buffering the AC, eg with an optocoupler? Or another relay? Or a small transformer?

    Any more information about the circuit as there might be other solutions.

    Actually, the odd thing about relay contacts is that they tolerate wear better with AC. It is quite common to set a relay rated for 240VAC and 24VDC. It seems that it is all about the migration of metal from the points. DC causes the migration in only one direction and thus only lower voltages have equivalent longevity; while AC high voltages just move the metal back and forth with the net effect being less wear.

    But mercury wetting certainly does resolve a lot of complexities.
  • jmgjmg Posts: 15,173
    edited 2012-11-19 15:47
    MJHanagan wrote: »
    The samples have a very strong PTCR factor so when measuring the resistance I need to use as little a current as possible to minimize Joule heating. I was hoping to make several measurements over a nominal 0.1 second time period to enhance the accuracy of the measurement (I am trying to accurately resolve differences in the sample's resistance over long periods of time of less than 1%). Even at a 10 mA current I can see the resistance increasing over time using a good quality electrometer. If I cannot find a suitable solution to the 10 mA relay contact problem I may need to opt for some faster measurements at 100 mA in order to minimize the self-heating. The single DPDTrelay certainly makes things simpler and safer compared to two separate relays and the nasty outcome if the Mains and the DC relays are both accidentially activated!

    You may want to add a tracking formula, if you seek highest precision, and that would let you measure sooner after the heat cycle too.
    Otherwise, you must be waiting sometime between heat and measure ?
    A tracking formula would also compensate for self heating effects. That assumes all samples have similar time constants, and self-heating asymptotes.

    - and another way to 'keep the relay happy' (and contact resistance consistent) , is to have a RC energy source, that charges when the sample is heating, and then on switch over that RC is across the contacts, to give a close-impulse but it will have little effect on your DC measure
Sign In or Register to comment.