How to make a Solid State Relay on ONLY if current flow is > 100mA?

Discussion in 'The Projects Forum' started by Xplode, Jul 29, 2014.

  1. Xplode

    Thread Starter New Member

    Sep 1, 2012
    19
    0
    I have a project in mind where I'll be using a wireless remote system to trigger a larger AC Mains device. For simplicity, I would like to use a SSR to handle the ~15A draw. There is a coil involved, so I like the spike protection built into the Solid State Relays.

    The wireless controller outputs a short (~50ms) pulse at 18VDC ~30mA once a second as it polls looking for its originally intended load. When activated from the remote, it connects directly to the 18VDC power source and the source current available will be > 1A (up to 3A I believe if the batteries are fully charged).

    I need a way to keep the voltage at the relay terminals below 1.5V (it operates 3-32VDC) when the current is at the low level, but over the 3V so the relay will trigger when the device is triggered and the full current is made available.

    I'm scratching my head looking at voltage divider networks and bypass resistors, but I can't seem to get my brain to come to a clear answer I understand.

    Thoughts/ideas?
     
  2. inwo

    Well-Known Member

    Nov 7, 2013
    2,433
    315
    Do you mean ignore the pulse?
    Or do you need something else?

    " when the current is at the low level"

    What low level?
     
  3. to3metalcan

    Member

    Jul 20, 2014
    228
    23
    Seems like you could bypass the relay coil with a moderate-size capacitor so it's insensitive to the short pulses (which would look like AC and therefore pass through the cap.) A constant voltage would charge the cap and become a steady voltage through the coil.
     
  4. NorthGuy

    Active Member

    Jun 28, 2014
    603
    121
    Apparently, when relay is open there's no current, and you cannot measaure current if it doesn't exist.
     
  5. Xplode

    Thread Starter New Member

    Sep 1, 2012
    19
    0
    @ inwo - I mean to ignore the pulse and not have the relay actually energize/change state. "low level" meant when the system was putting out the reduced 30mA only, and not the full available current that would be present when i press the button on the remote.

    @ to3metalcan - how would i go about calculating that sort of thing? I've guessed at the duration of the pulse by looking at a LED i hooked up. Its quick, but I don't have a scope to time it with..

    I seem to have it working with just a 27ohm resistor across the relay terminals, but it gets quite hot when I activate the relay as its trying to dissipate something like 12W (if my math was correct). I don't want the heat problem since the unit might stay on for a few minutes at a time... The capacitor idea seems elegantly simple, but I'm concerned the cap might potentially induce a delay when i press the button (even a half second would be nuisance)
     
  6. AlphaDesign888

    New Member

    Jul 27, 2014
    189
    10
    Yes. You need what is called an transconductance amplifier. You convert the current to an voltage and feed the voltage to an comparator to switch the relay on after it has detected the level of 100mA which must be translated into an voltage.
     
  7. inwo

    Well-Known Member

    Nov 7, 2013
    2,433
    315
    A voltage divider won't dissipate so much.;)
     
    to3metalcan likes this.
  8. Xplode

    Thread Starter New Member

    Sep 1, 2012
    19
    0
    @inwo - Ok, I understand basic theory with electronics... and normally once i put my head to it, I can sort out the math. But I'm getting a bit turned around with this one.

    How do I calculate a voltage divider that works properly with 30mA flowing (sense resistor would have to have <1VDC drop across it) and yet have ~5VDC when the circuit goes to full power and 18V is applied across them?

    I guess I was getting hung up on the idea that it was 18V @ 30mA, but if the source is regulating to 30mA, its probably not actually 18V is it - its probably varying it to hit the target of no more than 30mA... Hmm... I will play with this idea and see what I can come up with.

    In the mean time, I have connected an arduino and used it as a pseudo scope to see the pulses versus full on. the pulses are only about 10-15mS long, so much shorter than I thought.. which is good. I managed to get the arduino to detect the difference (wait for signal to be present for > 30mS) but I'd like to avoid a MCU in this particular application. too complex/costly.
     
  9. AlphaDesign888

    New Member

    Jul 27, 2014
    189
    10
    It is all OHMS law that you need. It is not simple in this case though.
     
  10. inwo

    Well-Known Member

    Nov 7, 2013
    2,433
    315
    Not enough known about your ssr to figure it out exactly.

    I just started with leaving the minimum voltage, plus a little, at the input. 1/5 of the 18 was a good guess, and works here.
    Picked series value for a few ma, as it only takes a couple to turn on an ssr.

    Use lower values if needed for reliable turn on.

    Then pick a cap for a good portion of a second time constant.

    I'll have to read OP again. I thought it was an 18v source. If it's a constant current 30ma (as you suggest), I'll have to rethink it.
     
  11. inwo

    Well-Known Member

    Nov 7, 2013
    2,433
    315
    Can't be constant current, or limited to 30 ma, else the 27 ohm would only dissipate 26mw.:confused:

    Where did you come up with the 30ma? Need 660ma+ thru 27 ohm for 12 watts.
     
    Last edited: Aug 1, 2014
  12. Alec_t

    AAC Fanatic!

    Sep 17, 2013
    5,791
    1,103
    Do the pulses stop, so that you get a steady 18V when you press the remote?
     
  13. Alec_t

    AAC Fanatic!

    Sep 17, 2013
    5,791
    1,103
    I think the voltage-divider/capacitor approach suggested above will be ok. Here's a sim:
     
    inwo likes this.
  14. Xplode

    Thread Starter New Member

    Sep 1, 2012
    19
    0
    They are ebay SSRs since I don't have a budget to be buying good branded units at $25 a pop.

    I posted my last post from my phone and didn't see the diagram inwo shared - that's basically what I was trying to work out with the resistors and I like the simplicity of it... I think I will investigate that route further today.

    I have ordered a handful of the same SSRs for this project, so hopefully I won't get any funny business with differences in turn-on voltage between them. It seems pretty reliable to not turn on if I'm below 1VDC so that's my target.

    inwo - I am honestly not terribly sure since I don't have a scope to test the voltage coming out. I have gotten a basic answer from the supplier that says the unit puts out a controlled 18V @ 30mA to check for load, but it absolutely will not put out more than 30mA so I'm leaning towards constant current - but for all I know it could be a simple inline resistor that is limiting to 30mA in a short circuit condition, and the current may drop a little as I add loads to the output terminals.

    I will have some more time tonight/tomorrow to mess about with it - I think I'll try hooking up a few various resistors and seeing if I can plot the voltage output using the arduino's analog inputs. It'll obviously be quite rough - but should give me a better idea how the unit behaves with varying loads.

    Thanks for the insight and help so far guys!
     
  15. Xplode

    Thread Starter New Member

    Sep 1, 2012
    19
    0
    Seems my browser gave me some grief and I just saw the 2nd page posts for the first time.


    Thanks for the simulation - I'd like the SSR to trigger a bit faster if possible - < 100ms would be ideal, but it needs to absolutely stay off with the 15ms Pulses that occur. What software is that? is there a free version I can learn to mess around in and test out my ideas? that looks like a great proof of concept tool.

    Once the signal is activated and goes to a proper 18V I can have the device output a steady 18 until i decide to disable it.

    inwo - there are two modes. (1) the pulsing which is max 18V and absolutely no more than 30mA. then there is the (2) full on mode which is direct connection to the 18V source with no current limiting. this is when the 27ohm resistor would be pulling .66 amps and that is about 12W...
     
  16. Alec_t

    AAC Fanatic!

    Sep 17, 2013
    5,791
    1,103
    LTspice. A free download from Linear Technology. The Help files and FAQ will get you started. Well worth the steep learning curve. There's a Yahoo LTS User Group for getting free libraries of additional models, plus assistance.
     
  17. Xplode

    Thread Starter New Member

    Sep 1, 2012
    19
    0
    Thanks for the help guys - I got it to where I think it will work fairly well, but some unexpected design requirements have changed and now I'll be forced to use a microcontroller to enable specific time-on requirements.

    Some of this will transfer over as I can use the analog inputs on the MCU to detect the correct voltage/current levels and distinguish from the pulses by waiting 20ms.

    Thanks for the help - this community is full of great people and great info!

    So I suppose I'll toss this out there... I'd like to fit everything into a deep 1Gang Electrical box... Any suggestions on a tiny 5V psu to cram in the box with the pro-mini arduino I'll probably end up using? (Could be anywhere from 5-13VDC) I'll only need probably 40-50mA of current to run it plus possibly an LED or two and the SSR.
     
Loading...