Author Topic: measurement error or bad part?  (Read 733 times)

0 Members and 1 Guest are viewing this topic.

Offline electrolustTopic starter

  • Supporter
  • ****
  • Posts: 562
  • Country: us
measurement error or bad part?
« on: July 15, 2020, 12:28:41 am »
i'm using a Coto Tech C347S SSR. Datasheet attached.

just testing it for on resistance, i measure >50M when the LED (switch) is off, and 0.2 ohm when the switch is on. just like the datasheet says. great for DC.

my test AC signal is a 1kHz sine wave, 100mVpp. with the switch on, i measure 100mVpp at the output. with the switch off, i measure 60mVpp. this also scales to 1V, where i measure 600mVpp. so, 60% of the signal is getting through.

i was initially sharing the ground with power and signal. i changed that to have a different signal ground, even though in my actual circuit they will be shared. then i went further and used a differential probe on the output signal. so only the signal and input probe share a ground. same results!

The C347s looks like it is a standard back-to-back MOSFET so it should be ok for AC signals, shouldn't it? I have the LED pins floating, no pulldown resistor, but this should be ok I think.

Even though I measured DC off resistance at >50M, I do notice that an RC filter with 1M / 190pF (the output capacitance of the C347S) has -3dB frequency of 840Hz. So I played with the frequency and I see that by 5kHz the SSR is passing the full signal even in the off state, and at 20Hz it's down to 20% of the signal. I guess I could do a bode plot later, but anyway it appears to be acting as a high pass filter in the off state.

this doesn't seem right, but at the same time, i can't imagine what's wrong with my measurement setup.
 

Offline wraper

  • Supporter
  • ****
  • Posts: 17952
  • Country: lv
Re: measurement error or bad part?
« Reply #1 on: July 15, 2020, 02:24:41 am »
You did not say a word about load resistance. As if you are simply measuring in series with multimeter/oscilloscope in series and no load at all.
 

Offline wraper

  • Supporter
  • ****
  • Posts: 17952
  • Country: lv
Re: measurement error or bad part?
« Reply #2 on: July 15, 2020, 02:29:56 am »
In the datasheet you can find output capacitance plot and at low voltage it's about 200pF.
Quote
this doesn't seem right, but at the same time, i can't imagine what's wrong with my measurement setup.
My gut feeling is that you have no load at all except multimeter/oscilloscope input resistance. And that certainly isn't right.
 

Offline electrolustTopic starter

  • Supporter
  • ****
  • Posts: 562
  • Country: us
Re: measurement error or bad part?
« Reply #3 on: July 15, 2020, 04:40:04 am »
it's correct, i have no load at all. however, on the far side of the switch, with the switch on I see 1:1 signal and with the switch off I should see 0 signal -- regardless of load. i mean, isn't the scope a high impedance load anyway? I also tried with 50 ohm and get the same results.

i realized something. when i tested this with CD4066 it worked as expected. now that i test with SSR i get this signal passing through. so i have some confidence in my measurement setup.

i notice in the data sheet that the block diagram doesn't show a diode in the mosfets. i just assume it's a simplification. it does show the 2 leads are connected to the 2 drains. the data sheet does specifically say, 80V peak AC voltage. It shouldn't say that if it doesn't work with AC.
 

Offline eblc1388

  • Frequent Contributor
  • **
  • Posts: 400
  • Country: gb
Re: measurement error or bad part?
« Reply #4 on: July 15, 2020, 10:35:05 am »
...i notice in the data sheet that the block diagram doesn't show a diode in the mosfets.

It actually does, visible with >500% magnification.

 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 17429
  • Country: us
  • DavidH
Re: measurement error or bad part?
« Reply #5 on: July 15, 2020, 12:31:18 pm »
it's correct, i have no load at all. however, on the far side of the switch, with the switch on I see 1:1 signal and with the switch off I should see 0 signal -- regardless of load. i mean, isn't the scope a high impedance load anyway? I also tried with 50 ohm and get the same results.

There is a fixed input to output capacitance which when combined with a high impedance load like a multimeter or oscilloscope, allows enough AC leakage to be misleading.  Real loads for power switches have much lower impedances so this leakage would usually be insignificant.

If lower leakage is a requirement, then a relay or a more sophisticated SSR is required.

Quote
i realized something. when i tested this with CD4066 it worked as expected. now that i test with SSR i get this signal passing through. so i have some confidence in my measurement setup.

A 4066 has much lower input to output capacitance, only a couple picofarads.

 

Offline electrolustTopic starter

  • Supporter
  • ****
  • Posts: 562
  • Country: us
Re: measurement error or bad part?
« Reply #6 on: July 15, 2020, 08:10:43 pm »
ah. good to know that my measurement is ok. i will either source a different part (like one from this thread https://www.eevblog.com/forum/projects/low-leakage-triac-or-solid-state-relay/), or change the design.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf