- #1
Jiggy-Ninja
- 309
- 1
I'm not that familiar with the inner workings of solid state relays, and I'm seeing an unusual anomaly in my circuit that I'm not quite sure how to explain.
For my job, I'm running tests on some sensors. One of the requirements for the test is to turn the power to the parts on and off periodically. One full power cycle (ON/OFF) is done every 2 minutes for 50 minutes, then the parts are left completely off for 30 minutes. The whole 80 minute cycle is repeated a specific number of times or until something breaks. To do this switching, I am using a solid state relay as a high side switch on the line coming out of the power supply.
The power is a 5V/5A basic bench top power supply. There are 18 sensors connected to the supply, each consisting on a pair of Hall sensors, 2 opto-isolator channels, and 1 small EEPROM. The SSR is a Crydom D1D40 (100V/40A rating, WAY overspecced for this job). It's DC input and DC output. Crydom website specifies that it has "MOSFET output". The input of the SSR has a 10k-ohm resistor across it, to prevent residual voltage across the input, thought this may be superfluous.
After a while of operating normally, the monitoring equipment starts to measure that Vcc isn't making it all the way up to 5V, and is staying down closer to 4.5V. After a while, it gets a bit more unusual.
At the beginning of the 80 minute cycle (50 minutes ON/OFF cycling), the Vcc line is being measured at only 0.7V!. That level remains stable for about 8-10 minutes, then over the next 40 minutes the level falls to 0.45V in what appears on the graph to be an exponential decay, or something quite similar. The parts then have power removed (SSR turned off) for 30 straight minutes, and the next cycle repeats the pattern EXACTLY (Within margin of measurement error, of course).
If it was just that, I could accept that the SSR was going bad without too much puzzlement. However, there is one final observation.
When I saw that the relay was ON (and giving power to the parts) but the measuring equipment was only recording a value of 0.7V Vcc, I measured across the output of the SSR with a voltmeter (set to voltage, not ohms or continuity; I know better than that), and read just a hair above 0V, which is what it should be if the switch is fully on. When I looked back at the other measuring equipment, it was recording full 5V Vcc! Somehow, just measuring the voltage across the SSR output had gotten rid of whatever gremlin was causing the problem.
Every observation given in this post has been verified with an oscilloscope monitoring the power supply above and below the switch. The voltage level above the relay remains stable, so the power supply does not seem to be causing issue. It's below the relay that the issues appear.
I have set up a similar testing circuit like this several times before, so it is not completely new to me. I have also double and triple checked the polarities of the input and output; they are correct. I have also done several other tests on the same parts using the same equipment; this is the first time I have ever seen this issue.
Has anyone ever had a similar problem using an SSR? The evidence seems to point clearly to the SSR as the root of the problem, and I have replaced it in the test circuit. However, I am responsible for running this test and determining if the parts pass or fail. I have already had one problem where my original reasoning led me to believe it was part failure, but further in-depth troubleshooting showed it was actually an error with our custom made and programmed testing boards. I wish to have a deeper understanding of this issue before I write it off as just equipment failure.
What kinds of things could cause these symptoms?
1) Damage to one or more of the parts? They are taking significant abuse (mostly thermal) to determine their robustness. Could failures of one or more sensors cause issues?
-----Point against this, all the parts are showing identical symptoms at the same time. This suggests that something external and common to all the parts is causing the problem, ie the SSR.
2) Misconfigured circuit? I've checked the wiring a few times, but there's still the possibility of overlooking something or some other bone-headed mistake.
3) Damaged SSR? What kind of damage would cause this? And why would appear to fix itself (at least temporarily) when measured with a voltmeter?
4) Something else that I'm overlooking entirely?
Those that have read this far, thank you. :)
For my job, I'm running tests on some sensors. One of the requirements for the test is to turn the power to the parts on and off periodically. One full power cycle (ON/OFF) is done every 2 minutes for 50 minutes, then the parts are left completely off for 30 minutes. The whole 80 minute cycle is repeated a specific number of times or until something breaks. To do this switching, I am using a solid state relay as a high side switch on the line coming out of the power supply.
The power is a 5V/5A basic bench top power supply. There are 18 sensors connected to the supply, each consisting on a pair of Hall sensors, 2 opto-isolator channels, and 1 small EEPROM. The SSR is a Crydom D1D40 (100V/40A rating, WAY overspecced for this job). It's DC input and DC output. Crydom website specifies that it has "MOSFET output". The input of the SSR has a 10k-ohm resistor across it, to prevent residual voltage across the input, thought this may be superfluous.
After a while of operating normally, the monitoring equipment starts to measure that Vcc isn't making it all the way up to 5V, and is staying down closer to 4.5V. After a while, it gets a bit more unusual.
At the beginning of the 80 minute cycle (50 minutes ON/OFF cycling), the Vcc line is being measured at only 0.7V!. That level remains stable for about 8-10 minutes, then over the next 40 minutes the level falls to 0.45V in what appears on the graph to be an exponential decay, or something quite similar. The parts then have power removed (SSR turned off) for 30 straight minutes, and the next cycle repeats the pattern EXACTLY (Within margin of measurement error, of course).
If it was just that, I could accept that the SSR was going bad without too much puzzlement. However, there is one final observation.
When I saw that the relay was ON (and giving power to the parts) but the measuring equipment was only recording a value of 0.7V Vcc, I measured across the output of the SSR with a voltmeter (set to voltage, not ohms or continuity; I know better than that), and read just a hair above 0V, which is what it should be if the switch is fully on. When I looked back at the other measuring equipment, it was recording full 5V Vcc! Somehow, just measuring the voltage across the SSR output had gotten rid of whatever gremlin was causing the problem.
Every observation given in this post has been verified with an oscilloscope monitoring the power supply above and below the switch. The voltage level above the relay remains stable, so the power supply does not seem to be causing issue. It's below the relay that the issues appear.
I have set up a similar testing circuit like this several times before, so it is not completely new to me. I have also double and triple checked the polarities of the input and output; they are correct. I have also done several other tests on the same parts using the same equipment; this is the first time I have ever seen this issue.
Has anyone ever had a similar problem using an SSR? The evidence seems to point clearly to the SSR as the root of the problem, and I have replaced it in the test circuit. However, I am responsible for running this test and determining if the parts pass or fail. I have already had one problem where my original reasoning led me to believe it was part failure, but further in-depth troubleshooting showed it was actually an error with our custom made and programmed testing boards. I wish to have a deeper understanding of this issue before I write it off as just equipment failure.
What kinds of things could cause these symptoms?
1) Damage to one or more of the parts? They are taking significant abuse (mostly thermal) to determine their robustness. Could failures of one or more sensors cause issues?
-----Point against this, all the parts are showing identical symptoms at the same time. This suggests that something external and common to all the parts is causing the problem, ie the SSR.
2) Misconfigured circuit? I've checked the wiring a few times, but there's still the possibility of overlooking something or some other bone-headed mistake.
3) Damaged SSR? What kind of damage would cause this? And why would appear to fix itself (at least temporarily) when measured with a voltmeter?
4) Something else that I'm overlooking entirely?
Those that have read this far, thank you. :)
Last edited: