0
\$\begingroup\$

I'm having a problem with load regulation errors on my board. When I engage our load with a relay, it causes a metering chip in our system to reset. I've placed a picture of the relay with the flyback circuitry below.

relay circuit

Some detail on the project:

The project details are sensitive, so we can only put so much online. But, let me know if you need any more specific details.

  • We are controlling LED light ballasts.
  • We are specifying
    • the ballasts may draw up to 16A
    • the ballast shall range from 85 - 277VAC, single-phase
  • We are using a relay to disconnect the lamp from the AC line.
  • The relay is in a normally closed configuration.
  • We are taking the 90-267VAC down to 5V with a simple AC to DC converter and using that to power our system.
  • We are metering the line using an IC meter.

Test Setup:

For the sake of having reproducible results, we thought to use a large wattage rated resistor on an actively cooled heat sink to act as a constant load.

  • The resistor is 500Ω rated for 150W
  • We are running the system off of a BK Precision power supply that seems to drift by maybe 0.5VAC @ 120VAC (spec'd at 0.2% ±0.6VAC)
  • The system draws about 0.249A of which about 25mA belongs to our circuit.

Symptoms:

  • When the relay is engaged, the load is disconnected, we see a ringing on the 5V line as well as on a regulated 3.3V line.
  • When the relay is disengaged, we see the same ringing, but much more exaggerated. This is primarily when the metering chip resets.
  • If we disconnect the load from neutral using a normally closed momentary switch, we see some 5V and 3.3V line voltage regulation errors. But, it very rarely causes a fault in our DC system as a whole.

So I was wondering if you can help me understand what I'm doing wrong in the circuit. I think it's related to the relay so that's the part of the schematic I included, if you need to see or know more, let me know.

Thank you very much.

\$\endgroup\$
  • 2
    \$\begingroup\$ How do you know supply is OK? Maybe its not and needs ultra low ESR output caps or cannot handle the load current. \$\endgroup\$ – Sunnyskyguy EE75 Jul 30 '17 at 1:23
  • 4
    \$\begingroup\$ How do you know it's a "brown out". Sounds more like EMI from the relay switching to me. Does it still happen with the load disconnected? If not (and assuming the 120VAC is not flaky) it's most likely EMI. \$\endgroup\$ – Spehro Pefhany Jul 30 '17 at 3:27
  • \$\begingroup\$ A 'brown out' is when your supply voltage drops significantly below its nominal value for some period of time. This doesn't seem to match the symptoms you're describing. \$\endgroup\$ – brhans Jul 30 '17 at 3:57
  • \$\begingroup\$ Scope ground is earth ground, neutral is earth grounded at panel you have a large ground loop when probing with scope \$\endgroup\$ – sstobbe Jul 30 '17 at 4:12
  • \$\begingroup\$ First question : are the brownouts related to the input side (relay coil and flyback/snubbing circuit) or the output side (AC switching)? To answer this, operate the relay without power on the AC side. Do you still see the brownouts? \$\endgroup\$ – Brian Drummond Jul 30 '17 at 13:24
0
\$\begingroup\$

There is nothing at all wrong with the drive circuit in your schematic, to the extent you have provided it, if anything it appears a bit overdesigned.

I suspect you are seeing EMI through your relay contact to coil coupling causing the symptoms you describe. This would likely be related to other connections to your board and to stray inductances and capacitances, including your power supply which probably provides a low-Z AC path to the mains. Generally a tight layout using a multilayer board with full ground and power planes can help a lot.

You cannot completely depend on the display of an oscilloscope in high EMI conditions especially if it is not a floating input type, so I do tend to discount your observations somewhat in that regard.

If you have proper supply decoupling, resets coincident with relay contacts opening or closing are most likely caused by EMI coupling to input pins (including, of course, any reset input, but not limited to that particular pin).

\$\endgroup\$
  • \$\begingroup\$ We believe you are right and that is like the most likely of the situation. We just engaged and disengage the relay with no load, and we did not have a problem. So if we have no load we dont have a problem and if we disconnect the load and reconnect the load outside of the relay we dont have a problem. We also scoped the 5V line and removed the ground so that we are only seeing AC, so no ground loop. With that we are still seeing the ring, but the voltage drop significantly reduced if we do not use the relay as intended. Lastly, I am not sure the proper way to get rid of the EMI, suggestions? \$\endgroup\$ – Silas Jul 30 '17 at 17:56
  • \$\begingroup\$ @Silas you can't get rid of the EMI entirely but you can make your circuit more tolerant of it. This is the nature of EMC (compatibility between the noise source and the affected circuit). It is a large subject with many books, some of them useful, and well-known practitioners of the art such as Howard Johnson (not a recommendation, never worked with him). \$\endgroup\$ – Spehro Pefhany Jul 30 '17 at 18:34
  • \$\begingroup\$ Ok, so we just tried a SSR and it fixed the problem. So you are absolutely right about the EMI. Thanks so much for the help. \$\endgroup\$ – Silas Jul 31 '17 at 19:29

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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