I have been using my DigUno with a relay on GPIO 19 (Q3) for the past few seasons. I updated WLED to the latest (14) and now the relay does not receive a signal to go high. Both the relay, and the power supply work independently, and the PS is powered if I connect the NC side of the relay, but not if connected NO. Before I rip the setup apart, has anyone any ideas?
All good suggestions. I tested the voltage change on the Q3 & Q4 - no change. Q1 & Q2 are pulled low, and there is no change there either. I am going to roll back to 13.0 âTokiâ and see what happens.
I came back to this relay problem, after about a year with some new understanding. I found, by experimentation, that the GPIO pin (19) on the ESP32 I was using still produces a logic signal, by oscilloscope, and that the relayâs I had been using to isolate the power supplyâs (coil 5VDC@8A 120VAC) were failing due to the contacts burning out. After the second attempt, I cut the failed relay apart and verified that the contacts had failed - burnt black. Now, why a 10 amp relay would fail, is beyond me; the power supply was barely loaded when the relays failed. These are the inexpensive ubiquitous âblue sealed boxâ dry contact relays. Has anyone been able to source a better quality relay to idle the power suppl?
So far I have only had 1 lil blue box fail and it got stuck latched where it would not cut the power. I was switching the AC side of psu. I think you are switching the DC side? You could try a DC/DC SSR. I am using a couple DC/AC SSRâs for a couple control boxes that sit right out in the snow/ice in winter.
What you may have to watch out for is the surge current when you power up the supply from âdeadâ (and down). The initial (very short) surge to charge the supply capacitors can often be 10x (or more) than the normal run current.
The disconnect current can cause a spark across the relay contacts when you shut things down if youâre under load. It doesnât take many of those to wipe out relay contacts.
SSRâs can be a reasonable solution, but be forewarned that you need to rate them at well byeond what your ânormalâ current reqtâs will be due to surge.
Really depends on the SSR you pick and what the mfg specs say.
For power supply/lines control, it much less of an issue than when youâre trying to do PWM dimming.
In the worst case, you could always use a levelshifter to get the signal up to 5V (yes I know, thatâs more complexity).
BTW, thatâs one application where I2C type shifters may be appropriate - relay/SSR drive.
Little to no speed requirement.
I received a couple of 20A relays today. I will test the cap. surge theory, but I suspect you are correct. The SSRâs might be a good fit if I can trigger them with logic level voltage. I have some cheapo level-shifters laying around.
For relay/SSR Power Supply drives from a GPIO, a simple 2N3904 and 2 resistors + blocking diode will work as a simple levelshifter (that also Logic inverts).
I received and installed the 20A relay this afternoon. So far so good. I am also going to experiment with the SSR. I have never used one before and I am curious.
I received the SSR yesterday. Having never used one before, I am impressed. It triggered easily with the logic level signal from the ESP32, but I think I will add a simple level shifter as divsys suggested. This UNO is remote from the router, and I am hedging all of my bets.
The beefier relay also worked well, but, oddly, was the same price.
I have 2 controllers using the SSRâs and have not had any problems with them switching @ 3.3v from the gpio. But I guess it canât hurt to add a shifter.
Check the specs of the UNO to be sure the relay pin on it is not already shifted. I donât use any Dig controllers and I know @Quindor has some of the pins on it shifted but am unsure of what ones.