Archived Forum

The voltmeter measured values are relatively stable at 251-252 V, but the voltage reported by the emonTx varies 181 - 248 V, is that correct?  And both do the same with separate adapters?

Have you tried measuring the a.c. output from the adapters, is that stable at around 11 V? If not, then the problem lies with the adapter or upstream of it. Maybe a stupid question: is the mains socket and/or any extension lead you are using good and clean?

If the a.c. out of the adapter is stable, I think you need to be looking for a faulty component or a bad joint in the voltage input circuitry, but first do basic voltage checks: that the 3.3 V supply is stable, and the mid-point voltage on the a.c. input socket (i.e with the adapter unplugged) is a stable 1.65 V (or close to that).

You say you have two emonTXes and two PSUs. Do you get these power fluctuations with both boards, and both PSUs?

If yes, then it sounds like the problem isn't there. Could be (as Robert says) the particular circuit you are plugged in to (a broken mains ring, perhaps, with heavy loads like washing machines and kettles?).

Or could it be that your mains IS quite dirty, but that the meter you are using to measure it is smoothing across several cycles ?

P.

This fault is a new one on me, I've never heard of anything similar reported here. At this stage, I wouldn't rule out anything, including a common assembly mistake on both boards.

We simply don't know enough at present - a logical systematic fault tracing exercise is called for to track down the problem.

Yes, I have two emonTxs and two AC PSUs, and all combinations appear to show unexpectedly low voltages reported from the test sketch, from time to time.  The test setup has the board powered only from the CP2102 programming module on the end of a USB wire connected to my laptop, for showing serial data.

The measured output of AC adaptors are 11.68V-11.71V and 11.67V-11.70V.

The 3.3V rail is 3.30V to 3.31V.

The centre pin of the voltage monitor is 1.59V-1.61V.

Currently I suspect that the programmer connection from the laptop has a slightly dodgy connection on some pins, and I am wondering if that is causing the duff measurements.  I have a test sketch (attached) which I will do more testing with.

When things are nice and stable, the output looks like this:

-0.24 3.97 248.16 0.02 -0.06 
Min: 246.80; Max: 249.22; Average: 248.49
0.32 21.21 248.33 0.09 0.02 
Min: 246.80; Max: 249.22; Average: 248.49

First, you need to tick "List" when you attach a file, otherwise no-one can see it. I've done it for you.

You can prove if the processor is resetting due to lost power by putting a print statement in setup( ). I can't see anything else on the programmer side affecting the voltage measurement.

To all intents you are monitoring continuously so at this stage, provided your programmer/power supply is not at fault,  my bet would be on a faulty socket or extension lead. It's almost certain that the emonTx is picking up dropouts that your meter isn't seeing, so a plan might be to put a low power (15 W or so) incandescent lamp across the mains you are measuring and see if you can detect flickering there. If that doesn't show anything, I'd suggest counting how many periods the voltage is below (say) 215 V out of how many checked in total.

I've been using a 60 incandescent bulb as a load.  I can't see flicker.  I agree that the mains voltage could be glitchy, as I am plugging in on an extension lead on the end of a rather strung out (but professionally installed) mains installation.  I hope to conduct some further tests but it won't be soon.  Thanks for the suggestions.