emonPi CM Firmware Update Oddities

[rdavies6 (Robin Davies)]

I recently updated my old emonPi to the latest CM firmware (from the original Classic firmware), The process was a little fraught…but over the last few days it’s seems to be working again (although sadly I did manage to inadvertently delete my old histori feeds…but c’est la vie!)

What I have noticed is the emonPi now to be measuring much higher power readings, and also a higher Vrms reading. I guess I shouldn’t be surprised at some difference, given the continuous nature of the new monitoring…but the readings are definitely much higher than I expected.

During the last Octopus Energy Saving session (29th Nov, 5.00pm-6.30pm), Octopus seem to think I used 0.29kWh, 0.26kWh & 0.26kWh of energy in each of the three 30 minute blocks. This is quite close to what my emonPi was reading with the old firmware (not exacty but the emonPi monitors the house, and in the garage there is some unmonitored circuits with a small amount of energy consuming equipment (garage door opener, maglock power supply, external light sensors, for instance). I don’t remember ever calibrating the old firmware…and the monthly cumulative power readings were within a couple of kWh of the bills each month (which I attributed to garage power usage).

With the new CM firmware, this is the power usage measured in the Savings Session window yesterday

The new firmware seems to be measuring about double our meter measured usage, and about double what I would have expected using the old firmware (over a number of years of using the emonPi with old firmware, I have a pretty good idea what the house energy usage is like with most devices turned off)

On the Vrms side…over a year the old firmware measured our mains voltage mostly around 232V, and nothing higher than 238V. Here’s a histogram of a year worth of readings.

With the new firmware I seem to be seeing a higher average voltage (although with just a couple of days readings that’s probably not a great sample population)

Is this behaviour expected in a move to the new CM firmware, or maybe there a default calibration or factor applied somewhere which I may have wiped out when re-configuring the new firmware in the emoncms.cfg file?

Any thoughts around what might be occurring?

Thanks

[TrystanLea]

Hello @rdavies6

I haven’t noticed this, but will do a quick sanity check test on an emonPi board here to see what it gives.
I wonder if it’s loading an incorrect calibration from the EEPROM…

It might be worth opening the Admin > Serial Config utility, stopping EmonHub, connecting to ttyAMA0 at 38400 baud and then first list the config with l and then resetting the config with r.

[rdavies6 (Robin Davies)]

Is this what you expect to see (i needed to connect at the higher baud rate too)

[rdavies6 (Robin Davies)]

Is there a chance that the type of CT clamp that’s being used is not correct in the new config? (That means I also have to figure out what the type of CT clamp I got with the emonPi was… )

[TrystanLea]

A quick test here and Im reading 247V on the emonPi vs 242V on an SDM120, and 728W vs 715W.

It might be worth adjusting the voltage calibration a little, e.g send via serial:

k0 263.52 

Save with

s

That brings the voltage and power in line with the SDM120 reference here, but doesn’t explain your 2x error?

I’m testing with an Ideal Power DB-06-09 power adapter here, what do you have there? Is your CT sensor SCT013 100A:50mA?

[Robert.Wall]

@rdavies6

The new firmware seems to be measuring about double our meter measured usage,

Why do you say that? Is it me going barmy, or looking at the last hour is 0.26 kWh over ½ hour not 520 Wh per hour, or 520 W? I think that makes your emonPi reading at roughly 500 W about 4% low, not 2× (high).

[rdavies6 (Robin Davies)]

Argh…it’s very possible it’s me going barmy then! Such simple maths… Sory

[john21042 (John Loesch)]

I have a related issue. Since the update (emonpi lpl cm v1.1.4) the emonpi message is showing a vrms of about 251 (it varies). Only problem is I’m in the US on 120v power. Is that barmy stuff contagious?

[Robert.Wall]

Serves you right for living in a country with a silly electrical system.

You need to change the voltage calibration back to the correct value for the a.c. adapter you’re using. It should be the same number as you had originally – the calibration factors haven’t changed.

This should help:
https://docs.openenergymonitor.org/emonpi/north-america.html#software-calibration

[john21042 (John Loesch)]

Resolved - The issue was I upgraded from an old version to EmonSD-01Feb24 which has the conf file set up for the OEM interfacer. I needed to use the Serial Monitor to change the voltage calibration in the eprom. Sure would be nice to have a single document on how to upgrade from old to new. However, I must say there is a wealth of good information on the forum!

[rdavies6 (Robin Davies)]

Just to resurrect an old thread…in the end I forgot about this topic until yesterday. Quick synopsis:

In November lst year, I upgraded my EmonPi setup to the continuous monitoring firmware. I had been monitoring VRMS, and prior to the upgrade I have been recording an average voltage of around 232V. The upgrade took me a while to get up and running, and about a week later when it came back online I noticed the voltage being measured was much higher. The attachment shows the VRMS data since Dec 1 '23 - and the average is around 251V. Trystan suggested recalibration, but life intervened and I never got around to doing this.

Recently had an EV chargepoint installed; the incoming mains goes straight from the meter into a Henley block, and then to a dedicated consumer unit for the charger. We’re have PME earthing here, so they installed a dedicated consumer unit for the charger, with overvoltage and PEN protection as part of the required standards.

Since the charger went live, I’ve now had the overvoltage switch trip twice; the overvoltage trip is set at 254V (UK voltage should be within -6%/+10% from 230V).

I’m beginning to think the voltage being measured by my EmonPi may in fact be accurate… and around the end of November something effected my electricity supply…and just happened to be coincident to me upgradeing the EmonPi firmware.

I suspect my chargepoint provider will be raising this issue with the DNO today. I think the DNO will want to put a recorder on the incoming supply, and I hope they might find the local substation might need an adjustment.

I’ve heard from the chargepoint provider that this is not an uncommon scenario as more of these overvoltage trips get installed and illustrate these issues.

[Robert.Wall]

@rdavies6

UK voltage should be within -6%/+10% from 230V

This, although technically correct, never ceases to amuse me, because in practice nothing has changed since our supply was 240 V. I don’t think every final distribution transformer in the country got its voltage changed, so I always say the centre voltage should still be 240 V.

The simple fact that there’s nothing below about 242 V says you’re spot on, assuming the emonPi is correct.

If you get a chance, take the opportunity to check the emonPi against their recorder, and if it’s reasonably close, produce your data as evidence of a long-term problem. It would be interesting to see at what time of day the peaks occur - I’d suspect late evening as the load on the network falls and generation is taken off.

[dBC]

My little corner of the Qld grid was running pretty high a few years back. The zooming/panning/averaging features of EmonCMS is great for seeing exactly when they adjusted it. In these parts the highest voltages are usually in the middle of the day - lots of power coming off rooftop solar and it often has to travel quite a way upstream before it can find a consumer.

(zoomed in on the adjustment)