The Victron blog about the Intersolar Europe 2023 show lists a new Victron-branded 3-phase energy meter called an “Energy Meter 3P”. It has an Ethernet port, but no resemblance to a Carlo Gavazzi ET24 or other unit with Ethernet.
The text states: “Energy Meter 3P offers faster measurements allowing three phase ESS (Energy Saving Systems) to respond more quickly. Featuring split-coil transformers for current measurement, rather than having to be wired in series with the distribution board, the new Energy Meter saves significantly on the installation time. Units are supplied with Current Transformers. (This product is not yet available to order, but coming soon).”
Any info on whether it’s made specifically for Victron, or is a white-labelled something else?
This is a Victron product, designed by Victron engineers, built in the usual places things are built.
It supports modbus-udp (aka ethernet) and CAN-bus. It will be SIGNIFICANTLY faster than anything else supported right now.
The CAN-bus connection means it can talk directly to a Multi-RS without needing a GX in the loop.
Will this meter be fast enough to avoid tripping those temperamental prepaid meters?
It will help, but remember that some meters are stupidly sensitive. The German code asks for pulling back 90% within 3 seconds. That is actually hard to do while still keeping a stable control loop… and it is too slow for some Conlog meters.
I’d imagine there is a PID algorithm on the feedback loop.
There are experts if that needs tweaking. I suppose you already know that.
The way it is done now for the Multis and Quattros, it is more like having a P-loop on top of another PID loop. A lot of the slowness has to do with the outer loop, which are limited to the 600ms (or worse) refresh rate of existing meters, but also, with managed batteries you have to calculate all loads (input, output), translate to a discharge current, then check that with the max discharge current on the BMS (and reduce if necessary), then re-split into input and output values, calculate the setpoint, and send it off.
So it is not ONLY the refresh rate of the meter. The freshness of power values from several other sources also play a role. These power measurements also don’t update at the same time instant, that is to say, every time you calculate a setpoint, some of your input values are already a little stale.
Then throw in a few propagation delays on top.
I wish it was as easy as tuning one PID loop…
Things will naturally improve where the meter sits on a CAN-bus directly to a Multi-RS, which also has a significantly faster microcontroller than the older Multis. In other words, I expect this meter to give you some real gains with a Multi-RS, but it will be a little limited with the Multis. It probably won’t get better than 90% in 3 seconds.
I know I am probably preaching to the choir, but there is also correctional philosophy to consider:
For instance, let’s say all 3 of the responses in the graph meet the 90% /3 s rule.
The stupid sensitive meter will also respond differently to each curve, probably tripping for the red trace but possibly resetting fast enough on another curve.
Wonder if this means Victron is slowely moving away from Carlo meters and what the reason for this will be? Part of it maybe stock issues, since I know stock on the 3 phase Carlo meters was a huge issue at some point during the covid years, even a bigger issue than Ubiquity gear. :smily
@plonkster Any idea how this will be priced vs say a Carlo EM540?
That was one reason (also why support for ABB was done around the same time), but another reason is that this is a VE.Can meter which simply integrates better with other VE.Can hardware that is coming in the future (the Multi-RS is already here).
Probably about on par. Don’t know, can’t say.
might there be a single phase option for domestic use at some point ?
You can always wire only L1, and it will work just fine.
Judging by the specs (and the color), the Carlo Gavazzi is still going to be a LOT easier on the wallet.
(Ethernet adds quite a bit to the BOM cost, and CAN is also not the cheapest option. Then development cost is also a thing).
true, agree… but for pure domestic users you then paying premium for 2 phases you know you will never use…
realise there is domestic users that know they might want 3 phase down the line, the “Omie’s” with big wood work or big CNC shops at home, but someone in the middle of a neighbourhood, well good chance he does not need that extra $$$
If you want an idea of cost, look at something like the Shelly 3EM Pro. That also has Ethernet (and Wi-Fi, and Bluetooth). It probably won’t be less than that, because I think Shelly does have higher volumes.
The Shelly is in some ways a slightly better meter (dedicated energy measurement IC, which jacks up the cost quite a bit), but unfortunately the shelly’s stock speed is quite slow, and while it is possible to significantly jack up the speed (I tested it up to 10Hz with custom firmware they built for me, for testing), it seems stock support for this feature is still far in the future. Sadly.
I used my ET340 as a energy meter using L1 & L2, L1 monitoring single phase grid connection and L2 monitoring my 6.5Kva diesel generator. This wokred well seeing that the generator was started by the Multi’s auto start/stop using a deepsea DS4520 controller and runniing it through a Auto transfer switch into the Multi. I could then clearly see when and how long the generator ran and how much energy it supplied.
It did remove the ET340 after I sold the generator though. It’s collecting dust in my office now. After I realized the generator only started for 10min maintenance runs for more than 8 months, after installing the eTowers, I decided it was not needed and sold it to utilize the space it was taking up better.
So if anyone is looking for an ET340, I’ve got one in it’s original packaging with Victron Usb to serial cable, staring me in the eyes everyday.
@FlashG where are you based? and what do you do for a living?
@plonkster , old buddy old pal… i am so glad these blue guys are coming to the party…
I love your product… and i am truly glad they are adapting…
Let me elaborate, i never since i started the training center had more than a 1 week (3 hour) training session on any other product. Victron training is booked for the next 3 sessions.
EDIT: How on earth do you train someone Victron in a 3 Hour session??
@JacoDeJongh I replied privately. Not sure if you’ve seen it.