A strange Multiplus/Node-Red behaviour

Modifications
, , , ,
1

I’m experiencing an issue with my Victron /node red setup.

My Node-Red is set to change the Multiplus state from “charger only” to “inverter only” after the night time charging cycle.

However, when I wake up the next morning, I always find the Multi in the “ON” state instead.

  • To clarify the situation I can explain how my charging cycle is setup:

  • Here is what it looks like:

  • 09-REVISED-Charge Profile copy
    09-REVISED-Charge Profile copy1701×867 278 KB

    1. My AC-In is always set to “ignore AC-In” which only becomes active when one of the Multiplus’s digital input is triggered to “Generator-Start” by the Cerbo’s relay.

      • NODE_RED

    2. An inject node sets the inverter to charge only mode at a specific time. (this also triggers the external transfer unit to change from island to grid)

    3. This causes the AC- out to disconnect passing the grid through as soon as the Gen-Start signal triggers the AC-In ignore to “off” (allowing the pass through)

    4. An external timed relay activates connecting the AC-In to the grid

    5. At this point Linx SOC, subject to the time setting, commands “Generator ON” finally completing the cycle.

    The opposite happens when charging needs to be interrupted avoiding high tariff times so:

    1. First the Gen is turned off by Cerbo

    2. Then Node Red’s Injects an “Inverter only” to the Multiplus

    3. Finally the external relay fully isolates the AC-In which had already gone back to “Ignore AC In”

    (The external relay is an extra precaution as I’ have noticed that, even when in Ignore AC In state, the VRM shows the AC-In voltage as active)

    The reason for this setup is that my solar shed is about 90ft from the grid service point and I cannot, or want, to divert the grid all the way there to come back to the property having to go through the Multiplus continusly.

    I want the property to remain with the original setup and in compliance.

    A transfer switch takes care of what the Multi would do if used conventionally.

    So, if Node-Red’s inject node sends the “inverter only” signal at the end of the cycle but in the morning I always find the Multi in the “ON” state instead, why?

    The question is, what else can override the inject’s “inverter only” command I ask?

    Any idea out there?

2

I think I might have found the culprit:

After having removed the SOC gate, the issue disappeared.

I now have to investigate why

3

I have a similar situation. Did you figure out what was happening?

4

No, I have not, but I know for sure now that Node-Red was not responsible.

During my tests, I have noticed a strange behaviour even when operating directly fom Cerbo’s console.

Every time I change the Multi switch state from “inverter only” to “charger only”, it immediately goes into the “on” state.

Having said that, I now have to run the same test without the AC out connected to verify that this is not caused by the external transfer switch.

I will report when this is done. In the meantime, you could confirm that this is or is not the case with your Multi.:wink:

5

Hmm how strange. I was testing a theory that it was the function node causing the problem, but like you I determined yesterday it was not the case.

Can you give me detailed steps to duplicate your tests?

Also should we contact some kind of victron support team?

6

Just use your Cerbo console, or whatever you use for direct GX control, and try the Multi Switch sequence: Inverter only > Charger only

If it operates correctly, it will accept the change then you don’t have my problem.

Conversely, as in my case, it will immediately jump to Inverter On State

You see, because I am using a Temporary external transfer switch (more on that later), the sudden change of state might cause a very brief arc between the contacts sufficient to trip the grid RCD. The loss of AC1 may cause the inverter to go back to the ON state, but this is only my speculation.

This will be fixed once I have completed the final system, where I will force AC1out to be interrupted a minute before the charging commences, thus avoiding any conflict in the line.

Run the above simple test and see if your system accepts the change of state correctly.

I’d like not to disturb the tech department until I have run out of options; the problem might be external, as I speculated earlier.

However, the fact that when in Charge-Mode only if the grid fails, the Master Multi (I have two in parallel) shows one yellow LED marked Float, that is something that I was not expecting and might be worth asking at some point.

7

Sorry for the delay. I had family visiting.

Ok so performed the following actions
✅Inverter was ‘On’. Selected ‘Inverter only’. Result ‘Inverter only’ after 2 seconds.
:cross_mark: Inverter was ‘Inverter only’. Selected ‘Charger only’. Result ‘On’ after 2 seconds.

✅Inverter was ‘Off’. Selected ‘Charger only’. Result ‘Charger only’ after 2 seconds.
:cross_mark: Inverter was ‘Charger only’. Selected ‘Inverter only’. Result ‘On’ after 2 seconds.

Its almost like when changing between ‘Inverter only’ and ‘Charger only’ it is forgetting to turn the current mode off.
Then the system notes that both are on and show the ‘On’ mode.

However, it is inconsitant. I then repeated the tests and it worked fine. I then repeated the tests again and the mode toggled for a few seconds before settle to ‘On’ (Which was incorrect)

Any thoughts on this?

8

I apologise for the delay too, in my case it is consistently doing it, every single time I go from charger only to inverter only, It systematically jumps to inverter on.

I’m not sure where to ask for a potential solution aor if this is the correct or incorrect behaviour of the inverter.

Thanks for letting me know your discoveries, I will continuously test and see if I also get an inconsistent outcome.