question

Pierre-Marie Baty avatar image
Pierre-Marie Baty asked

Bypass specific VE.bus error or raise tolerance level ?

Hello all,

Is it possible to force a MultiPlus-II GX to startup even in case of a VE.bus error #8 (ground relay test failed), or better, to raise the tolerance ? Before the usual "it's dangerous" advices start coming, allow me to contextualize.

I have fitted my house with a Multiplus-II GX with a PV array + battery storage working in ESS mode. Works well, a bit undersized though (I'll add some battery soon). I live in Corsica where we have frequent mains shortage in winter. My house is 30 years old and in spite of having partly rewired it anew, there are still parts of the installation that intermittently provoke a ground relay test failure on the Multiplus when it starts up. It is possible that these wires' insulation has eroded, or (as I rather suspect) have received sustained electrical overload by the former owner. Unfortunately these are the wires that are the harder to replace. I could isolate the cause of the problem to 3 circuits and their ground relay test failure is not systematic. Yet, the electrical installation has been verified by the legal diagnosis when the house was bought and was found "safety compliant" - for what it's worth.

It's being pretty much a hassle to wake up in the middle of the night after a power failure to cut the involved loads, shut down the Multiplus, turn it back on and wait 5 minutes for it to initialize and clear the VE.bus error #8, then turn the loads back up again and return to bed. Is there a way to raise the sensitivity of this trigger, or even to disable it completely ? I take full responsibility.

I am a good engineer and programmer, I already modified the Venus web interface to display in portrait mode on my orientation-locked phone, have installed a host compiler on the nanopi and brought a patched-up Midnight Commander of mine, and am starting to scour the root filesystem of the nanopi for the places where this next issue could be treated. I'm not afraid of putting a binary in IDA and patching some bytes if necessary, but I sure hope there's a much simpler solution for my problem.

Could a Victron engineer or a knowledgeable person hint me where to look for ? Thanks in advance

Multiplus-IIVenus GX - VGXVE.Bus
1 comment
2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

Pierre-Marie Baty avatar image Pierre-Marie Baty commented ·

*edited*

Erm. I suggest rephrasing the serial number version table description here: https://www.victronenergy.com/live/ve.bus:ve.bus_error_codes#error_8_-_ground_relay_test_failed

I went under the impression that my serial being not listed in the table, my device had the new hardware, where it's in fact the opposite. In my opinion an explicit table description such as "Devices whose models are listed here and whose serial number is higher or equal than those have the new hardware. Device whose models are not listed here already have the new hardware." -- would not be superfluous.

I made progress in understanding the device. The jargon doesn't help. I've found the firmware file which I believe operates the relay, it's unfortunately in binary form (a Pascal string, a bunch of values, then flat payload data). Without any documentation, I'd have to find the entrypoint, decompile it, get the microcontroller datasheet and reconstruct the electrical diagram if I wanted to understand what to change. No less a hassle than needing to hand-restart the system after each mains shortage.

My problem is still open. So my device has the old hardware and tests ground relay opening/closure by measuring a voltage difference. Which makes me think that for these hardwares a settable trigger value is necessary to check for ground relays and I would like to kindly request a firmware update with that feature.

Thank you for your consideration.

0 Likes 0 ·
1 Answer
markus avatar image
markus answered ·

Setting this grid code will disable the ground relay + test.

But the unit will be not compliant to any grid code then:

1638886909904.png


1638886909904.png (2.9 KiB)
21 comments
2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

Pierre-Marie Baty avatar image Pierre-Marie Baty commented ·

Hello Markus and thank you for your reply.

I am indeed considering this as a temporary solution. But I'm not sure I understand the implications of that choice. It's not that the ground relay should not be *used*, it's that it should not be *tested*, or at least not against the voltage difference value it currently uses.

All my load circuits are wired at the right size (ranging progressively from 16mm2 down to 6mm2, 2.5mm2 and 1.5mm2) and split into sections, each section protected with a differential switch ("interrupteur différentiel" in french - not sure about the translation), and each subgroup of loads protected with a correctly-sized residual current circuit breaker ("disjoncteur différentiel" in french - not sure I translate it right). All of these protection circuits are new. All the earth wires are tied together to the ground of the house - on both sides of the Multiplus. The mains input is additionally protected by a surge protector. I also have a 63A manual override switch for the case where the MP would fail for some reason.

If I set the "AC Neutral path externally joined" grid setting, I understand that I have to *actually wire* the neutrals between the input and the output of the Multiplus to be on the safe side ?

0 Likes 0 ·
markus avatar image markus ♦♦ Pierre-Marie Baty commented ·

Hi,

By your request, I left out the "usual "it's dangerous" advices" but this is indeed dangerous and might lead to a not safe installation.

After configuration, always confirm that the system is safe with routine electrical safety checks in both modes (eg test RCD operation with AC input connected and disconnected).

"If I set the "AC Neutral path externally joined" grid setting, I understand that I have to *actually wire* the neutrals between the input and the output of the Multiplus to be on the safe side ?"

This will be the case, yes.

But the final solution should be, to find and repair the supposed PE-N leak(s) inside your installation that triggers the Error8.


BR

0 Likes 0 ·
Pierre-Marie Baty avatar image Pierre-Marie Baty markus ♦♦ commented ·

...and I thank you sincerely for having done so and having provided me with a factual answer.

I don't see in my VEConfig utility any grid code called "Other: AC Neutral path externally joined (not compliant)". The only choices that have this label tag are Australian and New Zealand grid codes. Where did you take yours from ?

Also, it asks me for a password. Which was not documented anywhere... Had to extract it from the Delphi software itself: T******C. You might want to put that one on the community wiki perhaps... ?

And while I agree with you that finding and repairing those *SUPPOSED* PE-N leaks would be the way to go, they're not detectable with the tools that I have, they're not enough to make my protection circuits trip, and the electrician who made the diagnosis didn't find them either : it's not a trivial situation. So I *still think* this voltage-based detection should be made tunable. I'd be really surprised if I was the only person with that problem.

Thank you for your replies nevertheless. How do I obtain that famous "Other" grid code that you showed me ?

0 Likes 0 ·
markus avatar image markus ♦♦ Pierre-Marie Baty commented ·

Don't forget about the onreaders on a public community, that spend 20 seconds to this post and alter with grid code settings right after.

Therefore this password is not public and only for dealers and installers.

Which firmware version is installed on your MP2s? This screenshot was from a MP2 5k unit.

I agree with you, that this error triggers way too early on older MP2 models. I saw Error 8 because of a Voltmeter between N and PE, some surge protectors can trigger this error too.

0 Likes 0 ·
Pierre-Marie Baty avatar image Pierre-Marie Baty markus ♦♦ commented ·

Here you are:

Multiplus-II GX 48/5000/70-50 (product number PMP482506000)

Firmware version 470
VE.Bus version 2623470
MK3 version 1170210

Hardware version (serial number) is HQ2041BZF9A

Happy to see that we agree on that. I hope you can escalate that so that a firmware update adds this setting if possible. I can't stress out enough how annoying it can be when shortages are frequent.

I volunteer for testing, if necessary.

P.S. The ergonomy of this forum is not the most intuitive either. If I want to reply to a thread, I'm tempted to use the reply box at the bottom... fail. Instead it creates a fork in the discussion. In order to reply I have to click that tiny "Answer" blue link first, and type in a second reply box. IMO that could be improved.

0 Likes 0 ·
Show more comments