Victron - Grid Relay test fault

@TheTerribleTriplet, There are a host of difficulties getting a Multi to accept a generator’s sine wave. These can be overcome. The sine wave from an inverter type generator is normally in pretty good shape and usually acceptable. What I am talking about is not getting it working, but it being unsafe when it is.

@plonkster,

This may be the case, but if it is it represents an exception, not the rule for suitcase-type generators. I haven’t personally tested it.
But here is a discussion that people that claim to have tested the Ryobi say it is a V-0-V generator: 230 volt generator with a V-O-V earth connection

@Bobby,
That might explain why your generator survived. This also represents an exception to the norm.

No.
It appears to be something you might have got away with so far because of dumb luck.

Indeed, suitcase types are often V-0-V. And I think we can agree that that should be the first thing to test for. In the old days you’d just put an incandescent lamp between neutral and earth on the generator. It it was V-0-V, there’d be around 110V there and the lamp would come up at about a quarter of its normal brightness (and wouldn’t blow up since it is not a dead short). If it was simply floating (or already bonded) the lamp generally wouldn’t come on at all.

1 Like

HI Phil, I feel that you are over complicating things quite a bit in this spesific case. I am sure Ryobi has different models and topologies. Clearly in the OP’s case it is not a Center tapped Gen, and we can say this with confidence. If you try and understand how the Victron does the relay test, or self test for that matter, you would most probably understand what we are trying to tell you.

Not for one moment do I doubt what you are saying about center tap generators, but in this spesific case, I think you are wrong in your assumption.

I may very well be, it won’t be the first time.

About this I agree, I don’t understand how the Victron does the test or know what all is checked. But let me say that the reason this situation exists is not that I haven’t tried to find out.
The results of my research have not revealed anything but finding other people equally frustrated.
So I am very glad you understand how the relay test works.

This is what I have gathered, it may be wrong and is almost certainly incomprehensive.

There is a test on startup, attempted 3 times.
This test (the mechanism of it I am not sure) attempts to establish if the incoming Neutral is earthed.

  1. If the incoming Neutral is earthed the test passes and the internal grounding relay of the inverter remains open.
  2. If the incoming Neutral is not earthed the test passes and the internal grounding relay of the inverter is closed to provide an Earth.
  3. If there is a Neutral to Earth connection on the Load-side the test fails, regardless of the state of the presence of the incoming Earth the test fails.

Please can you fill in the blanks and correct my understanding.

In short, and as all know, I am not good with words or explaining stuff, so I will ask @plonkster to redo this attempt in his professional manner.

First it looks at what Grid code you use, the grid code will determine what results it will accept for each of the tests.

The main test that I am worried about and focus on the most is : (This is for South African Grid codes)

1: IT test if there is a bond between neutral ans earth on its input. It should be bonded for the test to pass.
2: It test if there is a bond between neutral and earth on its output, if there is, the test will fail because a bond on the output is not allowed.
3: It test to see if Live and neutral is wired correctly, if turned around it fails
4: The rest of the 8 tests I will leave for Plonkster.

For the Australian grid code for example, it will expect to see a bond before and after the inverter as its required to wire the input neutral with the output neutral in Aus…

With Bobbys system, it failed the test because it did not see the bond between the input neutral and earth because the earth was not connected. The moment he earthed the gen, the return path was completed and the bond visible to the inverter, hence the test passed. This is why we say, the gen must have an internal bond, Bobby did not create a bond, neither did the inverter, on testing for the bond the inverter detected is was present and continued to the next step of the test/process.

This is not correct, the inverter will not create a bond while the grid is present, the inverter only closes the Neutral Earthing relay during a grid outage.

If all 8 steps of the self test passes the requirements the Multi will still measure voltage and frequency and look at a few more parameters and only if all are within Grid Code specifications, the Multi will connect to the grid.

This is one of the major advantages of a Victron Multi and the newer Easy Solars, if it passes the relay/self test, You can be certain you dont have any wiring issues.

@plonkster , may I please ask you to use your communication skills to explain the other tests…

EDIT: I forgot to mention, if any of the Relay test fails, the output of the inverter switches off, and wont switch on till it passes all tests. (Except for in a offgrid setup)

3 Likes

Thanks, so the grid code used is all-important to the test results.
Very good, I didn’t know that.
I also see that the ground relay setting can be manually set if you were on a boat for example there would be no grid code and the earth relay becomes selectable manually depending on the shore power’s Earth availability.
This makes sense.
By the way, the ground relay setting is permanently checked for the ZA grid-code, which is a bit unintuitive, given your explanation.

The next question is the mechanism of the test. How do we know that a V-0-V gen doesn’t defeat these tests?
I think this is a reasonable question.
I ask because there seems to be a lot of people that use these types of generators and not a lot of people that know that the distinction even exists between generator types.
Victron themselves don’t seem to draw attention to the distinction either.
There is a fair amount of Victron documentation relating to generator compatibility and I don’t recall this distinction even getting a mention.

Correct, you are not allowed to disable the Ground relay function in SA. The grid code specifies that that relay must be used when the grid falls away so that a new Neutral earth bond can replace the one from the council. By graying that setting out, they prevent a client from disabling the function in a Grid code where it must be used. That tick is to switch the functionality on, not to switch the relay on.

For the rest of your question, I will let the pro answer you.

1 Like

Very good, another important distinction.

1 Like

I tried to see what I can share. The most juicy stuff is in internal documentation and obviously I cannot share that. So I will dumb it down.

It checks 8 things, but only 3 really matters to normal people.

  1. It checks that the input side is TN-bonded. Earlier models did this by passing a small DC current through the loop, later models use another method, which I believe is voltage-based.
  2. It checks that the output side is not-bonded. Again, earlier models passed a small DC current, later models do something else, voltage-based I believe (so it expects to see a voltage between neutral and earth).
  3. Then it closes the bonding relay (the grid is still disconnected). It again checks the output, and this time it expects them to be connected (otherwise the bonding relay failed).

Since it uses a voltage-based method to check bonding, it should fail the moment it tests input side TN bonding and sees a voltage between T and N (as it will with V-0-V), which should stop the relay test immediately and raise error 11.

2 Likes

The GUI helpfully decodes this for you.

error11-ccgx4

That 0x800805 is hexadecimal, and represents a 32-bit number (although the top byte is not used yet, so only 6 chars in the hex string).

Each of those bits mean something. If you want to know, you can decode the QML in the GUI. I can’t be bothered right now… :slight_smile:

0x80 = 10000000 <— reserved
0x08 = 00001000 <— Failed at step 8 (the first step)
0x05 = 00000101 <— error occured, relay test in progress

So when you call up support, they can look at that value and say: Aaah, you need to frobnicate the baz… :slight_smile:

2 Likes

Ok, so I can see that the voltage-based test will find an issue, but the earlier current-based test probably would not.
What is considered earlier?
Multiplus, Multiplus 11, Quatro? When was the watershed?

Multiplus-II. The older Multigrid in the aluminium case uses the current method (and it can sometimes interfere with RCDs). As far as I know the Quattros don’t do a relay test (they don’t have the double-disconnect) so none of this applies to them, and you can indeed connect a V-0-V to those units without raising any alarms.

2 Likes

I happen to have two of those that were purchased at a bargain price a while ago. Is there a way to convert them so that they are safe to connect to the house? If I remove the centre tap and earth neutral instead, will that release the magical smoke that makes all electrical things work?

Firstly, it would seem that not all suitcase-type generators are V-0-V, some appear to be a floating earth. So this must be established.
If you measure from the generator’s frame earth stud to Live and Neutral it will split the voltage 50/50 in a V-0-V generator.
A V-0-V generator would need physical wiring changes internally, these can be done if you can find the connection. Apparently, on some generators this is unfindable.
Then you’re stuck forget it.
In a floating generator, these two measurements could be all over the place for example 70/30 and these may fluctuate. (There isn’t a frame of reference to make the measurement besides tiny leakage current). This is much easier to sort out.
A Neutral-earthed generator is what you want.
Then you will read full voltage to Live and 0V to Neutral.
This is known as a standby generator.
This type of generator won’t have wheels and would typically have a bigger rating.
This is the right type to use in a permanent installation, the downside is it’s more expensive.

There are strict compliance regulations to doing this, and I would involve a professional.
It’s not a DIY job.
Getting it working is one thing, but ensuring your Earth Leakage protection and downline MCB’s will still trip are another.
You see the generator will have enough oomph to do damage but maybe not enough to trip your existing DB protection.
Most DIYers will stop after getting it working and not follow through with the responsible testing afterwards. That is why I am reluctant to give further detail other than how to establish what you have and to contact a professional.

1 Like

I am trained in the field, although admittedly somewhat rusty. I take your points though - it’s probably a bad idea to hook this up to the house, even if I manage to get it to work with the neutral earthed.

It most likely has very little in the way of protections built in. I haven’t actually tested how the cheap Ryobi behaves when overloaded: If the voltage sags it could be bad news for the electronics in the house.

I also suppose the more expensive standby generators will be designed in such a way that if/when they fail they don’t take the loads with them. I would be surprised if the Ryobi has such guarantees.

A little anecdotal story. I did my national service after going to college.
Our 16 man tents were supplied by 30V generators for lights.
We had great pleasure in replacing these 30V bulbs with 220V bulbs so that any overzealous officer who wanted to barge in and wake the troops while it was still dark in the morning would only achieve a tiny red glow and walk away puzzled.
I identified the camps spotlight day/night switch and we always equipped a guard with a lighter so he could put the camp into darkness so that we could hop the fence to enjoy the local nightlife.
Even then electrical training paid off.
Anyway I digress, our tent burnt down, I don’t exactly know why but certainly the wiring was destroyed in the fire.
These canvas tents burn in seconds.
The army got a local electrician to establish if the cause of the fault was electrical. The electrician proudly announced that it couldn’t possibly have been because the MCB on the generator hadn’t tripped. (It was running at the time, but no one was in the tent).

This statement totally overruled one silly troop who drew the opposite conclusion because MCB on the generator hadn’t tripped.

1 Like

Oh I remember those! When the SADF withdrew from Namibia in 1989, all those old things were sold on auction, and the farmers bought them in droves. They were mostly Hoffberg alternator/generator jobbies. Farmers would wire up three lorry batteries in series for 36V, and use those for lights.

The generator would run in the morning, so the lady of the house could do the washing, and so the big “minus 40” deep-cool freezer could run (these had large cooling tanks so they only had to run once a day for a few hours). This would then also charge the batteries, so that the house could have lights tonight.

Because of this prevalence of 32V incandescent lights, 36V battery systems, and army-surplus alternator/generator setups, 36V inverter systems were MUCH more common than 48V. That’s what I grew up with.

This was the first PV array, back in 1986 (ish). They are 33W modules made by a Japanese company called M-Setek. I believe some remnant of them is still around.

Note the lightning-arrestor in the background.

An anecdote of my own: A nearby farmer bought one of these at auction, on a trailer, and I got tasked with wiring it up. I must have been about standard 9, grade 11. It was a really neat 3-phase alternator, and it had to be integrated into his existing single-phase generator setup. Then the earth leakage started tripping, and through lots of trial and error we discovered that it stops happening if you remove the negative earth on the DC side of the trailer. It took many many years before I understood why that happened. Simply, some of the neutral working current would cruise over the bond on the existing installation, then over the negative battery cable out to the trailer, across the negative/earth bond on the trailer, then back on the earth cable, bypassing the RCD.

Of course, familiarity with these genset setups meant you knew names like Lister, Hätz, Yanmar, Kubota by heart… :slight_smile:

1 Like

Yes, this happened in 1989.

My father told me of some of the shenanigans they had. The mechanics would fill up an old rubber tube with acetylene (from a welding rig) and put it on a nearby anthill, and then they’d put a spark plug in through a small hole and wire it to a nearby vehicle. Of course the next morning some over-eager Samajoor would run out and start that vehicle, leading to a blown-up ant-hill, some frayed nerves, and a LOT of drilling of troops to punish them for the prank…

He told me of this one day when they were technically on pass, but just as they were about to leave a call came in for a Bedford truck that needed to be recovered (he was in the recovery team). So they decided to go out and do this one last job. Found the Bedford next to the road with a big pool of oil under it. They ask the driver: What happened? The driver says, I heard this loud bang… the engine was still running but when I saw all the oil I decided to rather stop it…

(It had put a leg out, broke a conrod and knocked a hole int he block. Those old engines were die-hard, they’d still run even after throwing a rod… :slight_smile: ).

2 Likes

Thanks to all who have discussed this issue .
I almost made progress from Marius from Greenpro who seemed to know what was going on within a You tube video i watched however
I have been unable to find any useful info on Victorn community and happened to find this group and topic

This 8/11 error is driving me crazy
It occurs even if I disconnect all the acc outs n, l and e, on the Multuiplus 11.
The error comes and goes and alas is back again after I restarted it
It wont do anything else but invert as it see that its ac in is "disconnected " . It happened with or without a SA grid code so i have removed the grid code and cant use ess but have been trying to set up Ignore AC in a Virtual Switch instead.
I thought my trouble was caused by the tick to change to on when Ie ignore ac current, when general system error occurs - but no, this did not help.
I would prefer to use ESS but im not making any progress in the intermittent ground relay errors resolution

Can anybody help me in Cape Town please ?

image