diy solar

diy solar

Growatt SPF5000ES doesn't trip RCD in batterymode

@mccljs : You are correct, it's not grounded to a phase but in the center, it's a star. Nice visualization and great first post.
@Bud Martin : Looks like you might have spotted it, I'll confirm later today.

I found a manual for the SPF5000ES in another forum which differs from my manual, the dry contact has different criteria to switch via a setting 24. It's located here https://solaradvice.co.za/wp-content/uploads/2021/01/User-Manual_Growatt-SPF5000-ES_2020_ENG.pdf

My manual has none of that, it switches the dry contact based upon battery condition.
It's my understanding that this setting was removed from the program at some point. You may or may not have this option.
My manual just shows that it switches with state of charge warning.
 
Just checked mine. There is no setting 24.
Hopefully, yours still has it.
 
I dont have it either.

I managed to find the buzzer and solder leads to it, not my finest work but I did it in situ.
1645873426088.png





EDIT: I managed to break the solder joints after I closed everything up. Kinda blows, I need both these things to be in production carrying the house load.
 
Last edited:
Hi everyone! I would like to add some experience to to this thread, and to invite everyone to keep on commenting on this issue.

This year, I have installed many Growatt SPF5000ES in different locations- and I had no trouble at all with all the earth/neutral bonding issue. Until now, every set up consisted in a single SPF5000ES connected to a 220V Grid (neutral distributed and grounded only on the energy provider transformer). Every set up was done with at least 2 RCD, one before the inverter in the main panel, and one after the inverter to feed the loads. Always the inverter ground terminals properly connected to the local earth rod. Never an issue with nuisance RCD tripping, and neither RCD blinding in battery mode. The bonding between Neutral and Ground while working in battery mode is done inside the inverter and seems to work just fine... I was pretty happy with these growatt until now.

Now, I am facing my first setup with paralell inverters (3 of them) and the RCD that is installed before the inverters keeps tripping inmidiatly after Grid is detected and energy is fed to the loads in bypass mode. I have tried with different RCD (types AC, A, B, inmunized, even a 300 mA RCD...) with no luck, until now I have not found what actually causes the main panel RCD tripping. For the moment, I have disconnected the main panel RCD, that is to say before the inverter, which is not totally bad, but leaves unprotected the installation between the mains and the inverter against earth faults.

I will keep on investigating, I have some theories to test:

1- The neutral/earth bonding internal relay does not work properly in paralell mode... would be odd, but you never know whith these guys

2- Leakeage of 3 inverters in paralell is too much for a single RCD (one may think that a 300 mA should do, but it actually does not....). I have found a Growatt note for SPH inverters (not the same as SPF, but should do...) suggesting that a separate 300mA RCD should be installed for each inverter while working in paralell, but I just don't buy it, and I can not do test it becaus I have no spare cables between the main panel and the inverters... One interesting test I did, was to shut down battery charging (by setting "only sun charges the battery in parameter 14 'OSO' "). In this way I supposed, that no high frequency currents would be generated, so leakeage should be minimum... even in this situation the RCD trips as soon the inverter tries to feed the loads with energy frome the grid. So, the leakeage problem doesn't seem to be asociated with high frequencies... strange


I think I could trick some way the RCD whith some external ground relay, but first I need to properly understand what is causing this. I am waiting for a leakeage current clamp meter recentrly bought to arrive, to further investigte this issue, and have many more tests to do...I will not surrender that easy

Please, if any one else has experience working with Growatt SPF paralell set-up and mains RCD tripping, I would be happy to know your experience
Regards,
 
Hi everyone! I would like to add some experience to to this thread, and to invite everyone to keep on commenting on this issue.

This year, I have installed many Growatt SPF5000ES in different locations- and I had no trouble at all with all the earth/neutral bonding issue. Until now, every set up consisted in a single SPF5000ES connected to a 220V Grid (neutral distributed and grounded only on the energy provider transformer). Every set up was done with at least 2 RCD, one before the inverter in the main panel, and one after the inverter to feed the loads. Always the inverter ground terminals properly connected to the local earth rod. Never an issue with nuisance RCD tripping, and neither RCD blinding in battery mode. The bonding between Neutral and Ground while working in battery mode is done inside the inverter and seems to work just fine... I was pretty happy with these growatt until now.

Now, I am facing my first setup with paralell inverters (3 of them) and the RCD that is installed before the inverters keeps tripping inmidiatly after Grid is detected and energy is fed to the loads in bypass mode. I have tried with different RCD (types AC, A, B, inmunized, even a 300 mA RCD...) with no luck, until now I have not found what actually causes the main panel RCD tripping. For the moment, I have disconnected the main panel RCD, that is to say before the inverter, which is not totally bad, but leaves unprotected the installation between the mains and the inverter against earth faults.

I will keep on investigating, I have some theories to test:

1- The neutral/earth bonding internal relay does not work properly in paralell mode... would be odd, but you never know whith these guys

2- Leakeage of 3 inverters in paralell is too much for a single RCD (one may think that a 300 mA should do, but it actually does not....). I have found a Growatt note for SPH inverters (not the same as SPF, but should do...) suggesting that a separate 300mA RCD should be installed for each inverter while working in paralell, but I just don't buy it, and I can not do test it becaus I have no spare cables between the main panel and the inverters... One interesting test I did, was to shut down battery charging (by setting "only sun charges the battery in parameter 14 'OSO' "). In this way I supposed, that no high frequency currents would be generated, so leakeage should be minimum... even in this situation the RCD trips as soon the inverter tries to feed the loads with energy frome the grid. So, the leakeage problem doesn't seem to be asociated with high frequencies... strange


I think I could trick some way the RCD whith some external ground relay, but first I need to properly understand what is causing this. I am waiting for a leakeage current clamp meter recentrly bought to arrive, to further investigte this issue, and have many more tests to do...I will not surrender that easy

Please, if any one else has experience working with Growatt SPF paralell set-up and mains RCD tripping, I would be happy to know your experience
Regards,
The simple fix for your situation is to remove the internal ground screw from two of the inverters. This should solve your issue.
 
Every set up was done with at least 2 RCD, one before the inverter in the main panel, and one after the inverter to feed the loads.
The upstream RCD should not be required since the operation requires creating a new earth bond in the inverter.
 
Latency on the internal relays, one will always be slower and cause earth fault?

I'd love it if you found out, currently I have no RCD before the inverters, only downstream. The inverters are tied to ground, so any mishap there will send it straight into mother earth and anything coming from the inverters is immediatly fed into an RCD before the power is distributed to the main panel supplying power to the house.
 
Latency on the internal relays, one will always be slower and cause earth fault?

I'd love it if you found out, currently I have no RCD before the inverters, only downstream. The inverters are tied to ground, so any mishap there will send it straight into mother earth and anything coming from the inverters is immediatly fed into an RCD before the power is distributed to the main panel supplying power to the house.
No, you are just creating a second neutral-ground bond downstream of the RCD (whenever on battery). So, your delay does not matter in general-- for the RCD to work you need the delay to be miliseconds, and for the neutral relay to work... you aren't going to have that.

@crossy might be able to make a recommendation on how to address it though. The only way I can think to do it is to kludge something together that would disable RCD in a battery mode and for a few seconds afterwards, but that would be a non-standard product.
 
If I am right and @drbytes is in the 220V world then any N-E link that's downstream of an RCD will cause interesting issues, anything from it working fine, to not staying on, to tripping randomly.

To get the RCD to trip when on battery you need a N-E link between the inverter and the RCD - I think that's been established by earlier posts.

Of course, that link will be downstream of any RCD that feeds the inverter, so when on bypass we have the issues associated with a N-E link downstream of an RCD.

The easy solution would be to have no RCD on the supply to the inverter. If your supply is TN-S or TNC-S this is pretty safe as there's a metallic path from your earth terminal to the transformer star-point. Any L-E fault will trip the over-current protection, the downstram RCD will provide user protection to everything fed by the inverter even when on bypass. Whether this is actually permitted by your local regs is another matter.

Otherwise, also as noted in earlier posts, you will need a relay (and some means to control it) to provide a switchable N-E link when on battery. That relay may be too slow opening when going from battery to mains to avoid the incoming RCD tripping so you will need a time-delay RCD on the supply to the inverter.

The inverter manufacturers really ought to have thought of this scenario and included their recommended solution in the manuals, but ...
 
I live in a land with 230 Pixies available at the electric tap.
However I have no Neutral, I have 3 x 230v. So 230V between fases, 135V between fase and earth. It's a crapfest.
 
I live in a land with 230 Pixies available at the electric tap.
However I have no Neutral, I have 3 x 230v. So 230V between fases, 135V between fase and earth. It's a crapfest.

Got it, 3-phase, 3-wire. All bets are off then :(

Our supply in Belgium was like that, 2-pole MCBs on everything and not an RCD in sight (early naughties).

EDIT You are going to have to create a "phantom" star-point which you can ground when on battery, an auto-transformer with the centre-tap grounded may fit the bill. How about trying a 220-110 "foreign stuff" transformer and grounding 1 side of the 110V outlet (check which one goes to the transformer centre tap).
 
Last edited:
Ah, I also live in Belgium.

You wouldn't want an auto transformer, it cannot create a true neutral, I checked with EREA and others.
I simply gave up on it and have implemented everything following the Synergrid regulations and I switch between grid and sun when offgrid installation runs out of juice

I would also need to run the whole house on the transformer, so there is humming, there is a loss of 1-2%, it's just a bunch of crap to add and costs incurred for a minimal return.
 
Ah, I also live in Belgium.

You wouldn't want an auto transformer, it cannot create a true neutral, I checked with EREA and others.
I simply gave up on it and have implemented everything following the Synergrid regulations and I switch between grid and sun when offgrid installation runs out of juice

I would also need to run the whole house on the transformer, so there is humming, there is a loss of 1-2%, it's just a bunch of crap to add and costs incurred for a minimal return.

Sounds like the sensible thing to do.

Actually, thinking about it, when you are on battery you actually have a floating or IT system, you should actually be able to get hold of either of the hots and not get a shock (no, don't try it). IT is actually "2 faults to danger" one hot to E changes nothing, a second hot to E should trip your RCD.
 
It's interesting, when the system is in ByPass it'll have two hots, rcd triggers when fault testing, when it's in solar/battery I have a true L+N, rcd also triggers. It switches in 50ms from L1+L2 to L+N. The house has 3x230 but in the main distribution panel one was terminated and the house was wired monophase. Old house ,..built before there was even electricity.

Oh yeah, my dad can't charge his car here when in ByPass, but when I switch to SBU the carger sees a real Neutral and it start charging.

The electric was added after the second WW and got another overhaul in the 80's. I've done the lastest overhaul, still need to get it inspected but can't be arsed to draw a eendraadschema
 
Back
Top