PIP-5048GE, PIP-5048GK and PIP-5048MG inverters (with unsafe SCC)

[Nietschy]

Welcome to the forum.

Thanks for the greetings

We have a separate topic for the GE/GK/MG models. But since this is a protocol issue, and the protocols are more or less uniform across the models, I'm happy to let this post stay in this topic.

I appreciate for you letting this stay here!
I know that there is a GK thread, but there is not much going on, and since you talk here about protocols, I thought I may put it here.

Perhaps you are using in-between values?
Switching back to battery mode is contingent on the battery voltage exceeding setting 13. Be aware of the effect of setting 29 (low DC cutoff voltage), as mentioned in FAQ question 2.

I tried this with multiple variations of the values, but with no luck.
I run the solution from ned-kelly https://github.com/ned-kelly/docker-vol ... eassistant (thank you for your work!)

Code: Select all

root@orgor:~# docker exec -it voltronic-mqtt /opt/inverter-cli/bin/inverter_poller -d -1 -r MUCHGC020
Mon Oct 14 06:23:43 2019 INVERTER: Debug set
Mon Oct 14 06:23:43 2019 INVERTER: Current CRC: F3 F1
Mon Oct 14 06:23:43 2019 INVERTER: cmd: MUCHGC020
Mon Oct 14 06:23:43 2019 INVERTER: buf2: MUCHGC020
Mon Oct 14 06:23:45 2019 INVERTER: MUCHGC020 reply size (7 bytes)
Mon Oct 14 06:23:45 2019 INVERTER: MUCHGC020: 7 bytes read: (NAK
Mon Oct 14 06:23:45 2019 INVERTER: MUCHGC020 query finished
Reply:  NAK

but as it seems, I am getting no where with it. I start to think that there is some error in the program with commands that are longer than 8 chars...
If you also say, that these commands are correct, I need to beef up some cpp wisdom of mine.

Also thank you for your help with the commands, since I am stuck with original firmware, I will see what parameters I can use while keeping the Pylontech protocol running.
I am running the Pylontech batteries that come with a serial interface to monitor the BMS and react accordingly. I thought this combination was a good fit, but since the reaction voltages you describe, are still for lead acid batteries, I start to question the integration level from MPP solar here...
Sadly many parameters (2, 26, 27, 29) are locked to certain values if you select PYL as battery type.

So I started by changing parameter 12 to 49V to get to the 2V difference you mentioned. I hope this gets me somewhere.

[coulomb]

If the PIP switches to line mode after a 110% load event, it stays there forever until I completely turn it off and on again

I just noticed that setting 13 has the note: "when selecting SBU priority or Solar first in program 01". Are you in either of these output source priorities?
[ Edit: oops. That's for an MS model; GKs could well be different. ]

[Nietschy]

I just noticed that setting 13 has the note: "when selecting SBU priority or Solar first in program 01". Are you in either of these output source priorities?

Yes I do the following:

In the morning when the sun comes up and the other 7.2kW inverter starts delivering:
- I change program 01 to USB
-- so I use my "free" solar energy for the house
- I change program 16 to SNU
-- so the batteries get charged by my "free" utility energy


In the evening when the sun goes down:
- I change program 01 to SUB
-- so I use the battery energy
- I change program 16 to OSO
-- so charging stops

The change is done automatically by homeassistant with the integration by ned-kelly and a few mods by me.

I do not understand what you try to tell me. So I hope you can clarify based on my configuration.

[coulomb]

I just noticed that setting 13 has the note: "when selecting SBU priority or Solar first in program 01". Are you in either of these output source priorities?

Oops! I just realised that I was using an MS manual, and the GKs are different. I had it in my head that only MKs (Kings) are very different with their double conversion, but in fact all the models with the high voltage SCCs behave differently. Sorry for the confusion.

Yes I do the following:

In the morning when the sun comes up and the other 7.2kW inverter starts delivering:

Oh! I didn't pay proper attention to your other grid-tied inverter. I had assumed that most of your panels would be connected to the PIP-5048GK. In fact, do you have any solar panels on the PIP-5048GK at all?

- I change program 01 to USB
-- so I use my "free" solar energy for the house
- I change program 16 to SNU
-- so the batteries get charged by my "free" utility energy

OK, an interesting setup. So it makes sense that during the day, the inverter essentially connects AC-in to AC-out. It's not clear to me whether the GK will blend any solar panel power that you might have with the mix of grid (hopefully little) and grid-tie solar power.

In the evening when the sun goes down:
- I change program 01 to SUB
-- so I use the battery energy

? But SUb prioritises Utility over battery. While there is utility available, the battery won't be used. I believe that you need SbU here. My understanding is that this inverter can't blend battery and utility power; it's one or the other. That's presumably why you can't get it out of line mode at night.

[Nietschy]

Hey coulomb thank you so much for your reply!

Oh! I didn't pay proper attention to your other grid-tied inverter. I had assumed that most of your panels would be connected to the PIP-5048GK. In fact, do you have any solar panels on the PIP-5048GK at all?

As a matter of fact no, not yet, maybe never.
The 24x 320W panels on my roof are more than enough for now.

? But SUb prioritises Utility over battery. While there is utility available, the battery won't be used. I believe that you need SbU here. My understanding is that this inverter can't blend battery and utility power; it's one or the other. That's presumably why you can't get it out of line mode at night.

Oh maaan, sorry for the confusion!
I mixed it up when telling you. Of course I use SbU at night!!!
Sorry again.

No it can not blend anything.
It is battery or bypass nothing else.
The only thing that was new for me, was that it can still overload when in bypass mode... But that is no big deal.


So... I have news from my change of rising the recharge value (parameter 12) to 49 volts...
It was a bad idea and its obvious now. The PIP recharged half of the night from AC in. No big deal, but useless for my project.

Soo.. since I can not get any modified firmware, I will go the custom battery setting route and kick my BMS integration with the Pylontech that way.
Would be great if the BMS communication would still work, but I doubt it.

I will try to set your suggestions out of your FAQ (lowering the cut off voltage a lot) and see where it gets me.
I think the BMS of the battery itself will cut off on its own when needed (hopefully!).

Thanks again!

If you have any more tips to get my charging speed changed, I would be happy to hear them.
I run an Ubuntu Server on an old laptop, so I can run anyting from shell, python to binarys of some sort. I can also compile c if needed.
So I guess the easiest route would be a binary for every change I need. So I can start "20A AC charing.bin" or "40A AC charging.bin" whatever needed.
What I try to say here is: I just need ANY linux way to change this setting with my USB (hidraw0 device) on my inverter...
I just do not want to ditch the homeassistant way of reading the data (feeding with MQTT...), becauuse I started to like it a lot.

[coulomb]

I start to think that there is some error in the program with commands that are longer than 8 chars...

Well, that's actually testable. Unlike the MS models (with firmware up to version 73.00 at least), the PE command can take a variable number of parameters. So you could send the 9 character PExxxxxxx command (literally with lower case Xs), which should have the effect of turning on the backlight 7 times This should be ACK'd, but have no other effect if your backlight is turned on already. Use the PDxxxxxxx command to convince yourself that these 9 character commands are actually being processed (by disabling the backlight), and of course a PEx command to turn it back on if desired.

[ Edit: Remember that it takes a minute or so for the backlight to actually turn off after no button presses. ]

I note that the protocol documents are somewhat confusing with respect to the PE and PD commands. For example, they seem to imply that you use upper case letters, and that PE and PD can somehow be mixed in one commamd; neither is true.

[ Edit: 8 character → 9 character ]

[Nietschy]

Well, that's actually testable. [...] This should be ACK'd, but have no other effect if your backlight is turned on already.

thanks!

did that. the result speaks for itself:

Code: Select all

root@orgor:~# docker exec -it voltronic-mqtt /opt/inverter-cli/bin/inverter_poller -d -1 -r PExxxxxxx
Wed Oct 16 04:58:30 2019 INVERTER: Debug set
Wed Oct 16 04:58:30 2019 INVERTER: Current CRC: 3A 80
Wed Oct 16 04:58:30 2019 INVERTER: cmd: PExxxxxxx n: 9
Wed Oct 16 04:58:30 2019 INVERTER: buf2: PExxxxxxx
Wed Oct 16 04:58:35 2019 INVERTER: PExxxxxxx read timeout
Wed Oct 16 04:58:35 2019 INVERTER: PExxxxxxx reply too short (0 bytes)
Reply:  
root@orgor:~# docker exec -it voltronic-mqtt /opt/inverter-cli/bin/inverter_poller -d -1 -r PExxxxxxx
Wed Oct 16 04:58:37 2019 INVERTER: Debug set
Wed Oct 16 04:58:37 2019 INVERTER: Current CRC: 3A 80
Wed Oct 16 04:58:37 2019 INVERTER: cmd: PExxxxxxx n: 9
Wed Oct 16 04:58:37 2019 INVERTER: buf2: PExxxxxxx
Wed Oct 16 04:58:42 2019 INVERTER: PExxxxxxx read timeout
Wed Oct 16 04:58:42 2019 INVERTER: PExxxxxxx reply too short (0 bytes)
Reply:  
root@orgor:~# docker exec -it voltronic-mqtt /opt/inverter-cli/bin/inverter_poller -d -1 -r PExxxxxx
Wed Oct 16 04:58:45 2019 INVERTER: Debug set

Wed Oct 16 04:58:45 2019 INVERTER: Current CRC: F0 2A
Wed Oct 16 04:58:45 2019 INVERTER: cmd: PExxxxxx n: 8
Wed Oct 16 04:58:45 2019 INVERTER: buf2: PExxxxxx
Wed Oct 16 04:58:45 2019 INVERTER: PExxxxxx reply size (7 bytes)
Wed Oct 16 04:58:45 2019 INVERTER: PExxxxxx: 7 bytes read: (ACK
Wed Oct 16 04:58:45 2019 INVERTER: PExxxxxx query finished
Reply:  ACK

I have just seen that the 8 chars do some timeouts, too.
So it gets worse the longer the command is.
Strange...
I guess I will have to try to fix it with the creator of this software. I looked into it, but do only get half of it.

[Prajna]

I have just seen that the 8 chars do some timeouts, too.
So it gets worse the longer the command is.
Strange...
I guess I will have to try to fix it with the creator of this software. I looked into it, but do only get half of it.

Hmm... are you using usb to talk to the inverter? It seems that usb (hid device) needs things in packets of 8 characters, which may explain the problem. I've been talking rs232 rather than usb so don't have experience with it. If you'd like to look at my code, in case it helps at all even though it's work in progress, its at https://github.com/prajna-pranab/converse

[josse]

Yes, with python and library hidraw on usb, everything longer than 8 bytes, you must split into more messages by 8 bytes. Problem is in Python library, not in inverter. Sending long messages only.

Look to example:
https://github.com/JosefKrieglstein/Axp ... /axpert.py
Find: def serial_command(command,device):

Not sure if it's last version, but serial_command is actual.

With this it's much slower, but ok for long commands.

[Nietschy]

Yes, with python and library hidraw on usb, everything longer than 8 bytes, you must split into more messages by 8 bytes. Problem is in Python library, not in inverter. Sending long messages only.

Look to example:
https://github.com/JosefKrieglstein/Axp ... /axpert.py
Find: def serial_command(command,device):

Not sure if it's last version, but serial_command is actual.

With this it's much slower, but ok for long commands.

Ohh maaaaan!
Thank you all so much for your help!

FINALLY! it works!!

This is the best thing I saw after a few afternoons of failure!

root@orgor:/usr/share/hassio/share# python auto_chg.py
MUCHGC030 : (ACK9
30

You are all perfectly right!
USB needs it to be split into commands of 8 bytes each. Since I am not really a big developer, I did not cope with writing it on my own.
But your script is perfect for my needs! I changed a few things I do not need, and et voilá! IT JUST WORKS!

Thank you all again for your help!
Also praya for your way, but I will stick to python for nor since my fronius export works on python, too. This way I can just copy the part from my tesla wallbox script into this one!

[Prajna]

@ Prajna or @coulomb if you can supply me with commands for my PIP would be just perfect!
Any help would be much appreciated!

Hi Andre, sorry I didn't see this till now. I'm still doing my best to figure out the protocol. It seems there are a number of different versions that all report themselves to be PI30 (protocol 30) but they all vary, either in the commands listed or the return values from the query commands. At the moment I am just working with the query commands (those starting with Q, except for VERFW, which is also a query command).

My inverter reports itself to be VMIII-5000 in response to the QMN command and the closest protocol manual I've found so far is the one titledPIP-GK_MK Protocol.pdf, which someone posted here recently.

According to that manual there are two setting commands: MNCHGC, to set the maximum charging current, and MUCHGC, to set the maximum utility charging current. One thing you should perhaps note is that there are only fixed values you can set and these are listed in response to the QMCHGCR and QMUCHGCR queries.

I hope that helps a little.

Edited to add max Utility carging current and QMN command

[birdibird]

This is great news!
I saw in https://powerforum.co.za/topic/4288-axp ... at-midday/ that there is a firmware update for the Axpert MKS II.
I've got the MPPSolar 5048MG (@71.40) so I guess can flash this Axpert MKS II 71.71 firmware, correct?

@coulomb : I would really appreciate (and with me some others I know with a 5048MG) if you could patch this firmware (for LFP)!

[coulomb]

I replied to this yesterday, but it doesn't seem to have come through.

I saw in https://powerforum.co.za/topic/4288-axp ... at-midday/ that there is a firmware update for the Axpert MKS II.
I've got the MPPSolar 5048MG (@71.40) so I guess can flash this Axpert MKS II 71.71 firmware, correct?

That's a great question. The 71.XX factory firmware seem to be a bit muddled. For example, 71.70 and 71.80 are for Axpert King / PIP5048MK models, yet version 71.71 seems to be for the Axpert MKS II with 64 V.

So despite the considerable gap from 71.40 to 71.71, it looks like it should work. I presume your model, being a PIP, has the 64 V option? (In other words, can setting 26 go past 58.4 V?)

I would really appreciate (and with me some others I know with a 5048MG) if you could patch this firmware (for LFP)!

As it happens, Weber and I are not particularly interested in the models with 450 V max SCCs. However, we are in communication with a German forum user who may be able to take on the patching of these models, starting with version 71.71. So there is hope!

[birdibird]

I replied to this yesterday, but it doesn't seem to have come through.

I saw in https://powerforum.co.za/topic/4288-axp ... at-midday/ that there is a firmware update for the Axpert MKS II.
I've got the MPPSolar 5048MG (@71.40) so I guess can flash this Axpert MKS II 71.71 firmware, correct?

That's a great question. The 71.XX factory firmware seem to be a bit muddled. For example, 71.70 and 71.80 are for Axpert King / PIP5048MK models, yet version 71.71 seems to be for the Axpert MKS II with 64 V.

So despite the considerable gap from 71.40 to 71.71, it looks like it should work. I presume your model, being a PIP, has the 64 V option? (In other words, can setting 26 go past 58.4 V?)

Yes, and I don't know of 5048 MG versions which don't do 64V, as they are quite recent, do you?

I would really appreciate (and with me some others I know with a 5048MG) if you could patch this firmware (for LFP)!

As it happens, Weber and I are not particularly interested in the models with 450 V max SCCs. However, we are in communication with a German forum user who may be able to take on the patching of these models, starting with version 71.71. So there is hope!

Ah, that's great. Is that the photovoltaik forum? Maybe you can connect us. I speak German (I'm Dutch) and can test if needed.

[weber]

So despite the considerable gap from 71.40 to 71.71, it looks like it should work. I presume your model, being a PIP, has the 64 V option? (In other words, can setting 26 go past 58.4 V?)

Yes, and I don't know of 5048 MG versions which don't do 64V, as they are quite recent, do you?

When a model has both 58.4 V and 64 V variants, we've noticed that so far MPP Solar have sold only the 64 V variant, so it's true that all PIP-5048MG are 64 V. But the equivalent Axpert MKS II 5K was available with both 58.4 V and 64 V variants. For example, the Mecer (rhymes with "spacer") SOL-I-AX-5NB was the 58.4 V variant (at least until recently). But I notice that the Voltronic Power website now only shows the Axpert MKS II 5K as 64 V.

Coulomb and I have reason to believe that the high frequency transformer has a turns ratio of 8:1 in 58.4 V variants and 7:1 in 64 V variants (of any 5 kVA model). And consequently, we suspect the boost-converter has to work harder in the 64 V variant when inverting.

As it happens, Weber and I are not particularly interested in the models with 450 V max SCCs. However, we are in communication with a German forum user who may be able to take on the patching of these models, starting with version 71.71. So there is hope!

Ah, that's great. Is that the photovoltaik forum? Maybe you can connect us. I speak German (I'm Dutch) and can test if needed.

I believe Coulomb meant "German (forum user)" i.e. of this forum, rather than "(German forum) user", although he may well be both. You can contact @derBastler by private message on this forum. You can see his posts here.

[birdibird]

I flashed 71.71 and everything went fine.
I will post my findings here if anything stands out.

If have sent @derBastler a pm

[coulomb]

Coulomb and I have reason to believe that the high frequency transformer has a turns ratio of 8:1 in 58.4 V variants and 7:1 in 64 V variants (of any 5 kVA model). And consequently, we suspect the boost-converter has to work harder in the 64 V variant when inverting.

In our investigations yesterday, Weber and I confirmed those turns ratios, but we found that the high voltage DC bus has no additional boost beyond the 8 or 7 times battery voltage. This means that the 64 V models can no longer produce a sine wave when the battery voltage falls below 48 V! Instead, the waveform is clipped top and bottom, and this gets worse as the battery drops further. In the 58.4 V models, the onset of clipping is a far more acceptable 42 V.

[ Edit: longer quote, major rewriting for clarity. Thanks, Weber. ]

[paulvk]

This means that the 64 V models can no longer produce a sine wave when the battery voltage falls below 48 V! Instead, the waveform is clipped top and bottom, and this gets worse as the battery drops further. In the 58.4 V models, the onset of clipping is a far more acceptable 42 V.

So it seems that increasing the battery voltage would be a good idea , for lead acid another cell or two and another cell for lithium.
But it means they are no good for nickel iron as a 48v battery needs 66v for a refresh charge the 64v gets them to about 95%.

[leo46]

Hello everyone,

My name is Florent, I'm from France.
I've bought a PIP-5048MG from MPPSOLAR 2 months ago, it worked well for a good month, and now i'm facing troubles.
When it's on heavy load (around 4500W) or sometime medium load (around 3000W) my main breaker of the house (500mA) fails. Normally this protection fails only when there is a problem with the ground (as i know).

I've asked MPPsolar but they are quite silent for the moment on the subject. No idea..
As you can see here, in october i ran very high sometime and had no failure :


And recently, even without an heavy load :


And today, normal load :


It's not happening at all when i'm on the grid instead of the inverter. With the inverter, it's not happening with the same load everytime. Sometime dishwasher, sometime oven, etc..
How my house is connected :
Grid -> 500mA -> distributor to A and B:

A : 30mA breaker -> inverter
B is a source selector : it has both output of inverter and the grid so i can run the house from one or the other.

Have you faced this problem with the ground ?

[ Edited Coulomb: Fixed image links; see Uploading Images ]

[paulvk]

I recently had a problem similar to this.
My 30ma protection circuit breaker was tripping with my washing machine
but only while running on the inverter
change to grid circuit with its own 30ma protection breaker all worked fine
the machine is 30 years old
I replaced the starting capacitor with a new one problem gone!
Why it was the capacitor I am not certain
but this would have caused a unbalanced inductive power factor.

I tested the washing machine on my portable 10ma safety unit
and it did not trip so not earth leakage

Note have been using the washing machine for years off the inverters

Regards Paul

[leo46]

Maybe I have some kind of same issue.
All the ups and downs you see on the graph is made by 2 pumps like this https://www.amazon.fr/Ribiland-Pompe-su ... 81&sr=8-41
They pump water in my underground to avoid my cave to be filled when it's raining a lot. And for a few weeks it simply doesn't stop raining, so they start and stop every few minutes for 15sec.

Maybe by their way of starting / stopping they create a problem with the inverter / ground that i don't have on the grid.

In this case i will try to put this 30mA breaker always on the grid.

[Frenelius]

Hi, I am new to this forum.

I am actually considering installing a 5048-GK, but I am unsure about the max solar current it can accept.

The solar panels I want to use are 10x Jinko Mono 400W, Voc=49.8V and Isc=10.36A, in 5s2p configuration, so in total: Voc=249V and Isc=20.72A

I have read that the theoretically 18A limitation on the solar array input is just the maximum that the inverter is going to take, BUT isn't the MPPT algorithm going to short circuit the panels once in a while for testing the maximum power possible?

Maybe the 18A limitation is due to that short circuit capability?

Should I consider the MK model in 2s5p configuration? I was avoiding it because I want to install the inverter at about 70ft from the panels, so I could reduce cable section with the GK.

Thank you!

[weber]

I have read that the theoretically 18A limitation on the solar array input is just the maximum that the inverter is going to take, BUT isn't the MPPT algorithm going to short circuit the panels once in a while for testing the maximum power possible?

Maybe the 18A limitation is due to that short circuit capability?

Welcome, Frenelius. No, it never short circuits the array. The maximum input current rating is just the maximum that the inverter is going to take. So 5s2p will be fine with the GK.

Should I consider the MK model in 2s5p configuration? I was avoiding it because I want to install the inverter at about 70ft from the panels, so I could reduce cable section with the GK.

2s5p would be correct for the MK. One reason to prefer the MK over the GK is that the GK requires a lethal array voltage but does not provide any insulation monitoring. But as you say, the MK will require a lot more copper.

Both the MK and the GK seem to have problems with dropping the PV array for no apparent reason, sometimes not reconnecting until the next morning or until reset (called "PV freezing").

[Frenelius]

Both the MK and the GK seem to have problems with dropping the PV array for no apparent reason, sometimes not reconnecting until the next morning or until reset (called "PV freezing").

Hi Weber,

Thank you for your comments.

I was not aware of that problem. I think I will do some more research before deciding which model to install.

I am considering installing the inverter closer to the panels so I could go for the MK.

Thanks!

[coiso]

Hello everyone,
I bought an EASUN SM II 5K inverter on aliexpress in August and was only able to plug it in for a day with batteries because, without batteries, the inverter only worked in AC mode, even in the sun.
One day later, with the batteries connected, the inverter loops and every 40 seconds restarts. At PV input, 220Vdc, but the software reads 600Vdc. Amazing! I complained to the manufacturer and he just sent me last Friday an MPPT PCB to replace. I replaced the module, but the problem remains, now with the aggravating reading of the 115Vdc battery input. I no longer expect the inverter to work with the help of EASUN, because they are very slow to respond and take a long time to make a decision, so I think I should start working and trying to fix it myself. Two errors are shown: 10 and sometimes 12 before and after changing the MPPT PCB. Any idea?
Thanks.