Jump to content

Axpert invertor parallel setup Can error #80


Recommended Posts

Hi 

Can someone please advise

I have 2 x 5KW Axpert invertors in parallel with 2 pylontech batteries and 9 x solar panels. Once a day, around Midday, the invertors  trip with error code #80 (CAN error). I am also running ICC software on the system

Have been running the system for 10 months with no previous issues

I have no clue as to what the issue may be?

Regards

 

Link to post
Share on other sites
11 hours ago, IrshadS said:

Once a day, around Midday, the invertors  trip with error code #80 (CAN error).

That's pretty weird. I assume you can restart the machines and all is well, until the next day?

It could be a problem with the paralleling cables (they carry the CAN bus from one inverter to another); perhaps one is loose, and in the midday heat, expansion causes it to work loose, but at other times of the day it hangs on just barely? Perhaps also clean the pins with alcohol or better yet with contact cleaner (I use CRC brand, I don't know what is available in South Africa). There could also be a chip on one of the parallel boards that is marginal, and fails at higher temperatures. Perhaps experiment with extra cooling for one or the other inverter to see if that seems to fit.

Are there days when it doesn't happen? If so, are these always cooler days?

If it was exactly at midday every day, there is some new buggy code in the latest firmwares that seems to be ready for hour by hour changing of output source priority (I think), and it seems to have an off-by-one error somewhere that sometimes causes odd things to happen at midday. Actually, now that I type this, I think it happens at midnight, not midday. But I suppose that the internal clock might be off by nearly 12 hours. This seems unlikely to trigger your symptoms, but I mention it in case other measures fail.

Link to post
Share on other sites
  • 2 weeks later...

HI Coulomb, 

Thanks for the reply

I have been monitoring his now for a week after, This definitely only happened when we were in a heatwave. Since then, outside temperatures have been cooler and no more errors.  

However, i was wondering if the CAN error is actually caused by raspberry Pi that is getting "too hot" instead of the invertors, I only assume this as the error is a communication error and not overheating or something to that fact?

Regards

 

Link to post
Share on other sites
12 hours ago, IrshadS said:

I have been monitoring his now for a week after, This definitely only happened when we were in a heatwave. Since then, outside temperatures have been cooler and no more errors.  

That's progress of a sort.

12 hours ago, IrshadS said:

However, i was wondering if the CAN error is actually caused by raspberry Pi that is getting "too hot" instead of the invertors,

No, the CAN bus is only between inverters. Your Pi would connect only via RS-232, and would not be causing this error.

Can you tell if the error originates from only one of the inverters? If so, you could swap the parallel cards (fair bit of work) to see if it's the parallel card or something inside the inverter-charger. But I suspect you can't tell which machine initiated the error.

You could try using a heat source (e.g. hair dryer) when it's not playing up on a cooler day to see if one of the parallel boards seems sensitive to heat. If so, just buy another paralleling kit from your supplier and replace the sensitive paralleling board.

Link to post
Share on other sites
  • 9 months later...
  • 1 month later...

I'm operating my system with a Synerji/Axpert MKS 5K (5kW, 74.40) inverter. So far it works fine. But with electric cooking the cook sometimes manages to overload the inverter. It then correctly switches to utility bypass. In order to prevent this I plan to parallel it with a second unit that would also serve as spare in case one unit fails. I made a submission on Internet with all the suppliers in SA. The only concrete offer comes from Herholds in Bloemfontein. But now they informed me that the new units on stock run a newer firmware without specifying it. I wonder if it would be compatible with my existing unit. Herholds promised to inquire and inform me.

Has anybody experience with this? Would it be possible to flash the 74.40 into it?

Link to post
Share on other sites
16 hours ago, Beat said:

Would it be possible to flash the 74.40 into it?

You'd have to get a Synerji 74.40 firmware file to flash the other inverter. Or a new firmware update file to flash your existing firmware. I don't think it would be safe to flash an Axpert 74.40 image into the Synerji. They do seem to have functional differences.

It's possible that you could parallel the two inverters with slightly different firmware (say 74.40 and 74.50), but your supplier should be able to tell you if it will work, and supply a firmware update file if it won't. It would be pretty poor if you couldn't parallel two machines you bought a few months apart. But I have no idea how any of that works with Synerjis. I agree it's far from ideal.

Link to post
Share on other sites
6 hours ago, Coulomb said:

 It would be pretty poor if you couldn't parallel two machines you bought a few months apart. But I have no idea how any of that works with Synerjis. I agree it's far from ideal.

Thank you for your considerations. I'll wait for the information from the Supplier.

Link to post
Share on other sites
  • 2 weeks later...
On 2020/11/12 at 1:58 PM, Beat said:

 But now they informed me that the new units on stock run a newer firmware without specifying it.

Surprise, surprise!

Today I got the new Synerji MKS 5K delivered - only 2 days after ordering it. I must praise that company Herholdts for its rapidity.

My inspection results: According to the seal It left the factory on April 2019 one month before my first unit.

2. It runs firmware 74.40. There should be no compatibility problem.

3. It appears that these units are factory equipped with parallel board and cables. I suspected this as my first unit was delivered with parallel board. But I didn't want to take the risk and ordered the parallel kit in addition. It was delivered along with the unit. Now I have a spare parallel kit.

It looks like these people do not really know what they have on stock. But on the other hand I must praise them. After I got informed of compatibility I ordered immediately. The consultant handed it over to the sales department. They sent me a new quote. Surprise: It was some 10% lover than the former and did not mention cost of delivery. When asked about they confirmed that delivery is included. With this I hurried up to order and pay by EFT before they eventually might change their mind. That was on Monday - today I got delivery.

It remains to me to install and commission the unit.

Label-2.thumb.JPG.45fa853a6839ed5c8028cdab8248e461.JPG

Link to post
Share on other sites

That label looks like a genuine Voltronic Power model. Either Synerji is getting very good at copying, or these are actually made in the Voltronic Power factory.

I wonder if they got tired of all the niggling problems they had and decided it was easier and more profitable to go legal. That can happen 🙂

Edit: or maybe the Voltronic Power legal eagles too them to court, and now they are forced to be a legal reseller.

Edited by Coulomb
Link to post
Share on other sites
7 hours ago, Coulomb said:

That label looks like a genuine Voltronic Power model. Either Synerji is getting very good at copying, or these are actually made in the Voltronic Power factory.

I wonder if they got tired of all the niggling problems they had and decided it was easier and more profitable to go legal. That can happen 🙂

Edit: or maybe the Voltronic Power legal eagles too them to court, and now they are forced to be a legal reseller.

As a matter of fact the consultant at Herholdts was getting (wrong) advise from Voltronic. It looks like they are buying from Voltronic. I think the wrong advise referred to the Voltronic stock whereas the Herholdts stock still contains some older units.

By the way - where is Axpert in that picture? Their internet site says that HQ is in Taiwan but manufacturing in China.

Link to post
Share on other sites
14 hours ago, Beat said:

By the way - where is Axpert in that picture?

The word "Axpert" often doesn't appear on the label; Axpert is what Voltronic themselves call their off-grid products. I judged it as genuine based on the criteria in my Do I Own a Clone post.

14 hours ago, Beat said:

Their internet site says that HQ is in Taiwan but manufacturing in China.

Yes, R&D in Taiwan, and manufacturing in mainland China.

Link to post
Share on other sites
  • 2 weeks later...

Interesting experience:

The two MKS 5K work well together in parallel mode. I programmed the second unit exactly the same as the first one. When commissioning the choice of master and slave is done by the units themselves arbitrarily. In my case the new unit has become the master.

All my solar panels are still connected to the older unit that has become the slave. The master unit works without panels solely from battery. Now we head two miserable weather days such that the batteries got almost emptied. Switching a heavy load (Geyser) on provoked the inverters to  switch to utility bypass. So far well synchronized. However since the master unit has no panels connected it believes to be at night and charges the battery from utility whereas the slave charges them from the panels (program 16 CSO). I wonder if I could program the master without panels to 16 OSO to prevent him to charge from utility?

Link to post
Share on other sites
19 minutes ago, Beat said:

I wonder if I could program the master without panels to 16 OSO to prevent him to charge from utility?

Certainly. Charger source priority is one of the settings that is particular to an inverter; that setting won't be sent to the others.

For cases like this.

Link to post
Share on other sites
On 2020/11/26 at 5:14 AM, Coulomb said:

That label looks like a genuine Voltronic Power model. Either Synerji is getting very good at copying, or these are actually made in the Voltronic Power factory.

I think the latter is the case.

Normally a company puts in the product label of its own brand product its brand name first. The way its written in this label: "Model Name: Synerji 5kVA/5kW 80A" below the title "INVERTER CHARGER" tells me that apparently Voltronic is manufacturing and selling this product for other brand names such as Axpert and Synerji, perhaps even others.

Link to post
Share on other sites
  • 3 months later...
On 2020/12/07 at 11:34 AM, Beat said:

Interesting experience:

The two MKS 5K work well together in parallel mode. I programmed the second unit exactly the same as the first one. When commissioning the choice of master and slave is done by the units themselves arbitrarily. In my case the new unit has become the master.

All my solar panels are still connected to the older unit that has become the slave. The master unit works without panels solely from battery.

In this configuration they work well. Lately I had a complete shutdown because of load shedding as reported in another thread. When restarted this time the older unit with the PVs connected became the master. In this configuration it repeated the bad behavior when batteries are full in the afternoon. It disconnects the PVs and reconnects them only the next morning unless I disconnect and reconnect them manually. I thought that this has to do with the master/slave configuration. So I shut them down again to eventually restart them in the original configuration. On the following sunny day I observed that they again behave correctly, that is when batteries are full the MPPT reduces power delivery to the actual load on the system, maintaining floating charge voltage. The battery current goes to zero. As soon as the load increases the MPPT delivers it and vice versa.

It appears that the one and the same unit behaves differently when in master configuration than in slave configuration. Bizarre! Perhaps@Coulombhas an explanation for that? I wonder how the master unit would behave when I connect the planned additional PV panels to it.

The relevant settings are:

01: SBU, 26: 53V, 27: 52.5V, 29: 42V, 31: SbE. The batteries are 300Ah 48V LiFePh.

Firmware version 74.40.

Edited by Beat
Link to post
Share on other sites
5 hours ago, Beat said:

It appears that the one and the same unit behaves differently when in master configuration than in slave configuration. Bizarre! Perhaps@Coulombhas an explanation for that?

Well, when a slave, it charges according to instructions from the master. The two have presumably identical firmware, so the difference has to be settings, or the fact that one has PV and the other not.

Two possibilities come to mind. One, does the one with PV connected have the solar balaning setting disabled? I suspect that setting doesn't get auto transferred.

Two: what is the voltage of your panels about an hour before noon? Maybe the voltage is too high, and the warning doesn't get transferred to the master. As I type this, it doesn't seem likely, as power limiting happens in the solar charge controller.

Okay, a third possibility. I assume that the master uses its own battery voltage reading. Do the two inverters agree about the battery voltage, with only the occasional 0.1V difference? If not, battery calibration might fix it. You'll probably have to check the battery voltage on the inverters' LC Displays.

Link to post
Share on other sites
7 hours ago, Coulomb said:

Two possibilities come to mind. One, does the one with PV connected have the solar balaning setting disabled? I suspect that setting doesn't get auto transferred.

Thank you for your response.

I stated: "01: SBU, 26: 53V, 27: 52.5V, 29: 42V, 31: SbE." Setting 31 is the solar balance. It is set to ENABLE (default).

[Quote=Coulomb]"Two: what is the voltage of your panels about an hour before noon? Maybe the voltage is too high, and the warning doesn't get transferred to the master. As I type this, it doesn't seem likely, as power limiting happens in the solar charge controller."[/Quote]

The voltage of my panels remain between 100 and 110V at full power, confirmed with my multi meter. It raises to 129V at full sun with no load. It is as a matter of fact interesting. They have been sold to me as 335W Canadian panels. I connected them 3S3P. Puzzling is the fact that they are 2x72 cells but behave like 2x60 cells. I got sometimes a maximum of 3.4kW out, if I dare to believe the display. That would yield 377W per panel.

[Quote=Coulomb]"Okay, a third possibility. I assume that the master uses its own battery voltage reading. Do the two inverters agree about the battery voltage, with only the occasional 0.1V difference? If not, battery calibration might fix it. You'll probably have to check the battery voltage on the inverters' LC Displays."[/Quote]

I just got the readings from the LCDs: Master 50.3V, Slave 50.6V. Seems to me within measuring inaccuracy.

I remember now: My first unit installed as stand alone came factory configured as master [HS] and with parallel board installed. I didn't care about it. But it had the same bad behavior until I reconfigured it as stand alone [  ] according to your recommendation to another poster. I attributed it then to settings changes I made at the same time. I now understand that that the behavior change is due to the configuration away from master.

Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...