On 2022/06/26 at 9:31 PM, Stormpie said:

the bus-voltage climbs to 640v

Yikes. The IGBTs are likely only rated at 600 V. Is the 640 V measured by a multimeter, or reported by a QPIGS (etc) command?

But since it works fine with only battery connected, I suspect some sort of leakage to earth with the Solar Charge Controller. Leaky panels can do this. Presumably, you had no panels connected, so that points to a leakage on the SCC (whether you use it or not). Perhaps try disconnecting the SCC board as a test.

Edited June 27, 20223 yr by Coulomb

Sorry for late reply i had a crazy week. I have it back in front of me. To clarify the bus voltage is measured with a multimeter on the bus+ bus- connections.

i removed the scc.
I just measured it again and it actually toppes out at 669V, the error08 and it sounds the alarm as soon as the bus passes 600v. 

So this issue(or at least my inverter) it not related to the scc.

if it start it on battries and it can run for hours. Assoon as i cannect ac and it goes to bypass, and the igbt turn off, the voltage shoots up in a matter of 5seconds .

 

And thanks for everyone’s replies so-far.

OK, with the SCC disconnected, I'd suspect capacitors to earth (as in chassis earth). These are for EMI suppression. If these are leaky, it could cause the same problem as leaky panels.

But perhaps the inverter (as in the DC-AC converter) needs to be running for this to happen. If so, perhaps @Igubu is correct and you have to prevent or severely restrict utility charging with appropriate settings.

 

There might a few reasons -

these axpert inverters really

weren’t great, but if you read the 4Y + history of error 8 and ALL the rebrands using voltronic powers’ pcb, they all had the same issue, utility charging has a design flaw and the bus V shoots up to 600+. There is a long forum post history about it, many things were blamed- seemed to happen when it was raining, so panels and joins were mentioned, but coincidentally was because of rain batteries ran low and utility charging were used. 
 

Some clever engineers explained why, how to physically modify the pcb, or add a “push button bypass” to reset the issue manually- but it all came back to the same old pcb design flaw. When I say all- any of the axpert,rct,mustek,growatt etc rebrands - they all used the same mainboard with that 145V mppt limit… maybe newer models have fixed it. 
 

Only workaround that was confirmed by many in the field to help , was limit grid charge to as low as possible- or disable if it’s an option. The logic of grid charging at 2A when raining is flawed for sure- you’ll never get the batteries charged during these multiple loadshed days, but that’s what you were stuck with.

I remember a year or so ago somebody sat with the same issue- the best advice- sell the inverter for somebody that uses it for ups mode only, and buy a decent one- small sunsynk 5kW hybrid comes to mind…even the newer versions of this- but if you consider a kodak 5kW for R14K- save up and buy a hybrid sunsynk for 20k- worth it for many many reason,  but not tue correct page to explain the differences and advantages of proper bi-directional hybrids vs “off grid” ups style cheapies like axpert/kodak.

If you are desperate to fix this- and cannot disable/limit grid charge, scroll through the history of error 8 and, somewhere the post with engineering changes required by somebody in Saudi I think was posted- but this involves opening up, and soldering- some surface mounts, so obviously for somebody that’s comfortable doing that kind of thing..

22 hours ago, Stormpie said:

Sorry for late reply i had a crazy week. I have it back in front of me. To clarify the bus voltage is measured with a multimeter on the bus+ bus- connections.

i removed the scc.
I just measured it again and it actually toppes out at 669V, the error08 and it sounds the alarm as soon as the bus passes 600v. 

So this issue(or at least my inverter) it not related to the scc.

if it start it on battries and it can run for hours. Assoon as i cannect ac and it goes to bypass, and the igbt turn off, the voltage shoots up in a matter of 5seconds .

 

If you scroll to page 9 - the “workaround” from HermanD was given:

”Hi Smirresa,

Yes, setting 11 set to 2A and setting 16 set to CSO.  In CSO the inverter will only charge from utility in bypass mode when there no solar.  If you are in bypass mode, but there is solar (cloudy day with batteries low), then the AC charge will not occur.  If you are running from batteries and no solar (at night time) then the AC charge will not occur.  AC charging will only work when you are in bypass mode and no solar.”

Also on the same page 2 posts of the pcb and some arabic text (use google translate” and pictures of a possible fix- but nobody tried and replied that it worked ever- so would try HermanD’s workaround first.

 

 

3 hours ago, Igubu said:

Also on the same page 2 posts of the pcb and some arabic text (use google translate) and pictures of a possible fix- but nobody tried and replied that it worked ever

That "fix" seems to me to be gagging the messenger, rather than fixing the problem. It would unbalance the voltage dividers feeding the op-amp that measures bus voltage, so that it is more sensitive to common-mode voltage. That may not be too serious (without thinking too hard about it), but the other effect will be to under-report the bus voltage slightly. The effect would be small: I estimate about 1 in 24 (4.2%), since there are 6 resistors of about 400 kΩ average, and you are adding 100 kΩ to that. It seems that 4.2% is way too small to prevent the error and allow >600 V onto the bus (perhaps only for a short time). I could well be misunderstanding how this is supposed to work. But brave and/or desperate owners may care to prove me wrong.

Hi just an update, i have set the charge to solar only and 2A AC-charge an for fun 10A total charge. While on batteries only doing this change the inverter went to Error8. So it can actually happen on battery only as-well🤯. Now it keeps going to error8 before almost everytime.

@Coulomb i agree with you. My 2cents on that fix is the following. If you have a slight overvoltage you can do it, but my inverter the voltage is higher than than the igbt and for that fact the caps will handle. So if i cheat it it is going to pop the main igbt’s

I am going to change the step-up transformer tomorrow even-though i don't believe it will make a difference.

17 hours ago, Stormpie said:

I am going to change the step-up transformer tomorrow even-though i don't believe it will make a difference.

Err, wow. That's hard core. The firmware won't be expecting that. I predict massive problems. But if you're fed up with fault code 08 (i.e. the machine is unusable anyway), then perhaps we can learn something from this.

On 2020/08/02 at 7:19 PM, APV said:

Hallo @charlez

Good job!  Have you watched it working around 18h15 and 07h15 (depending on the sun)?  I especially like the morning startup where the bulb first starts glowing and then the relay kicks in, the bulb goes off and the inverter switches to solar...

Your wires exiting the board on the left top side, look a bit thin to me.  Remember that the solar panel current flows through all these wires.  For 20A, you should use at least 2.5mm².  Maybe your wires are that thick and it is just the photo making them look rather thin?  Of course the current will only be high when your panels are delivering maximum power, in other words in bright sunlight and the inverter delivering maximum power...

Hi @APV please check if the below components are fine to use; I am unsure about the AC or DC ratings, the varistor and the zener diode?:

On 2022/11/29 at 3:08 PM, fnaude57 said:

Hi @APV please check if the below components are fine to use; I am unsure about the AC or DC ratings, the varistor and the zener diode?:

Please can anyone confirm if above components can work in die circuit board APV designed?

3 hours ago, fnaude57 said:

Can anyone confirm if above components can work in die circuit board APV designed?

Most seem fine, but I'd change the varistor to something a fair but lower in voltage, e.g. V385LA40BP for max continuous DC rating of 505 V (from memory). Panels are DC.

The zener will probably work, but I'd go for a slightly higher wattage, e.g. 2 W, perhaps up to 5 W. Those relay coils can pack a whallop.

thx

Hi all. 

Being Cash strapped , i have a Kodac OGX 5.48 , the one with 450VDC solar controller 

After reading all the posts regarding these inverters acting up if you put solar on them , I have an idea. This is usually where things get dodgy.

 

If i were to use a Crydom D4D12 DC SSR, ( see note : a* )rain sensor , and day/night switch , to only connect my 4 x 450w string under the following conditions:

1 - sun is up

2 - Eskom grid failure

3 - it isnt raining 

I am only interested in saving a cycle on my batteries during daytime loadshedding , bare essentials ( + - 750W) , and not energy saving , as per se.

My only other plan is to use a standalone solar controller , directly on the 4 x 200ah Mecer battery bank. 

However - i am not sure the inverter's charging will play nice with the standalone also trying to charge the bank during grid ON and sun up conditions.

If this post is stupid enough to give other readers the same crazy ideas , i will delete it. 

Thoughts ?

Note a : It seems like breaking only one leg isnt going to solve anything , might have to use 2*

Edited December 20, 20223 yr by Slowtoker
details..

I had the same issue with my Mecer SOL-I-AX-5M4. Bought a couple of weeks ago from Mustek in SA. Read this thread and tried setting charge priority to Solar First. Started with AC Max charge amps as 2A as well and problem seemed to be solved. Changed charge current to 10a and still problem seems to be solved. So I will be increasing my charge current even more to see what happens.

As a side note, I have 1 x JA Solar 545w panel which is not enough for this inverter as min solar volts are 120v. (adding more panels soon) So my solar switch was disconnected (both wires I believe) and I still had this problem every time loadshedding ended.  My battery would charge for about 15-30min before error F08. Depended how far the battery was depleted I guess. I got the idea that as soon as the inverter changed from bulk charge to float, the error F08 showed up. It gave a long beep and restarted itself. After that the charge current would drop from 40A to 20A and all went well to charge battery to 100%, but then this will repeat itself after every load shedding.

I am using a Sunsynk 5.32 battery and a coms cable I made myself.

Thanks for this thread and thanks for this solution! I guess this is something the manufacturer needs fix in a firmware update maybe.

Edited May 2, 20233 yr by PeterPan

Glad you have a workable “workaround”- at least 10Amps gives you 500W approx- so should be able to fill up the sunsynk batt in 8 hours or so..  if you can get 20Amps min your golden!

 

yes it’s incredible that the same issue exists years later.

 

out of interest: the diy comms cable..? Does the inverter not have can bus interface and are you using rs485 or usb? I know the sunsynk batt CAN cable is just a normal flylead- unsure about the inverter, not up to date with them.

 

* found a link! Think rs485 might work as well with sunsynk batt - unsure. You’d need at least 3x 545 panels to work nicely

https://www.mecerpc.co.za/pdf_downloads/20/206710_PDF1.pdf

Edited May 2, 20233 yr by Igubu

Hello All,

I have solar assistant monitoring my system(solar/battery/axpert 11kw) with the mqtt server enabled and I have a linux server that runs a python script every couple of minutes to query and modify the settings based on the time and/or battery soc. The system has been running fine since I installed but I made a change to the script last night to set the solar charge max to 10A at the end of the cycle which triggered F08 which I verified from the log outputs of the mqtt script.  I've commented the solar charge max changes from the code so lets see if that makes a difference tonight.

Edited August 8, 20232 yr by ddmbr