Hello,
I have 2 inverters expert king 5kw link in parallel and one find battery voltage more low the the other (same battery). Every time there are 0.5 different for example 47.8 instead of 48.3. But sometime there are 46.5v instead of 49v. In this case it's problem because inverter swith in bypass whereas battery at 75% power left. When i take voltage with multitools it's exactly the same that the highest.
anyone have an idea what can do this .?
Thans a lot for any idea .
Regards

On 2022/02/01 at 6:21 PM, revanfreeman said:

Every time there are 0.5 different for example 47.8 instead of 48.3.

For a fixed difference like this, you can calibrate the one that disagrees with a trusted multimeter; it looks like in your case, it's the one that reads low.

On 2022/02/01 at 6:21 PM, revanfreeman said:

But sometime there are 46.5v instead of 49v.

Sometimes it's hard to compare because loads are coming on and off all the time. But if you routinely find that one inverter's battery voltage reading is intermittent, then that suggests a hardware fault. If you're hardware inclined, it's likely to be the string of resistors that forms part of the voltage divider, and it's fairly easy to spot them.

I suggest trying the calibration first. This post describes how to calibrate a model that doesn't have a removable display. However, you have a removable display, so you'll need to remove it and use the connection at the inverter directly. NOTE: the pinouts here (the cable connecting the removable display and the main inverter) are NOT the same as the RS-232 port. So you'll need a special adapter. I've made such an adapter, but I can't recall if or where I've documented it. It's probably in email, but I lost all my email history recently. The adapter itself is at a friend's place, I think. If I can find my notes, I'll document it.

Edit: Here is the pinout for the required adapter.

Not terribly convenient, I know.

Edited February 2, 20223 yr by Coulomb

Thanks a lot for your answer, it's very good lead to try to resolve it  !!
I think in my case, there is a problem with the hardware. This is the decreasing of 2 or 3 volts from time to time, not always, that is annoying. 
There are not load during test, and the other inverter is very stable. The decrease lasts until the total restart (shutdowns battery) or the complete charge battery.
 

I will try to contact the seller, and depending on the answer, I will try to find the bug.

Your explanation about calibration is useful. I think i will try to fix the difference with that method too 
Regards

15 hours ago, revanfreeman said:

I think i will try to fix the difference with that method too

Cool. Obviously, fix the intermittency first.

Hello,
Finaly the inverter it still haven't read 2 voltage too low for now (after totaly restart, shundown batterie) 
so, i tried to fix 0.5 too low voltage. I have buy usb serial cable and i have follow your link to do custom rj45 cable to make adapter.
it's working and i have fix my inverter. But with some problem, indeed i needed to reverse TD and RD pin. 
Maybe i have bad anderstanding because serial cable supose to naturaly reverse pin ou just my inverter is different (5048mk pip solar)

it's like this for me, to considerate serial d9 to out computer and not out serial cable

+----------+--------+--------+------------+
| Function | RJ-45  | Serial | Function   |
| (display)| Pin #  |  D9    | (computer) |
+==========+========+========+============+
|  SGND    | 1 ---- |   5    |    SGND    |
|   TD     | 6 ---→ |   2    |     RD     |
|   RD     | 5 <--- |   3    |     TD     |
|  +12V    | 2      |  N.C.  |            |
|  -12V    | 3      |  N.C.  |            |
|          |        |        |            |
| Inv Sw   | 7,8    |  N.C.  |            |
+----------+--------+--------+------------+

Maybe help other people
Thanks a lot for help
Regard

3 hours ago, revanfreeman said:

indeed i needed to reverse TD and RD pin. 

I probably screwed up the diagram; sorry. I've added a note to that effect and a link to your post. 

3 hours ago, revanfreeman said:

it's working and i have fix my inverter.

I'm glad you got it working.

Edited February 17, 20223 yr by Coulomb

I do have also two axpert kings in parallel. Depending on what batteries you have the master inverter, with pylontech batteries, adjustment number 5 should be on PYL. And the slave inverter should be on user, not on PYL.

 

Hi Guys. Could I revive this thread? I have a similar problem.

I have 2 x RM Axpert Kings in // showing different VDC and VAC readings on the LCDs. I bought them about 1yr apart and had to get the FW updated so they were the same version number to work in //.

When a check the VDC on the inverter terminals, both read the same ( within 0.1) vdc, but inverter 1 (HL) shows the correct (higher) voltage 48.9vdc while inverter 2 (SL) shows incorrect (lower) voltage 47.8vdc. The same happens on the VAC (grid) side, the multimeter gives 248vac on both sets of input terminals, while inverter 1 shows 247vac and inverter 2 shows 242vac.

I note this ratio 47.8/48.9 = 0.98 and 242/247 = 0.98

Any ideas how I could correct this problem?

Edited March 12, 20232 yr by IanThomson
more info added

22 hours ago, IanThomson said:

Any ideas how I could correct this problem?

That's weird that they went out of calibration after being paralleled. I'll have to think about whether I can conjure up a reason for that to happen. Unless perhaps you're near salt water or a particularly wet climate? In that case, moisture may have affected the measurement resistors.

As for a fix, I don't know of any practical way of adjusting the AC-in voltage reading, other than adjusting the resistance of the measurement strings.

The battery voltage calibration can be adjusted by software, but the easy to use commands don't get through the display firmware. Weber and I are working on a fix for that right now. If you're in a hurry, you can bypass the display with a special cable, as explained in this post. Then perform the battery calibration as per this post.

But by far the easiest and safest way would be to wait for our fully patched firmware.

Hi Coulomb. Thanks so for making contact. Maybe I did not explain fully / clearly.

I bought inverter 1 (SOL-I-AX-5.RM King 5K RM - 48V 4000W MPPT) about a year ago. I then wanted to add a second inverter in //, but needed to get the FW updated on the old so they would work in //. Once installed, I then noted that the voltage readings on inverter 2 (new) were low compared to inverter 1 (old). The VDC is my main concern, as it triggers a panic as the inverter sees the battery pack voltage going too low, I have program 05 set to usr. The inverter BMS did not want to talk to my Dyness 4.8 bats in PLY mode.

I did try fixing the VDC reading using "Calibrating your inverter's battery voltage reading" but got 'nacks' using Weber's nudge method and Tera Term. I assumed when nudging the reading, I should send the command as ASCII and not HEX (i tried both). I have not tried using your HEX route.

Thanks for the help.

 

6 hours ago, IanThomson said:

I did try fixing the VDC reading using "Calibrating your inverter's battery voltage reading" but got 'nacks' using Weber's nudge method and Tera Term.

Yes, that's because the display doesn't forward the useful commands (ATE without the trailing 0, 1, or 2, or PSAVE) to the DSP. It does forward the ATE0/1/2 series of commands, so that's something. It's just way less convenient.

On 2023/03/14 at 2:14 PM, Coulomb said:

Yes, that's because the display doesn't forward the useful commands (ATE without the trailing 0, 1, or 2, or PSAVE) to the DSP. It does forward the ATE0/1/2 series of commands, so that's something. It's just way less convenient

Thanks for the reply. You have lost me a bit there, here I take very small steps.

Are you suggesting I should use your HEX route, as per your answer to the original post?

16 minutes ago, IanThomson said:

Are you suggesting I should use your HEX route, as per your answer to the original post?

The issue isn't hex versus ascii.

You send commands to the King's display. The display has its own computer chip. The display is supposed to forward the commands to the main DSP (computer chip in the inverter itself), fitting it in with various other commands it is sending continuously. When the DSP sends back a reply to the display, the display should forward the response to the PC. If the display doesn't see a response in a certain time, it send back a NAK (Negative Acknowledge, basically a thumbs down).

But not every command gets passed on from the display to the DSP. For technical reasons, only some are passed through. Voltronic decided not to pass on the BTA or the PSAVE commands. In our fully patched King firmware, we will be passing those commands on. However, our last attempt crashed, we'll have another go soon.

There is the option of bypassing the display and talking directly to the DSP, but that requires a special cable. All the details are in an AEVA post that I linked to four posts ago.

Thanks for the explanation, I am now back on the same path. I am not sure the display is removable. I hope I have not waisted your time.

The RM inverters I am using are these "SOL-I-AX-5.RM"   https://www.mecerpc.co.za/Mecer-SOL-I-AX-5-RM-p-249017.php 

8 hours ago, IanThomson said:

The RM inverters I am using are these "SOL-I-AX-5.RM"

Ah, the Rack Mounted (RM) versions. I haven't studied these in any detail. My working guess has been that there is a display board in there, but it uses the old MKS displays due to a lack of front panel space. This is supported by the existence of display firmware version 15.05, which seems to be specifically for the rack mounted Kings.

So yes, I don't know if you can bypass the display., sorry. The Rack Mounted Kings also use different firmware, which we won't be fully patching, due to our inability to test it. Right now, our display firmware is crashing, and we don't yet know why. Hopefully we get to it tomorrow, though there are other things to temp us away. But it highlights the difficulty of performing complex patches on models that we don't have access to.

The ATE0/1/2 commands (Weber recently uploaded the official Chinglish instructions) looks like your only option, sorry. It's attached at the end of the AEVA calibration post.

Edit: When they say "set the voltage to be 56 V" they mean adjust the voltage of the bench power supply that you are using to supply power to the battery port of this machine. If you don't have such a power supply or it's not convenient, you can still calibrate the battery (perhaps not as precisely) by waiting for the battery to become full, wait for the loads to become very light, measure the battery voltage with a trusted multimeter, and use that value instead of "056.00" in the ATE1 command. Then wait until the battery is as discharged as you dare, again wait for the loads to be very light, again measure the battery  voltage, and use that value instead of the "044.00" they instruct. Of course, you'll need to calculate a new CRC, I use this site and read the two bytes from the CCITT (Xmodem) result. Remember to increment one or both CRC bytes by 1 if they happen to result in 0D, 0A, or 28 (carriage return, line feed, or open parenthesis).

Edited March 16, 20232 yr by Coulomb

Thanks for all the good work Coulomb, enjoy the coffee, when you get a break.