Hi all,

First post for me

Last week I got a team to come in and install a solar system for me. I went with 16 330W panels in an E-W configuration and a Goodwe 5048ES inverter. The one odd thing I did was I got hold of two LG Chem Telecoms Spec batteries (I work in the industry). These batteries are essentially the same as the LG Chem RESU 6.5kWh units but I got them for much much less. The problem with these batteries is that they are not designed to communicate with residential inverters. I am assured that this isn't a problem. They told me to just treat them essentially like a lead-acid unit, charge them to 57.2 (92% SOC) and discharge to around 51V (40% SOC) for on-grid cycling and when there is an outage run it down to around 47.6V (10% SOC), the internal BMS will manage the rest and I can connect a computer to it if necessary. Seemed easy enough and, by all accounts, I have set the system up like this.

After the system was installed all seemed well initially but the Goodwe unit only works on SOC for setting the cycling points. That shouldn't be a problem, in theory, I don't need it to be exact. The problem I am having is the inverters SOC calculation is not just off by a bit, it is completely and unusably wrong. I have attached some charts that I pulled from the system.

Looking at them you will see that the SOC is almost working backward. When the battery is discharging and the voltage dropping the SOC remains relatively constant, then when the sun rises the battery starts to charge, the voltage increases and inexplicably the SOC starts decreasing!!! Then around 11 am a large load kicks in and the battery starts discharging and the SOC jumps back up to 96%.

I am completely stumped. I thought this was a basic calculation based on Ah in and Ah out but this seems to be behaving very bizarrely and, unfortunately, the SOC figure is core to the functionality of the system. It has been through 2 charge and discharge cycles with not much sign of improvement.

The installer, despite knowing my requirements ahead of time, is claiming that he is out of his depth and cannot assist. I have contacted Goodwe who put me in touch with a local distributor who has been helpful but not really taken the time to actually understand the issue. They just told me to change the systems battery capacity to represent usable capacity (which I had done already).

Is there anyone out there that is using an inverter like this without battery coms or with a lead acid battery that could assist?

I have yet to try the following:

• running the system on more cycles and seeing if it corrects over time
• setting the system to "default" which doesn't let you set the capacity and hoping for the best
• setting the system to the LG Chem RESU 13 and hoping it works without coms
• riping the unit out and getting a much cheaper Axpert inverter which seems to perform functionality no problem 

Thank you in advance for any advice that you may have!

Daranger.

 

 

 

Maybe the inverter has some "reset" code built in. It expects to be running lead-acid chemistry, so when the voltage doesn't drop as much as would normally be expected, maybe the inverter has a "psych, back to 90%" moment.

One way would be to simply ignore the SOC from the inverter 🙂

22 hours ago, Daranger said:

setting the system to the LG Chem RESU 13 and hoping it works without coms 

I would definitely try this, or any other way to tell the inverter to use a Lithium charge curve.

Gave this a try, unfortunately still not working correctly. Completely baffled. 

My first post as well... 😉

I also have a Goodwe 5048D-ES with 8 x Allgrand 12V 200Ah GEL VRLA Batteries (GEL 6-GFM-200/12V200Ah) in a 4S2P configuration (48Vdc, 400Ah) connected. I have bought the batteries a couple of months before the invertor at a good price, thus did not really  then have an option to buy something else with the Goodwe later on. I have used the "Self Define" option to set the Ah, cycle and float voltages etc and is sure it is correct (On-Grid DOD set to 10% (@SOC 90%)). 

Same problem you have with the SOC measurement. It has been running in the backup mode the last couple of months and I only recently installed the first string of solar panels and then switched to the "General Mode". The battery SOC in backup mode never go beyond 96% either even when fully charged. The moment you start drawing high current, the SOC hits 100%. When it starts charging, the SOC falls to 80% or so. The SOC graph is discontinuous and bears no correlation with the effective charge being moved in either direction. I get the impression that it correlates more to voltage than current, but in essence, the SOC measurement is a real stuff-up.

Although set to only discharge down to 90% SOC in general mode when on-grid, it nearly drained my batteries close to 100% DOD (about 45Vdc) when I luckily got home and switched it back to the back-up mode. I still have to discuss it with the supplier but I believe it is a bug or just plain crappy software/algorithm running on it. It measures the battery current and voltage accurate enough but do not base the SOC on the battery current integrated over time, which is quite easy to do and not rocket science. One would think that after so many years in the market that basics like this would have been sorted…

Edited May 23, 20197 yr by Heinz Hirsch

Where are the Goodwe people to answer the questions?

enquire about a firmware update. I had a similar issue on the early version, different firmware, same prob, the update did help....

Mike, I had WiFi connectivity problems when I initially installed the system. Has upgraded on 7 March 2019 to the latest versions for WiFi, Master and Slave firmware at the time, which fixed the Wifi problem. PV Master report Firmware version as 15154, which might still be the latest, what version do you have? Maybe we should down grade to an earlier version...

Daranger, as a temporary workaround, I've set the inverter to run in "Economic Mode" see settings below. My night load is fairly stable, thus from 23:00 to 07:00, the battery runs the house load which is approximately 20% of the battery capacity. From 07:00, the solar panel charge the battery and 0% grid power allowed for charging. Excess solar power above charging helps powering the load during the day as well. Not entirely what I want but better than getting the batteries drained beyond reasonable limits. Battery power is not used until 23:00 again.

Edited May 24, 20197 yr by Heinz Hirsch

On 2019/05/23 at 9:21 PM, The Terrible Triplett said:

Where are the Goodwe people to answer the questions?

I have the same inverter but never had battery charge/discharge problems. Did you discuss this with Goodwe?

Fuenkli, I have emailed the supplier yesterday, will see what he comes back with. What batteries do you use, Lead Acid type without a BMS?

23 hours ago, Heinz Hirsch said:

What batteries do you use,

Lead Acid without BMS. I set the parameters using the battery "self define" option. The battery stats reported by the inverter perfectly match the Victron bmv values. 

I would also contact Goodwe China directely

Fuenkli, do you have a new or older model Goodwe? What firmware version are you running on it?

I have newer 5048D-ES model, serial number starts with 9.

1 hour ago, Heinz Hirsch said:

Fuenkli, do you have a new or older model Goodwe? What firmware version are you running on it?

mine is an old (2017) model. Serial number starts with a 3.  I had problems with the mppt's (noisy and one mppt always produced less power). Goodwe China send me a new control board with firmware 1212. Now it is running perfectly. Would you be prepared to give me access to your station on the SEMS portal? I might be able to spot the problem. 

Heinz, I'm sorry that you are experiencing the same issue but at the same time pleased to hear that it is not just me . I like your workaround but obviously not a long term solution. I have the same 5048 serial starting 9 and running latest firmware (15158).

I have spent some time trying to understand what it is doing. Essentially it seems to have no correlation to the voltage at all, the SOC only responds to instantaneous current, also no correlation with current over time. When there is a charge current the SOC drops, the larger the current the larger the drop. When there is a discharge the SOC stays at 92 or 96%. I have tested this on nearly every profile available and the results are identical.

I have been working with Goodwe's local support team as well as my installer. Goodwe is saying that they think the new firmware (1515) is the problem and they are recommending that I downgrade it to the 1414 version. Heinz, here is a link to a document that will help you and direct you to the correct firmware: https://portal.wholegreen.co.za/reseller/docs/Full Firmware Update Guide.pdf

Unfortunately, my laptop has only usb-c ports and I cannot get these to work with the inverter despite having the correct cables. Goodwe is suggesting that I try get hold of an older windows 7 laptop and use that 😕.

Edited May 29, 20197 yr by Daranger

1 hour ago, Fuenkli said:

Would you be prepared to give me access to your station on the SEMS portal?

Thanks Fuenkli, will send you a pm with the SEMS portal info.

 

1 hour ago, Daranger said:

Goodwe is saying that they think the new firmware (1515) is the problem and they are recommending that I downgrade it to the 1414 version. Heinz, here is a link to a document that will help you and direct you to the correct firmware: https://portal.wholegreen.co.za/reseller/docs/Full Firmware Update Guide.pdf

Daranger thanks for the feedback, will definitely try the earlier version. As soon as I managed to install it, will let you know if there was any improvement.

On 2019/05/30 at 12:04 AM, Daranger said:

here is a link to a document that will help you and direct you to the correct firmware: https://portal.wholegreen.co.za/reseller/docs/Full Firmware Update Guide.pdf

Those files seem to be saved incorrectly to dropbox. [ Edit: at least for my computer. Seems it's fine for others; see next two posts. ] What I had to do to get sensible file contents is this:

• At the (e.g.) .out or .exe file, right click and select Copy link address (Google Chrome; other browsers will have a similar option)
• Open a new tab (ctrl-T for Chrome) and paste in the address to your browser's address bar.
• At the very end of the just-pasted URL, you should find "dl=0". Change the trailing '0' to a trailing '1' and press enter. You should have a file save dialog in a second or so. Save to a suitable folder.

How this could ever have worked is beyond me, so maybe it's something crazy about my computer. I found the latest master.out file to be 570 KB (in Windows explorer). The Ezflash_D.exe should come to 1,834 KB. If you can click on it and eventually get it to run (Windows warned me about executing .exe files downloaded from the internet), then your download technique is good. If you end up with a .html file, you have not successfully downloaded it.

Edit: It's confusing about the ARM vs non-ARM processors. All the *master.out and *slave.out files seem to be C2000 (Texas Instruments DSP) executable files, but the .bin files in the DataSend folders seem to be ARM binaries. The documentation says to tick the ARM box for EZflash, which makes no sense. It says to update three processors: master, slave, and the ARM processor that the DataSend tool accesses. Would there really be three processors to update?

Edited June 3, 20197 yr by Coulomb
Added "to your browser's address bar"

Strange, I could get the files no problem.

I had the installer come over yesterday and downgrade the system to 1414. Sadly I am seeing no change, the SOC is behaving exactly the same and is unusable. This is becoming really frustrating and I am worried that this lack of control is going to damage the batteries. This is such basic functionality and I am getting no joy from Goodwe.

Is there another inverter that someone would recommend rather than this one? I picked the Goodwe one because I need the IP65 functionality.

Sadly to report, but experienced the same...

First upgraded to the latest 151508 (had 151504, got the 08 patch from my supplier) with no difference to SOC behavior. Then downgraded the firmware to 141406 and SOC behavior still the same. Did the "kettle/geyser test" (high load) for both firmware versions and SOC value jumps in the wrong direction to 100% or close. In my book it would classify as a design defect (software wise) and should be recalled... Think the hardware is pretty solid (one massive heat-sink, no fans etc.) but who ever wrote the firmware should be fired.

Daranger I agree, if an inverter cant work with the most basic and oldest of battery technologies, it should not be on the market, IMHO.

Coulomb, I also could get the files without any problems, have used Firefox and Chrome.

I have a goodwe on site with 4 x 260 ah agm batteries. It is only used as backup on setting 3, (5048 em) . after being back to the supplier that i will never sopport again is it took 6 weeks for a responce i got it back still having a delay on changeover. Goodwe in the end send me a list of around 10 batteries only that is compatible with the unit. No AGM on the list and they will not look at the issue any further. its only just over a year old but i am going to scrap it at my cost and put a victron in for the client.

7 hours ago, Heinz Hirsch said:

Coulomb, I also could get the files without any problems, 

Well, I could get some contents, but it was all HTML which I'm sure won't work. Were you able to run the .exes and load the .out files?

 

 

1

13 hours ago, Heinz Hirsch said:

Sadly to report, but experienced the same...

... Think the hardware is pretty solid (one massive heat-sink, no fans etc.) but who ever wrote the firmware should be fired.

Sorry to hear that Heinz. I agree the hardware so far is great. Completely silent and so far very reliable. The frimware and user control on the other hand.

Can you PM me your email address so I can add you to my conversations with the local Goodwe team? It might help if there are two of proving that there is an issue. Mostly they keep trying to find fault with my batteries which is clearly not the problem here.

I would be happy with two solutions:
1) find a way to fix the SOC calculation or
2) insert user-controlled voltage cutoff values on the app for the on-grid and off-grid discharge values. This is even easier to do (even Axpert offers this).

17 hours ago, MartinCoetzee said:

and they will not look at the issue any further.

...well, that's not very nice... 🤔

16 hours ago, Coulomb said:

Were you able to run the .exes and load the .out files?

Yes, used EzFlash_D.exe for the .out files (ARM box must be ticked) and DataSend,exe to load the software patch afterwards. However, DataSend  would hang up during programming, thus after a restart of my wife's laptop (luckily it has a USB 2.0 port) and closing WiFi link and unnecessary software, it would run stable.

10 hours ago, Daranger said:

Can you PM me your email address...

Yes, no problem, will do.

Just an update for everyone here. We are moving at snails pace with Goodwe but I am still trying.

I have asked my installer if they would be able to swap the unit for the Solis hybrid inverter. Does anyone have experience with this unit? It has a nicely laid out user define section for the batteries which is what I need.

 

This is a bit of lead down to read this sort of stuff because i'm planning to by a Goodwe unit because of there back up but seems to me there is a bit of a struggle there to.  

2 hours ago, Gerlach said:

This is a bit of lead down to read this sort of stuff because i'm planning to by a Goodwe unit because of there back up but seems to me there is a bit of a struggle there to.  

Dont be too disheartened. If you use the product with a standard battery like Pylontech I am sure that you will have a better experience than we have.

My first post here.  Thank you for a great forum.

I had a system installed 2 days ago.  Goodwe GW5048D-ES, 2 x Pylontech 3.5kWh batteries and 6 panels totalling 2kW.

Everything seems to be working fine, but the installer was reluctant giving me control to change my settings myself - Only the SEMS Portal.

He made a mistake and set my battery to 30% DOD max (on grid), which makes the system not very efficient.

The company I bough from told me to use PV Master to change settings, but I struggle for hours to get it to connect to my system.

SEMS Portal connects no problem (WiFi connected to home WiFi router.  I found EZManage which also connects no problem, and I can get into the setup, but for some or other reason I cannot set the DOD % and it also dont have the Pylontech 3000, only Pylontech 2000B in the setup.

I somehow managed to get  the PV Master to connect last night and changed the percentages, but tonight it is the same story:  No connection.

Any ideas how I can get it to connect more reliably or something else I can use to play around with the settings?

Thank you in advance