Install and Operation of the SUNGOLD 10KW 48V SPLIT PHASE SOLAR INVERTER

[Kenny_]

Well after all that hoorah with Comm Cables yesterday I wake up too damn early today find Solar Assistant reporting the batteries still at 98%....I'm going what?? and find my inverter in fault/error state 58. I put in the modified Comm Cable and it goes away. WTF? It worked fine with the standard cable for almost a week and now it errors with it. I don't get it.

 

[Kenny_]

Not sure how advisable it is (or how secure) but is it a safe thing to do to post link to my last 24 hours charts from Solar Assistant?

Just noticed that it allows this.

 

[Lighthouse Beacon]

Well after all that hoorah with Comm Cables yesterday I wake up too damn early today find Solar Assistant reporting the batteries still at 98%....I'm going what?? and find my inverter in fault/error state 58. I put in the modified Comm Cable and it goes away. WTF? It worked fine with the standard cable for almost a week and now it errors with it. I don't get it.

Yeah you need to use the custom cable. I had an interesting thing happen yesterday.
I caught an error 59 and Powerwall reported an OVP error….

Had a lot of power coming in from panels and house wasn’t using much, batts were at 99.96% SOC and inverter started trying to dump the power into batts. This caused the batts to go into over current protection mode, when it did that it suspended the bms comms causing inverter to report a bms failure error 59.

My solution was to turn off solar, and let the batt discharge slowly back to correct SOC. I then set the inverter to not charge over SOC% of 98%, and it hasn’t happened again. I think 100% SOC is just to close to overcurrent BMS protection configuration if you’re getting heavy sun

 

[ChrisG]

Yeah you need to use the custom cable. I had an interesting thing happen yesterday.
I caught an error 59 and Powerwall reported an OVP error….

Had a lot of power coming in from panels and house wasn’t using much, batts were at 99.96% SOC and inverter started trying to dump the power into batts. This caused the batts to go into over current protection mode, when it did that it suspended the bms comms causing inverter to report a bms failure error 59.

My solution was to turn off solar, and let the batt discharge slowly back to correct SOC. I then set the inverter to not charge over SOC% of 98%, and it hasn’t happened again. I think 100% SOC is just to close to overcurrent BMS protection configuration if you’re getting heavy sun



View attachment 176224View attachment 176225

Odd the inverter didn’t throttle back PV with low loads and batts full.

 

[Kenny_]

Yeah you need to use the custom cable. I had an interesting thing happen yesterday.
I caught an error 59 and Powerwall reported an OVP error….

Had a lot of power coming in from panels and house wasn’t using much, batts were at 99.96% SOC and inverter started trying to dump the power into batts. This caused the batts to go into over current protection mode, when it did that it suspended the bms comms causing inverter to report a bms failure error 59.

My solution was to turn off solar, and let the batt discharge slowly back to correct SOC. I then set the inverter to not charge over SOC% of 98%, and it hasn’t happened again. I think 100% SOC is just to close to overcurrent BMS protection configuration if you’re getting heavy sun



View attachment 176224View attachment 176225

Sounds right. Not sure what my threshold is at but watching it ... it does slow to a crawl when it is over 90-95%

just checked ... Param 60... mine is set to 100%...default I presume.

 

[Kenny_]

This morning....just starting to get solar/pv in....dropped to about 55-60 % overnight from basically full charge on my two 5.1 Kw batteries

 

[Hedges]

Had a lot of power coming in from panels and house wasn’t using much, batts were at 99.96% SOC and inverter started trying to dump the power into batts. This caused the batts to go into over current protection mode, when it did that it suspended the bms comms causing inverter to report a bms failure error 59.

My solution was to turn off solar, and let the batt discharge slowly back to correct SOC. I then set the inverter to not charge over SOC% of 98%, and it hasn’t happened again. I think 100% SOC is just to close to overcurrent BMS protection configuration if you’re getting heavy sun

Do your settings have to be based on "SoC", and they get that from communications with BMS?

BMS protects based on cell voltage.
I'd think inverter should be programmed with a max voltage. Part way up the knee, there would be some headroom.

 

[Kenny_]

Do your settings have to be based on "SoC", and they get that from communications with BMS?

BMS protects based on cell voltage.
I'd think inverter should be programmed with a max voltage. Part way up the knee, there would be some headroom.

Seems right. Maybe an issue with the BMS in the battery?

 

[Hedges]

I'm not sure BMS default setting are well though out for interaction with dozens of different no-name inverters.
The AC coupled ones (typically quality big names) are going to have to shove a few seconds of current into battery.
All the rest, don't know how quickly their MPPT ramps down charging.

I think batteries ought to be balanced at higher voltage, charged a bit lower, leaving headroom. But balance may drift.
If it was open-loop, I'd set charge inverter/SCC charge voltage above the BMS balancing point but well below 3.65V per cell.
Closed loop, you should be able to set BMS to request a lower charge voltage.
If cell balance improves over time, then maybe turn up higher.

I'd think reducing magnitude of load dump would avoid having cells run so far. For instance, a dryer rewired for 120V instead of 240V into heating element would be 1/4 as large a step in power when it shuts off. Whatever your biggest load is. If balance gets better, then can try full power again.

An ideal system might have multiple batteries on separate busses. Could periodically charge one to full in order to rebalance, while operating from the rest. But most inverters have a single battery input.

 

[Kenny_]

I'm guessing that too is the reason for shipping a 'standard' rs485 cable with the inverter rather than one that works specifically with their own batteries....but still...

I'm very bothered by that cable situation and the error last night. I mean it has been work fine, no errors for a week and then overnight for the first time an error 58... switched to the modified cable and all is happy again.

I have not tried switching back to the standard supplied with inverter cable yet. May not as long as things are working.

 

[Hedges]

Now that you have made it work intermittently, try reconnecting the ground wires you cut previously.
Differential signal are subject to noise taking them outside allowed common-mode voltage range. They may find a return path through ground wires (big loop picking up EMI), and voltages may just drift around.
Unlike Ethernet which is transformer coupled, RS-485 relies on circuit "ground" references being within a couple volts.

 

[Lighthouse Beacon]

Sounds right. Not sure what my threshold is at but watching it ... it does slow to a crawl when it is over 90-95%

just checked ... Param 60... mine is set to 100%...default I

I'm not sure BMS default setting are well though out for interaction with dozens of different no-name inverters.
The AC coupled ones (typically quality big names) are going to have to shove a few seconds of current into battery.
All the rest, don't know how quickly their MPPT ramps down charging.

I think batteries ought to be balanced at higher voltage, charged a bit lower, leaving headroom. But balance may drift.
If it was open-loop, I'd set charge inverter/SCC charge voltage above the BMS balancing point but well below 3.65V per cell.
Closed loop, you should be able to set BMS to request a lower charge voltage.
If cell balance improves over time, then maybe turn up higher.

I'd think reducing magnitude of load dump would avoid having cells run so far. For instance, a dryer rewired for 120V instead of 240V into heating element would be 1/4 as large a step in power when it shuts off. Whatever your biggest load is. If balance gets better, then can try full power again.

An ideal system might have multiple batteries on separate busses. Could periodically charge one to full in order to rebalance, while operating from the rest. But most inverters have a single battery input.

Yes that was my thoughts. I went with SOC because it’s the only setting I can manipulate without going into “user defined” batt settings and I haven’t gone through a did a custom setup but it would be the best setup, their program defined L16 LFP settings are too aggressive imo, so by changing ID 60 Charging Cutoff SoC to 97% I’m staying away from that 100% OVP protection threshold. House was using cooking gear in kitchen (wife) when she stopped that load is when inverter pushed excess power to batts instead when batts were near full. That spike is what pissed the BMS off is my theory.

I’ve documented it and followed up with SG support “Cindy” to hear her thoughts. I’m sure she won’t care:

“It’s working now? Then shut up!”

 

[42OhmsPA]

I bulk to 55.2v and float at 54.8v. I don't see any need to take them to the top of the knee. I reprogrammed my JBDs to read full at this point, should be able to reprogram the SGP BMSs as well if you download the PC software. This would allow you to use SOC to 100 percent and not hit the top of the knee.

 

[Lighthouse Beacon]

I bulk to 55.2v and float at 54.8v. I don't see any need to take them to the top of the knee. I reprogrammed my JBDs to read full at this point, should be able to reprogram the SGP BMSs as well if you download the PC software. This would allow you to use SOC to 100 percent and not hit the top of the knee.

My soc limit on inverter program isn’t stopping the batt apparently, batts reporting 98% SOC and it’s pumping more amps to it now.
I’ve got a lot of power coming in on panels now and nowhere to put it. Inverter showing batts full and not charging. BMS on batts showing it’s taking charge and over 97% soc

 

[Lighthouse Beacon]

My soc limit on inverter program isn’t stopping the batt apparently, batts reporting 98% SOC and it’s pumping more amps to it now.
I’ve got a lot of power coming in on panels now and nowhere to put it. Inverter showing batts full and not charging. BMS on batts showing it’s taking charge and over 97% soc




View attachment 176254
View attachment 176253

Yeah OVP and erroring out code 59 right now. It did the same thing again. I just disconnected the battery to leave the house. I don’t trust it. I got to get a game plan together here

 

[42OhmsPA]

My soc limit on inverter program isn’t stopping the batt apparently, batts reporting 98% SOC and it’s pumping more amps to it now.
I’ve got a lot of power coming in on panels now and nowhere to put it. Inverter showing batts full and not charging. BMS on batts showing it’s taking charge and over 97% soc




View attachment 176254
View attachment 176253

Start running things...
Situations like that are why I bought a water distiller and steam generator for a mini sauna to use as dump loads.
I just started a load of laundry a few minutes ago, a dryer cycle should keep me from getting to float today. I had the bank down to 40% before I started generating power this morning.

Setting 57 may help you with reducing charge current, I'm not sure exactly how it works.

 

[42OhmsPA]

Yeah OVP and erroring out code 59 right now. It did the same thing again. I just disconnected the battery to leave the house. I don’t trust it. I got to get a game plan together here

Are you sure the setting 60 value stuck?
Set it ~90% and see what happens or ditch comms and run off voltages.

 

[Lighthouse Beacon]

Are you sure the setting 60 value stuck?
Set it ~90% and see what happens or ditch comms and run off voltages.

Yeah I checked it stuck, I think BMS comms is having priority over inverter settings? Makes no sense. BMS comms is starting to look more like a problem than a solution. Getting frustrated

I bulk to 55.2v and float at 54.8v. I don't see any need to take them to the top of the knee. I reprogrammed my JBDs to read full at this point, should be able to reprogram the SGP BMSs as well if you download the PC software. This would allow you to use SOC to 100 percent and not hit the top of the knee.

 

[42OhmsPA]

Yeah I checked it stuck, I think BMS comms is having priority over inverter settings? Makes no sense. BMS comms is starting to look more like a problem than a solution. Getting frustrated

Take a breath and walk away. It's a learning experience. I wish I could be more help but I never dealt with closed loop.

 

[Kenny_]

I bulk to 55.2v and float at 54.8v. I don't see any need to take them to the top of the knee. I reprogrammed my JBDs to read full at this point, should be able to reprogram the SGP BMSs as well if you download the PC software. This would allow you to use SOC to 100 percent and not hit the top of the knee.

I just asked about the software....there are a couple listed on the website but neither specifically lists the 10k48 .. sent a message...haven't heard....haven't heard about my shipment/order either, but I see labels created at Fedex this morning....

 

[Kenny_]

Now that you have made it work intermittently, try reconnecting the ground wires you cut previously.
Differential signal are subject to noise taking them outside allowed common-mode voltage range. They may find a return path through ground wires (big loop picking up EMI), and voltages may just drift around.
Unlike Ethernet which is transformer coupled, RS-485 relies on circuit "ground" references being within a couple volts.

I was considering doing just that...I'm very confused about why it was 'appearing' to work properly with the full cable...
Will give it a go! Thanks!

Okay reconnected the green wires....still working (checked the full cable while I was at it...still giving an alarm)....no alarm with modified cable and green/ground wires connected.

 

[Lighthouse Beacon]

Take a breath and walk away. It's a learning experience. I wish I could be more help but I never dealt with closed loop.

Ok I’m back home. I’m about to ID 32 SLA default (disable BMS comms)

Then ID 9 Bulk Battery Charging 56.2

And see what happens. I’ve still got a decent amount of power coming into panels

 

[Kenny_]

Ok I’m back home. I’m about to ID 32 SLA default (disable BMS comms)

Then ID 9 Bulk Battery Charging 56.2

And see what happens. I’ve still got a decent amount of power coming into panels

Good Luck! mine is cranking right now (at least as far as my panels/system goes) about 2.5 k or so mostly into batteries...back up to about 80% SOC

 

[42OhmsPA]

Ok I’m back home. I’m about to ID 32 SLA default (disable BMS comms)

Then ID 9 Bulk Battery Charging 56.2

And see what happens. I’ve still got a decent amount of power coming into panels

Don't forget 08 to USER.
If you still have issues keep dropping 09.
You could have cells that are out and balance and 56.2v might cause a runner to trigger ovp. I'm not sure what kind of balancing the sgp BMS has... Cell voltage changes quickly when you get up above 3.5v from what I've seen.

 

[Lighthouse Beacon]

Don't forget 08 to USER.
If you still have issues keep dropping 09.
You could have cells that are out and balance and 56.2v might cause a runner to trigger ovp. I'm not sure what kind of balancing the sgp BMS has... Cell voltage changes quickly when you get up above 3.5v from what I've seen.

My next step is to go user if this doesn’t work. It appears to have worked but I thought the same thing last time. I think it should take the change of ID9 with my batts set to L16



I went:
ID32 SLA
ID9 55.2
ID11 54.8

Its appears to be behaving, I will have to see tomorrow at peak sun what happens.

For now BMS Comms are disconnected.
In all honesty i am not even using batts but for UPS backup, I just wanted to use this system basically an isolated grid tie

I think the BMS failure error I was getting was compile errors stacking from the BMS to the inverter. I think the language between these two systems contains bugs. Inverter brain and battery brain may work better independently.

Sorry for high jacking your thread @Kenny_