diy solar

diy solar

Solis RHI 5kw + Pylontech US500 not stopping discharge at defined SOC

Glad I'm not the only one with this issue. I tried opening a ticket with Solis and they said they changed something and also offered an update. I had a look though the settings and found that Grid Standard Precision has been changed from "Voltage 0.1V, Time 0.01s, Frequency 0.01Hz" to "Voltage 1V, Time 0.1s, Frequency 0.1Hz". Not sure why they changed this and was curious if anyone had any insight into that!?

I accepted the update and am now running firmware version 14003F on my RHI-3P10K-HVES-5G. Interesting at first that seemed to solve the problem of battery drain once hitting minimum SOC but have now noticed a few occasions where it has not occured/odd behaviour. Few imaged attached. 1 is before update (~50w being pulled), 2 is after (Nothing being pulled once hitting minimum SOC), 3 is the behaviour I'm seeing now, patches of nothing and with random areas where it pulls from battery. It also now only pulls 38w when it decides it needs to.

Since we are now in the winter months I have just bumped minimum SOC to 20% with force at 10% to give the bit more leeway and prevent force charges.
 

Attachments

  • 1.png
    1.png
    7.2 KB · Views: 4
  • 2.png
    2.png
    16.4 KB · Views: 4
  • 3.png
    3.png
    12.1 KB · Views: 5
@jamesy311 Welcome! I'm not familiar with the RHI-3P10K-HVES-5G - am guessing it operates on quite different firmware to the LV single phase models. You're the first UK-based person on here (that I'm aware of) that is running the HV version of Solis, so can't help, but maybe @Solar Guppy (who has a HV Solis in the USA) might chip in. Either way, it would be good if you keep this thread updated with what you find out, thanks.

What batteries are you using with the RHI-3P10K-HVES-5G? And how much PV input? If you have a tariff with a cheap night rate (like economy 7 or one of the Octopus ones) it may be worthwhile to charge up the batteries at night anyway, this time of year, to maximise your use of low priced energy - just a thought?
 
@jamesy311 Welcome! I'm not familiar with the RHI-3P10K-HVES-5G - am guessing it operates on quite different firmware to the LV single phase models. You're the first UK-based person on here (that I'm aware of) that is running the HV version of Solis, so can't help, but maybe @Solar Guppy (who has a HV Solis in the USA) might chip in. Either way, it would be good if you keep this thread updated with what you find out, thanks.

What batteries are you using with the RHI-3P10K-HVES-5G? And how much PV input? If you have a tariff with a cheap night rate (like economy 7 or one of the Octopus ones) it may be worthwhile to charge up the batteries at night anyway, this time of year, to maximise your use of low priced energy - just a thought?
I've got a 14.2kwh Pylontech force H2 with 10kw of panels. I was luckly to get a contract before everything went down the toilet so until mid 2024 I have pretty good import rates. Once that time comes closer I plan on doing some research into the best plan for solar/batteries. I've had others suggest charging overnight so I will keep that in mind.
 
0.65A at around 48V would be about 30W, which IIRC is about the power the Solis uses (assuming no PV input) to keep it running - that sounds about right to me. I would also assume that the Solis info screen would only show battery power being inverted and not just the DC consumption to keep it running.

You could set the re-charge SOC to say 10% to avoid the on-off charge/discharge.
FWIW, my Solis RHI-5K-48ES-5G(v1) firmware 00A2/0032 Hybrid inverter (which is about 1 year old) seems to gobble about 40 or 50W when there is no solar. This means the darn thing is consuming over 300kWh per year, which does not compare well with the analysis from my installer which cites 'standby power' as 8kWh per year.

As an electronics/firmware/software engineer, I really fail to see why it is not possible for these inverters to just shut their eyes and go to sleep once minimum SoC is reached, just monitoring PV input and battery status - those jobs should take less than 1W, I'd have thought - there is no need to even track the grid. 40W in standby is an astonishingly high amount - yet seems to be tolerated by the marketplace.

I'm sure if I had schematics of the inverter and access to the firmware, I could do a far more frugal job! (but that, of course, would then need certification...)
 
Last edited:
FWIW, my Solis RHI-5K-48ES-5G(v1) firmware 00A2/0032 Hybrid inverter (which is about 1 year old) seems to gobble about 40 or 50W when there is no solar. This means the darn thing is consuming over 300kWh per year, which does not compare well with the analysis from my installer which cites 'standby power' as 8kWh per year.
Are you sure that much? Both my RHI and my new EH1P measure 29-30W idle. Others have measured similar and @Lazy_pete's recent picture shows 0.03kW draw from battery...

see https://diysolarforum.com/threads/ct-inaccuracies.45436/post-923930

As an electronics/firmware/software engineer, I really fail to see why it is not possible for these inverters to just shut their eyes and go to sleep once minimum SoC is reached, just monitoring PV input and battery status - those jobs should take less than 1W, I'd have though
It is possible via writing to a ModBus register to turn the inverter off. However, doing so will mean that the backup supply will also shut down - maybe that is not of concern to you. Details for doing that are in one of @peufeu's postings - I'll see if I can find it.
 
I'll see if I can find it.
Here are the relevant postings...




and also mentioned here...

 
FWIW, my Solis RHI-5K-48ES-5G(v1) firmware 00A2/0032 Hybrid inverter (which is about 1 year old) seems to gobble about 40 or 50W when there is no solar. This means the darn thing is consuming over 300kWh per year, which does not compare well with the analysis from my installer which cites 'standby power' as 8kWh per year.

As an electronics/firmware/software engineer, I really fail to see why it is not possible for these inverters to just shut their eyes and go to sleep once minimum SoC is reached, just monitoring PV input and battery status - those jobs should take less than 1W, I'd have thought - there is no need to even track the grid. 40W in standby is an astonishingly high amount - yet seems to be tolerated by the marketplace.
Yes.

This standby power is with the inverter running, all power stages included, and ready to switch the backup output from mains to inverter mode in milliseconds in case of a blackout.

Obviously it is totally dumb to keep doing that when the battery is discharged which means if there's a blackout, it won't be able to provide any energy to the backup output anyway.

I guess they forgot! Or maybe it was to save cost on relays...

My code is querying the inverter via modbus. It checks battery SOC and voltage on solar MPPT inputs. If voltage is low it means it's night and PV is not producing, then if SOC is low, it issues a modbus command to turn off the inverter. Then it really sleeps, the only thing active is the display and buttons which you can use to turn it back on manually. In this state it uses very low power (~1W). When MPPT voltage rises in the morning, it issues another modbus command to turn it back on.

So yeah it has to be done "manually" lol

Of course, when it's off (via modbus) the backup output turns off! So I had to add a DPDT contactor to switch the backup loads between mains and backup output (its coil is driven by the backup output).

This contactor also allows to keep backup loads powered if the inverter shuts down or takes a trip to the service center, so it's useful.

Since the backup output can be controlled via modbus too, by default I leave it off, and the contactor power loads from mains. So the coil would only be energized during a blackout.
 
my installer which cites 'standby power' as 8kWh per year.
He probably read the spec sheet, but that's the thing, no matter the manufacturer, they never say what "standby power" means. Is it with the thing on, doing nothing, but ready to take over in case of a blackout, or is it off?... who knows... although one thing is sure, they're always quoting the lowest number among the possibilities ?

as a result all these AIO inverters get away with terrible ACTUAL "it's on but it's doing nothing useful" power consumption, like the DEYE also uses 50-something watts.
 
I might have found an easy workaround solution to the leaking min SOC problem thanks to this great thread! ?
I have a RHI-6K-48ES-4G green inverter marked "Zero Co2", in Italy. My installer, the local Pylontech support and I have been fighting with my settings for months. (Local Solis support never really tried.)

I applied the new solution at 9:30 pm, while I was in my usual Self Use mode, with Charging from Grid Allowed, Optimal Income on STOP and, if I recall correctly, Backup off. OverDischarge was set at 20% and Force Charge at 17%. My inverter was apparently drawing about 88W out of my 3x Pylontech 3000 battery bank which was at 16% SOC.

Contrary to usual, this time I did not switch off the batteries and the inverter for the night but, inspired in particular by the smart observation of meteoloco, I just set:
- "Optimal Income" to RUN under Self Use>Time of use,
- "Charge limit" to just 1,5A, and
- A charging time of 1' (9:32 pm - 9:33 pm)
.
All the other charge and discharge times were left at 00:00 and all the remaining settings untouched as before.

Now is 00:58 am, and since 9:33 pm, my display has been showing a 16% battery, charging with 19W inflow, and a grid power correspondingly higher than the house loads.

1704760147713.png

I will update as soon as anything changes <?>or in a couple of days to confirm that no more changes were necessary</?>​

PS The only strange thing is that at 11:24 pm my Solis app reported a Fan-Alarm F011. It was just a "Hint Alarm" which was automatically resolved at 11:29 pm. But I never had Fan troubles before, the PdC was not running, I only had the TV and a few lights on (the inverter app graphic does not show any peak low at that time - see below) and the temperature downstairs tonight is well above 0°. I cannot imagine why the Fan could have wanted to start in these conditions. (Should the alarm come up again, I will write a new message asking about it.)
1704760843046.png
 
Yes, based on the last 5 days (with very variable weather) I'd say that my workaround to the original problem raised by K-Tech. Earlier I had frequent BMS alarms raised by the inverter and frequent protection episodes + audio alarms from one or more Pylontech batteries, repeatedly reaching extreme and dangerous low SOC levels.
To avoid this, while in "Self Use" mode, you just need to choose "Allow" under "Charge from Grid" and Optimal Income to "Run" under "Time of use". If you do not want to use the expensive grid energy to charge your batteries, also set the "Charge limit" to a low value (e.g. 1,5 A) and limit the Charging time to 1 minute in the evening, sometime after sunset. All other charge and discharge times should remain at 00:00. After that 1 minute time, the system will continue to keep your batteries safe by recharging them with a mere 20W until when your PV charges them again above the Over Discharge SOC.
BEFORE (Jan 5, cloudy day): / (AFTER Jan 8, cloudy. Initial 0% due to prior BMS error/protection):
1705229582731.png 1705229598210.png

SUNNY DAYS (e.g. yesterday NB W-facing PV panels):
1705229969102.png

The BMS alarms and protection episodes started well before I managed to get the meter installed, but I should probably add that my system includes an Elastrom Meter connected to the inverter, enabling the routing of the current in the right direction according to the current load and production.
 
Last edited:
I see what you're saying about SOC calc and voltage. Pylontech require max 30% DOD for warranty purposes, but I'm starting to realise that maybe their "warranty" isn't worth much. Also, it seems the BMS determined that 30% SOC was at around 48.6V last night, but then continued to discharge until I woke this morning when realised it was at 12% SOC at 48.2 and switched off until the sun came up. Also to confirm, the LEDs that indicate battery state were still showing discharge at this level, whereas if it was functioning properly the BMS should have put it in idle until charge was available.

As far as I am aware the idle state should be no inverter to battery/cell connection but BMS still active and reporting voltage/SOC etc.

Maybe I'll try setting under Voltage to 48.6 and see what happens
I K-Tech, I am not sure if you are still looking for a solution. If yes, I believe I just found an easy workaround that appears (explained at page 6 of this thread) to be working well. I am not entirely sure that this will work in the UK in winter, too. I think it should, and I hope so. Please let me know your results if you try it.
 
To avoid this, while in "Self Use" mode, you just need to choose "Allow" under "Charge from Grid" and Optimal Income to "Run" under "Time of use". If you do not want to use the expensive grid energy to charge your batteries, also set the "Charge limit" to a low value (e.g. 1,5 A) and limit the Charging time to 1 minute in the evening, sometime after sunset. All other charge and discharge times should remain at 00:00. After that 1 minute time, the system will continue to keep your batteries safe by recharging them with a mere 20W until when your PV charges them again above the Over Discharge SOC.
I'm not sure that setting a charge time is necessary. It may be that just enabling Charge from Grid would be sufficient to allow the inverter to top up the batteries when they reach Force Charge SOC.
 
Good morning guys. I've the same problem descibed in the post, so i'm making some test.
I'm working on the Battery Reserve Switch, putting it ON, with 21% reserved SOC. The first issue is when the SOC reach the trigger, 21%, the Solis start to charge the batteries (4x Pyl us2000c) whit a very high power, about 5kW, and this overload my grid feed, nominal 3kW, and the supplier breaker open. So now my question : is it possible set the value of the maximum current/power absorbed from the grid, like in forcecharge SOC (setted at 750W)?
 
So now my question : is it possible set the value of the maximum current/power absorbed from the grid, like in forcecharge SOC (setted at 750W)?

Yes in the TOU menu you can set charge and discharge battery current values as well as the time it is enforced
 
100A there is 5kW (at 50V). If you want it to take 3kW set it to 60A.

There is also another setting for Max grid power when Force charging that can be used to limit the automatic charge when the battery gets too low.
 
All sounds so much firmware issue

Got a Solis S5-EH1P with 2 Dyness A48100.
At night, as a result of normal discharge of the batteries, when the battery reaches the SOC I’ve specified in the Solis to stop discharging and use the grid, the batteries “leak” 62w constantly until it gets to the force charge parameter and then starts charging.

View attachment 119277

My SOC setting is 16%. But if it were 20% it would be also ignored. :(
This has to be software fixable. Counterproductive.

Only once, a week back, saw the battery icon with a grey color in 13% and apparently idling and 0 kw as it should be triggered by the inverter.

I bet you there are loads of people with this issue but haven’t noticed. Please do report if you hear anything towards a solution here.

Solis is very obscure on regards firmware versions and models, if they release a solution it won’t be effective in all inverters unless they apply a firmware update remotely and individually. I just hope it can be fixed toying with settings, no luck so far for me.

Keep us updated
How did the solis ticket go?.
 
Yes, based on the last 5 days (with very variable weather) I'd say that my workaround to the original problem raised by K-Tech. Earlier I had frequent BMS alarms raised by the inverter and frequent protection episodes + audio alarms from one or more Pylontech batteries, repeatedly reaching extreme and dangerous low SOC levels.
To avoid this, while in "Self Use" mode, you just need to choose "Allow" under "Charge from Grid" and Optimal Income to "Run" under "Time of use". If you do not want to use the expensive grid energy to charge your batteries, also set the "Charge limit" to a low value (e.g. 1,5 A) and limit the Charging time to 1 minute in the evening, sometime after sunset. All other charge and discharge times should remain at 00:00. After that 1 minute time, the system will continue to keep your batteries safe by recharging them with a mere 20W until when your PV charges them again above the Over Discharge SOC.
BEFORE (Jan 5, cloudy day): / (AFTER Jan 8, cloudy. Initial 0% due to prior BMS error/protection):


SUNNY DAYS (e.g. yesterday NB W-facing PV panels):


The BMS alarms and protection episodes started well before I managed to get the meter installed, but I should probably add that my system includes an Elastrom Meter connected to the inverter, enabling the routing of the current in the right direction according to the current load and production.
This fix is working for me on a S5-EH1P5K-L.
Thank you for posting, its been a challenge to find a solution.
 
This fix is working for me on a S5-EH1P5K-L.
Thank you for posting, its been a challenge to find a solution.
Glad to ear it! However, I am afraid that the challenge continues and it might not be such a permanent fix as I thought:
- I have to go back and adjust the charging time settings about twice per month. This might depend on the changes in the meteo conditions and sunset times.
- At least once, the "Allow charge from grid" seems to have switched back to "Do not allow" by itself.
- In the last week, I also received 3 BMS alarms on different days, while the SOC was well above the lower threshold. When checked, I found that one of the three batteries had all LEDs flashing, including the red, and the beeping alarm on, while the other two were powering my house normally. The flashing battery was not always the same (twice the top, once the bottom one). Once I restarted the system, the flashing battery seemed to be full (all green LEDs on), and balanced the charge with the others in a little while. This behavior was much more frequent before I applied my fix, and it seemed to have disappeared for more than a month until recently.
 
Back
Top