Advertisement
Help Keep Boards Alive. Support us by going ad free today. See here: https://subscriptions.boards.ie/.
If we do not hit our goal we will be forced to close the site.

Current status: https://keepboardsalive.com/

Annual subs are best for most impact. If you are still undecided on going Ad Free - you can also donate using the Paypal Donate option. All contribution helps. Thank you.
https://www.boards.ie/group/1878-subscribers-forum

Private Group for paid up members of Boards.ie. Join the club.
Hi all, please see this major site announcement: https://www.boards.ie/discussion/2058427594/boards-ie-2026

Sunsynk (DEYE) Hybrid Inverters

14143454647

Comments

  • Registered Users, Registered Users 2 Posts: 1,958 ✭✭✭tnegun


    Not once TOU is enabled and Grid Charge is unticked for the active program



  • Registered Users, Registered Users 2 Posts: 241 ✭✭serox_21


    Thanks,

    I'm going to enable grid charge and up the amps to 100 in battery settings.

    I'll use the ToU menu to limit the charge to 2500W until oct/nov, after full 5000w charging power.



  • Moderators, Home & Garden Moderators Posts: 7,270 Mod ✭✭✭✭graememk


    The tou power is for discharge not charge.

    Charge rate is set in amps where you ticked the box

    I've had weird issues with the tou page when I had it spread out over 2 days, I'd be having it start at midnight and end at midnight



  • Registered Users, Registered Users 2 Posts: 241 ✭✭serox_21


    @graememk if you are running your inverter with overnight charging can you validate bellow settings for a 2-6AM grid charging (up to 2500W for now)

    If I lower in battery setting the grid change A will it also limit the max charging power in ToU menu?

    IMG_20250730_111509.jpg IMG_20250730_111528.jpg


  • Registered Users, Registered Users 2 Posts: 241 ✭✭serox_21


    Or does anyone can provide some screenshots with ToU menu and other settings to make it work for charging during the night



  • Advertisement
  • Moderators, Home & Garden Moderators Posts: 7,270 Mod ✭✭✭✭graememk


    All looks good.

    The power on the TOU is only for the battery discharging. The charging power is set in the battery settings.

    It should charge overnight with the above settings



  • Registered Users, Registered Users 2 Posts: 241 ✭✭serox_21


    Thanks,

    I did find a informative video which explains how Zero export to load and CT work, including the charging and Time of Use menu.

    All I need to understand is how to export from battery to the grid just before 2AM (like 12-2AM)

    The rest of the day battery should power the house if no solar available and not export to the grid.

    Solar Sell option should allow for battery export to grid based on time of use.

    I might need home assistant integration for above scenario.

    Waiting for esp32 boards, so I can connect invertor and battery to HA.

    Thanks @graememk



  • Registered Users, Registered Users 2 Posts: 241 ✭✭serox_21


    A quick update.
    Inverter charged the battery as defined in ToU with a 50A current.
    This morning I was exporting before 8AM. Usually it would export later after 10 - 100 depending on the weather(It was cloudy most of this or last week).



  • Registered Users, Registered Users 2 Posts: 23 Blakes77


    1.jpg

    Anybody else seen this weird charging issue ?? JK BMS is asking for 55.2 volts (as seen on the right) but the DEYE only gives a maximum of 54.7 volts (as seen on the left). What ever I set in the BMS (by changing Vol CELL RCV) the DEYE always gives 0.5 volts under that! I found another post on the web discussing the exact same issue with no solution. If I untick TOU then the issue goes away but then the battery won't discharge to load. Ticking Disable Float mode on the battery settings makes no difference. Struggling to see how I can make this work with the JK BMS Absorption/Float because the battery voltage never gets to Vol CELL RCV so it doesn't enter absorption and then after the timer switch to Vol CELL RFV. Maybe I can calibrate the JK to read 0.5 volts over the real voltage? I love the big charge/discharge capability of the DEYE but the settings on these inverters are a nightmare compared to SOFAR.



  • Moderators, Home & Garden Moderators Posts: 7,270 Mod ✭✭✭✭graememk


    Li-BMS screen is what is reported to the inverter by the battery, Via CAN. (everything on that page is supplied by the bms)

    What the inverter actually reads is on the first screen. Also the voltage is a target, it wont be at that unless the battery is full

    also 3.45/cell is very conservative, Especially if charging at high currents.



  • Advertisement
  • Registered Users, Registered Users 2 Posts: 241 ✭✭serox_21


    I noticed same behavior, that the battery voltage rarely reaches the RCV values, and the inverter will changes with around 0.5V lower than what the BMS is asking.

    The problem I have with this is that after a few days SOC will show around 80% even though the cell voltage is up to 3.4x which is 100% charged.

    As a workaround from time to time I increase the RCV value to 3.5x which will force the charge, then lower it down after a few minutes to intended RCV value (in my case 3.48), forces the BMS to enter float mode after set time(1 hour in my case).

    This way the SOC is reseted to 100%

    Have anyone contacted Deye with this behaviour?



  • Moderators, Home & Garden Moderators Posts: 7,270 Mod ✭✭✭✭graememk


    from what ive observed, the JK needs regular full charges (with a reset to 100%) to stay in line.

    I dont get to full often, and my controlling BMS is a ZEVA bms, But the last time I got to full, I was at my set 58V.

    I dont faff around at 3.4x, My minimum max charge would be 56/57v. It spends so little time there it doesn't matter.

    Keep the RCV to 3.5x

    The full charge voltage of 58v or whatever you set it at, will only be seen when the battery is at that voltage, otherwise it would be pulled down to the battery voltage.



  • Registered Users, Registered Users 2 Posts: 6,555 ✭✭✭championc


    I just have mine set to battery voltage mode, so there no connection between my JK and the Inverter. The inverter calculates it's own SOC



  • Registered Users, Registered Users 2 Posts: 23 Blakes77


    Ok so its not just me. Looks like there's some wonky code in the Deye charging algorithm. I have the JK settings as per Off Grid Garage and this is not working for me, batteries sit on Bulk charge all the time and never reach target because of the deye -0.5 volt offset. So batteries never reach 100% SOC, don't balance and no bulk/absorb/float. Wish I knew this before I swapped out my old Sofar inverters to Deye!

    There is a thread on this on the DIY solar power forum here:

    https://diysolarforum.com/threads/deye-weird-voltage-behavior-when-time-of-use-is-enabled.80908/

    Its mentioned you can use an ESPHome module that's been developed to apply an offset to the CANBUS set-points from the JK to BMS

    I'm going to try messing with the voltage calibration in my JK, and I'll try changing the Deye to battery voltage mode, which all feels like a big step backwards



  • Moderators, Home & Garden Moderators Posts: 7,270 Mod ✭✭✭✭graememk


    Bump the charge voltage and never look back. There is no charging algorithm. It just charges at a constant current, until it reaches the set voltage- on the inverter, not the one it gets told by the bms, then switches to constant voltage.

    Next time you get your battery close to that voltage, dont look at the libms screen check the first screen, that will show you the voltage the deye is reading. You could even verify via a multimeter on the terminals of the deye.

    There is a voltage drop when power goes through a diode, My controlling bms is a ZEVA bms, and it actually doesnt go "through" that bms, No diodes, just a fuse and an isolator. I see the full voltage on my pack.

    I must see if theres a difference on the packs that are on a JK bms

    Off grid garage is way too conservative anyway. TLDR: up the voltage.



  • Registered Users, Registered Users 2 Posts: 60,115 ✭✭✭✭unkel


    I always use voltage as the SOC calculation of whatever BMS is brutal and prone to errors. In particular stopping discharging as it thinks SOC is too low, which means buying very expensive day rate electricity from the grid. No thanks!

    "Make no mistake. The days of the internal combustion engine are definitely numbered" - Quentin Willson, 1997



  • Registered Users, Registered Users 2 Posts: 2,689 ✭✭✭SD_DRACULA


    Agreed, if you have more than one pack or do not charge to 100% daily forget about using the SOC, go for voltage mode instead.



  • Moderators, Home & Garden Moderators Posts: 7,270 Mod ✭✭✭✭graememk


    Been experimenting, watching my 2 jk packs.

    And they now are tracking each other soc wise (still has the same downward drift)

    The way I got them to track is by either increasing the ah setting on the battery that is moving faster. If the ah is set too low, that battery will take more charge than the other battery and it's SoC will run away from the other one.

    Everyday for a 100% charge (with a reset to 100%) is a bit much but once a week should be fine.



  • Registered Users, Registered Users 2 Posts: 60,115 ✭✭✭✭unkel


    What is the benefit of comms though? Voltage is so much easier and less error prone and the BMS will still protect the battery

    "Make no mistake. The days of the internal combustion engine are definitely numbered" - Quentin Willson, 1997



  • Moderators, Home & Garden Moderators Posts: 7,270 Mod ✭✭✭✭graememk


    Makes automation easier, indicated voltage from a battery not at rest, can vary wildly, as the load has an effect on it.

    SOC drift on a JK, At reset all cells were above 3.6x V
    15 days since a full charge, with minor use, (roughly about 25% per day) ended up with a 16% drift in soc



  • Advertisement
  • Registered Users, Registered Users 2 Posts: 60,115 ✭✭✭✭unkel


    Agree with all that…..if the battery is small. Most of our home made packs are at least 16kWh now, often multiples of that. So load has little to no effect on voltage and voltage doesn't matter for automation of export if it is in the middle somewhere. The LiFePO4 curve is not all that flat either, significant difference between 52.0V, 52.5V, 53.0V

    "Make no mistake. The days of the internal combustion engine are definitely numbered" - Quentin Willson, 1997



  • Registered Users, Registered Users 2 Posts: 23 Blakes77


    I think I’m making progress. I'm suspicious of the Deye because when I don’t have TOU ticked it gives me the charge voltage the BMS asks for, and when I do have it enabled then it gives me 0.5 volts less than what I ask for. This is important if you’re using a JK Inverter BMS because its the same parameter used to request the charge voltage from the inverter that is also used to trigger the change from bulk to absorption and initiate the SOC reset. So if you don't reach target, the JK will not work as intended. I found it really noticeable when I was trying to charge to 55.2 volts because the Deye will charge up to exactly to 54.7 volts and then sit there until sundown. I saw exactly the same behavior when trying to charge to 55.8 volts, the Deye charged up to exactly 55.3 volts. I think what is going on is the Deye is just being conservative in its charging and you need to push it more into the area of exponential voltage growth, 56+ volts seems to work (as graememk suggested). It is important to know also that the Deye seems to discharge the battery if the battery voltage is over the requested charge voltage. So with the target of 56 volts, when the battery voltage gets near 55.5 my Deye seems go to into a loop of charge/standby/discharge, the voltage nudges up and after some minutes the JK will get a blip of 56 volts and trigger its timer for absorption and SOC reset. When the JK switches the charge voltage request down to the float voltage, my Deye will discharge the battery to reach this target. Yesterday was full sun all afternoon, my batteries were 100% by noon but then discharged by the Deye to 92% by late afternoon. So today I have also increased the float voltage setpoint to prevent such a big discharge. Maybe this is a waste of time and I should just charge on battery voltage but I will continue to tweak setpoints and post findings later if any interest.



  • Registered Users, Registered Users 2 Posts: 483 ✭✭Nuphor


    I've a Deye SUN-8K-SG01LP1-EU (single phase) inverter that I bought about a year ago now but not installed. I bought it thinking I might expand from my existing Deye 5K with extra battery charging capacity, but in the end I don't think I need it. Thinking of sticking it on Adverts, if anyone in this thread was interested in an upgrade.



  • Moderators, Home & Garden Moderators Posts: 7,270 Mod ✭✭✭✭graememk


    I Might have a chance to test the JK bms communication on sunday to see if I can replicate what your getting. I am getting full voltage on mine. (but with the zeva bms in control)

    on the first battery screen, what voltage is the Deye reading?

    Interesting how it discharged the battery so it would reach the float voltage, Must test that sometime and see where the power goes.



  • Registered Users, Registered Users 2 Posts: 23 Blakes77


    In the JK I've got Bulk voltage set to 56 (Cell RCV = 3.500) and float voltage set to 54.7 (RFV volts = 3.420). The JK did hit the bulk voltage around noon, waited the absorption timer, reset SOC to 100%, then changed the charge voltage in the Deye to 54.7, all exactly what should happen. The Deye then slowly discharged the battery to grid at around 150 watts (PV was over 2KW all afternoon and house load only around 300w) until the battery voltage reached 54.2, exactly 0.5 volts under setpoint! ending up at 97% SOC. Not what should happen! See pics attached. The Deye is reading 54.21v and the JK 54.19v.

    Float1.jpg Float2.jpg


  • Registered Users, Registered Users 2 Posts: 443 ✭✭jasgrif11


    @SD_DRACULA Do you use voltage with your battery or do you take the percentage from the Smart Shunt? Is there a big difference between the JK reported SOC and the Shunt?



  • Registered Users, Registered Users 2 Posts: 2,689 ✭✭✭SD_DRACULA


    Yeah I ignore jks, they are wrong after one cycle even.

    The shunt also get out of whack though so it is not perfect, I usually just drain them until 50v which is around 10%



  • Registered Users, Registered Users 2 Posts: 241 ✭✭serox_21


    I've purchased a esp32D board and max485 ttl converter to connect my Deye inverter to home assistant but not sure how compatible this 2 are.

    ChatGpt say they are incompatible esp32D board runs at 3.3V while max485 at 5V.

    Feature

    Spec

    ESP32 Compatibility

    VCC

    Requires 5V to operate

    ✅ OK if powered from ESP32's 5V (VIN) pin

    Logic Levels

    RO pin outputs ~5V

    ⚠️ Too high for ESP32 (max 3.3V) — can damage RX pin

    DE/RE/DI inputs expect ~5V

    ⚠️ ESP32 3.3V signals may not be reliably detected

    Does anyone uses this setup to connect Deye invertor to home assistant?

    IMG_20250806_130101.jpg IMG_20250806_125838.jpg


  • Registered Users, Registered Users 2 Posts: 23 Blakes77


    The easier way is to use a USB to RS485 converter and connect to the RS485 port on the Deye. If you chop a standard ethernet cable in half, the RJ45 ethernet plug goes in the top left socket on the deye and the blue and white wires onto the RS485 converter. From memory white is the A, blue is the B. Same would also probably apply to your ESP32. There is a good home assistant add-on from kellerza that works perfectly.



  • Advertisement
  • Registered Users, Registered Users 2 Posts: 23 Blakes77


    Today I tried running on battery voltage and guess what…. The Deye still has the exact same behaviour. With TOU ticked the actual target Deye charge voltage is 0.5 volts less that what is set in Absorption V. The trigger to start the absorption timer seems to be (Absorption V – 0.5 volts) so it does start the timer which seems to be 30 minutes. Then it switches to Float V, where the actual target charge voltage again has the offset (Float V – 0.5 volts) ! If you switch off TOU then the 0.5 offset is not used. Its totally baffling. Also to point out that charging on battery voltage, when it switches from Absorption to Float it seems to discharge the battery at max watts until the float voltage is achieved! When I was running of ‘Lithium’ this discharge rate seemed to be limited to 150 watts. As far as I know the charger should not actively discharge the battery between Absorption and Float! this seems like total nonsense!

    On the plus side I've got my second Deye connected and working in parallel despite the weird charging!



Advertisement