June 26, 20242 yr Author On 2024/06/25 at 2:14 PM, Cslling said: Would it be possible to share this even its abit old please. On 2024/06/11 at 10:37 AM, Stas said: Hello, Can you please share the program Please find link Regards Edited June 26, 20242 yr by ZS1SCI
July 18, 20241 yr Author Good day Update time ! XD Deye was kind enough to send me the LVESS software. This means I can reverse engineer the application and remove RS485 cell monitoring from my project all together (since it requires a RS485 USB device per battery). If we plug our analyser into the INTER-CAN BUS we can get all the information for all the cells in the parallel pack. This would then only require one CAN device reader, hopefully not the pricey CANALYST-II that I got to be able to work with this software. If all goes well home assistant support via ESPhome would be my first prize , but for now I'll poke around and find the correct commands to send to be able to gather the data from any arbitrary CAN device. Huzzah Ps. I will upload this software to my GitHub profile, you may find the software there sometime later today.
July 25, 20241 yr Can you please share this 1.16 LVESS software (privately ok) ? I guess it is working over some USB2CAN adapter.
July 27, 20241 yr Author On 2024/07/25 at 11:18 AM, Sfinx said: Can you please share this 1.16 LVESS software (privately ok) ? I guess it is working over some USB2CAN adapter. I would gladly share it publicly, as Deye isn't very concerned with my battery balance issue. To quote their email. Quote Our battery balancing strategy is passive balancing, which will be activated when the voltage difference between individual battery cells reaches 30mv and it is charging... This is bad since my one pack is now at about 150mV when fully charged. So no keeping the BMS at constant voltage to have other cells catch up ;/ Here is a sample of BMS in Voltage mode at 55.4V Anyhow link in signature Huzzah Edited July 27, 20241 yr by ZS1SCI would moderator move topic to batteries please...
July 29, 20241 yr On 2024/07/27 at 7:22 AM, ZS1SCI said: This is bad since my one pack is now at about 150mV when fully charged. I'm pretty sure you can return this battery to them. They claimed 10 years of warranty. BTW: Still struggling in connecting this CAN app to battery using Canable slcan adapter. The app just says it "Open can failed". My USB-CAN device is present at COM16. But this stupid app do not have the COM port selection dialog. The china soft is shit.
July 30, 20241 yr Author Mmmmh I also have slcan device, the software doesn't work with it ongelukkig... You can use can monitor 3000 and check the extended frame ids, there is lots of info there, when I'm at my pc I'll share the xml for the app with the addresses ive mapped. Regarding the battery, they(deye) claim 200mV is normal operating condition and not a concern, what bugs me is the bms charge voltage of 58.4V, that seems somewhat excessive, I'd prefer 56.8V but this cannot be changed in software... I've asked for updated fw with 56.8V but they keep on claiming this is fine without really caring that the 58.4V is outside of their specified operating voltage range. Considering the resistance I get from asking for a lower charging voltage and lack of control on the BMS. We might be better suited rolling our own. Please se Henry's response on cell difference Cheers
July 30, 20241 yr Author Just adding here Constant voltage mode(56.4V) does not allow the BMS to balance cells. I've tried all sorts of current limiting and CV to try and balance pack2 to somewhere in the normal <50mV. changing any alarm and protection settings voids the warranty ;/ Edited August 6, 20241 yr by ZS1SCI life
July 30, 20241 yr Author On 2024/07/29 at 7:32 PM, Sfinx said: I'm pretty sure you can return this battery to them. They claimed 10 years of warranty. BTW: Still struggling in connecting this CAN app to battery using Canable slcan adapter. The app just says it "Open can failed". My USB-CAN device is present at COM16. But this stupid app do not have the COM port selection dialog. The china soft is shit. For intercan bus comms(it has more cell information) check the wiring its different than PCS CAN. (IN/OUT) ports Then you need to download cangaroo app from CANABLE site which links to github This will Initialise the device first for use in CAN Monitor 3000. Start measurement (F5) Set Bitrate to `250000` -> OK You'll see messages pop up in cangaroo, then close the app and fire up CAN Monitor 3000. File => Load Tree => open InterCan Bus.xml Selecting SL-Can/ Lawicel => set to listen only and 250 => open You should see the cell values without the proprietary software. Hope it helps Cheers InterCan Bus.xml Edited July 30, 20241 yr by ZS1SCI grammar slammer
August 4, 20241 yr Are you able to connect over PCS CAN bus port ? I'm able to see the batteries over IN port but not the PCS one
August 5, 20241 yr Author Hey No I have not been able to connect with PCS over the Deye software... I just got a bump in the right direction from `Adminius` over on GitHub. Using ESPhome on ESP32 + MCP2515, we can now get all PCS data and cell information via two CAN cables. MCP2515 => 250K for InterCAN bus ESP_CAN => 500K for PCS bus For use in automations in home assistant Just finalising the `yaml` for ESPHome then I will commit. Cheers Edited August 5, 20241 yr by ZS1SCI
August 5, 20241 yr Author 15 hours ago, Sfinx said: Are you able to connect over PCS CAN bus port ? I'm able to see the batteries over IN port but not the PCS one This might also be that PCS is at 500K and not 250K like InterCAN
August 6, 20241 yr Author Good day For those interested I made a new repository for esphome using an ESP32+MCP2515 to listen to both CAN busses on these batteries. https://github.com/Psynosaur/esphome-deye-batteries Huzzah
August 6, 20241 yr On 2024/08/05 at 10:01 AM, ZS1SCI said: This might also be that PCS is at 500K and not 250K like InterCAN Thanks, seems like brain dead deye s/w is not able to use properly its own can bus protocols
October 20, 20241 yr @ZS1SCI Hello, I'm trying to get information from my Deye SE-G5.1 Pro battery but have had no luck so far. Perhaps you can help me. I bought USB2CAN adapter (China version described in https://github.com/tixiv/CAN-Monitor-qt) connected it to CANIN port and tried to use CANMonitor with XML files from your repo https://github.com/Psynosaur/Deye_Battery_Serial_BUS/tree/main/CAN2JSON/CANInformation I'm getting some data, but that data can't be decoded. I would appreciate it if you can help me.
October 21, 20241 yr Author 22 hours ago, Jcomp said: @ZS1SCI Hello, I'm trying to get information from my Deye SE-G5.1 Pro battery but have had no luck so far. Perhaps you can help me. I bought USB2CAN adapter (China version described in https://github.com/tixiv/CAN-Monitor-qt) connected it to CANIN port and tried to use CANMonitor with XML files from your repo https://github.com/Psynosaur/Deye_Battery_Serial_BUS/tree/main/CAN2JSON/CANInformation I'm getting some data, but that data can't be decoded. I would appreciate it if you can help me. Hi Looking at you screenshot it seems you're connected to the InterCAN port at 250K. You should connect to the PCS port at 500k to be able to see data for this template, wiring is different I hope this helps PS. your welcome to try the home assistant version of the logger https://github.com/Psynosaur/esphome-deye-bms Edited October 21, 20241 yr by ZS1SCI
October 21, 20241 yr @ZS1SCI Thank you very much, now I finally got some data from my battery. What I'm really interested in is individual cell voltages which are not available via PCS as far as I can understand. I've changed the wiring to pins 1-2 and used InterCAN port at 250K with InterCAN_Bus.xml from this post. Got some data but not Cell information. Is it possible to somehow get Cell information? P.S. esphome-deye-bms sounds interesting, but I'm a newbie in all this stuff: ESP32, MCP2515, and HomeAssistance sound like "magic" keywords to me right now 🙂
October 21, 20241 yr Author 1 hour ago, Jcomp said: @ZS1SCI Thank you very much, now I finally got some data from my battery. What I'm really interested in is individual cell voltages which are not available via PCS as far as I can understand. I've changed the wiring to pins 1-2 and used InterCAN port at 250K with InterCAN_Bus.xml from this post. Got some data but not Cell information. Is it possible to somehow get Cell information? P.S. esphome-deye-bms sounds interesting, but I'm a newbie in all this stuff: ESP32, MCP2515, and HomeAssistance sound like "magic" keywords to me right now 🙂 Hi, see if you can find a lot of frames starting with either 0B or 0C or 0D depending on SOC of pack It might be that your battery has other frame ID's for them For only one battery I would expect these frame ID's E 4008001 E 4018001 E 4028001 E 4038001 Perhaps see what you can find and report back
October 21, 20241 yr I'm getting very few frames and only three of them change raw data over time Nothing starts with either 0B or 0C or 0D. I've tried another program that came with USB2CAN adapter and the result is the same: Could the reason be that my Deye BMS firmware is pretty "old" - 2004? Edited October 21, 20241 yr by Jcomp
October 22, 20241 yr Author 8 hours ago, Jcomp said: I'm getting very few frames and only three of them change raw data over time Nothing starts with either 0B or 0C or 0D. I've tried another program that came with USB2CAN adapter and the result is the same: Could the reason be that my Deye BMS firmware is pretty "old" - 2004? Good morning Yes indeed there was a lot added with newer FW: Here is a changelog from 2004 to 2007, I am currently on 2008. 1. Modify the filter time of battery empty calibration; 2. Add the function of reading historical data and historical events on the 485 serial port host computer; 3. Optimize SOC, SOH calculation 4. Optimize the host to request forced charging from PCS; 5. Add the 485 serial port upgrade function; 6. Optimize the AFE initialization timing and speed up the power-on speed; 7. Optimize the record of upgrade events; 8. Add AFE temperature collection and reporting, event recording; 9. Add fault light language strategy; 10. Modify the full charge request voltage to 3.65*16=58.4V; 11. Optimize PCS converging strategy Please find latest FW attached, this should make the InterCAN bus more chatty Huzzah LVESS0224626N02_2018.bin
October 22, 20241 yr Thanks for sharing the changelog. > 10. Modify the full charge request voltage to 3.65*16=58.4V; This is what I'm trying to escape, I don't want to get an overvoltage alarm and I believe 3.6 is more than enough for LiFePO4. The dilemma: upgrade and face a potential issue with overvoltage or live with the current firmware and inability to read cell voltage data/connect to BMS with LVESS software. And there is no easy way to downgrade battery firmware, right?
October 22, 20241 yr Author What inverter are you running it with? Im using ESP in the middle to alter the message for charge voltage. We can downgrade if we had 2004 fw. Deye didn't want to provide the old binary as it was too buggy... 2004 had 57.6V charge voltage request The behaviour of the BMS is to set soc to 100% when any cell hits 3.6V and then switch off the charge mosfet until 99% soc is hit, then allowing charging again. With 58.4V this is sped up quite a bit and imbalanced cells reach this quickly whilst others are still 3.44V. With charge mosfet off after this there is no equalization time. Anyways it's not like the balancer is up to the task, I'm adding a Neeeeeeeeey 4th generation balancer to each of my packs to fix the imbalance and void the warranty 😁 Anyways good luck deciding This was at 55.2V charge voltage request Edited October 23, 20241 yr by ZS1SCI picz
October 24, 20241 yr On 7/30/2024 at 7:54 PM, ZS1SCI said: I've tried all sorts of current limiting and CV to try and balance pack2 to somewhere in the normal <50mV. changing any alarm and protection settings voids the warranty ;/ Have you tried to use equalization function? ( which is basically for AGM but you can used it for lithium if you modified the voltage ) so this feature will allow charging voltage without current so it could Triger balancing for enough time
October 24, 20241 yr Author 1 hour ago, esmail-kassir said: Have you tried to use equalization function? ( which is basically for AGM but you can used it for lithium if you modified the voltage ) so this feature will allow charging voltage without current so it could Triger balancing for enough time Hi Lol yes here is the balancer in action... I do not believe it is a 200mA balancer, considering it can only balancer 4 cells per time that would probably be more like a 50mA balancer(200/4). Intervention is required.
October 24, 20241 yr On 2024/10/22 at 8:52 PM, ZS1SCI said: What inverter are you running it with? Im using ESP in the middle to alter the message for charge voltage. We can downgrade if we had 2004 fw. Deye didn't want to provide the old binary as it was too buggy... 2004 had 57.6V charge voltage request The behaviour of the BMS is to set soc to 100% when any cell hits 3.6V and then switch off the charge mosfet until 99% soc is hit, then allowing charging again. With 58.4V this is sped up quite a bit and imbalanced cells reach this quickly whilst others are still 3.44V. With charge mosfet off after this there is no equalization time. Anyways it's not like the balancer is up to the task, I'm adding a Neeeeeeeeey 4th generation balancer to each of my packs to fix the imbalance and void the warranty 😁 Anyways good luck deciding Hello ZS1SCI, Thanks for info, I'm running Deye sun-5k-sg03lp1-eu inverter. Looks like it's time to buy ESP and my Neeeeeeey balancer is already on the way from China 🙂
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.