@snmya that is a really good question indeed. Unfortunately I have no answer as I did never do a hardware analysis on the device. If bluetti implemented in its hardware design the corresponding necessary features, it could be possible to include this in future firmware updates. If the design of the hardware does not include this then no.
@BLUETTI_CARE: can you ask your developers if they see a possibility to tie dummy load to the charge/discharge schedule? It would be enough to disable the dummy load in discharge setting and enable dummy load in charge setting (schedule). If you want to be extra fancy: Do an additional control switch in advanced settings called “Solar autonomy mode” which will tie grid input dummy load to charge/discharge schedule in UPS settings.
Hi @ftrueck, It seems that after switching to “permanent open connection,” the issue has been resolved — is that correct?
When you mentioned, “If I check the input side (grid) and output side (load) in the Bluetti app, I see that the values for voltage and frequency are exactly the same,” — that’s expected. When using grid charging and there is an active AC load, the system enters bypass mode, and the voltage and frequency will sync exactly with the grid.
If your goal is to disconnect from the grid during non-charging periods, but still allow charging at other times, you’ll need to control the charging power by limiting the current. A display of 0W means the input is 0W, which already meets this objective.
When you mention the “actual power draw is around 90W,” could you let us know how the cables are connected, the power rating of the load, and where the power meter is located? Ideally, please record a short video, upload it to a cloud drive, and share the link with us — that will help us replicate the issue more accurately.
Currently, we do not support the use of a “dummy load” as you described.
Many thanks to snmya for the input and discussion on this matter.
@ftrueck Hello! A followup question about HA: can it work with Bluetti via WiFi, not Bluetooth? If not HA, maybe there’s another lib that can do that (like query SOC at least)? Just found out that Bluetooth module present but not accessible on my router ((
Unfortunately not. I did not find any project that uses bluetti API via web to access the values.
Unfortunately the unit is again not accessible via bluetooth. I do not know what is happening there but the unit simply refuses to react to bluetooth at all. I can not even determine specific conditions that lead to this loss of connection. It just happens randomly it seems.
I hope my drawing makes things a bit more clear.
What I tried to describe is: AC200L shows 0W grid usage while tasmota plug shows 90W power draw with PF of 0.2. That means the AC200L is drawing power from the grid but the app shows 0W. The problem I have is: the 90W are actually billed because they are Active power which count as normal power usage. It is just like plugging in a 90W light bulb for 24/7. When the bluetti app shows 0W, I have to assume that it does not draw ANY power from the grid. The reality in my case is different: It draws 90W 24/7 which is too much and not acceptable. I refered to a dummy load because I assume that is how the inverter is synched to the grid when he is in bypass mode. If this is not switchable or not even possible to disable bypass mode that is actually very unfortunately.
Hi @ftrueck, Regarding the issue of Bluetooth connection failure, please troubleshoot according to the following steps.
Here are some basic troubleshooting steps for connection issues:
Bluetooth Connection Troubleshooting:
- Check Device Settings:
○ Verify if the Bluetooth function on your device is enabled. Long press the AC+DC power buttons to check if the Bluetooth icon is displayed on the top right of the LED screen. If not, switch to the P06 page using the DC button and press the AC button to enable the Bluetooth function. - Restart the Device.
- Restart Bluetooth:
○ Turn off and on the Bluetooth function on both the device and your phone. - Offline Mode:
○ Log out of the account and try offline mode to directly connect via Bluetooth. If it connects, the issue may be with the phone signal. - Unbind and Rebind:
○ If the device is already bound, unbind it and try reconnecting. - Reinstall the App:
○ Try reinstalling the app.
As for the problem of power loss, it indeed seems to be quite high. We will conduct tests and try to reproduce the issue, and we will give you a reply next week.
When I have the Bluetooth issue my AC200L becomes unresponsive to button presses. It only starts reacting again when I disconnect PV, Grid AND the B300K that is also connected to it.
I have no idea how this makes the device unresponsive but I can not press any button on the device until I disconnect the mentioned things. Then I usually power off the AC200L and then I start it again and the issue is solved.
Regarding the power loss tests: Just to make sure it does not get lost in conversation: My AC200L has a B300K connected to it. Maybe this is relevant for the tests, but I do not know for sure.
Anyways: Thanks for your help, I highly appreciate it.
Hi @ftrueck, You are welcome, it’s my pleasure.
This kind of issue is indeed rare. We still would like to know your firmware version, as it will help us with testing and reproducing the problem. We look forward to your reply.
Hi @ftrueck, Thank you for your feedback. When the power station is connected to the grid, it will inevitably draw a small amount of electricity from the grid, which is unavoidable and something we cannot resolve.
Regarding the issue of unresponsive buttons, could you please try to reproduce the problem, record a video of it, upload the video to a cloud storage service, and then share the link with us? We would like to conduct a further analysis of this matter.
Thanks for the reply. Regarding the unresponsive buttons: The problem is I can not reproduce it. When it happens, the situation is there. If I could reproduce it easily you would already know unfortunately.
Regarding the power draw: I know that it is technically necessary. But I have a drain of 90W which is a bit high for just tracking the grid. Having said that: I know there is a relay on the grid side. I can hear it, when I restore grid power to the unit. I can see in the app that before the relay clicks, the power, voltage and amps in the grid info page are all zero. After the relay click they refresh and show actual values. So in theory there is a way to disconnect the unit from grid physically through the relay. That is what I was referring to. The idea is simple: When the time schedule says discharge, then disconnect the unit from grid via the existing relay on the grid side of the unit. If the schedule says charge: reconnect via relay. The only requirement would be that the relay can be controlled via firmware, which I hope. Then this would be possible. Additionally a setting to control this behavior and all is fine.
Hi @ftrueck, It seems that the issue of unresponsive buttons is an occasional problem, which was just a temporary freeze. Now this problem has been resolved.
Regarding the relay issue, you’re right. When the current is quite low, say 0.1/0.2A, the relay is disconnected, and it will still show 0W. A certain degree of self-consumption is unavoidable, and it does exist in the AC200L. Currently, we cannot solve this through firmware updates.
If you agree, we can also replace a unit for you. Although we don’t think replacing one will solve all problems, you can test with another unit. Please contact us via private message and provide the order information so that we can arrange it for you.
I think i figured out what the problem is. The AC inverter in all AC200L units is not working the way I was hoping. It can not handle capacitive loads correctly.
Here is how i figured it out:
- I went to my older AC200L unit that I have bought 3 years ago.
- I connected my laptop with its power supply (switch mode as this is common today) to the AC200L.
- I had set the unit to PV priority mode.
- I turned on AC on the AC200L and watched my power monitor
Here is what happened:
- Any power draw below 80W was passed through 1:1 to the grid.
- Power factor on the laptop power supply was between 0.8 and 0.99 (variations happened)
- Power factor on the grid side was between 0.1 and 0.2
- Watt draw was the same on both sides
- I then connected an air de-humidifier (compressor, so ohms load).
- The power draw on the load side of the AC200L was now 380W but with a power factor of 1.0 (laptop still was connected)
- Grid side of AC200L immediately dropped to 8W power draw
- Power factor at grid side dropped to 0.1 which I deem as normal operation and what I wanted to see
So as a conclution:
- Power factor on load side matters
- Load type on load side matters
- normal grid behavior only happens with inductive or ohms loads
I then went to my AC200L unit in the office and did more checks.
I thought: Ok, maybe it needs more load for the inverter to work correctly, so i also connected my PC to the Load side of my AC200L.
Here is what happened:
- Power draw on load side was 380W
- Power factor on load side was approx 0.8
- Grid side shows now 110W draw instead of 85W
- Power factor remains at 0.2 on grid side
- I then disconnected my PC as it only made the problem worse
- Then I added the same air de-humidifier (compressor)
- Power draw on load side of AC200L now was also 380W
- Power factor on load side now was 1.0
- Grid power draw fell to 7W
- Grid power factor was 0.1
- Then i removed the de-humidifier and connected a normal fan
- Power draw was now 195W
- Power factor was 0.85 - 0.9
- Grid power draw was 85W in comparison to 94W without fan
Conclusions:
- Same behavior on the newer unit and the older unit
- power factor on load side affects Grid side
- The worse the power factor of the Load side, the higher the power draw on the grid side
- inductive loads and ohms loads smaller than the capacitive loads do not compensate enough
- only when the inductive / ohms load is equal or above the capacitive load you have a reduction in grid power draw
So this looks like the inverter of the AC200L can not compensate the power factor of loads connected to it. In standalone mode (isolated from grid) this is no issue. As soon as you have grid connection the power draw is significant and leaves a hole in the electricity bill. The higher the load and the worse the power factor of the load, the higher the power draw on the grid side. As I tested on 2 x AC200L units that are in different age and they show exactly the same behavior it must be the inverter and as far as I can see this is not fixable in any way other than replacing the inverter in the AC200L with a better version that can handle mixed or pure capacitive loads better.
I think with my testing procedure bluetti should be able to reproduce the issue and confirm that the inverter is the issue.
Thank you to Bluetti for engaging with this post, and thank you to ftrueck for providing a simple test procedure.
.
Offsetting my PC running costs was one of the main reasons for my recent Bluetti purchases - the AC200L and Elite 200 V2 - so this topic is of great interest to me.
.
Some quick tests:
The PC setup draws about 180W at desktop, Power factor of around 0.77.
.
AC200l, 90% charge, PV priority mode (150w incoming), AC Eco mode off.
- Reports about 230W AC output, and I measure about 45W- 50W draw from the grid. (Of note, the fan was running for the full test.)
Elite 200 V2, 98% charge, PV priority mode (150w incoming), AC Eco mode off.
- Reports about 240W AC output, and I measure about 45W-50W draw from the grid while the app shows 0W (Power factor 0.15). (Of note, the fan was not running for the full test.)
So, good news, the Elite 2200 V2 is no worse than the 220L!
More seriously, I confirm ftrueck’s findings and find the Elite 200 V2 shows the same behaviour.
-
So, what does this mean to me, someone who knows little about all this power stuff?
Correct me if I’m wrong but what I see is the PC setup draws 180W, the Bluetti delivers (discharges @220W) and is drawing an additional 50W from the mains. For 180W useful power, I’m using 220 + 50 = 270W of power that will need to come from PV + grid? -
For the AC200L I could use a smart socket driven by Home Assistant to reduce the grid draw. Note: reduce and not cancel out as I’m certain there will be times I need grid + PV. Also, I’ll need to do something about the fan noise.
-
For the Elite 200 V2 the Home Assistance option is currently unknown, as I’ve been unable to connect to the V2.
-
The Home Assistant connection may not be 100% reliable, especially when disconnected from the grid. I’m investigating (something to do with sleep modes?).
-
Do I have this right? The Energy company charges for actual power; it does not charge for apparent power (actual power = apparent power * power factor). The solar battery (generator) is providing apparent power which when topped up from the grid is charged as actual power?
Looks like I need to adjust my offsetting calculations! :(
.
Thank you for confirming my tests so far.
In my case I just came up with the idea to test what happens when I allow the unit to charge while AC is on and I found another interesting behavior:
- Bluetti app shows roughly 700W grid draw
- While my power socket shows about 600W grid draw,
If I do a little Math: (698 + 81) - 153 = 626. That would come close to the real power draw but still a difference of 26W. The app shows the numbers in a very weird way anyways.
@ ftrueck you’re welcome 
Did you just find 100W, if so this is the first good news I’ve heard today!
.
On communication with the Bluettis, of course, I would like everything reliably available at all times. Not possible for various reasons, but here’s an idea for BLUETTI_CARE -
If the device is discharging and/or charging, then every time period it broadcasts an information packet.
- Time Period: about every minute would be (more than?) enough, and time accuracy is not important.
- Broadcast: I’m thinking a UDP broadcast packet. The downside of UDP is that it is not guaranteed to be received. The upsides are: UDP broadcasts are limited to the local network subnet; no open network ports. And yes, I prefer network over Bluetooth but I’m open to persuasion.
- Information: Unique device identifier, some sort of sequence number (perhaps?), all the data - though I think battery charge% would be a very good start (if % < 20 then enable smart plug regardless of time of day, and PV priority will do the rest).
For that it would be cool if bluetti would open and document their api that they use for their app. Even if it would be read only. All this info is already transferred to them and there would be the perfect place to get the infos. But i guess they will not do that because of extra traffic and potential risks. Fully opening the bluetooth access would be my preferred way.
I could imagine a little bluetooth device connected to one of the usb ports of the powerstation transmitting the infos into the wifi. Actually for the moment a RPI with bluetti_bt would probably do exactly that.