There should not be 60 volts hot-to-ground. If the voltage is real (and not an artifact due to how the measurement was taken), I’d ask Bluetti for replacement. When I initially tested my first AC300 and measured similar hot-to-ground voltage, I contacted Bluetti support, and Bluetti was going to replace my unit. Then I realized my measurement was false (because I had a receptacle tester inserted in one of the AC output outlets, after removing the receptacle tester, hot-to-ground voltage was basically zero) and I did not end up replacing that unit.
Why are you switching over to Standard UPS at night? Wouldn’t the PV Priority act the same at night…if not enough solar…it switches to GRID? Sorry if this is a basic question and I am just not understanding the difference of the modes.
.
I have a AC300/2xB300 on PV only…not even plugged in to the GRID for the last 2+ months. But this week (11/13/23) has been the first time I have had 2-3 days of rain/clouds…although sun peeked out, along with cloud charging got to about 86%…combined with mild temps and not needing AC or heater, and still starting out today at 69%. That is better than yesterday’s start…and the sun may be out more today.
.
But, I am sure 3+ days of clouds, and higher demand is around the corner. I have the GRID ready to plug in…have seen one person put a WIFI controller on the GRID power, which I plan to do as well. I have my unit already set to UPS set to PV Priority, 10%" for if and when I do plug it in.
.
Thanks!
An ARM update fixed the older AC300 SN2138000218536 screen issue.
This unit now demonstrates hot to ground multimeter readings of 5v-72v/9v-76v no grid/grid (low-high results based on probe position).
The newer 223000180501 AC300 unit now demonstrates hot to ground 4v-59v/9v-117v readings no grid/grid and as previously noted tripped GFCI when a neutral plug was inserted (unsafe for planned hookup to transfer switch as you explained).
Updated Multimeter testing chart after ARM firmware update:
bluetti electrical.pptx
One B300 battery remains unable to demonstrate that it updated with eitherAC300, reporting “no new firmware” as pictured in prior post.
Residual issues:
- Need to replace the newer AC300 with a unit that has a floating neutral, due to incompatibility with the home integration transfer switch if charging thru the grid while outputting to the transfer switch.
- Do hot to ground results suggest abnormalities with both AC300 units requiring new units (I was very careful with the last round of testing-no other AC outlets were being used, I recorded the highest and lowest reproducible results)?
- 1/4 B300 units not updating properly.
*As per your recommendation I will test any AC300 unit attached to GFCI prior to connecting to the transfer box.
You have already helped define/resolve significant electrical issues and educate in the process. Of course additional insight with my residual issues is welcome.
Good luck with your setup.
Thanks again for helping with mine.
You are welcome. For your low-high hot-to-ground voltage measurements (such as 5-72 volts), does the measured voltage hold steady (such as staying at 72 volts for several seconds, or is it only a very short transient voltage that quickly drops near zero? On my AC300, as I move the probe positions, there is some small transient voltage fluctuation that quickly disappears.
As for the issue of your oldest B300 not receiving the same BMS firmware update as your newer B300 units, this could be normal. I think Bluetti makes their units with different hardware versions, such that what’s the latest firmware version available for one unit isn’t necessarily the same as that of a different unit. The B300 units don’t have serial numbers that can be retrieved via the app or screen. The serial number is on a sticker at the bottom of each unit. Maybe given the serial number of the oldest B300, Bluetti would be able to confirm the latest BMS firmware version available for that unit.
Heres a video of my hot-ground testing from today: 20231115_210953.mp4
As you can see the results are reasonably stable (and reproducible as todays results are essentially identical to my table of testing performed 2 days ago).
Thanks for your advice regarding the B300. I suspect you are correct that the B300 that won’t update may not need updating as it is significantly older than the other 3 B300’s I own and may have different hardware… Although I have learned not to be too quick to assume model year correlates with alternate electrical circuitry/function(i.e. the AC300 OLDER-NEWER MODEL NON-FLOATING VS, FLOATING NEUTRAL HYPOTHESIS which now appears unsupportable as you noted).
In your video, it looks like the 10-volt measurement is due to poor contact with the probe and does not reflect the actual voltage (even though the reading was stable). It looks like the actual voltage is 120 volts (as shown later in your video). To make sure I’m getting good reading (for sockets that are challenging to probe), sometimes I would plug an extension cord into the outlet (without plugging in anything at the other end of the extension cord). Then I might either probe the other end of the extension cord or pull out the extension cord’s plug very slightly (so that the plug’s connectors are barely exposed) and probe the plug’s exposed connectors.
Retested with an extension cord for hot to ground and results were consistent,
bluetti extension cord hot to ground.pptx
The retested voltages seem strange to me. The voltages have varied substantially (and with some strange values) to make me wonder if there could be a problem with the multimeter.
If you do try the “pull out inserted plug slightly and probe the exposed contacts” approach, be very careful. Depending on the length of metal part of your probe, you might accidentally touch the metal probe to two exposed contacts at the same time or accidentally touch the metal probes to each other, resulting in shorting the exposed contacts. Make sure you have good lighting (so you can clearly see where the probes are relative to the three exposed contacts), go slow, and be very steady with the probes.
I won’t do that testing at this time.
Were you planning on using grid charging with your split phase setup or are you just using solar"?
Have you changed your plans as a result of the “old vs new non-floating vs floating” hypothesis having been challenged?
This was the Bluetti-Care response:
BLUETTI_CARE
November 16
@bluetti_barrylzimmermanhotmailcom
bluetti_barrylzimmermanhotmailcom:
One B300 battery remains unable to demonstrate that it updated with either AC300, reporting “no new firmware” as pictured in prior post.
We checked the firmware version of all battery packs, and all of them are the latest ones. If the battery is shipped from the factory at different times, the latest version of the firmware will be inconsistent. Please don’t worry.
bluetti_barrylzimmermanhotmailcom:
This unit now demonstrates hot to ground multimeter readings of 5v-72v/9v-76v no grid/grid.
The newer 223000180501 AC300 unit now demonstrates hot to ground 4v-59v/9v-117v readings no grid/grid and as previously noted tripped GFCI when a neutral plug was inserted(unsafe for planned hookup to transfer switch).
The ground wire is floating. Therefore, the voltage data measured with a multimeter is not accurate, which is normal. The tripping problem may be due to the GFCI’s ability to detect ground, and you may need a ground-neutral bonded plug.
My plan hasn’t changed (split-phase with grid connected). When the replacement for my second AC300 arrives, I’ll probably first try the Pro/Tran 2 with one or two circuits before doing the full wiring.
The Bluetti-Care response (the varying hot-to-ground voltages are expected because of floating ground, GFCI tripping means you might need neutral-ground bonding plug) seems wrong to me. Your video showed hot-to-ground voltage of 120 volts; that is not a floating ground. As for the suggestion to fix the GFCI trip problem by using neutral-ground bonding plug, you already are using such a plug, and the plug is what actually exposed the problem. When you wire AC300’s AC output to the Pro/Tran 2, it will have the same effect as the neutral-ground bonding plug, and plugging in a neutral-ground bonding plug won’t fix the GFCI trip (because neutral and ground will already be bonded together through the Pro/Tran 2 connected to the main service panel, which is where the neutral-ground bonding exists). In fact, plugging a neutral-ground bonding plug while connected to Pro/Tran 2 would be unsafe because there would be two neutral-ground bonding points, resulting in current flowing via the ground wire.
Did you get email response from the Bluetti technical staff? I had good experiences with that avenue so far (have not used Bluetti-Care).
Bluetti_care recommended replacing the OLDER unit. Bluetti service.has been in contact, I await an updated response from them.
It does seem that only the newer unit has “floating neutral” circuitry in view of :
- GFCI tripping only after neutral plug insertion into the newer, but not older AC300.
- Receptacle testing on the newer unit showing “open ground” without grid and “normal circuitry” with grid attachment, older unit shows open ground on and off grid.
*Maybe multimeter testing is unreliable in view of my inexperience
Your newer unit does not have “floating neutral”; if neutral is floating, the receptacle tester would show “open ground” and not “normal circuitry.” A unit that trips GFCI is unsafe; the GFCI is meant to protect you. If AC300’s AC output exhibits “normal circuitry” (neutral and ground bonded together) and you connect the AC output to Pro/Tran 2, you would end up with two points where neutral and ground are bonded together, resulting in current flowing over the ground path (unsafe).
I wonder if Bluetti at one point fixed the neutral bonding problem but later started making units with the problem again. I received the replacement for my second AC300. The replacement is a brand new unit with newer serial number yet demonstrates the same problem. Of the four AC300 units I’ve tested so far, the oldest and the newest (in terms of serial numbers) are the problematic units. The middle two units are the ones that don’t have this problem.
The receptacle tester does read “open ground” when the AC300/B300 is tested as a standalone unit. It only reads “normal circuitry” when connected to a “neutral (to ground) plug” or the grid (thru the main panel which is neutral to ground bonded). Wouldn’t a ground plugged newer unit appropriately trip a GFCI outlet? Thanks in advance for the upcoming explanation of how my logic is backwards .
Are you asking whether the tripping of GFCI by the newer unit (when neutral-ground bonding plug is inserted) is correct/desired behavior? The answer depends on how the unit is intended to be used. If the unit will always be connected to grid and the unit will not be used with a non-neutral-switching transfer switch (such as Pro/Tran 2), then the behavior would be a good thing (because there is only one point where the neutral-ground bonding occurs, and the neutral-ground bonding is achieved without needing to add neutral-ground bonding plug).
However, if the unit is to be used with grid connected and also a non-neutral-switching transfer switch connected, then the behavior would be undesirable (because there would be two points where neutral-to-ground bonding exists, resulting in current flowing on the ground connector during normal operation). During normal operation, no current should flow on the ground connector. The ground connector is intended to carry current only when a short fault occurs.
Yes, I am postulating that the GFCI tripping of the newer unit with neutral-ground bonding plug insertion is desired behavior. It is consistent with the newer unit- transfer switch being compatible by virtue of their shared floating neutral circuitry.
I am considering switching to an AC500 based hybrid system, hoping for more consistent circuitry between 2 units of the same model.
What is your next move?
1 Like
It could be that the newer unit’s behavior is intended by Bluetti, but they don’t seem to have provided this information to their support staff. They previously responded to you that the ground wire is floating (to explain the varying voltages you measured), which seemed to indicate that they expected the floating-ground behavior. I also asked Bluetti technical staff multiple times about the floating ground/neutral behavior (even when grid power is connected) I observed with my first AC300 and was told that the behavior is normal. I sent video to Bluetti showing the GFCI tripping and 120V hot-to-ground voltage of the new replacement unit they sent me, and they offered to do another replacement.
I wired up a few circuits (including some AFCI breakers) in my Pro/Tran 2 310A and decided to test both AC300 units for real before doing any more replacement. I connected each unit individually (single phase) to GFCI grid input and the 310A. Both AC300 units (even my first one that did not exhibit the GFCI tripping with neutral-ground bonding plug) tripped the GFCI input as soon as the AC300 output is connected to the 310A (I tried two different GFCI outlets, both tripped), which was disappointing. I then repeated the same procedure but with non-GFCI grid input. The good news is that neither AC300 burned up, and I was able to go through repeated cycles of charging the batteries from grid input up to 3000 watts, stop charging the battery, and start charging again with grid input while maintaining power to the 310A’s output loads (and the more sensitive AFCI breakers I added didn’t trip either). I then proceeded with full split-phase grid-connected operation with the 310A, and that worked too (despite the two AC300 units exhibiting different neutral-bonding behaviors).
The testing did show a problem with the older AC300. If the grid input power is disconnected and then restored, often the older AC300 will exhibit “H-Bus Voltage High” fault and turn off the AC output (after clearing the fault, then the AC output can be enabled again). This problem does not occur with the newer AC300. So I’m now intending to replace my first AC300 (instead of replacing the newer AC300).
Today I temporarily disconnected grid input power to my older AC300 (while the unit was driving Pro/Tran 2 310A transfer switch in single-phase mode). When I reconnected grid input power, there was a loud pop and some smoke. The unit remained on (and continued to charge the batteries with PV input power), but AC output could no longer be enabled (each time I attempt to enable AC output, the unit will fail with over current protection failure even if all the AC output outlets are empty). I think this might be the “burned up” failure some people previously encountered. I’m returning the unit for replacement.
1 Like
Unexpectedly my journey with the AC300 came to an end. AC300 did not work out for my use case (two AC300 units in split-phase configuration simultaneously connected to grid input power and driving Pro/Tran2 transfer switch), but I wish that all vendors would stand behind their products like Bluetti did (continued responses to email queries spanning multiple months, 5 replacements, and full return/refund). At no time did Bluetti suggest that there must be something wrong with my environment or equipment. Hopefully the following information will be of use to others.
I encountered three types of AC300 units.
- AC300 units that can exhibit “H-Bus Voltage High” fault or pop/smoke failures when connected to Pro/Tran2 and the grid input power is lost and then restored (I’ll refer to these units as “Type 1”)
- AC300 units that don’t exhibit the above failures. When connected to grid input power, the AC output measures 0V between hot and ground (I’ll refer to these units as “Type 2”).
- AC300 units that don’t exhibit the above failures. When connected to grid input power, the AC output measures 120V between hot and ground (I’ll refer to these units as “Type 3”).
Clearly the “Type 1” units are unsafe (even in individual single-phase configuration) to be connected to both grid input power and Pro/Tran2 at the same time. Bluetti replaced these units upon request.
When used in individual single-phase configuration, both “Type 2” and “Type 3” units performed 24/7 without problems, with both grid input power and Pro/Tran2 connected at the same time.
The unresolved failure (which led to the eventual full return/refund) was combining two AC300 units in split-phase mode, with both split-phase grid input power and Pro/Tran2 connected at the same time. The configuration works reliably when driving 120V circuits. But when powering my level 2 EVSE, the AC output would randomly turn off for a few seconds and then turn back on (each AC300 showed 2500-2900W load, no fault/warning reported), even though grid input power is stable. The random AC output power off/on occurrences were too disruptive to power critical circuits.
Because my split-phase configuration consisted of a “Type 2” unit and a “Type 3” unit, the suspicion was that one of these two types is incorrect. Changing which unit is master vs. slave did not resolve the problem. Bluetti initially indicated that the “Type 3” units are incorrect and replaced the unit. Later, Bluetti indicated that the “Type 2” units are incorrect and that “Type 3” units are correct. Unfortunately each time a unit is returned and a replacement unit sent, the replacement unit ended up being either the same “type” or being a defective “Type 1” unit. Despite going through 5 return/replacements, I could not obtain a matched pair to confirm whether having two units of the same type would resolve the random AC output power off/on occurrences. At this point, I requested return of both units, and Bluetti refunded as requested.
In response to the topic of this thread (using AC300 to drive Pro/Tran2 loads full time with PV connected full time), my experience is that it worked fine (even with grid power connected full time in UPS mode) for powering 120V circuits. I’ll note also two others learnings from the experience.
- I underestimated the effect of idle power consumption. With the grid input power connected, I did observe about 50W power consumption (as others noted) with AC output on but driving no load. The 50W idle power consumption is 1.2 kWh per day (with two AC300 units in split-phase mode, 2.4 kWh per day), which probably explains why my actual power saving was less than I expected.
- I was initially surprised to find that I could disconnect both the neutral and ground connections between AC300 and Pro/Tran2, and the AC300 could still power the loads in the Pro/Tran2 as long as the grid power input is connected to the AC300. I think the AC300 is a common-neutral inverter (the grid input neutral is shared with the AC output neutral), which means the return current (on both the grid input side and the AC output side) could flow over either (or both) paths back to the main service panel. At first it feels surprising that the grid input return current could flow via the AC output neutral connector (and the AC output return current could flow via the grid input neutral connector), but it makes sense that with the two parallel neutral paths, electricity does not care which path is labeled “input” vs. “output.” I think this is the reason why connecting AC300 to GFCI grid input and Pro/Tran2 always trips the GFCI (because the GFCI detects that the current on the hot connector does not equal the current on the neutral connector).
1 Like
That is quite the journey you’ve taken! I’m sad to hear that the only resolution for you was to return the AC300 systems, will you consider the AC500 instead? I understand that these issues you had have been resolved in that unit (AFAIK).
I have one of the initial type of AC300s, where you cannot connect input and output to the house/grid neutral (my first unit was damaged due to that, and replaced by Bluetti. I’ve never used grid charging since until unplugged from my transfer switch). So I wonder if it’s possible to use grid power to charge if I only used the hot wire and no neutral in a custom power cable. I don’t think I need to, but it’s an interesting question.