Replies: 17 comments 142 replies
-
|
Aha, no, this does seem to respect max charge/discharge current, and min soc. It doesn't respect max soc. |
Beta Was this translation helpful? Give feedback.
-
|
Ah, the Min Soc thing doesn't actually work. That's a shame. |
Beta Was this translation helpful? Give feedback.
-
|
Aha, using register 44013 (Pwr_limit PV) seems to work. When this is > pv_power then PV isn't being limited, but when it's < pv_power, PV is limited. The relationship is a bit odd, in that pv_power will happily be higher than pwr_limit PV, but the >/< relationship does seem to hold. |
Beta Was this translation helpful? Give feedback.
-
|
@abuesen, can you see whether register 44013 is accessible on your H3, please? |
Beta Was this translation helpful? Give feedback.
-
|
It is possible to read. PV-Power is zero at the moment (dark :). |
Beta Was this translation helpful? Give feedback.
-
|
Perfect thank you! Yes it's write-only, so that's expected. |
Beta Was this translation helpful? Give feedback.
-
|
Thank you for working this out! I set the remote timeout to 35s and then use a HA automation to periodically write to the power demand register as long as the SoC is within limits and my override boolean is true. This way, any HA failure or SoC out of range will revert to normal operation after 35s. And when turning on emulation mode (name is a hangover from the meter emulation): Edit: Remove lower SoC limit and fix time pattern. |
Beta Was this translation helpful? Give feedback.
-
|
I'm using Tony's HA Modbus integration hence am not able to. I've got it working anyway, unless there something I should be aware of when directing writing? |
Beta Was this translation helpful? Give feedback.
-
|
Ah right, I assumed you were using foxess_modbus, since you were asking for help in the foxess_modbus repo. Tony's integration is just a fork of StealthChesnut's old integration, and lacks many of the features which this integration was created to add. I've still got no idea why you're disabling force discharge by setting a watchdog of 5s, rather than just... turning off remote control by setting 44000 to 0. |
Beta Was this translation helpful? Give feedback.
-
|
Ah, Setting it directly to 0 will stop it right away, I'll update my automation, thank you. I'm using Tony's just cause of all the dashboards etc it gives me and don't really want to go and reset it all up. |
Beta Was this translation helpful? Give feedback.
-
|
I mean, set the "Enable" register to 0, not the watchdog register. |
Beta Was this translation helpful? Give feedback.
-
I don't know the other things you do, it seems you control the FoxESS completely yourself. Just for the EV I started exactly at the same point: I worked on controlling the FoxESS completeley myself. Anyway, here's my solution: During investigation of my mission it became possible to set the WorkMode (Register 41000) of FoxESS with foxess.modbus what was marked as "doesn't work" before. Of course you can use any other integration writing this register, too. Just as an Idea. |
Beta Was this translation helpful? Give feedback.
-
|
I was thinking about exposing this as extra Work Mode options. That seems fairly intuitive (as the Work Mode doesn't have any effect when we've got remote control enabled anyway), and it masks the fact that we need to fiddle around with Work Mode under the hood to stop a force charge. |
Beta Was this translation helpful? Give feedback.
-
|
Just catching up on this - can also confirm setting remote control power allows forced charge/discharge 😍 Agree with your suggestion of extra work modes @canton7 . This way automations can act as 'charge periods' I.e. switch work mode 'charge' between 23-5 and 'discharge' between 22-23. |
Beta Was this translation helpful? Give feedback.
-
|
It is working perfectly for my needs. 👍 I agree with the reasoning behind implementing this under work modes. |
Beta Was this translation helpful? Give feedback.
-
|
We now expose work modes for all inverters (I think?). Having said that, maybe we haven't got to the bottom of whether it's available for the H1 over lan - I'll look into that. |
Beta Was this translation helpful? Give feedback.
-
|
I haven't checked to be fair, I just looked at the compatibility table in the Wiki. |
Beta Was this translation helpful? Give feedback.
-
Beta Was this translation helpful? Give feedback.
-
|
I'm working on getting a beta out, hopefully next week sometime |
Beta Was this translation helpful? Give feedback.
-
|
Some opinions please: Charge power My current plan is to not make the charge power configurable. Instead, we'll behave the same as charge periods, where the charging rate is limited by the Max Charge Current setting. Discharge power I'm planning to have the discharge power configurable, and mean the power that the inverter feeds into the grid while the Work Mode is set to Force Discharge. Questions First, does that sound sensible, and fit how you'd want to use Force Charge/Discharge? Secondly, what should happen if:
Thanks! |
Beta Was this translation helpful? Give feedback.
-
|
Great, two people and two different opinions 😣 One limiting battery charge, one limiting house supply. We already have a way of limiting battery charge, and that's Max Charge/Discharge Current. I'm inclined to avoid adding an extra control which does the same thing (in a different unit), if only because the Force Charge/Discharge power will only apply to Force Charge/Discharge, but Max Charge/Discharge Current applies all of the time, and having two controls for the same thing in different units which apply at different times is just confusing for everyone. I think for the first release I'll avoid adding a Force Charge Power setting. If we get more requests for one, we can reach a consensus on what it should actually mean. |
Beta Was this translation helpful? Give feedback.
-
🤣 Be happy to get only two from two. If they are lawyer you'll get five from two. I wonder if there is a difference between force charge power and Max charge current? I thought the last one includes PV while the first one only respects power from grid or house (if any other generator). My 2ct opinion: If you calculate the charge/discharge power in the code it will be wrong for one or another. Needless work. So just offer this entities plain, as they are. Set a sensible default on activating force charge/discharge. |
Beta Was this translation helpful? Give feedback.
-
|
@canton7 I presume if I don't set the work mode to Force Discharge, but continue to use the write register service on a timer, I can continue to drive Active power directly? I'm in the same boat as @art103 with regard to wanting to limit supply. I use the battery charge limit now, but it's a poor proxy and means I have to set both charge period times and vary the charge limit (and make sure I don't set the Active Power register during a force charge period as this cancels the rest of the force charge period). Being able to set active power (+ve or -ve) values seems to work well. What about the new work mode being called "Direct Control" and then just have number.active_power which can be +ve or -ve for discharge or charge? We can provide some sample automations for working with Octopus flux or IO. |
Beta Was this translation helpful? Give feedback.
-
|
Actually, my plan of having the discharge current by the house discharge relies on having the load power available, which isn't easily obtainable for 3-phase inverters... |
Beta Was this translation helpful? Give feedback.
-
I like this idea too. |
Beta Was this translation helpful? Give feedback.
-
|
I'm transitioning my automations to using this and found an interesting interaction between this and the old force charge schedules. If, during a force charge schedule, you set the Active Power register, the Active Power register takes precedence, however, when the Remote Timeout expires while still within the force charge period, it doesn't revert to force change, but just drops back to the work mode. I don't think this is an issue, other than being aware that setting Active Power cancels the remainder of the force charge period. |
Beta Was this translation helpful? Give feedback.
-
|
@art103 Can you confirm that you're able to read all the following registers please:
Thanks! |
Beta Was this translation helpful? Give feedback.
-
|
44000-44002 Yes. |
Beta Was this translation helpful? Give feedback.
-
|
Early on in my playing with this I had as a one off (just using the dev tools to call a service) set Remote Timeout to 35 seconds. Today when updating another automation I discovered that Remote Timeout was much larger. I don't think my inverter has rebooted, so I wonder under what conditions it changes and also how often we should set it? @canton7 How frequently are you setting it? |
Beta Was this translation helpful? Give feedback.
-
|
I'm setting it whenever I enable remote control. |
Beta Was this translation helpful? Give feedback.
-
|
Right, give v1.10.0b1 a try, see how that works out for y'all. |
Beta Was this translation helpful? Give feedback.
-
|
ok, so ignore that - I had my max charge current still set at 1.7A |
Beta Was this translation helpful? Give feedback.
-
That shouldn't matter -- it should just cope with that. I'd still be interested in seeing your logs! |
Beta Was this translation helpful? Give feedback.
-
|
Do you still want me to set logging?, or accept that I was being stupid ;) |
Beta Was this translation helpful? Give feedback.
-
|
Setting the max charge current shouldn't break anything -- something's still wrong with my control loop. So yes, very interested in seeing your logs please! |
Beta Was this translation helpful? Give feedback.
-
|
ok 👍 |
Beta Was this translation helpful? Give feedback.
-
|
Ok, here's the logs for a couple of cycles and stripped out all of the normal 'reading registers stuff' - I set the max charge current back to 1.7A (it was left like that early this morning from my charge profile, it would normally be much higher)
|
Beta Was this translation helpful? Give feedback.
-
I've no idea to be honest. I haven't seen it in any of the register maps. |
Beta Was this translation helpful? Give feedback.
-
|
The startup voltage isn't a very good indicator as unless the strings tracker can find an effective power point you can get a reasonable voltage ~120V but with no power at all, that said I have also seen ~30W of power in full moonlight - perhaps use pv_power instead where a combined output of less than 50w is 'no sun' |
Beta Was this translation helpful? Give feedback.
-
|
The problem is, if I'm importing too much, PV power will drop to 0 (which is precisely the situation I'm trying to avoid with my little control loop). I need to know when we might be clipping PV and therefore need to activate the control loop, as distinct from the usual case of it being night so I don't need to worry and can just import at max. |
Beta Was this translation helpful? Give feedback.
-
I think that's fine here. That probably means the sun is up, so we can turn on the control loop, with the effect that the battery charge current will sit a little bit below the max (to give the PV some space to start up). I'm mainly trying to avoid the confusion of "My battery is charging at less than the power I set" from people charging at night. |
Beta Was this translation helpful? Give feedback.
-
|
that sounds fine, thanks - i'll update to b4 |
Beta Was this translation helpful? Give feedback.
-
|
v1.10.0b3 is out. I'm trying a different approach to force charge when the sun is shining, by controlling for battery charge power. |
Beta Was this translation helpful? Give feedback.
-
|
Force Charge is working well and balancing the charge between PV and grid, even when I restricted the charge current to 1.7A (~700W), where it started to export the excess solar - only a couple of hundred watts as it's not the best day for PV generation today. |
Beta Was this translation helpful? Give feedback.
-
|
actually just to qualify that, it's been one of those days with solar, lots of low clouds - it's there, it's not, its back again. |
Beta Was this translation helpful? Give feedback.
-
|
Not sure whether this is a daytime thing yet (i'll try later when there is no sun), but if I set 2kw Force charge it ignores that and follows whatever I do with the charge current - if I set that to 10A, it's charging at 4kw, is that correct ? |
Beta Was this translation helpful? Give feedback.
-
|
Glad to hear the dynamic scaling is working well. I switched approach to reading a register telling me how much charge power the inverter wants to supply the battery with, and controlling the import power so that the actual charge power is about 200W less, which seems to be a much more stable way to approach the problem. The limit is the maximum inverter input power (there's a discussion on another thread about what that should actually mean). So if you have 2kW of PV and a limit of 2kW, yes it'll charge at 4kW. |
Beta Was this translation helpful? Give feedback.
-
|
ok, yep makes perfect sense, thanks 👍 |
Beta Was this translation helpful? Give feedback.
-
|
It looks like the H1 over LAN is struggling to read the 44xxx registers, even though it seems to be able to write them, see #529. @Carl-mgb66 Are you able to use modpoll to try: Thanks! |
Beta Was this translation helpful? Give feedback.
-
Ok, I'll try again tomorrow and see if I can see anything else. Initial look suggests it complained the integration setup was taking more than 10 seconds. I get that list of versions if I turn off beta versions, with beta switched on I only see v1.10.b versions and main. |
Beta Was this translation helpful? Give feedback.
-
|
Thanks! The logs should still be there, if you click the button to download full logs. |
Beta Was this translation helpful? Give feedback.
-
|
OK, so I've tried with v.1.10.0b5 again. I had to turn on Show Beta Versions (not surprisingly) and when I did I got the following list of versions in HACS: After restart I get: But I also enabled debug logging and in the logs I get Hope that helps. For now I'll revert back to v.1.9.0 |
Beta Was this translation helpful? Give feedback.
-
|
Ah fun, that's helpful thanks! |
Beta Was this translation helpful? Give feedback.
-
|
Yeah, it looks like HACS just shows the 5 most recent versions in that drop-down (+ main). |
Beta Was this translation helpful? Give feedback.
-
|
@art103 You're using a H1 over LAN aren't you? Have you seen this issue? #529 |
Beta Was this translation helpful? Give feedback.
-
|
Aha, nice thanks. And to confirm, you're able to read that register with |
Beta Was this translation helpful? Give feedback.
-
|
No. |
Beta Was this translation helpful? Give feedback.
-
|
Sorry, I lost track of your last message! Too many threads in here... Interesting that you're able to do a 16-register read, but the numbers look utterly questionable. Right, let me add support for not having anything above 44002 available... |
Beta Was this translation helpful? Give feedback.
-
|
I confess I'm only getting enough context to reply to @art103 tags - I lost track ages ago 😄. |
Beta Was this translation helpful? Give feedback.
-
|
In |
Beta Was this translation helpful? Give feedback.
-
|
v1.10.0b6 is out. The main change is that this should unbreak the H1 over lan. @art103 / @Carl-mgb66, mind giving it a try? |
Beta Was this translation helpful? Give feedback.
-
|
Just installed v1.10.0b6 all looks OK, thanks for fixing this. |
Beta Was this translation helpful? Give feedback.
-
|
Glad to hear, thanks! You should have a force charge selector to enable force charge/discharge, and a force charge max soc input as well. That's made me think: although force discharge does respect min soc, you can't set that on a H1 lan, so I should probably make you a force discharge min soc as well. |
Beta Was this translation helpful? Give feedback.
-
|
Saw the selector etc. Probably won't be using it, but good to see it's there and all's working. |
Beta Was this translation helpful? Give feedback.
-
|
@abuesen could you give me your Manager Version please? I've got a report of someone with an H3 failing to write register 44000, and you clearly could write that, so I'm trying to work out what's going on. |
Beta Was this translation helpful? Give feedback.
-
|
I'm not at home but I think it is still the same as at January 9th: #498 (reply in thread) Master : 1.83, Slave : 1.03, Manager : 1.66 (If you add them as entites I could look the actual versions ;-P ) |
Beta Was this translation helpful? Give feedback.
-
|
Forgot it if I came back, sorry.
|
Beta Was this translation helpful? Give feedback.
-
|
The version is still 1.66. The answer of |
Beta Was this translation helpful? Give feedback.
-
|
@abuesen You said 358 here, and 385 in #509 (comment). Which of those is it? (358 is 0x166 in hex, which would match 1.66, so that would make sense). |
Beta Was this translation helpful? Give feedback.
-
|
358 is correct. 385 was a typo. |
Beta Was this translation helpful? Give feedback.
-
|
@canton7 I'm noticing a new issue. I have two 10.5kW K series setup in parallel mode.
|
Beta Was this translation helpful? Give feedback.
-
|
@canton7 Our comments crossed
So it sounds like it wraps internally rather than give an error on modbus write ;-( |
Beta Was this translation helpful? Give feedback.
-
Thanks - from the data sheet, there are 4 possible values - 6000W (rated) 6600W (Apparent) 26.1A (rated) and 28.7A (apparent). Because I'm export limited to 4850W I can't tell if the force discharge will go up to 26.1A, but I certainly have seen force charge in the past at 6.2kWish, so 26.1A * current voltage would seem like the "real" limit.
|
Beta Was this translation helpful? Give feedback.
-
|
Out of curiosity, can you force-charge at 6.2kW using Force Charge from the scheduler (rather than charge periods)? I don't think the 26.1A can be related to the charge current? It's in the "AC INPUT AND OUTPUT" section, and it's "per phase" which implies that it's for AC. It's called "Output Current", so it must be AC output. The only value there which looks relevant is "Max AC Input Power", but I know for sure that my H1-3.7 can't charge at 7.6kW! |
Beta Was this translation helpful? Give feedback.
-
No:
OK, so I wonder if when it was force charging from charge period and I was seeing 6.2kW if that is some artifact - when I was using charge periods, I would control the charge rate by changing the battery charge amps limit. I'll edit my local files and have a play tonight and get back to you (and also confirm my memory of charge period). It may be that remote control has the same limit as the scheduler. |
Beta Was this translation helpful? Give feedback.
-
|
@canton7 I just tried the scheduler, it has a maximum value of 6000 and throws an error if you try and enter 6200 - using 6000 mine shows battery charge at 5,808W If I discharge at '6000' it shows battery discharge at 6,202W ( I assume losses in the conversion through the VBus?) |
Beta Was this translation helpful? Give feedback.
-
|
I've successfully been using automations and the foxess_modbus.write_registers service for a couple of months now to force charge / discharge by writing fixed values to 44002. I'd ideally like to be able to write values to 44002 which are driven by a sensor value / value template rather than a fixed value. Is this possible? |
Beta Was this translation helpful? Give feedback.
-
|
We support force charge / discharge using the force charge / discharge work mode options. Why not use the supported way, rather than manually writing registers? |
Beta Was this translation helpful? Give feedback.
-
|
Hi canton, I'm trying to write a register using the value from another CT clamp I have in home assistant to remove one more clamp and cable run from the meter cupboard. |
Beta Was this translation helpful? Give feedback.
-
|
Cool, we provide charge / discharge rate entities you can set to set the charge / discharge rate, if you're using the supported mechanism |
Beta Was this translation helpful? Give feedback.
-
|
Thanks, that's great, I need to play about with it a bit more but I've got an automation doing roughly what I wanted: |
Beta Was this translation helpful? Give feedback.
-
I'm having a poke at the remote control registers, as a means of achieving force charge/discharge.
Notes:
Defaults when I started poking:
Remote Enable sems to be the master on/off for the whole remote control system. When set to 0, Remote Not Active Reasons is 1 (EnSet).
Remote Timeout is a watchdog, in seconds. It's set to Remote Timeout Set when you write to Active Power (but not Remote Enable, interestingly), and then Remote Timeout Countdown counts down to 0. I assume once it hits zero, remote control is disabled with Remote Not Active Reasons 2 (HostOffline).
Active Power is the inverter output, signed, in Watts. Positive is discharge. This is power going through the AC interface, so if it's 0 then any solar generation goes into the battery, while simultaneously any house load comes from the grid.
There seems to be a mix of whether these registers are available as input/holding:
So, the WR registers are available on both, but the RO registers are input-only. Makes sense. The other interesting thing is that the input registers can be read as a block, but the holding registers need to be read one-by-one.
This "registers can only be read one-by-one" thing is new, and popped up recently with the H3 in #503. I wonder how many registers we've missed because we were always scanning in blocks...
So, this looks like a viable way to control force import/export from automations (and probably more convenient to use than charge periods!). The only thing is that there's no concept of min/max soc here. The fact that it control inverter power and not import/export power is slightly annoying, but probably not too much of a problem in practice: people are going to want to import a lot, or export a lot, where that amout is much greater than generation / house load. The only thing is if we want to limit battery charge current, that is dependent on house load.
I'm not sure of the best approach to integrate this. We may want to intelligently control this (with a small watchdog timeout), and respect the min/max socs and/or min/max charge currents ourselves.
Beta Was this translation helpful? Give feedback.
All reactions