Nkunzi

NO WAY!

Hi all I'm able to read values (QPIGS) but get NAK when I try and do eg. POP02. Could it be the firmware is read-only? Many thanks.

@Steve87 yeah if you're interested on your recommendation I tried Caprica Solar for a repair and got pretty bad service. They said they could repair, great, I paid upfront, naive bunny that I am. Then I could just not get any updates out of them. A month came and went, then suddenly they could not repair it anymore. For some reason it would take days to get my money back. After 2 calls to the owner I finally got my money back, pleasantly surprized by this actually. This company is best avoided in my opinion.

The Pi is rock solid without any extra grounding while using a power source somehow from mains. Have you got a link for the buck power supply you use, please, also what pin does it feed from on the RJ45 if that's not too much trouble? Thanks

So I've tried a battery bank's USB output into the Pi, pretty sure enough power. When I'm powering the pi via mains (say USB port from a PC) it's all good, also using a PC with a built in serial port, also via mains, data is all good. My normal operation ie not on the test bed is with the inverter AC output, onto a USB cell phone charger, and then into the pi's power usb. Both the stand alone battery bank and from the inverter AC output options give me gibberish. I get what you're saying, there is a ground as part of the RS232 connection, but I'm guessing maybe it floats up a bit via the USB adaptor or something. Anyhoo, it's not tooooo bad, I can get enough data out for sensible charts, I might try bypassing the USB adaptor sometime like I say. If I can get back to my original question, are you still seeing no usage if not on bypass? For me it looks like when on battery I see all usage, even low, but when on mains it seems to be only above a certain usage. I guess I'll just live with it as well.

@Logic can I pick your brain on something? I'm reading the serial port with an R Pi + usb to serial device. When I power the Pi off mains all is good. When I power it with a battery, the data back goes all over the show, just rubbish data. It helps a bit if I ground the Pi to the inverter (I use that little earth bolt at the bottom of the chassis). Have you experienced anything like this with your Arduino, any tips and tricks? I can get the data in most of the time by checking against the CRC but I have to hit it a few times, typically, which I do every minute but like I say it usually takes a few tries. I could do what you did, straight off the Pi's GPIO serial port + the 'Max' voltage converter, might try that sometime.

https://www.youtube.com/watch?v=PdMWBv8kAo0&t=25s&ab_channel=WowThings OK so it looks helluva hazardous dealing with that lead. I would do it with a bit more protective gloves and what not. I was thinking casting thin lead sheets, and rolling an anode and cathode in a cylinder form, maybe something like fibreglass sheet between them to provide insulation. I'd then put these in some kind of tube (PVC won't work, apparently not resistant to sulphuric acid) and join them up in series. I think these guys coat the lead plates with lead sulfate and lead dioxide, that seems to me will be the trickiest part. I'm guessing just using lead and sulphuric acid and charging it will cause these compounds to form by themselves, but I don't know if enough of it will form to give you a working battery. Anyone have any ideas?

Hmm good thought, I don't see anything that jumps out though, looks more like below a certain use. It started showing again a little bit while I'm working on the PC, guess I'm using the physical hard drive so it jumps up a bit. Thanks bud, I do know how to use a soldering iron kind of thing but I'll first try my luck with my vendor, hopefully still in warranty. While I've got you, what do you make of this charging current? Lead acid x2 12v, supposedly 120AH each. I got the feeling it's not giving enough amps for long enough, I bumped up bulk to 30V and float to 29 to try and get more amps for longer, but doesn't look like enough to my untrained eye at least. I've got the unpatched 00043.05 I can try your patch see if it's any better. Muchos gracias

Seems to be something to that theory, I was on shedding last night 10 to 00. Don't know why I'm getting something in the day, maybe when usage is above a certain level even though I'm on bypass (should be on bypass all the time unless shedding). Doesn't seem to be the best inverter in the world, does it? 😂

Thought I'd ask a new question here, specifically for @Logic because you have the same model and we've been talking about it in another thread. What I'm seeing is I'm often getting 0 apparent and active power reported by the inverter. I'm still testing it right now so it's on my desk and it's got a PC and few peripherals on it so not drawing a lot of power. I do sometimes get it showing some values, but mostly 0. Here's my rough Grafana chart of it. I'm speculating it's not registering anything under some value, maybe 100W? Curious if you see similar behaviour, do you ever use yours on fairly low power use and it's still reporting values? Many thanks.

No worries thanks! If you can get the principle right and are willing to share your findings, that will be great. If I do need this I'll try and implement on Pi / Python anyway. Thanks for your help.

Hmm OK. I'm not very clear on what the bug actually does or how the inverter's bulk and float voltages play into the inverter's algorithm for delivering bulk then float charging. I'd presume it has a time parameter of its own and stays (if it is working right) at the higher voltage and a high current, then drops down to the lower float voltage. But let's say it does have the bug, charging current is then low, i.e. lower than mcur, so the first IF will not catch it and do nothing. Let's assume float and bulk are different, so the second IF does not catch it either. So nothing at all happens. I'm thinking time based might be the only way to cater for all possible states. Voltage might be high even though it hasn't spent enough time in bulk and you'd not be able to know it by just looking at settings and charge current at any given time. What do you reckon?

Nice setup. Bonus points for the Class A Private IP, not just Class C like me What did you build the web page on, just a dark theme and custom PHP or is it some kind of framework? Does it run on your ESP? I do know the ned-kelly repo, I've been plagiarizing https://github.com/JosefKrieglstein/AxpertControl/blob/master/axpert_tmp.py for comms to the inverter with Python which works OK for me. I've got Mosquitto and Node Red running on Pi in the house, all still very basic. Hmm, whole thing seems a bit tricky. I don't have experience with this problem, but let me try think it through. What do you assume starting conditions to be, say it's discharged and now it needs to bulk charge. So say you're on 22, bulk is 26 and float is 24, mcur is 10A. Let's say it's not suffering from the bug, charge is higher than mcur so you don't do anything. If charge current is < mcur though, that's when you have to set float and bulk the same and you do that. Now float and bulk's the same, so it should kick into bulk mode. Charge current increases. How long must it bulk charge? For a certain time or until it hits another voltage? Either way, once it hits that time or voltage, I presume now you set float lower again and it drops from bulk to float mode. Float is now lower, and charge current is also low, as it's supposed to be. However, you need to not kick it into bulk again. If you just do a comparison again, charge current is low and float is lower than bulk, you'll set bulk and float equal again which is not right. How about a time based flag that it does not do this again, maybe? Slight tangent, can you actually read the settings for float and bulk from the inverter, or do you just keep it as a value in a variable? I looked at commands https://forums.aeva.asn.au/uploads/293/HS_MS_MSX_RS232_Protocol_20140822_after_current_upgrade.pdf and QPIGS does not seem to give it, nor any other query. Thanks

6 months sitting on a rack waiting for my finances to afford batteries So can you set charging, float, etc on the fly, in principle at least? As I was telling @Coulomb when I try PBFT etc I get NAK. You also mentioned ESP32 to try alleviate your problems, have you got something that uses the correct RS232 voltages on an ESP board or do you have to run it through a voltage converter? Just curious. I've got a Pi zero for logging metrics, still hoping I don't need to mess around with setting voltages to solve the floating charge bug, still too early to say if I really suffer from that.

Yeah, it's worrying that it's so far off. Bit of background, I bought the inverter beginning of the year and it was never used, was saving up for batteries. So a bit of an anticlamax that everything is not exactly perfect.