@gazzaman2k Thank you very much.
If this chart is correct, then doesn’t DIYBMS need to also display percentages less linear?
The DIYBMS chart should be more exponential shouldn’t it?
I should note I have Victron gear including a Victron shunt and BMV-700. Should I use and rely more on my BMV-700 and shunt for accurate SoC?
LiFePo4 cells have 3.2V all time, without discharge, charge. Not reponsable this to SoC. SOC correctly indicates coulombmeter, - smart shunt, BMV700 etc.
Hello,
I’m looking for 20 of the V4.4 modules. The site store seems to be out of them. Is there anywhere else that I can get them? or would getting them made in JLCPCB be the best bet with Stuarts links, files and instructions?
JLCPCB wants to have an additional charge for “two designs”. The reason they gave me is the V-cut for the thermal separation:
They said the distance to the board edge should be at least 10mm.
Just in case someone else runs into this problem…
The two designs are because of the snap off thermal sensor on the bottom left of the board, rather than the cut out.
I had 30 boards made 2 weeks ago, never raised a problem with the hole/slot cuts!!
I thought so too! but the support specifically said it was this long V-cut.
I also ordered the identical boards a few weeks ago and there was no issue. I mentioned that to the support, and they said that they will “punish the colleague” because he missed it. WTF.
So just try again?
I would say: cancel order and try again…
Has anyone an idea to get an ESP32 with LAN connection work on the controller?
I found one but it seems it is not electrical compatible. Less pins and diffrend configuration.
We have steel beton between the router and the controller.
And we do not want to set a repeater. Cable connection is more reliable.
Can someone please guide me to documentation relating to MQTT for the DIYBMXv4ESP32?
Thank you.
So I’ve been around here for awhile but I finally got around to working on my setup again and I have a problem with OOS and possibly bad CPU’s. Does anybody know how I can debug this reliably? I have desoldered and installed a lot of new cpus and I almost have an error free board but I’ll get random Oos. Is there debug code that can tell me which module communications stopped at? If communications couldn’t finish its trip?
It’s a unidirectional loop, so no there cannot be any debug code to show where comms stop, at least not when it’s completely broken.
Otherwise, well, you already have per-module packet counters and can infer from them where an intermittent error is located.
I was thinking about implementing like a timer in the code to trip if communications weren’t received in a certain amount of time, then add it to the information on the next comms trip to be recorded in the log. Seem viable?
You don’t even need a timer. Just check the sequence number for gaps.
Very likely this won’t work as there are not enough free IO pins on the ESP32 to support Ethernet adapters, at least without dropping some features.
Well, I’ve got LAN, by way of connecting the optocouplers to a RP2040 and plugging that into a Raspberry Pi (running Victron’s Venus image; no CAN bus required).
Still needs some refinement, as well as a bit of hardware design, before it’s ready for the public. If anybody would like to preview this thing, DM me.
How is performance on the RP2040 with lan? Is it via SPI or a native phy built into the RP2040?
There is no LAN on the RP2040. It’s connected to the host via MicroPython’s built-in USB-serial driver and exchanges msgpack-etized command/status messages with it.
Any web, mqtt, or other front-end stuff is the host computer’s problem. Right now I’m even tunneling the serial messages to the cell modules to the host and process them there, because debugging that code is a lot easier on a “real” computer; obviously that’s on my “fixme ASAP” list. As is connecting a better ADC setup (i.e. less noisy and less input current) than the RP2040’s built-in one, for voltage and current measurement.
Damn…just when I touched up my crappy handsoldering with a proper soldering station from my son…i managed to let the magic smoke come out of capacitor c4 on my controller board. I obtained some 4.4 boards which had allready been assembled bar the attiny’s. I had the communication leads attached to 1 module before i attached the battery connection (@3.86v) all other components seem ok, do you guys think it ok to just solder in a new c4 component?
I have had no luck programming a shunt yet…i have a arduino uno connected to my laptop and connected to the shunt. But i no succes yet, tomorrow I try again
So I just replaced CPU’s #0,1,6,10,11
No more crc errors, now only OOS.
Also I’m using the 9k6 comms speed.
