DIYBMS v4

[kiwihacker (Marc N)]

Hi all.

Thanks again to Stuart and anyone else contributing to the project. I have just got my test-bed up and running and all seems to work well. Hand soldering worked well for me though I think I’ll have to give the fry-pan method a go when I get time to do another batch.

To anyone looking to source parts/boards for this project in New Zealand (or possibly Aussie as well), msg me as I have enough extra parts to make quite a few boards. I hate getting ripped off so ordered in quantities to get decent pricing.
I am happy to offer them at as close to cost plus shipping that I can (without losing money myself). I have to work out exactly what that is, but I’m sure it’s much better than trying to source say quantities of 10 for ATTINY/Voltage ref/etc.

Msg me if you’re interested. I am happy to offer to those further abroad if no one from my neck of the woods gets back to me (and you’re willing to pay for postage).

Thanks,
kiwihacker

[neil_l12 (Neil L)]

Very happy I’ve managed to build my first few modules, its been in interesting journey with surface components. A few times the BMS boards wouldn’t boot into normal mode and this turned out to be a dry joint somewhere, solved this by blowing the rework air gun over the entire board to remove any traces of solder paste or flux, re-flowing all the joints solved the problem.

Rework station cost £40, best buy ever.

Target is 20 BMS boards, 14 for my powerwall and the rest for testing/spares.

[stuart]

Nice work

[lanzecki (Simon)]

Mark, Try connecting them one at a time. Bet you’ll find the faulty one. Check the Opto-isolator, on any you find faulty. Is it the right way around?

[Jonisonveapa (Jon)]

hello
i have 10 attiny chip for sale if anybody need them ~?

[lanzecki (Simon)]

@stuart

The serial protocol has a CRC check in it to prevent issues with corrupt packets being received.

That been said, I have also experienced the issue you mentioned but that’s normally when I’m randomly unplugging and reconnecting the modules, sometimes one needs a nudge!

Thinking this is a code error rather than hardware issue.

Hi Stuart,

Any thoughts about this? I’m having to reset random BMS’ half a dozen times in the evening to keep them working

Interestingly (Or possibly not), when charging, particularly when the BMS’ are balancing they will not lock up. Discharging doesn’t seem to make a difference.

I did disconnect each BMS for a day to see if there was any difference (reducing my battery down to 6s). Still I note the lock up tends to be (but not always) at the front of the BMS’s (IE BMS 0, 1, and 2).

As I said previously I have the influx DB logs going back weeks if they are of any use?

Thanks

[stuart]

@lanzecki

Interestingly (Or possibly not), when charging, particularly when the BMS’ are balancing they will not lock up. Discharging doesn’t seem to make a difference.

When they lock up is the green LED fully lit up (not flashing) ?

[lanzecki (Simon)]

Until This weekend just gone yes, I had left it for a couple of weeks powered up, and restarting the BMS at fault when it happened.

This weekend I’ve been doing some load and discharge tests, so the BMS’ have been working instead of just monitoring.

I did notice some occasions where they locked with the LED off.

I only noticed this as the http: monitor would complain.

[stuart]

@lanzecki

I did notice some occasions where they locked with the LED off.

Ok, diagnosing and fault finding intermittent issues is always challenging.

I suspect the module is hanging whilst waiting for data to arrive or in the middle of a data receive and is stuck thinking more data is going to arrive - only a guess at the moment. Need to review the code.

[lanzecki (Simon)]

If you would I’d really welcome it.

Anything I can do to assist?

[stuart]

Some what ironically, I’ve tried for an hour to get a module to hang and it won’t !

When do your modules hang during charging or discharging ?

[lanzecki (Simon)]

Hi Stuart, thanks for the quick consideration of my issues. I do recognise that you don’t have to, and aprechiate the effort you’ve put into all versions of your BMS, making it available to us, and providing support.

I’ve had 7s pack (just 1P) during testing. I wonder if that’s potentially the cause?

I charged the 7s pack and left it all plugged in for a day or so. Andafter that I started to take some mha from the pack (down to 4.0V). That’s when it started locking up.

I then left the pack alone, connected to the BMS’ and controller, feeding data to the database. I did this for 2-3 weeks. Randomly BMS’ would lock up.

A few days ago, I removed the last BMS to reduce the Pack to a 6S pack (Still waiting for a charger to arrive that will cope with 7s.

Once down to 6s, I started charging, and noticed the BMS balancing, and not locking up. It did lock up only once during charging. I did a small discharge and don’t recall a lock up, but I only discharged for a few hours, and at a very low current.

Once charged closer to full (About 4.1V per Series). I left the pack alone for a while. I was getting lock up’s as usual.

Generally I find that it’s the first 3 BMS’ that lock up. Be it one or two. Never more then 2. Less often it’s cell 5 or 6.

I cannot tell you or a definite way to get it to lock up.

Note that I’ve used the pre compiled code. I work in IT, and wasn’t able to work my way thru the ‘manual’ way to upload it. I will admit that I didn’t break by arse to compile the code as there was precompiled code available.

This generates a question. Is the pre-compiled code an older version then the code that requires compiling? Maybe I should bite the bullet and work out how to do it with the un-complied code… Problem is I have no idea where to start, and cannot find an idiots guide.

Thanks.
Simon

[stuart]

@lanzecki

Is the pre-compiled code an older version then the code that requires compiling?

That code is the current version.

I’ve never had the modules lock up during normal usage, on a couple of occasions I’ve had a module hang with the green light on - normally during me plugging in cables or randomly unplugging stuff.

I think this happens at the exact same time as the module is trying to take a voltage reading - it puts the LED on, starts the reading process but never finishes. The ATMEL chip hasn’t hung or the in-built watchdog timer (8 seconds) would have rebooted the chip.

I’m running a 4S1P test environment on these at the moment and left them running. Internally the code does keep a count of any corrupt data packets that the modules receive, I’ll add some code to expose this on the web page just to ensure its not a data corruption issue.

[lanzecki (Simon)]

Thanks Stuart,

You are a star. Bet you will have to wait a while though

[stuart]

@lanzecki

You are a star. Bet you will have to wait a while though

I’ve left the BMS running for a week to try and catch a lock up. I finally got 1 module to hang at some point yesterday.

What I noticed is that the module before the one that locked up received a corrupt data packet (the only 1 in a week of running). I’m thinking that this packet was then passed down the chain and the next module to receive it locked.

More investigation to do, but at least I’ve got a train of thought to work with.

[Edoardo_Conti]

Hi, great job Stuart. Congrats!

I’ve been reading about this project for a while and I decided to build my own. I’ve been meaning to buy the PCB from jlcpcb with the “SMT Assembly” service (for both save money and lack of right equipment).

Unfortunately it requires two additional files besides the Gerber file:

• BOM
• CPL (Pick&Place File)

The first one is available thanks to Stuart but I didn’t find CPL anywhere.
Is it possibile to get it?

[super2222 (Steve Desrochers)]

there you go just make shure everything is ok
v4 edit.csv (3.9 KB) v4-top-pos.csv (1.7 KB)

[Edoardo_Conti]

Thank you so much!

Unfortunately there are several components unfound. Do I have to search for alternative components with specs as close as possibile to the original ones?