This is the repo I used for the Leaf modules.
I see what you are saying. I had thought as the comms cables had been working for a long time that that couldn’t be the issue but as the problem occurred when I plugged in the comms cable it seem like that is likely the problem. I guess when I moved it two exposed parts that weren’t touching came in to contact with each other and when I then connected it back up it had created a dead short. I will go and check all of my cables with a multimeter and visually to see if I can see the issue.
i have the original version and they are in the github i just checked, there seems to be a v2 version of the leaf boards now mine are the original version
@hotear you added a pack that was previously separated and on a second controller board. My first guess is RX from the last connected pack is connected to RX of the already connected packs. So basically somewhere you have either RX to RX or TX to TX connected. Sometimes else to look at is your buck converter, most of these devices have the negative in and out connected this result in a high potential when TX is connected to a TX or rx to rx there will not be a insulation if that occurs.
Sorry to hear about the problems Nick.
I’m not sure what the answer to you’re problem is, I’ve not seen others report such dramatic failures.
I have reports of damage to modules when people connect the Rx and Rx connections together, this would create a short.
No worries Stuart. I have some avenues to look down now so hopefully it’s one of those. I might video my next attempt (when my new Wemos D1 mini arrive later today) just so if it happens again I have something to show everything I have checked. Every time I lose a cell module it fills me with dread that I might have to solder another ATTINY (they are just so small), I find it quite stressful.
In my humble opinion, the number of errors is not critical, but the response time except in an installation of many modules 50 or more is high, how many modules do you have?
The error rate percentage over 800,000 packets isn’t that great.
I’m surprised that you are getting crc errors though, that is most likely corruption on the data cable from nearby electronic devices.
Are you using twisted cables?
Yes i am using twisted cables and short distances.
Ok that’s good.
Do you have a screen shot of the modules page?
OSS errors are “out of sequence” which means that the controller was expecting message number “X” to turn up and it didn’t - “Y” turned up instead, so we lost a message/reply.
Given that the “bad packet count” is zero for all the modules - you have an issue with CRC errors and OOS errors from the last module back to the controller.
How does that look for cabling ?
I’m not saying this would definitely work, but you could consider putting a ferrite ring on the last cable to the controller - something like this (just put the cable through it and wrap around a few times). Make sure you order them with a suitable internal hole size, if you want to try this.
I’m trying to place an order for v4.40 modules, but the CMFB103F3950FANT thermistors are out of stock. Does anyone knows for a replacement that can be used?
@SeRiusRod if you want one or two days, the OEM shop will have kits of the complete V4.40 boards available, including all the JST connectors, ATTINY chips (which are in short supply) shop link
1 Like
So the last cable looks like this :
The orange cable is twisted (that’s the way it’s made) but i did add a small portion to both ends which is not, unsure whether that might be the issue ? The cable is in total 50cm since it as to get to the controller on the other side of the pack. Maybe it’s a combination of long cable and not twisting the last part. In the other end i also have a simmilar cable but it’s not close to other cables (that could make potentiall noise ?)
Its difficult to say, but it does look like noise is getting into the cables on the last connection into the controller. Is this one of the newer controller boards?
No the controller is the old one.
c95968 es ok for thermistor R19 y r23
2 Likes