Badger Orion Endpoint GIF2014W-OSE not working with -R 223

[ddffnn]

My city water meter RF module was recently replaced with an "Orion Endpoint."
I had been able to receive and decode readings from the old module using -R 223, but that's no longer the case.

I'd like some suggestions for reverse engineering the data I'm receiving from the new module because I believe it's quite different from the old one that worked with -R 223.

This is the new module:

I'm quite confident that I've been able to detect the correct pulses using rtl_433 -f 915.8M -s 1M -M level -R 0 -X 'name=badger,modulation=FSK_PCM,short=10,long=10,reset=1000'.
I have a cable going out to my meter pit with a magnetic antenna base right next to the RF module, but with no antenna attached.
I'm seeing over 30dB SNR so it's gotta be the meter pulses I'm catching right?

It seems that while there are transmissions every 20s or so, the payload only updates once per hour, on the hour.
This is based on known water usage throughout the day.

I've looked at the comments in the source code for protocol 223, but I think the protocol has changed.

I've attached a complete CSV from the above command, but it's got a bunch of other device transmissions.
I've also attached a filtered (regex '76317631.*b8531') version that I believe only has my meter readings.

filtered.csv

complete.csv.gz

Thank you in advance!

[zuckschwerdt]

The codes look promising but not like we expect for the Badger ORION. The start looks like a sync word of 7631 7631 or d8c5 d8c5 or 273a 273a or 89ce 89ce (shift and invert however you like…) and the end looks like what could be a CRC-16. But alignment is not clear to me and I can't see if there is a 3of6 coding present?

[ddffnn]

Yes, I agree on all those points. Also, it's longer than we would expect. I've been trying different antennas, frequencies, and dongles, but I've never seen my water meter using 223. I kept thinking I'm doing something wrong, but it sure seems like the protocol had just changed.

I plan to read the analog display of the encoder today. It used to match with the old module so I can only assume it will with this one also.

[ddffnn]

I looked at the meter today to read it directly and realized it's new too. The old meter was mechanical, but this new one is ultrasonic with a nine-digit digital display. It alternates showing total volume and instantaneous rate.

Anyway, I read the meter manually three times, and we only used water during a few periods so there were only 7 unique payloads.

{273}555555555555557631763175f08ecd76b6a526f438c99d3c11ab85576716ec3e9fa70
{274}555555555555557631763175f08ecd76b6a526f438c99d3e91ab85576716ec3e90570
{272}555555555555557631763175f08ecd76b6a526f438c99d2791ab85576716ec3e8336  [4614.36]
{277}555555555555557631763175f08ecd76b6a526f438c99d0711ab85576716ec3e8c0420 [4620.64]
{275}555555555555557631763175f08ecd76b6a526f438c99d0110ab85576716ec3e89420
{273}555555555555557631763175f08ecd76b6a526f438c99d0e902b85576716ec3ee2508 [4695.78]
{275}555555555555557631763175f08ecd76b6a526f438c99d56102b85576716ec3eac658

I messed around in BitBench but I'm starting to wonder if there is some 3of6 still in use or something similar. I can't find the serial number or get increasing values, let alone values that match my manual readings.

[ddffnn]

@nickovs Since you did the work for protocol 223, do you happen to have any suggestions for this one?