search

strigeus / sparsnas_decoder

Decoder for IKEA Sparsnäs using RTL-SDR
65 stars 37 forks source link

Problem handling drifting #8

Closed sm3ulc closed 6 years ago

sm3ulc commented 6 years ago

Have spent some hours trying to get my head a problem with a Sparsnäs not cooperating at all. Very hard to get hold of the base frequencies and when detected it only works for a short while and FreqErr is indicating drifting but not sure on which side. Sharing this in hope someone else might shed some light on the issue.

Including some data from a classical black dongle. Have also tried with a V3 with TXCO which usually is rock solid. No better luck with that one though. Tried to "pre-heat" the dongles before measuring the frequencies. Got a bit more stable results on the fq numbers but not much better decoding.

rtl_sdr -f 868000000 -s 1024000 -g 10 -d 0 - | ./sparsnas_decode Found 1 device(s): 0: Realtek, RTL2838UHIDIR, SN: 07001060

Using device 0: Generic RTL2832U OEM Found Rafael Micro R820T tuner Sampling at 1024000 S/s. Tuned to 868000000 Hz. Tuner gain set to 8.70 dB. Reading samples in async mode... [2018-04-11 15:40:05] 1235: 2223.9 W. 2.094 kWh. Batt 100%. FreqErr: 0.73 [2018-04-11 15:40:20] 1236: 2190.1 W. 2.101 kWh. Batt 100%. FreqErr: -1.35 [2018-04-11 15:40:35] 1237: 2228.2 W. 2.108 kWh. Batt 100%. FreqErr: -0.82 [2018-04-11 15:40:50] 1238: 2215.4 W. 2.116 kWh. Batt 100%. FreqErr: -0.62 ... [2018-04-11 15:47:35] 1265: 201.0 W. 2.300 kWh. Batt 100%. FreqErr: -5.89 [2018-04-11 15:47:50] 1266: 120.4 W. 2.300 kWh. Batt 100%. FreqErr: -6.46 [2018-04-11 15:48:05] 1267: 85.9 W. 2.300 kWh. Batt 100%. FreqErr: -6.53 [2018-04-11 15:48:20] Bad: 11 04 74 07 0F A2 30 19 43 D4 56 29 60 47 D0 AA C5 DC [2018-04-11 15:48:35] Bad: 11 04 75 07 0F A2 30 19 43 D4 57 2A E5 47 D0 AA C4 DC [2018-04-11 15:48:50] Bad: 11 04 76 07 0F A2 30 19 43 D4 54 2A E5 47 D0 AA C4 DC

vincic commented 6 years ago

How long did you run the --find-frequencies command? When I run it for more then a minute I got much better results and smaller FreqErr.

sm3ulc commented 6 years ago

Usually 60s. Tried just now with the TCXO-dongle with 300s pre-heat and then 300s measure. I have done some important finding that exactly the same packets are received on dongles run simultaneous, i.e. same packets are received bad on both sides and vice versa.

The Sparsnäs receiver seems to work through this "interference". Have also looked on the spectrum with a sdrplay but se no obvious problem. Hoping to get another Sparsnäs to test with.

sm3ulc commented 6 years ago

Found probable issue local interference.