TGBT41A_912_22 - MR tests
Goals
Tests
Blanking time, messages
- Frequency switch calculations: MR_Mode10SP
- 1.0 secs is ok, 0.9 triggers it - blanking should be 2x38ms + 2x10.4ms - so this is correct.
- Config tool correctly reports only one message for each Bank.
nchan tests
- SB: MR_NchanModeTest. Combine different combinations of nchan, subband, etc. All works as expected
key / value tests
- SB: MR_NchanModeTest. Combine different combinations of upper and lower case.
- Keyword are case independent. After some work, so is the value "bank", but not "beam".
- correctly catches unknown keywords, and fails to validate correctly
Add "subband" to the restfreq dictionary.
Remove if3freq offsets
- Create scheduling block MR_IF3FREQ
- Scan 6 - 32768 channels, reduce to 4096
- Scan 7 - 4096 channels
- Spectral resolution is 5.7 kHz, so put in test tone at 9000011 kHz = 9000.011 MHz.
- Scan 8 - mistyped test tone
- Scan 9 - no sign of the test tone
- Scan 10 - put it at 9005.0 MHz to see if I see it at all. Mistyped again!
- Scan 11 - still don't see it - Fsky if offset. - switch to topo (although that shouldn't be it?)
- Scan 12 - still no line!! Could it be getting dwarfed by 3000K system temperature? I wouldn't have thought so, but maybe
- I was looking at the noise source!!! Reconfigure for Bank B.
- Scan 13 - see it easily.
- Scan 14 9000.011 MHz, don't see it. Maybe too close.
- Scan 15 - 9005.85938 - I don't see anything. Forgot the A-ToneB
- Scan 16 - Excatly in channel 1024 as expected. But did this mode have an offset? I think it was only modes 1,2,3
- Scan 17 - Mode 1 no test tone - as expected
- Scan 18 - 9375MHz - appears at 9375.0 / 256 - as expected. This means no IF3FREQ offset.
- Scan 19 - MR_DeltaFreqTest. Configure a 9000.0MHz test tone, and deltafreqs of:
restfreq = [ {"restfreq": 9000.0, "deltafreq": -9.375, "bank":"B"},
{"restfreq": 9000.0, "deltafreq": -7.03125, "bank":"B"},
{"restfreq": 9000.0, "deltafreq": -4.6875, "bank":"B"},
{"restfreq": 9000.0, "deltafreq": -2.34375, "bank":"B"},
{"restfreq": 9000.0, "deltafreq": 2.34375, "bank":"B"},
{"restfreq": 9000.0, "deltafreq": 4.6875, "bank":"B"},
{"restfreq": 9000.0, "deltafreq": 7.03125, "bank":"B"},
{"restfreq": 9000.0, "deltafreq": 9.375, "bank":"B"}]
- Lines come out in channels: 410, 826, 1242, 1626, 2458, 2874, 3290, 3674
- Frequencies: 8.999995 (much less than a channel width) in each case
- Had hoped for 409.6 and multiples thereof, but seem to get:
- 410, 413*2, 414*3, 406*4, 409*6, 410*7, 411*8, 408*9.
- Why not 410 in every case?
- try a smaller, whole multiple, i.e (23.4375 / 4096) * 128 = 0.732422
- Scan 20 - test for Melinda
- Scan 21 - test for Melinda
- Scan 22 - above test - forgot A-Tone B again!
- Line is at: 1536, 1664, 1792, 1920, 2176, 2304, 2432, 2560
- which is 12, 13, 14, 15, 17, 18, 19, 20 time 128.
- Thus they are appearing exactly where they should, to the channel.
- LO2? is set to 13250.
Conclusions
- Everything tests correctly, apart from Beam 1 case sensitivity, which is an old issue (nothing to do with these changes).
- We should keep this on the list of things to do.