Calibration does not seem to lock in

[dtheriault]

Setup is OSC feeding 135Mhz signal directly to dongle @ -40dBm. Not using a mixer.+

rtl_hpsdr -i 192.168.2.23 -a 1 -r 1 -C 135000000

Launched cudaSDR, looks like initially I am off -14.460Kc. Once things settle down, we seem locked, but then rtl_hpsdr bounces around.

[09:13:49] cal update, rcvr 1 old offset -48 new offset +0 new freq 134985754 [09:13:54] cal update, rcvr 1 old offset +0 new offset -48 new freq 134985706 [09:13:58] cal update, rcvr 1 old offset -48 new offset +16 new freq 134985770 [09:14:02] cal update, rcvr 1 old offset +16 new offset -96 new freq 134985658 [09:14:07] cal update, rcvr 1 old offset -96 new offset +0 new freq 134985754 [09:14:15] cal update, rcvr 1 old offset +0 new offset -16 new freq 134985738 [09:15:29] cal update, rcvr 1 old offset -16 new offset +16 new freq 134985770 [09:15:33] cal update, rcvr 1 old offset +16 new offset -48 new freq 134985706 [09:15:37] cal update, rcvr 1 old offset -48 new offset +48 new freq 134985802 [09:15:41] cal update, rcvr 1 old offset +48 new offset -48 new freq 134985706 [09:15:46] cal update, rcvr 1 old offset -48 new offset +64 new freq 134985818 [09:15:50] cal update, rcvr 1 old offset +64 new offset -48 new freq 134985706 [09:15:54] cal update, rcvr 1 old offset -48 new offset +64 new freq 134985818 [09:15:59] cal update, rcvr 1 old offset +64 new offset -48 new freq 134985706

Don't understand the 14.4kc offset, thinking that's an artifact from cudaSDR.

[n1gp]

I saw the bouncing back and forth about a bin size 24khz, but not since adding an additional check for that in the cal routine. Yours is bouncing at +-96Khz every 4-5 minutes. Head scratch...

I haven't tried this on an ARM board yet, just the Intel i7. I can have the cal enabled and I still get spots. I'm sure Skimmer looses some in between the freq changes. Are you able to spot and cal on the ARM?

[dtheriault]

Will check once I get past compile issue. I updated all my workspaces this morning with your latest commits but get compile bug. Unmatched {} believe. Just have to find it.

---

-dt-

On Feb 23, 2015, at 10:05, n1gp notifications@github.com wrote:

I saw the bouncing back and forth about a bin size 24khz, but not since adding an additional check for that in the cal routine. Yours is bouncing at +-96Khz every 4-5 minutes. Head scratch...

I haven't tried this on an ARM board yet, just the Intel i7. I can have the cal enabled and I still get spots. I'm sure Skimmer looses some in between the freq changes. Are you able to spot and cal on the ARM?

— Reply to this email directly or view it on GitHub.

[dtheriault]

I am seeing spots... but not very often. Still off freq, but close, maybe too high by 300Hz.

Revealing myself as a Hermes version 2.6 rcvr. My IP Address: 192.168.2.24 My MAC Address: 00:1e:06:c2:0d:a2 Was discovered by 192.168.2.3 Sent discovery acknowledgement. Setting hpsdr output rate to 96000 hz Requested 7 Activated 7 actual rcvr(s) Set rcvr 5 to freq 10000000 with offset 0 Set rcvr 6 to freq 10000000 with offset 0 Set rcvr 7 to freq 10000000 with offset 0 Received Start command Set rcvr 1 to freq 10000000 with offset 0 Set rcvr 1 to freq 7045500 with offset 0 Set rcvr 2 to freq 10000000 with offset 0 Set rcvr 2 to freq 10145500 with offset 0 Set rcvr 3 to freq 10000000 with offset 0 Set rcvr 3 to freq 14045500 with offset 0 Set rcvr 4 to freq 10000000 with offset 0 Set rcvr 4 to freq 18113500 with offset 0 Set rcvr 5 to freq 21045500 with offset 0 Set rcvr 6 to freq 24935500 with offset 0 Set rcvr 7 to freq 28045500 with offset 0 [13:21:43] cal update, rcvr 2 old offset +0 new offset -2440 new freq 10143060 [13:21:43] cal update, rcvr 1 old offset +0 new offset -2976 new freq 7042524 [13:21:44] cal update, rcvr 3 old offset +0 new offset -2880 new freq 14042620 [13:21:44] cal update, rcvr 4 old offset +0 new offset -2792 new freq 18110708 [13:21:44] cal update, rcvr 7 old offset +0 new offset -2880 new freq 28042620 [13:21:44] cal update, rcvr 6 old offset +0 new offset -2976 new freq 24932524 [13:21:53] cal update, rcvr 7 old offset -2880 new offset -2832 new freq 28042668 [13:22:06] cal update, rcvr 7 old offset -2832 new offset -2792 new freq 28042708 [13:22:10] cal update, rcvr 4 old offset -2792 new offset -2744 new freq 18110756 [13:22:26] cal update, rcvr 2 old offset -2440 new offset -2392 new freq 10143108 [13:22:31] cal update, rcvr 7 old offset -2792 new offset -2744 new freq 28042756 [13:22:40] cal update, rcvr 3 old offset -2880 new offset -2832 new freq 14042668 [13:24:06] cal update, rcvr 4 old offset -2744 new offset -2696 new freq 18110804 [13:24:40] cal update, rcvr 2 old offset -2392 new offset -2344 new freq 10143156 [13:24:45] cal update, rcvr 3 old offset -2832 new offset -2792 new freq 14042708 [13:25:33] cal update, rcvr 7 old offset -2744 new offset -2696 new freq 28042804 [13:36:15] cal update, rcvr 7 old offset -2696 new offset -2648 new freq 28042852 [13:36:40] cal update, rcvr 3 old offset -2792 new offset -2832 new freq 14042668 [13:36:53] cal update, rcvr 7 old offset -2648 new offset -2600 new freq 28042900 [13:37:15] cal update, rcvr 7 old offset -2600 new offset -2648 new freq 28042852 [13:37:24] cal update, rcvr 6 old offset -2976 new offset -2928 new freq 24932572 [13:37:28] cal update, rcvr 6 old offset -2928 new offset -2976 new freq 24932524 [13:37:58] cal update, rcvr 7 old offset -2648 new offset -2696 new freq 28042804 [13:38:46] cal update, rcvr 7 old offset -2696 new offset -2648 new freq 28042852 [13:39:07] cal update, rcvr 7 old offset -2648 new offset -2696 new freq 28042804

Got this when I first started w/o skimmer running.

Revealing myself as a Hermes version 2.6 rcvr. My IP Address: 192.168.2.24 My MAC Address: 00:1e:06:c2:0d:a2 Was discovered by 192.168.2.3 Sent discovery acknowledgement. Setting hpsdr output rate to 96000 hz Requested 7 Activated 7 actual rcvr(s) Set rcvr 1 to freq 10000000 with offset 0 Set rcvr 2 to freq 10000000 with offset 0 Set rcvr 3 to freq 10000000 with offset 0 Set rcvr 4 to freq 10000000 with offset 0 Received Start command [13:16:59] cal update, rcvr 2 old offset +0 new offset -2488 new freq 9997512 [13:16:59] cal update, rcvr 3 old offset +0 new offset -3000 new freq 9997000 [13:16:59] cal update, rcvr 6 old offset +0 new offset -3000 new freq -3000 [13:16:59] cal update, rcvr 4 old offset +0 new offset -2904 new freq 9997096 [13:17:00] cal update, rcvr 5 old offset +0 new offset -2928 new freq -2928 [13:17:00] cal update, rcvr 1 old offset +0 new offset -3000 new freq 9997000 [13:17:03] cal update, rcvr 3 old offset -3000 new offset -2928 new freq 9997072 [13:17:04] cal update, rcvr 5 old offset -2928 new offset -2856 new freq -2856 [13:17:08] cal update, rcvr 5 old offset -2856 new offset -2792 new freq -2792 [13:17:09] cal update, rcvr 7 old offset +0 new offset -3000 new freq -3000 [13:17:12] cal update, rcvr 3 old offset -2928 new offset -3000 new freq 9997000 [13:17:13] cal update, rcvr 5 old offset -2792 new offset -2744 new freq -2744 [13:17:13] cal update, rcvr 1 old offset -3000 new offset -2952 new freq 9997048 [13:17:17] cal update, rcvr 4 old offset -2904 new offset -2856 new freq 9997144 [13:17:17] cal update, rcvr 5 old offset -2744 new offset -2696 new freq -2696 [13:17:17] cal update, rcvr 1 old offset -2952 new offset -3000 new freq 9997000 [13:17:21] cal update, rcvr 5 old offset -2696 new offset -2624 new freq -2624 [13:17:26] cal update, rcvr 5 old offset -2624 new offset -2576 new freq -2576 [13:17:30] cal update, rcvr 5 old offset -2576 new offset -2504 new freq -2504 [13:17:34] cal update, rcvr 5 old offset -2504 new offset -2464 new freq -2464 [13:17:39] cal update, rcvr 5 old offset -2464 new offset -2392 new freq -2392 [13:17:43] cal update, rcvr 5 old offset -2392 new offset -2344 new freq -2344 [13:17:43] cal update, rcvr 7 old offset -3000 new offset -2952 new freq -2952 [13:17:47] cal update, rcvr 5 old offset -2344 new offset -2296 new freq -2296 [13:17:51] cal update, rcvr 3 old offset -3000 new offset -2952 new freq 9997048 [13:17:51] cal update, rcvr 4 old offset -2856 new offset -2816 new freq 9997184 [13:17:51] cal update, rcvr 5 old offset -2296 new offset -2248 new freq -2248 [13:17:56] cal update, rcvr 5 old offset -2248 new offset -2200 new freq -2200 [13:18:00] cal update, rcvr 5 old offset -2200 new offset -2136 new freq -2136 [13:18:09] cal update, rcvr 5 old offset -2136 new offset -2064 new freq -2064 [13:18:09] cal update, rcvr 7 old offset -2952 new offset -2904 new freq -2904 [13:18:12] cal update, rcvr 2 old offset -2488 new offset -2440 new freq 9997560 [13:18:17] cal update, rcvr 5 old offset -2064 new offset -2016 new freq -2016 [13:18:25] cal update, rcvr 3 old offset -2952 new offset -2904 new freq 9997096 [13:18:26] cal update, rcvr 5 old offset -2016 new offset -1968 new freq -1968 [13:18:35] cal update, rcvr 1 old offset -3000 new offset -2952 new freq 9997048 [13:18:39] cal update, rcvr 1 old offset -2952 new offset -3000 new freq 9997000 [13:18:39] cal update, rcvr 7 old offset -2904 new offset -2856 new freq -2856 [13:18:43] cal update, rcvr 4 old offset -2816 new offset -2768 new freq 9997232 [13:18:47] cal update, rcvr 5 old offset -1968 new offset -2016 new freq -2016

[dtheriault]

And sometimes when running, I never see it enter the calibrate mode.

Revealing myself as a Hermes version 2.6 rcvr. My IP Address: 192.168.2.24 My MAC Address: 00:1e:06:c2:0d:a2 Was discovered by 192.168.2.3 Sent discovery acknowledgement. Setting hpsdr output rate to 96000 hz Requested 7 Activated 7 actual rcvr(s) Set rcvr 1 to freq 10000000 with offset 0 Set rcvr 2 to freq 10000000 with offset 0 Set rcvr 3 to freq 10000000 with offset 0 Set rcvr 4 to freq 10000000 with offset 0 Set rcvr 5 to freq 10000000 with offset 0 Set rcvr 6 to freq 10000000 with offset 0 Set rcvr 7 to freq 10000000 with offset 0 Received Start command Set rcvr 1 to freq 7045500 with offset 0 Set rcvr 2 to freq 10145500 with offset 0 Set rcvr 3 to freq 14045500 with offset 0 Set rcvr 4 to freq 18113500 with offset 0 Set rcvr 5 to freq 21045500 with offset 0 Set rcvr 6 to freq 24935500 with offset 0 Set rcvr 7 to freq 28045500 with offset 0

I'm getting spots, but were about 3kc too low. This is starting with

rtl_hpsdr -i 192.168.2.24 -a 1 -r 7 -C 125000000 -l 125000000

I'll try wwv

[n1gp]

I just checked in some changes hopefully to fix the cal not working sometimes and the flip flop issue. And also added s check for weak cal signal with warning.

[n1gp]

The recent checkin has been working well for me tonight on my OC1. I have CAL and 7 dongles enabled at 48Khz. Spots are coming in at ~ 100Hz or less.

I added a bin field as I want to make sure your bin size is the same as mine.

[12:10:15] cal update, rcvr 3 old offset -1852 new offset -1828 new freq 3566422 bin 23 [12:13:57] cal update, rcvr 5 old offset -1266 new offset -1172 new freq 10121578 bin 23 [12:14:01] cal update, rcvr 5 old offset -1172 new offset -1289 new freq 10121461 bin 23 [12:26:27] cal update, rcvr 1 old offset +1429 new offset +1312 new freq 1824062 bin 23 [12:26:36] cal update, rcvr 1 old offset +1312 new offset +1406 new freq 1824156 bin 23

[dtheriault]

Ok, my BW is set to 96Khz in skimmer. Ran with Cal as 125000000. Got this run...

[21:45:57] cal update, rcvr 4 old offset +0 new offset -2696 new freq 10142804 bin 23 [21:45:58] cal update, rcvr 3 old offset +0 new offset -2832 new freq 7042668 bin 23 [21:45:58] cal update, rcvr 2 old offset +0 new offset -2272 new freq 3543228 bin 23 [21:45:59] cal update, rcvr 6 old offset +0 new offset -2904 new freq 18110596 bin 23 [21:45:59] cal update, rcvr 7 old offset +0 new offset -2880 new freq 21042620 bin 23 [21:45:59] cal update, rcvr 5 old offset +0 new offset -2792 new freq 14042708 bin 23 [21:46:03] cal update, rcvr 3 old offset -2832 new offset -2880 new freq 7042620 bin 23 [21:46:03] cal update, rcvr 7 old offset -2880 new offset -2856 new freq 21042644 bin 23 [21:46:04] cal update, rcvr 5 old offset -2792 new offset -2768 new freq 14042732 bin 23 [21:46:07] cal update, rcvr 4 old offset -2696 new offset -2672 new freq 10142828 bin 23 [21:46:07] cal update, rcvr 3 old offset -2880 new offset -2976 new freq 7042524 bin 23 [21:46:08] cal update, rcvr 7 old offset -2856 new offset -2832 new freq 21042668 bin 23 [21:46:08] cal update, rcvr 5 old offset -2768 new offset -2744 new freq 14042756 bin 23 [21:46:12] cal update, rcvr 3 old offset -2976 new offset -3000 new freq 7042500 bin 23 [21:46:13] cal update, rcvr 5 old offset -2744 new offset -2720 new freq 14042780 bin 23 [21:46:17] cal update, rcvr 5 old offset -2720 new offset -2696 new freq 14042804 bin 23 [21:46:20] cal update, rcvr 1 old offset +0 new offset -2976 new freq 1842524 bin 23 [21:46:22] cal update, rcvr 5 old offset -2696 new offset -2672 new freq 14042828 bin 23 [21:46:25] cal update, rcvr 3 old offset -3000 new offset -2952 new freq 7042548 bin 23 [21:46:25] cal update, rcvr 2 old offset -2272 new offset -2296 new freq 3543204 bin 23 [21:46:26] cal update, rcvr 7 old offset -2832 new offset -2792 new freq 21042708 bin 23 [21:46:26] cal update, rcvr 5 old offset -2672 new offset -2648 new freq 14042852 bin 23 [21:46:30] cal update, rcvr 3 old offset -2952 new offset -2928 new freq 7042572 bin 23 [21:46:31] cal update, rcvr 5 old offset -2648 new offset -2624 new freq 14042876 bin 23 [21:46:35] cal update, rcvr 6 old offset -2904 new offset -2928 new freq 18110572 bin 23 [21:46:38] cal update, rcvr 4 old offset -2672 new offset -2648 new freq 10142852 bin 23 [21:46:39] cal update, rcvr 3 old offset -2928 new offset -2904 new freq 7042596 bin 23 [21:46:40] cal update, rcvr 7 old offset -2792 new offset -2768 new freq 21042732 bin 23 [21:46:40] cal update, rcvr 5 old offset -2624 new offset -2600 new freq 14042900 bin 23 [21:46:44] cal update, rcvr 5 old offset -2600 new offset -2576 new freq 14042924 bin 23 [21:46:49] cal update, rcvr 5 old offset -2576 new offset -2552 new freq 14042948 bin 23 [21:46:52] cal update, rcvr 3 old offset -2904 new offset -2880 new freq 7042620 bin 23 [21:46:58] cal update, rcvr 7 old offset -2768 new offset -2744 new freq 21042756 bin 23 [21:46:58] cal update, rcvr 5 old offset -2552 new offset -2528 new freq 14042972 bin 23 ^CSignal caught, exiting!

freq_offset -2976, -2296, -2880, -2648, -2528, -2928, -2744

Quit when I had all 7 offsets and I added them into the .conf file and turned off the Calibrate entry.

Requested 7 Activated 7 actual rcvr(s) Set rcvr 1 to freq 10000000 with offset -2976 Set rcvr 2 to freq 10000000 with offset -2296 Set rcvr 3 to freq 10000000 with offset -2880 Set rcvr 4 to freq 10000000 with offset -2648 Set rcvr 5 to freq 10000000 with offset -2528 Set rcvr 6 to freq 10000000 with offset -2928 Set rcvr 7 to freq 10000000 with offset -2744 Received Start command Set rcvr 1 to freq 1845500 with offset -2976 Set rcvr 2 to freq 3545500 with offset -2296 Set rcvr 3 to freq 7045500 with offset -2880 Set rcvr 4 to freq 10145500 with offset -2648 Set rcvr 5 to freq 14045500 with offset -2528 Set rcvr 6 to freq 18113500 with offset -2928 Set rcvr 7 to freq 21045500 with offset -2744

So far, so good... First CQ post on 40m came in -100Hz but heck, that's really good ! I think you're onto a solution here Rick !

[dtheriault]

While spots haven't been coming in for CQ's, I've gotten a bunch on 160m, 80m and 40m now that are all right on. I set receiver to CW, BFO 650, BW of 450, and spots are right where they're listed. <100hz easily.

[n1gp]

Glad your seeing better results. My results are about the same. One of the dongles seems to need more correction than the rest. Actually it's #4 right in the middle. Maybe the temperature is more in flux there? I hope this at least provides a solution for the low end medusa not needing to crack open the dongles and add a tcxo.

[dtheriault]

Unfortunately the propagation god is not favoring me today.  Spots are dribbling in, had one dongle -400hz but all the others still look right on.  

Would any of recent changes affect the sensitivity?  For instance, 18093 there's bit of a pileup but I'm seeing no decoders on that band at all.  Like to blame propagation but usually Alex's code can pick up few stations calling in a pileup w/o issue.

I'll fire up cudaSDR to take look.  I'm seeing on avg. 33.8Mb/s.  whoops, 12m just reported me .600 off... 

On Feb 25, 2015, at 06:38 AM, n1gp notifications@github.com wrote:

Glad your seeing better results. My results are about the same. One of the dongles seems to need more correction than the rest. Actually it's #4 right in the middle. Maybe the temperature is more in flux there? I hope this at least provides a solution for the low end medusa not needing to crack open the dongles and add a tcxo. — Reply to this email directly or view it on GitHub.