External radio IF offset not updated / applied when changing modes

[thedillybar]

External radio IF offset for different modes works at startup. But if the mode is changed while SDRconsole is running, the new offset is not updated/applied.
For example, if started in USB (SSB-U), the SSB-U offset is persistent for the entire session regardless of what mode is later selected.

Workaround: Anytime the mode is switched, restart SDRconsole.


Running v3.2, RTL-SDR with External Radio, Tracking on.

[Max]

External radio IF offset for different modes works at startup. But if the mode is changed while SDRconsole is running, the new offset is not updated/applied.
For example, if started in USB (SSB-U), the SSB-U offset is persistent for the entire session regardless of what mode is later selected.

Workaround: Anytime the mode is switched, restart SDRconsole.


Running v3.2, RTL-SDR with External Radio, Tracking on.

Strongly suggest update to 3.3 Beta. It's fully tested and many many changes since 3.2 Release version. Update (over the top of existing install - no need to uninstall) and see if issue persists then report back.

https://www.sdr-radio.com/download#Beta

73

Max

[thedillybar]

When I run 3.3 beta with the same settings, I get no signals at all...just some noise around -140 dBm. Compared the log files, messed around with settings for 15 minutes and no joy.

[thedillybar]

I did some digging in the v3.2 logfile. Looks like SDRconsole is trying to set the center frequency, but the RTL-SDR isn't cooperating.

Code: Select all

18:51:23.598: Radio RTL SDR (USB): Frequency> Set 64.448.950 (async), Now 0, Direct 0, PPM 0, status -9 after 22 ms

At startup, the frequency is set and returns status 0.

Code: Select all

18:48:00.726: Radio RTL SDR (USB): Frequency> Set 64.452.480, Now 64.452.480, Direct 0, PPM 0, status 0 after 413 ms

So this is probably isolated to generic RTL-SDRs.

[thedillybar]

Well, the solution was to plug it into a different USB port. More information at the link below.
[nturl]https://groups.google.com/g/gqrx/c/ddVx6sCWI5U
[/nturl]

Still no joy with v3.3 but will save that for another day.

[jdow]

If you have a radio that changes its IF frequency depending on mode I suspect you will have extreme annoyance while running SDRConsole due to the issue you have. The SDRC in panadaptor use presumes the radio has a fixed IF. I am wondering if you are misreading some of the "Receiver" displays in SDRC.

It would help if you showed us a screen shot of everything in use, the definition page for your front end, and information about your external radio. More information is good. Show as many of the settings as you can. Please provide any pictures as full size attachments showing the full SDRC windows as the status bar is usually quite helpful.

(rtlsdr? I hope you can grow out of that disaster soon. They are PITA, and not the bread kind of pita, either.)

And do make sure you are running the latest 3.3 beta. That has all but a very few recent changes and is more reliable than the earlier versions.

{^_^}

[thedillybar]

I'm running a pretty straightforward setup. It's normal, as I understand, for IF offsets to be different by mode -- thus why most SDR suites have settings. I am running an Icom 756 Pro III with 64.445 MHz IF output. SDRC changes the center frequency of the SDR whenever the mode is switched, to adjust for these offsets -- see previous comments with details from the log file.

Just figured out V3.3, it's working. Had to change the gain to Auto -- I guess that's the default in V3.2 but not V3.3.

[jdow]

Some of this is train of thought typing. Please bear with me and slog through to the end. ICOM may do something funkity. I wish my Pro II was still functional so I could try to take some measurements. I never used it that way. I'd have used a tap off the antenna feed or a separate antenna.

SDRC moves filters. It does not move the center frequency. It has what I consider a defect in that the Receiver windows report their center frequency rather than the possibly phantom carrier frequency as is convention. You might be seeing that and misinterpreting it, in which case you may not be operating where you think you are.

I see from the picture that you have set some utterly preposterous offsets. All receivers I am aware of that use up conversion for the first IF have no offsets. The reason for this is obvious once you consider what the offset might be used for, USB sideband filters. Tune high for LSB and tune low for USB. (That presumes low side up conversion, reverse for high side mixing.) A typical IF is near 70 MHz. Temperature drift for quartz in filters is not unusually 25 ppm or more. That would mean about 1,750 Hz drift, totally compromising the SSB filtering capability. So they handle that selection in the second IF at more reasonable frequencies for filters. SDRC does not fiddle itself around to change where the tuned frequency lives. Thus you should have zeros where you have the outlandish 4000 Hz and 7020 Hz values. I have no idea what drove you to those settings other than a misinterpretation of how an SDR works.

An SDR can setup filters with very good characteristics with center frequency offsets that can exceed the filter width either way. This is like having nearly infinite crystal filters with various bandwidths and offsets. The Pro III is not that different from the Pro II I had. It's IF is digital and somewhat rude and crude about it. I suspect it might design all its filters to have the same center frequency and offset the tuning to perform sideband selection. Since the VFO is effectively the first conversion oscillator that would shift things around at the 64.455 MHz IF. (64.455 or 64.445? Whatever) So you would need to use offsets. In that case the offset should be near or a little more than half the width of the selected Pro III IF filter. That cannot be saved because it is not known by SDRC. It could be. But, I'd not have designed that in. It appears Simon did not, either. (2.7 kHz filter would have an offset of 300 Hz plus 1350 Hz, for example.)

If your rtl dongle is old it is likely to drift a lot. I seen one old one drift several kHz as it warms up. Newer ones should not have that problem. I wonder if you are seeing some spectrum inversion effects from the rtlsdr's high side mixing. (I have 6 or so of them, some old some new, and have grown to seriously hate the little buggers for their performance issues.)

Try inverting the spectrum and see if that brings some sense to the offset values you must set in. That is the only thing I can argue myself into as the source for some of your problem. Oh, check WWV in DSB mode for its center frequency. The 64.455 kHz IF setting may need some tweaking. The SSB offsets should be the same for the same filter width selection.

{^_^}

[thedillybar]

Thanks for the detailed response. I should add that I have the spectrum inverted and with the current settings it is performing 100% as expected.

The "utterly preposterous offsets" were adjusted solely to make it behave correctly. As I'm writing this, I'm realizing I do have a typo in the first IF -- it should be 64.455 MHz not 64.445 MHz. With the incorrect value 64.445 MHz, the offsets are +4000 SSB-L and +7020 SSB-U. I've changed it to the correct 64.455 MHz and changed the offsets to -6000 SSB-L and -2980 SSB-U. The result is the same...behaving correctly in both modes.

I don't understand the details of the SDR frequencies as much as you do, maybe I am mis-using the term "center frequency" but as I've posted previously the logfile shows SDRC setting the RTL-SDR frequency each time I change modes from LSB->USB and vice versa.
Setting mode to USB (IF frequency + USB offset)

Code: Select all

Radio RTL SDR (USB): Frequency> Set 64.452.020, Now 64.452.020, Direct 0, PPM 0, status 0 after 229 ms

Setting mode to LSB (IF frequency + LSB offset)

Code: Select all

Radio RTL SDR (USB): Frequency> Set 64.449.000 (async), Now 64.449.000, Direct 0, PPM 0, status 0 after 45 ms

I do notice the temperature drift, although it's not a big issue for me, especially since I am mainly using the SDR for the waterfall display, not actually to receive audio. If it's 50Hz off, when I click it tunes to the nearest 500Hz therefore close enough for tuning -- even if not close enough for receive audio.

I suspect the filters in the 756pro3 are applied after the first IF. When I change from 3.0k to 2.4k to 1.8k on the rig, there is no change in the SDR output/receive.

I guess you're suggesting the USB/LSB offsets should be symmetrical? Ultimately what I need is 64.449.000 for LSB and 64.452.020 for USB. Rather than my current IF and different offsets, would it make you feel better if I used the following...
IF = 64.450.510
LSB = -1510
USB = +1510

[jdow]

I suspect I want to dig around what the Pro series uses as the digital IF. I suspect they fudge things a little. You might see some more expected numbers and settings results if you use the real IF you found.

If that front end is an old rtlsdr such as an E400 tuner or one of the early R820T units from the commercial digital receiver market then you could have a crystal that far off. A ham repeater output may be stable and accurate enough to tell you something useful about the dongle's calibration comparing what you read and it's nominal frequency. Of course, it does not have to be an old one. The VCXO in one of my rtlsdr.com dongles has gone WAY off frequency.

And with numbers made to look sane I wonder if I can get Simon to comment on the offset not being applied when you startup. That seems to be what you have as a problem. Does that sound right to you?

Oh, and I do believe the Pro series all used the first oscillator as the sole tunable oscillator so any offset it uses would have to be done in that oscillator. That's not the way I'd do it. But their digital IF is so low that it may be constrained that way.

{^_^}