radio-jove@groups.io | LINEARITY

#13613

I was intrigued by the calibration of the standard RJ setup.? While I do not use that setup, I decided to measure what I use for other purposes, including RJ.? So, have a read of the attachment.? I'm using the AirSpy Discovery HF+ to test the linearity.? The system is diagrammed to start and then, I ran an HP generator up by 10 dB steps ultimately into Radio SkyPipe.? I was somewhat disappointed at the non-linear behavior of the system.? Please have a read of the attachment.

CONCLUSION:? Yes, calibration is needed if linear response to the input is required!!!

Dave - W?LEV

LINEARITY of the AIRSPY DISCOVERY HF.docx

#13614

Dave

The non-linearity isn’t in SDR Console for sure. You’re using a single frequency stepped calibration where we use Jim Sky’s stepped wide band noise calibrator spanning ~8 MHz. If you have a known calibrated power meter you could check your HP output. That would help to tell you where the non-linearity lies.

Larry

_______________________

Show quoted text

On Wed, May 21, 2025 at 9:48?PM W0LEV via <davearea51a=[email protected]> wrote:

I was intrigued by the calibration of the standard RJ setup.? While I do not use that setup, I decided to measure what I use for other purposes, including RJ.? So, have a read of the attachment.? I'm using the AirSpy Discovery HF+ to test the linearity.? The system is diagrammed to start and then, I ran an HP generator up by 10 dB steps ultimately into Radio SkyPipe.? I was somewhat disappointed at the non-linear behavior of the system.? Please have a read of the attachment.

CONCLUSION:? Yes, calibration is needed if linear response to the input is required!!!

Dave - W?LEV

#13618

I've confirmed a number of times using a number of professional spectrum? analyzers that the signal generator is putting out what the knobs indicate.? The non-linearity is not due to the HP8648C.? My personal "target" for the non-linearity would be in order: 1)? VB Cable and? 2)? SkyPipe.

I can test SkyPipe as I did the HP8648C, but VB Cable has no hooks other than links between applications within the PC.? Possibly Jim Sky could comment on the linearity of SkyPipe?

I well realize the difference between white noise cal and my single signal tests.? But either will expose non-linearities in the system.? I have a known professional noise generator source which I could use which I used to calibrate the old RJ setup which employed only Radio SkyPips.? That was far easier and quite intuitive. ? Personally I am not a fan of judging colors.? No two pairs of human eyes view subtle color differences and/or intensities identically.? But they sure make pleasing and attractive plots.? Just my two-cents.

Any additional suggestions as to how best to sleuth the non-linearities I exposed?

I have an SDRPlay RSP1.? Early on with our experiments with those at LTO (Little Thompson Observatory), here at home for me and Terry's (Dr. Terrance Bullette) experiments, we sadly discovered it was sorely prone to IMD and overload.? That's why at LTO (and personally) we collectively placed the SDRPlay products on the shelf where they still reside.? I've only recently obtained an RX-888 MKII, but, man, does it eat the 5-vdc power like a Class A amplifier at L-Band! ? ?

Dave - W?LEV?

?

Virus-free.

Show quoted text

On Thu, May 22, 2025 at 2:00?AM Larry Dodd via <101science=[email protected]> wrote:

Dave
The non-linearity isn’t in SDR Console for sure. You’re using a single frequency stepped calibration where we use Jim Sky’s stepped wide band noise calibrator spanning ~8 MHz. If you have a known calibrated power meter you could check your HP output. That would help to tell you where the non-linearity lies.
Larry

_______________________

On Wed, May 21, 2025 at 9:48?PM W0LEV via <davearea51a=[email protected]> wrote:
I was intrigued by the calibration of the standard RJ setup.? While I do not use that setup, I decided to measure what I use for other purposes, including RJ.? So, have a read of the attachment.? I'm using the AirSpy Discovery HF+ to test the linearity.? The system is diagrammed to start and then, I ran an HP generator up by 10 dB steps ultimately into Radio SkyPipe.? I was somewhat disappointed at the non-linear behavior of the system.? Please have a read of the attachment.

CONCLUSION:? Yes, calibration is needed if linear response to the input is required!!!

Dave - W?LEV

--

Dave - W?LEV

#13620

Hi Dave,

The detection method selected in RSP makes a big difference in how it processes the input data, and as such, whether the output is linear with respect to its input.

--
Dave

Show quoted text

On 5/22/25 10:44, W0LEV via groups.io wrote:
I've confirmed a number of times using a number of professional spectrum
analyzers that the signal generator is putting out what the knobs indicate.  The
non-linearity is not due to the HP8648C.  My personal "target" for the
non-linearity would be in order: 1)  VB Cable and  2)  SkyPipe.

I can test SkyPipe as I did the HP8648C, but VB Cable has no hooks other than
links between applications within the PC.  Possibly Jim Sky could comment on the
linearity of SkyPipe?

I well realize the difference between white noise cal and my single signal
tests.  But either will expose non-linearities in the system.  I have a known
professional noise generator source which I could use which I used to calibrate
the old RJ setup which employed only Radio SkyPips.  That was far easier and
quite intuitive.   Personally I am not a fan of judging colors.  No two pairs of
human eyes view subtle color differences and/or intensities identically.  But
they sure make pleasing and attractive plots.  Just my two-cents.

Any additional suggestions as to how best to sleuth the non-linearities I exposed?

I have an SDRPlay RSP1.  Early on with our experiments with those at LTO (Little
Thompson Observatory), here at home for me and Terry's (Dr. Terrance Bullette)
experiments, we sadly discovered it was sorely prone to IMD and overload.
That's why at LTO (and personally) we collectively placed the SDRPlay products
on the shelf where they still reside.  I've only recently obtained an RX-888
MKII, but, man, does it eat the 5-vdc power like a Class A amplifier at L-Band!

Dave - W?LEV



<>
Virus-free.www.avg.com
<>

<#DAB4FAD8-2DD7-40BB-A1B8-4E2AA1F9FDF2>

On Thu, May 22, 2025 at 2:00?AM Larry Dodd via groups.io <>
<101science@... <mailto:[email protected]>> wrote:

    Dave
    The non-linearity isn’t in SDR Console for sure. You’re using a single
    frequency stepped calibration where we use Jim Sky’s stepped wide band noise
    calibrator spanning ~8 MHz. If you have a known calibrated power meter you
    could check your HP output. That would help to tell you where the
    non-linearity lies.
    Larry

    _______________________



    On Wed, May 21, 2025 at 9:48?PM W0LEV via groups.io <>
    <davearea51a@... <mailto:[email protected]>> wrote:

        I was intrigued by the calibration of the standard RJ setup.  While I do
        not use that setup, I decided to measure what I use for other purposes,
        including RJ.  So, have a read of the attachment.  I'm using the AirSpy
        Discovery HF+ to test the linearity.  The system is diagrammed to start
        and then, I ran an HP generator up by 10 dB steps ultimately into Radio
        SkyPipe.  I was somewhat disappointed at the non-linear behavior of the
        system.  Please have a read of the attachment.

        CONCLUSION:  Yes, calibration is needed if linear response to the input
        is required!!!

        Dave - W?LEV*
        *




--
*Dave - W?LEV
*

#13621

Thanks, Jim, for once again making me read the description of Radio SkyPipe.? I've searched the www for other strip chart applications and found none as useful (and mostly free).?

However, please note that many, maybe most, of us are no longer using the sound card.? As my diagram which I sent out yesterday illustrated, using any of the SDR applications (my favorite has become SDR Console), we need SW link between that application and RSP.? For many of us, we use VB Cable.? Not that it's perfect, but.....? At LTO one long afternoon a couple of years ago, Terry and I sat down to sleuth out our data stream.? We caught VB Cable dropping chunks of data which plays havoc with additional data analysis.? Yes, it's a consumer product aimed at the audio world and we're attempting to use it for real science.? Reminding us.....that we are still amateurs.

Are there differences with RSP between porting directly from a sound card? to porting through a virtual audio cable application to RSP?? Many of us, myself included, no longer involve the sound card.? If I truly need a sound card, I use the Beringer UMC 202HD.??

In rather extensive testing and experimenting, I've discovered most of the PC and consumer directed sound cards, be they internal or external, do not perform to the advertised specifications.? One of the M-Audio external cards produced so much RFI I could not use it for radio astronomy or much of anything else.? Sound Blaster and many ASUS products are especially guilty of that.? I spent too much $$$ on Sound Blaster cards believing their published specs until I did some quantitative testing on the bench.? I finally went with Beringer which in some aspects exceeds and is true to their advertised specifications.? Buyer beware!

Dave - W?LEV

Virus-free.

Show quoted text

On Thu, May 22, 2025 at 6:53?PM Dave Typinski via <davetyp=[email protected]> wrote:

Hi Dave,

The detection method selected in RSP makes a big difference in how it processes
the input data, and as such, whether the output is linear with respect to its input.

--
Dave

On 5/22/25 10:44, W0LEV via wrote:
> I've confirmed a number of times using a number of professional spectrum
> analyzers that the signal generator is putting out what the knobs indicate.? The
> non-linearity is not due to the HP8648C.? My personal "target" for the
> non-linearity would be in order: 1)? VB Cable and? 2)? SkyPipe.
>
> I can test SkyPipe as I did the HP8648C, but VB Cable has no hooks other than
> links between applications within the PC.? Possibly Jim Sky could comment on the
> linearity of SkyPipe?
>
> I well realize the difference between white noise cal and my single signal
> tests.? But either will expose non-linearities in the system.? I have a known
> professional noise generator source which I could use which I used to calibrate
> the old RJ setup which employed only Radio SkyPips.? That was far easier and
> quite intuitive.? ?Personally I am not a fan of judging colors.? No two pairs of
> human eyes view subtle color differences and/or intensities identically.? But
> they sure make pleasing and attractive plots.? Just my two-cents.
>
> Any additional suggestions as to how best to sleuth the non-linearities I exposed?
>
> I have an SDRPlay RSP1.? Early on with our experiments with those at LTO (Little
> Thompson Observatory), here at home for me and Terry's (Dr. Terrance Bullette)
> experiments, we sadly discovered it was sorely prone to IMD and overload.
> That's why at LTO (and personally) we collectively placed the SDRPlay products
> on the shelf where they still reside.? I've only recently obtained an RX-888
> MKII, but, man, does it eat the 5-vdc power like a Class A amplifier at L-Band!
>
> Dave - W?LEV
>
>
>
> <>
>
> <>
>
> <#DAB4FAD8-2DD7-40BB-A1B8-4E2AA1F9FDF2>
>
> On Thu, May 22, 2025 at 2:00?AM Larry Dodd via <>
> <101science=[email protected] <mailto:[email protected]>> wrote:
>
>? ? ?Dave
>? ? ?The non-linearity isn’t in SDR Console for sure. You’re using a single
>? ? ?frequency stepped calibration where we use Jim Sky’s stepped wide band noise
>? ? ?calibrator spanning ~8 MHz. If you have a known calibrated power meter you
>? ? ?could check your HP output. That would help to tell you where the
>? ? ?non-linearity lies.
>? ? ?Larry
>
>? ? ?_______________________
>
>
>
>? ? ?On Wed, May 21, 2025 at 9:48?PM W0LEV via <>
>? ? ?<davearea51a=[email protected] <mailto:[email protected]>> wrote:
>
>? ? ? ? ?I was intrigued by the calibration of the standard RJ setup.? While I do
>? ? ? ? ?not use that setup, I decided to measure what I use for other purposes,
>? ? ? ? ?including RJ.? So, have a read of the attachment.? I'm using the AirSpy
>? ? ? ? ?Discovery HF+ to test the linearity.? The system is diagrammed to start
>? ? ? ? ?and then, I ran an HP generator up by 10 dB steps ultimately into Radio
>? ? ? ? ?SkyPipe.? I was somewhat disappointed at the non-linear behavior of the
>? ? ? ? ?system.? Please have a read of the attachment.
>
>? ? ? ? ?CONCLUSION:? Yes, calibration is needed if linear response to the input
>? ? ? ? ?is required!!!
>
>? ? ? ? ?Dave - W?LEV*
>? ? ? ? ?*
>
>
>
>
> --
> *Dave - W?LEV
> *
>
>
>

--

Dave - W?LEV

开云体育

LINEARITY - MEASUREMENTS