Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Diogo Nunes Pereira]

Hello.

I'm testing out a SignalHound SA44B a friend recently bought and I'm getting some unexpected signals in my TV band sweeps.

Attaching picture: yellow line is SA44 sweep compared to tinySA (cyan) and RF Explorer scans (magenta).

Same antenna setup on all. I've changed RBW and span settings and noise block remains there. Also find it odd various blocks have such a similar shape...

Am I doing anything wrong, going nuts, or is the SDR giving me a bad reading?

Cheers,

[Henry Cohen]

Hello.

I'm testing out a SignalHound SA44B a friend recently bought and I'm getting some unexpected signals in my TV band sweeps.

Attaching picture: yellow line is SA44 sweep compared to tinySA (cyan) and RF Explorer scans (magenta).

Same antenna setup on all. I've changed RBW and span settings and noise block remains there. Also find it odd various blocks have such a similar shape...

Am I doing anything wrong, going nuts, or is the SDR giving me a bad reading?

What is your antenna setup? Having a small whip directly connected to the SH will pickup some of the LO and, and more importantly, mixer leakage. Use a remote antenna at least 1 meter from the SH.

[Rui Lisboa]

Hello.

I'm testing out a SignalHound SA44B a friend recently bought and I'm getting some unexpected signals in my TV band sweeps.

Attaching picture: yellow line is SA44 sweep compared to tinySA (cyan) and RF Explorer scans (magenta).

Same antenna setup on all. I've changed RBW and span settings and noise block remains there. Also find it odd various blocks have such a similar shape...

Am I doing anything wrong, going nuts, or is the SDR giving me a bad reading?

Cheers,

Yes Diogo. LO of the SA44 is @10.7. I recall also seeing the image freq @21.4.

[Diogo Nunes Pereira]

What is your antenna setup? Having a small whip directly connected to the SH will pickup some of the LO and, and more importantly, mixer leakage. Use a remote antenna at least 1 meter from the SH.

Antenna was out of the room window with a 5m cable run but I guess LO's and mixers are involved in producing this effect...

SignalHound support was fast and helpful. Here's what I learned today:

"Unfortunately, the SA44B is working as designed. It uses a software image rejection algorithm, that creates a spurious response when there is a signal both 21.4 MHz above AND 21.4 MHz below the signal you’re trying to measure. There is more information in section 2.2.1 of the SA44B user manual.

Our BB60C and BB60D and SM series have hardware-based image rejection and do not have this weakness."

And here's what the manual states:

" The USB-SA44B does not have hardware-based image rejection, instead relying on a software algorithm
to reject image responses. The algorithm mixes the incoming RF with two distinct local oscillator
frequencies, typically spaced by 21.4 MHz and up to 100 milliseconds, and rejects responses not present
in both. This algorithm has some limitations:

1. A signal must be present for both captures to be displayed. Pulsed or swept signals, which do
not stay at any given frequency for this duration, will be rejected as potential image or spurious
responses.
2. An analog modulation envelope may be clipped, since certain frequency components of the
modulation envelope may not be present at both times. Most digital modulation tends to spread
energy evenly across its bandwidth and is relatively immune from this effect.
3. Two RF input signals, spaced by 42.8 MHz, will generate a spurious response halfway between
the two RF input signals. This spurious response will not be present when a 200 kHz span is
selected. Broadband signals which exceed 42 MHz cannot be accurately measured with the USBSA44B because of this effect."

Point 3 hits the mark: those blocks of artifacts are 21.4 MHz above and below occupied DTV Channels... as Rui stated:

Yes Diogo. LO of the SA44 is @10.7. I recall also seeing the image freq @21.4.

So, my question now is... what can I do to rely on the readings of this unit? I seems broadband sweeps of the TV band are out of question, unless I run 200kHz sweeps one after the other. Hundreds of CSV files. I wonder what other users of this SA are doing with it... only looking at single signals/modulation?

Thanks.

[Rui Lisboa]

That’s exactly why I disconsidered the SA44. Well I guess you can always carry a tunable pass band and filter what they didn’t 😅.

[Diogo Nunes Pereira]

That’s exactly why I disconsidered the SA44. Well I guess you can always carry a tunable pass band and filter what they didn’t .

Yeah. Somehow I’ve missed the memo… I imagine all spectrum analyzers have limitations - this one’s in particular are just now hitting me in the face.

Tunable band pass filters you say… then tune in blocks of 40 Mhz one after the other. It would reduce the amount of needed scans two-hundredfold. 

I know little about filters or the way to tune them though. The Smith chart in my nanoVNA is a stranger to me…

But from you saying this I’m thinking about the fixed/tracking filters selection in WWB equipment profiles editor… most high-end receivers there have that tracking option. Suppose tunable filters on front end helps those units become top notch.

Also imagine filters is what SignalHound support engineer refers to when talking about hardware-based image rejection.

[Jason Glass]

Yeah. Somehow I’ve missed the memo… I imagine all spectrum analyzers have limitations - this one’s in particular are just now hitting me in the face.

Tunable band pass filters you say… then tune in blocks of 40 Mhz one after the other. It would reduce the amount of needed scans two-hundredfold. 

I know little about filters or the way to tune them though. The Smith chart in my nanoVNA is a stranger to me…

But from you saying this I’m thinking about the fixed/tracking filters selection in WWB equipment profiles editor… most high-end receivers there have that tracking option. Suppose tunable filters on front end helps those units become top notch.

Also imagine filters is what SignalHound support engineer refers to when talking about hardware-based image rejection.

FWIW, back before I purchased my BB60C, I used a USB-SA44B with an inserted sweepable bandpass filter having 1% passband.  I engaged MAX HOLD and swept the filter from below START frequency of my span of interest to above its STOP frequency.  This worked perfectly to eliminate images.

[Andrew Broughton]

I've never seen or heard of this issue before. (Or really understand it, but I'd like to!)
Would I see this issue when using a Shure receiver to do a scan?
Is there a way to test my Siglent to see if it would show this issue?

[Diogo Nunes Pereira]

FWIW, back before I purchased my BB60C, I used a USB-SA44B with an inserted sweepable bandpass filter having 1% passband.  I engaged MAX HOLD and swept the filter from below START frequency of my span of interest to above its STOP frequency.  This worked perfectly to eliminate images.

That's a wonderful workaround... just have to find some of those (cheap) tuneable bandpass filters. 
I knew (from reading it here) that you used the SA44B, just wasn't aware that it had a little help in the front-end from your tunable friends...

I was flabbergasted at the looks of my wideband sweeps with this unit. Still trying to understand how can I trust the readings (even narrowband) on it, knowing that on a clean TV channel frequency (let's say 546MHz here in Barcelona for example) there's fictitious dirt caused by the existence of real dirt at +- 21.4MHz.

I understand you are happy with the BB60C... how was life with the SA44B before?

I've never seen or heard of this issue before. (Or really understand it, but I'd like to!)
Would I see this issue when using a Shure receiver to do a scan?
Is there a way to test my Siglent to see if it would show this issue?

I'm with you Andy; really would like to understand it better. I've bookmarked this page on the topic of RF Mixers: https://www.electronics-notes.com/articles/radio/rf-mixer/rf-mixing-basics.php Hopefully our panel of experts can shed some light on the topic or at least some other resources to get me out of my ignorance.

I've also never seen this when using Shure (or any other pro-brand) receivers or the AXT600.

[Rui Lisboa]

I'm with you Andy; really would like to understand it better. I've bookmarked this page on the topic of RF Mixers: https://www.electronics-notes.com/articles/radio/rf-mixer/rf-mixing-basics.php Hopefully our panel of experts can shed some light on the topic or at least some other resources to get me out of my ignorance.

I've also never seen this when using Shure (or any other pro-brand) receivers or the AXT600.

Cool article about mixers.
But this is only a part of the equation which leads us into the very nature of receiver design. Perhaps Henry can help us as he is the most eloquent person summarising complex topics into accessible inputs. 😅
But please do search for the topic: superhet image frequency. Historically there has been different ways to deal with this both in the front and back end of the receivers. Double or triple superheterodyning are a common solution for solving this issue (although they don't serve this exclusively).

The sort of apparatus we use for our tasks put on high efforts to eliminate such artefacts. Mainly on higher end equipment.
You've probably seen Image Rejection of X dB on the spec sheets.
On the AXT600 you just refer Shure states figures higher than 110dB (typical). The lower I've found for these top tier receivers is 70dB.
SH's SA44B is a big wide open window with little hardware filtering. Its priority is perhaps single signal analysis and not taking wide scans of the RF landscape. You can do it but as Jason said it takes a certain effort and investment which at least for me threw me out the train.

[ProSoundWeb Community]