EEVblog Electronics Community Forum
Products => Test Equipment => Topic started by: markus_jlrb on October 14, 2023, 05:20:00 pm
-
Dears,
sorry to raise this issue, but If I'm not made a serious error I have some
observations that someone could verify to confirm or decline.
While playing with a SSA3032X-R by measuring some output spectra of
a LiteVNA64, I observed some irregularity in the AVG100 trace measurement.
First of all I was thinking this is an issue of the LiteVNA but after verifying the
center and span parameters I attached a second LiteVNA (last pic) and found
the same irregularity at the AVG100 trace measurement.
Would be nice if someone of the Siglent guys could comment on this findings.
Thanks in advance for your[attachimg=1][attachimg=2][attachimg=3][attachimg=4][attachimg=5][attachimg=6] help and effort.
Markus
-
Similiar observation here.
The displayed noise at the lower frequency side is visible higher than at the higher frequency side.
Not so pronounced as in your example, but visible.
Source here is my frequency standard, an OCXO, used as external reference.
I suppose this noise distributon is not from the OCXO.
Analyzer is an SSA3075X-R, firmware 3.2.2.6.0R7.
Hope other users will check and report their findings.
-
I looked the signal of my OCXO with my scope and found I have measured the rectangular output of
my OCXO by mistake.
Attached is the measurement with the sinus output now, which looks much better.
I assume the differences were caused by the transient response of the filters in the analyzer to square waves.
Does your LiteVNA output sinus or rectangular signals?
I assume rectangular signals.
-
Thanks kladit for your effort of verification.
Indeed my VNA output at lower frequencies is square shaped.
I will check with a sinus formed signal generated by my SDG6022X
to verify your findings.
Keep you informed on the result.
Thanks
Markus
-
Sorry to report the same observation with a sin input at the SAS30kX-R FE.
10MHz at 1Vpp + 13dB passive ATT at SSA input channel and AVG100 Trace view.
Any ideas whats wrong?
Markus
-
No idea at the moment.
Do we have the same firmware?
Here
SSA3075X-R FW 3.2.2.6.0R7
SDG6052X FW 6.01.01.36R8
-
Here two screenshots, same conditions.
If we have the same firmware, I assume an hardware defect now.
-
What type of coax cable are you using.
Here it is double shielded Huber&Suhner ENVIROFLEX 400.
Any switched power supply running near the cables?
Your noise is higher even at the high frequency side, therefore the quesitions.
What does your noise marker show at 10 MHz without any signal and with 50 Ohm terminator cap at the input?
-
Here an fft of the same signal made by an SDS2054X HD.
Noise here is lower because we have no 20dB attenuation in front like in the analyzer.
No higher noise at lower frequency end here like at the analyzer.
-
@kladit
While my LiteVNA64 measurement I had used a SUHNER SUCOFORM 141 20cm cable with N to SMA connectors.
While measuring with the sinus output I had used a RG58 C/U MIL C17 N to BNC cable.
What makes me wonder is the fact that while sin sig is on I could see the AVG100 irregularity, but when turning
off the signal from the gen. the AVG100 trace looks fine???
Markus
-
I think it is a software error, both your devices show the same behavior, I suspect the FFT scan is segmented.
It would be very helpful if your used both the same settings, i.e. RBW, VBW, Span etc.
You can also switch the device to not use the FFT, maybe try that.
Maybe @tautech can say something about it
-
Maybe @tautech can say something about it
Little sorry.
Source = SDG6022X @10 MHz 0dB
Results from SHA852A and SVA1032X attached.
-
The noise distribution the analyzer shows is correct and has to do with the Rayleigh function.
(dl4raj you remembered the right thing)
I am anything else but not a mathematician, so I can not serve with a simple answer, but the internet is informational about.
google "rayleigh noise spectrum"analyzer "
An interesting article is "hp_sa_and_noise_5966-4008e_an_1303.pdf" and one of keysight, lost here in the shuffle
at the moment.
What still is not to explain is the difference of the the amount of noise shown
between our instruments. They should show not more than less of a dB difference.
We should agree on setup for comparison. Same frequency, span, attenuation, bandwidth, signal level, cables etc.
RG58 is not double shieded, and MIL is meaningless in this context.
You have not answered if we have the same firmware in our instruments.
Your noise was much higher even at the high frequency side, the cause of that has to be investigated first.
I have sent you my phone number, look at your messages, so I think we can advance the progress of solving this mystery.
-
Dears,
I have checked the FW Version -> SW1: 3.2.2.6.0R8, SW2: 20230621-1513 20221101-1700, HW: 01.01.04
In addition to my past measurements with the LiteVNA-64, I have now used my SDG6022X at 10MHz and 100MHz
to document my findings again at span of 50, 100, 200 and 500 kHz.
Please see the attached screen shots taken from the SSA.
Markus
[attachimg=1][attachimg=2][attachimg=3][attachimg=4][attachimg=5][attachimg=6][attachimg=7][attachimg=8]
-
Sorry the 10MHz 50KHz pic got lost during atteching it.
-
Does result = same after Default Preset and set up the SA again ?
-
@tautech,
and all,
I have preset the SA as requested.
Compare the screenshot of SA mode vers RTA mode.
What do you think?
See the small blue line in between the noise in RTA mode.
this is the AVG100 line - no irregularity there.
This is a proof that FW has an issue in simple SA mode.
Am I right?
Thanks to all for your effort.
I have reported this findings to Siglent Germany pointing
to this EEVBlog thread.
Markus
-
Sorry again, have an issue with attaching pic's.
Second attempt of RTA screenshot.
Markus[attachimg=1]
-
Now the difference is not so big as before.
Use markers to make the results numerical comparable.
What do you use for the external 13 dB attenuation?
Here two mini-circuits bnc attenuators 10 dB + 3 dB in series.
-
See my measurement constellation.
SDG6022X in 50Ohm mode and 0dBm output level.
Markus
-
I will check the result with the extern ref as you did. (SDG ref out to SSA ref in)
-
The effect is obviously a bug. For these short spans the analyser is likely to work in FFT mode and could suffer from an initialisation or numerical truncation / rounding issue. Maybe the average of the trace is not calculated over all the data that is being displayed.
-
ext. ref. of 10MHz had no impact on the previous measurement result.
I'm of the same opinion that the FW of simple SA mode has an issue in averaging the trace.
Thanks for confirmation.
Markus
-
@tautech,
and all,
I have preset the SA as requested.
Compare the screenshot of SA mode vers RTA mode.
What do you think?
Sorry no X-R model here but pointed this thread to Hendorog when at lunch with him as he has SSA3075X-R.
When he has time he will check.
-
@tautech,
>Sorry no X-R model here but pointed this thread to Hendorog when at lunch with him as he has SSA3075X-R.
>When he has time he will check.
Any progress concerning my findings?
Markus
-
The noise distribution the analyzer shows is correct and has to do with the Rayleigh function.
....
I don't think the Rayleigh function explains this problem and applies in this case.
Please check the behavior of the disabled FFT sweep.
-
@KeBeNe
>Please check the behavior of the disabled FFT sweep. ???
Sorry not catch you - could you please provide a more detailed procedure.
Thanks in advance.
Markus
-
@KeBeNe
I also wrote:
"Your noise was much higher even at the high frequency side, the cause of that has to be investigated first."
I still hope one of the owners of a comaprable R&S analyzer can setup the same and show us the results.
@Markus
please post a screenshot and use markers and noise markers to make the results numerical comparable.
-
I don't have a comparable analyzer, yet it can provide RT-SA at narrow spans like 100 kHz.
Attached is the 100 MHz example. At a VBW of just 3 Hz averaging is rather slow, so I made do with 10 averages instead of 100:
SH_SA44_PN_100MHz_-18.5dBm_Span100kHz_AVG10
-
From my point of view not the noise level is to be discussed but the avg function.
Noise could be indeed higher at left side of the carrier but nevertheless the left side
trace have to be smoother when 100 times the noise was run into the avg function.
I will compare next week, if I have time, the same measurement at a SSA3021X (without+)
and report the result.
Markus
-
@perfoma01,
remarkable low noise, your SH_SA44 looks much better than the SSA3075X-R.
Does its software exist for linux or is it windows only?
Here my screenshots, 100 MHz signal originates from SDG6052X.
-
remarkable low noise, your SH_SA44 looks much better than the SSA3075X-R.
Well, the SA44 has a very different mixing scheme, so it is hard to beat at lower frequencies and narrow spans. In general, it has quite some limitations, but whith the right settings it can outperform most of the "big boys" in certain situations.
Does its software exist for linux or is it windows only?
There is at least a Spike version for 64-bit Ubuntu.
Here my screenshots, 100 MHz signal originates from SDG6052X.
My signal source has been a SDG6052X as well. For the following screenshots this wasn't even necessary, as the SA44 happens to produce a strong spur at exactly 100 MHz :-DD
The attached screenshots show the noise floor around 100 MHz with and without preamplifier.
SH_SA44_Noise_100MHz_PA-on
SH_SA44_Noise_100MHz_PA-off
-
@KeBeNe
>Please check the behavior of the disabled FFT sweep. ???
Sorry not catch you - could you please provide a more detailed procedure.
Thanks in advance.
Markus
I think you could switch between FFT and accuracy under sweep settings.
What do you hope for when the measurement is made with an R&S Analyzer?
I could offer a measurement with an SA4401A VNA and an FSIQ or SMIQ + FSIQ.
-
Just because I had one out and not yet had access to an X-R, a quick check with my SNA5004A in SA mode.
-
What I saw in the pictures of @kladit that the correction is off, is that the default setting on the R model?
-
Now with access to SSA3075X-R I can certainly replicate the issue.
To be reported.....
-
Thanks tautch for your effort and confirmation of my findings.
So now I can laid back and wait until Siglent fix the issue.
Markus