Quick comparison Siglent 3032X-R vs Rigol RSA3030N

[ehoernchen]

Hi,

as I had the chance to get a hand on both SAs at the same time I wanted to compare them directly to make a decision for one or the other.

First  the disclaimer: I'm not an expert in Spectrumanalysers just used them in the past and want to get now one in my privat lab. This is also not a professional product review. It's just my personal view and could easily be wrong and also missing important aspects.

Lets start. First picture shows both on the bench.

The 3032X powers up quicker then the 3030N, something in the range of 30 s versus 150 s. Could be the Rigol does some self calibration during the start up as I hear relays clicking.
The sweeptime for lower RBW is also faster on the Siglent, I will show some examples later.
Overall the Siglent seems more responsive and faster sometimes. The Rigol had also some hickups at the beginning. E.g. after first power up the touch screen was not working and also the VNA function did not work well. After a second power up both issues did not show up again, at least up to now (1day).
Both units had the Firmware version running with which they were deliverd. The Siglent had the second last (R6) and the Rigol had even a newer on as on the Web.

Now first a look on the external IF. There the Rigol seem to offer more. There more USB host IF, an IF out, an additional Trigger In/Out, an HDMI output and a small speaker. Nice.

Both have also the possibility for control over the LAN. The Siglent seems a bit more handy. Screenshots are automaticly saved and not just a frozen picture which you have to copy. The Rigol also had here some hickup as the welcome screen was just partly filled after returning to the page.

I measured also the 10MHz ref out with my counter. Siglent 0.05 ppm off and Rigol 0.4 ppm. Siglent 0.011 Hzpp over 1 min, Rigol 0.016 Hzpp. The Rigol has the option for a more stable clock source. Both seem to be accurate enough.

Thats it for now, I will add some measurements later. All with the lack that I don't really have high quality calibration standards.

Theo

[807]

You might get more response/views if you had posted this in the Test Equipment section. 

[tautech]

Both units had the Firmware version running with which they were delivered. The Siglent had the second last (R6) and the Rigol had even a newer on as on the Web.

Welcome to the forum.

Have you updated SSA3032X-R firmware to latest version V2.0.0.6.0R7 ?
https://int.siglent.com/upload_file/zip/firmware/Spectrum_analyzer/SSA3000X-R_V3.2.2.6.0R7_EN.zip

[ehoernchen]

Good point, yes, I have in the meantime updated to R7, although I dont know the difference.

[ehoernchen]

You might get more response/views if you had posted this in the Test Equipment section. 

Probably right, I have just used this board as another SSA thread was posted here but after looking I see most of the SSA threads are posted in Test Equipment.

[tautech]

You might get more response/views if you had posted this in the Test Equipment section. 

Probably right, I have just used this board as another SSA thread was posted here but after looking I see most of the SSA threads are posted in Test Equipment.

You as the OP can move this anywhere you like with the Move Topic feature at the foot of the page. Only you can see this in this thread.

Good point, yes, I have in the meantime updated to R7, although I dont know the difference.

Within the zip file download is a revision history PDF to compare what has been changed, added or fixed.

[ehoernchen]

Now first Noise floor (again its just what I observe on my two units, others may be different):

The Siglent seems to have a lower noise floor (Sig NF1) compared to the Rigol (Rig NF1).

Siglent has -128 to -129 dBm/Hz in the mid of the span
Rigol has about -120 dBm/Hz.

After power up I looked more that they had similar noise floor but the Siglent attenuator is set to 20dB default and the Rigol to 10dB.

The SSB noise gives the opposite picture. I used a Siglent SDG60xx as Source at 450Mz with amplitude close to 0 dBm.
The Siglent had about -96 dBc/Hz (I expected a bit better) at 10kHz offset. (Sig SSB1)
The Rigol had about -108 dBc/Hz, really good. (Rig SSB1)
Similar picture at 100 kHz offset. -95 dBm/Hz vs -110 dBm/Hz. (Sig SSB2 and Rig SSB2)
At 1MHz offset Siglent showed -113 dBc/Hz and Rigol showed -118 dBc/Hz, now closer. (Sig SSB3 and Rig SSB3) But very strange I could not set the Rigol to 100kHz RBW because then it showed nothing at all. (Rig SSB3a).

At 10 MHz the noise floor was dominant.

More to follow...

[tautech]

Keep RBW/VBW as 1:1.
Narrow spans permit lower RBW settings without excessively long sweep times.

You can see this here:
https://www.eevblog.com/forum/testgear/siglent-sha800a/msg5051896/#new

You do have a max of 10001 points available but it is always a compromise between accuracy and sweep time.

[ehoernchen]

You as the OP can move this anywhere you like with the Move Topic feature at the foot of the page. Only you can see this in this thread.

Thanks for the hint I did the move.

Within the zip file download is a revision history PDF to compare what has been changed, added or fixed.

Yes, but this history just contains all the changes of V3.2.2.6.0R7 compared to the former version V3.2.2.5.1R1. As I had V3.2.2.6.0R6 I don't know the difference between R6 and R7, but not so important.

[ehoernchen]

Now Tracking Generator.

The pictures "SigSpek" and "SigScope" show the Siglent Generator at -10 dBm (10 dB Attenuator in the connection cable). Measured with the Rigol SA and with the Scope. Looks clean.

The pictures "RigSpek1 and RigSpek2" and "RigScope" show the Rigol Generator at -20 dBm. Measured with the Siglent SA and the Scope. Looks short time clean but if I select MaxHold shows that the signal jumps, also visible in the Scope picture. Looks a bit strange but may be no issue when used for the VNA function.

Both where configured with Zero Span.

I will add VNA and Realtime function next days and then I have to choose one and send back the other.

If somebody has a suggestion what else I should look on just let me know.

[ehoernchen]

The realtime capability comes initially with 25MHz BW on the Siglent and 10MHz on the Rigol.

The two pictures show a WLAN spectrim in the density view on both with 10 MHz Bandwidth.
To look at a complete channel the 25 MHz span comes handy.

On the Rigol you have more options to setup the density display like pallette and density hues.

The Rigol has also more display combinations but it has no 3D display.

Otherwise the function seems quite similar and both offer up to 40MHz Span for Money.

[ehoernchen]

Now the VNA function.

First two pictures show the TF of a filter on Siglent and Rigol. I like the split screen of the Rigol. The Siglent seems to need to show both on top of each other.

SigVNA2 and RigVNA2 show that the Siglent seems to have a bit more dynamic range.

Initially also here the Rigol showed some hickup. E.g. the S21 was displayed just to the center frequency and then only max was displayed. (RigVNA3). After restarting it disappeared. But I saw it sporadicly again.

Also the reflection measurement on the Rigol puzzled me (no pictures). With an open connector it showed a S11 close to 0, but I expected close to 1. Is there an external reflection bridge necessary?

That was it on the comparison. I will send one back in the next days.

[tautech]

The Siglent seems to need to show both on top of each other.

Yes, you need View/Stop/Hold a trace then activate another and select the measurement type required.
With the 4 traces available you can have Log Mag, SWR, Phase and a Smith chart each in a different trace color and with common to all markers.
This does get seemingly busy for the inexperienced but to have them all correlated on one display can be useful.

BTW, did you notice for more precise traces/measurements 10001 points can be selected at the expense of greater sweep times. 200 points is the default minimum.

[ehoernchen]

BTW, did you notice for more precise traces/measurements 10001 points can be selected at the expense of greater sweep times. 200 points is the default minimum.

Not sure, is this done by increasing sweeptime?

One other thing which puzzles me on the Siglent is the Reflection measurement. In the doc and on the product webpage it’s listed but I can not find it.

[tautech]

BTW, did you notice for more precise traces/measurements 10001 points can be selected at the expense of greater sweep times. 200 points is the default minimum.

Not sure, is this done by increasing sweeptime?

No, an increase in sweep time is the penalty of selecting a large # of datapoints.
You can select the points to suit your particular needs (accuracy vs sweep time) from the Measure menu.

One other thing which puzzles me on the Siglent is the Reflection measurement. In the doc and on the product webpage it’s listed but I can not find it.

Reflection measurements are related to SA mode where an external reflection bridge would be used, typically for instrument isolation from the DUT where bridges themselve offer additional attenuation and where addition pads can more easily be added to attenuate strong signals.

In VNA mode, Port 1 (the Stimulus port) has an inbuilt bridge (not as robust as an external bridge) but very suitable for measurement of passive DUT's, like antennae where reflection measurements tell us what we need of antenna performance and for adjustment, if required.

[ehoernchen]

You can select the points to suit your particular needs (accuracy vs sweep time) from the Measure menu.

In the Measure menu I dont find it. There is "Channel Power",  "ACPR" and so on.
In the Sweep menu I find "Points" = 751 at the bottom, but its grey, I can not modify. (Picture) Manual tells a max of 751 but not how it can be modified.

Reflection measurements are related to SA mode where an external reflection bridge would be used, typically for instrument isolation from the DUT where bridges themselve offer additional attenuation and where addition pads can more easily be added to attenuate strong signals.

I dont find it in the SA mode to select it.

[807]

Now Tracking Generator.

The pictures "SigSpek" and "SigScope" show the Siglent Generator at -10 dBm (10 dB Attenuator in the connection cable). Measured with the Rigol SA and with the Scope. Looks clean.

The pictures "RigSpek1 and RigSpek2" and "RigScope" show the Rigol Generator at -20 dBm. Measured with the Siglent SA and the Scope. Looks short time clean but if I select MaxHold shows that the signal jumps, also visible in the Scope picture. Looks a bit strange but may be no issue when used for the VNA function.

Both where configured with Zero Span.

I will add VNA and Realtime function next days and then I have to choose one and send back the other.

If somebody has a suggestion what else I should look on just let me know.

So RigSpek2 shows the Rigol with tracking generator set to 100 MHz at zero span. Wonder where all that spurious is coming from! The waveform is severely distorted. The Siglent waveform looks perfect.

[tautech]

So RigSpek2 shows the Rigol with tracking generator set to 100 MHz at zero span. Wonder where all that spurious is coming from! The waveform is severely distorted. The Siglent waveform looks perfect.

This is RigSpek2 but mislabeled as it's from the Siglent analyzer and certainly not zero span.



I see some broadcast and possibly LoRa signals.

[ehoernchen]

Sure the picture is from the Siglent as I measured with the Siglent the Rigol TG. The Rigol was set to zero span to be at a fixed frequency. The Siglent which measured the Rigol was set to 1 GHz span to measure also the harmonics. So "RigSpek" means the Spektrum of the Rigol TG measured on the Siglent.

[807]

Sure the picture is from the Siglent as I measured with the Siglent the Rigol TG. The Rigol was set to zero span to be at a fixed frequency. The Siglent which measured the Rigol was set to 1 GHz span to measure also the harmonics. So "RigSpek" means the Spektrum of the Rigol TG measured on the Siglent.

So there seems to be a problem with the Rigol's tracking generator on zero span. It looks like it's still sweeping. Did you try it on zero span on any other frequencies?

[ehoernchen]

Yes I tried it also at other frequencies. If you measure NOT with Max hold it looks like a stable carrier, just once and a while a spurious line. Then I measured with the scope and saw that once and a while the signal changes. Then I used the Max Hold and saw the displayed spectrum.