SIGLENT SSA3032X input related spurious

[rfspezi]

Here are some more interesting spurious from a 70 and 80 MHz AM-modulated (50%, 1MHz) signal.
Seems there is something happening around 75 MHz. 

[RFDUK]

What is the signal source?

[rfspezi]

What is the signal source?

It's a SDG2122X.
Can't exclude that the sideband effects come from it.
Would be great if somebody else could verify the measurement with a different source?

Varying the carrier of the modulated AM-source shows that the sidebands (and only them) get mirrored at 75 MHz.
The carrier spurious always occurs 60 MHz below the true carrier frequency.

[PA2HK]

Yes, interesting topic. I also have a new Siglent SSA3021X, manufactured in 2017, that I have recently bought about 4 weeks ago. Must say, very nice unit offered at a good price. Indeed.

I have ran three tests at 144Mhz, 433Mhz and 1285Mhz analog to what we have seen before in this thread.

SSA system information:
SW1: 1.2.8.3
SW2: 20170512-1
SW3: 00000001
HW: 07.03.00
Calibration: 2017-09-19

Signal source is Agilent E4421B generator connected to the Siglent SA via 2 meter Coleman RG213 interconnect. Needless to say, my unit also shows spurs and unfortunately does not seem to match the specification of minimal -65dBc.

145MHz, spur -60MHz with -61dBc


433MHz, spur -61MHz with -61dBc


1285Mhz, spur -60MHz with -60dBc

[RFDUK]

I requested a refund and that has been agreed by Batronix ASAP to allow me to make other arrangements. I'll post confirmation that loop works effectively as soon as funds are returned.

It's been an interesting ride and I'm wiser for it with regards to buying new 'economy' test instruments .... but also a little stressed and late on some projects

I'll be interested to read how this pans out technically speaking for other folks, good luck 

[cncjerry]

There must be something else going on with reference to the modulation.  I can't believe the analyzer is smart enough to demod the signal and spit the sidebands out all over the place.  It looks more like IMD, no?

If you move the 60Mhz spur into the center and zoom in on it with a narrow(er) span, lower the RBW, you should see the AM mod on the original spur assuming it isn't in the noise, I would think. 

[Orange]

I requested a refund and that has been agreed by Batronix ASAP to allow me to make other arrangements. I'll post confirmation that loop works effectively as soon as funds are returned.

It's been an interesting ride and I'm wiser for it with regards to buying new 'economy' test instruments .... but also a little stressed and late on some projects

I'll be interested to read how this pans out technically speaking for other folks, good luck 

Sorry to hear that this is the 'solution' for you. I'm somehow disappointed in Siglent and Batronix to let this happen. The analyser has great potential, and the first models did fine on this aspect. My own analyser does roughly 65 dBc. It is meeting the published specs (just), but I can live with it. Let's face it, where do you find a new analyser that does 3.2 GHz, with TG for 1700 euros....
My other analyser, a Rigol 815TG, does not come near the Siglent in terms of IMD and freq. range and sensitivity.

Please let us know what you next analyser will be.....

[DonRon]

Hi!

I bought my SSA3021 in December 2017.
Made the test in the range betwenn 100 and 200 MHz and observed the same results as PA2HK - Delta is around - 61 dBm.
Also measured the noise level, it's -97 dBm at 100 MHz.
My unit is HW 07.03.00 and calibration date is 2017-09-18

Cheers,

Ronald

[rfspezi]

Let's face it, where do you find a new analyser that does 3.2 GHz, with TG for 1700 euros....

...not 1.700 but 2.700€ for the 3.2GHz version.

I did another measurement using a 50% 1MHz AM-modulated 100MHz sine (SDG2122X).
The around 75MHz mirrored sidebands are most likely coming from the function generator.
But what i don't understand are the non 1MHz but 1.5MHz spaced sidebands around the 40MHz spurious?

[Emo]

OK,

Also took the measurements

SW1 1.2.8.5a
SW2 20170512-1
SW3 000000D1
HW 07.03.00
Calibration 2016-12-19

RF gen Philips PM5390A
Note; This generator delivers some spurs on 80Mhz-90Mhz below and above depending output frequency

Edit Compared to PA2HK, the HW is the same, SW 1 differs, but calibration date is later. Could someone at Siglent have made some different routine for calibration??

[Orange]

Let's face it, where do you find a new analyser that does 3.2 GHz, with TG for 1700 euros....

...not 1.700 but 2.700€ for the 3.2GHz version.

I did another measurement using a 50% 1MHz AM-modulated 100MHz sine (SDG2122X).
The around 75MHz mirrored sidebands are most likely coming from the function generator.
But what i don't understand are the non 1MHz but 1.5MHz spaced sidebands around the 40MHz spurious?

You need to extend it's possibilities in a clever way, and turn a 2.1GHz into 3.2GHz, and for sure you get it for 1700, but don't tell Siglent :-)

[TurboTom]

Let's face it, where do you find a new analyser that does 3.2 GHz, with TG for 1700 euros....

...not 1.700 but 2.700€ for the 3.2GHz version.

I did another measurement using a 50% 1MHz AM-modulated 100MHz sine (SDG2122X).
The around 75MHz mirrored sidebands are most likely coming from the function generator.
But what i don't understand are the non 1MHz but 1.5MHz spaced sidebands around the 40MHz spurious?

That's one of the results of the way the SSA works. The 3rd IF is not demodulated (rectified) as in an analog SA but it's digitized and analyzed via FFT. This takes place in segments of approx 1.25MHz (as we understand it currently). In these "bins", the "60MHz Spur" appears to be mirrored, i.e. an increasing carrier frequency lets the mirror spur decrease in frequency (as long as it stays within the bin). Have a look at rf-loop's post where he very nicely explains and illustrates this peculiarity: https://www.eevblog.com/forum/testgear/siglent-ssa3032x-input-related-spurious-99091/msg1391448/#msg1391448

At a modulation frequency of 1MHz, the side bands (and the carrier as well of course) spread over at least two adjacent bins, maybe even three, depending on the location of the carrier within "its" bin. Assuming the carrier is right at the centre of its bin (625kHz from either edge), the lower sideband's spur will be reflected to 375kHz apart from the lower bin's lower edge (which equals 875kHz span to the lower bin's upper edge) and the upper sideband's spur will be just mirrored, i.e. it will be located 875kHz apart from the upper bin's lower edge.

Now you've just got to add up the increments which is 875kHz + 625kHz = 1.5MHz, just like you reported. Depending on the location of the carrier, the arrangement of the spurs may even become asymmetric -- must be very confusing to watch if the operational principle of the SSA is unknown to the user.

Siglent should really have published some in-depth information on this instrument, pointing out its operation principles, block diagrams and some general figures like frequencies of the local oscillators and IFs to make understanding and interpreting such spurious signal easier for the user. Gone are the times of really detailed, comprehensive documentation of HP / Agilent equipment of the 1980s to early 2000s (probably also before that interval, but I haven't got much experience with such dated equipment). Then again, Siglent isn't what HP once was... 

[rfspezi]

Would be really interesting to do the test again with FFT turned off.

[RFDUK]

A couple of weeks ago I was waiting on a refund, but I had a change of heart when Batronix passed on confident reports from the factory regarding isolation of the problem.

Just received a new SSA3032X Manufactured Jan 2018 and all is well. The spur is looking very much like the earlier instruments at -71dBC, it's fixed.

Noise floor compared to the earlier instruments posted in the thread remains 3 to 4dB inferior. Checked against Siglent data sheet and this machine easily meets minimum spec and meets typical floor all the way up to 3.2 GHz, so no problem with that, very happy.

Winner winner .... you know the rest. Thanks to all who added useful information and technical interest to this thread.

[Orange]

A couple of weeks ago I was waiting on a refund, but I had a change of heart when Batronix passed on confident reports from the factory regarding isolation of the problem.

Just received a new SSA3032X Manufactured Jan 2018 and all is well. The spur is looking very much like the earlier instruments at -71dBC, it's fixed.

Noise floor compared to the earlier instruments posted in the thread remains 3 to 4dB inferior. Checked against Siglent data sheet and this machine easily meets minimum spec and meets typical floor all the way up to 3.2 GHz, so no problem with that, very happy.

Winner winner .... you know the rest. Thanks to all who added useful information and technical interest to this thread.

So the question we all have is 'What is causing this'

[rf-loop]

A couple of weeks ago I was waiting on a refund, but I had a change of heart when Batronix passed on confident reports from the factory regarding isolation of the problem.

Just received a new SSA3032X Manufactured Jan 2018 and all is well. The spur is looking very much like the earlier instruments at -71dBC, it's fixed.

Noise floor compared to the earlier instruments posted in the thread remains 3 to 4dB inferior. Checked against Siglent data sheet and this machine easily meets minimum spec and meets typical floor all the way up to 3.2 GHz, so no problem with that, very happy.

Winner winner .... you know the rest. Thanks to all who added useful information and technical interest to this thread.

Good!

The main reason was the previous major suspect as I write yesterday here (yesterday edit) :

https://www.eevblog.com/forum/testgear/siglent-ssa3032x-input-related-spurious-99091/msg1393911/#msg1393911

So, the "guilty" has been found and terminated. Everyone can breathe and sleep again and relax.

[Orange]

A couple of weeks ago I was waiting on a refund, but I had a change of heart when Batronix passed on confident reports from the factory regarding isolation of the problem.

Just received a new SSA3032X Manufactured Jan 2018 and all is well. The spur is looking very much like the earlier instruments at -71dBC, it's fixed.

Noise floor compared to the earlier instruments posted in the thread remains 3 to 4dB inferior. Checked against Siglent data sheet and this machine easily meets minimum spec and meets typical floor all the way up to 3.2 GHz, so no problem with that, very happy.

Winner winner .... you know the rest. Thanks to all who added useful information and technical interest to this thread.

Good!

The main reason was the previous major suspect as I write yesterday here (yesterday edit) :

https://www.eevblog.com/forum/testgear/siglent-ssa3032x-input-related-spurious-99091/msg1393911/#msg1393911

So, the "guilty" has been found and terminated. Everyone can breathe and sleep again and relax.

Relax ?.....
So anyone with this issue needs to contact Siglent right ?

[rf-loop]

A couple of weeks ago I was waiting on a refund, but I had a change of heart when Batronix passed on confident reports from the factory regarding isolation of the problem.

Just received a new SSA3032X Manufactured Jan 2018 and all is well. The spur is looking very much like the earlier instruments at -71dBC, it's fixed.

Noise floor compared to the earlier instruments posted in the thread remains 3 to 4dB inferior. Checked against Siglent data sheet and this machine easily meets minimum spec and meets typical floor all the way up to 3.2 GHz, so no problem with that, very happy.

Winner winner .... you know the rest. Thanks to all who added useful information and technical interest to this thread.

Good!

The main reason was the previous major suspect as I write yesterday here (yesterday edit) :

https://www.eevblog.com/forum/testgear/siglent-ssa3032x-input-related-spurious-99091/msg1393911/#msg1393911

So, the "guilty" has been found and terminated. Everyone can breathe and sleep again and relax.

Relax ?.....
So anyone with this issue needs to contact Siglent right ?

I meant that the cause of the problem was discovered. It would be a pretty stressful situation if the cause of the problem was not yet found. This, and only this, was the cause of my commentary. And it may not be unclear to anyone that I'm more interested in technology than trading, even if I sometimes do it.

I have also previously said that Siglent made this mistake and Siglent is responsible for it.
I think it is quite unequivocally said.
If I had sold a spectrum analyzer that is part of a potential problem set, I would contact the buyer or buyers and make a request for service. This is no more special than the fact that new cars are called to be repaired several times for many different reasons and constantly.
I do not take a stand on how someone else works. I will only take a position on what and how I do.

Just do not buy from sellers who do not have real after sales customer care.
Who ever suffer this isssue, imho, first place to contact is seller, not Siglent. Seller need handle this and seller seller cooperate with Siglent as/if nessessary and how need in different situations. This is main road - or least sales network need develop (and limit) so that it go to this way sooner or later. Better if yesterday..
If I have problem with my Sony TV I do not call Sony factory in Japan and I do not ask in some forum do I need really call Sony.  In the old days there was no ambiguity about where to go if the purchased item was not ok.

[rfspezi]

I am a bit concerned about how Siglent will correct the faulty units.
I assume they repair the units at their local service centers - so... will they be recalibrated after the modification or simply retuned to the sellers afterwards?

[Performa01]

I am a bit concerned about how Siglent will correct the faulty units.
I assume they repair the units at their local service centers - so... will they be recalibrated after the modification or simply retuned to the sellers afterwards?

Distributors are advised to send the analyzers back to the factory for repair (= replace the aluminum shielding block).

[PP7BB]

Is it possible to reveal marking of first new unit manufactured without this bug?
Some distributors might offer units not modified by Siglent. I hope Siglent will mark old units with new number. Without it we will not know if unit is free of this bug (newly manufactured or fixed by Siglent) or sit in wharehouse as it was produced with bug.

[rfspezi]

I am a bit concerned about how Siglent will correct the faulty units.
I assume they repair the units at their local service centers - so... will they be recalibrated after the modification or simply retuned to the sellers afterwards?

Distributors are advised to send the analyzers back to the factory for repair (= replace the aluminum shielding block).

How do you know?
Got any source?
To me it's more likely they send the replacement parts to their local service/repair centers and let them have the ssa's repaired.
That's why i fear there will be no calibration after repair.

[GDK_ATL]

Has anyone had any luck getting this fixed by either the dealer or SIGLENT? I bought mine from Saelig in October of 2017. I queried Saelig about the problem and they replied that:

Siglent is aware of the problem that you mentioned and are putting together a solution with the factory on this. The problem is not on all of the SSA3000X spectrum analyzers so part of the process is putting together a test to verify the unit has a problem and then getting everything in place to take care of the units that have problems. As soon as the details come in Siglent I will let everyone know how they will proceed going forward.

From reading about the issue here, I was under the impression that the problem had been identified by Siglent. Is that not the case?

[tautech]

Has anyone had any luck getting this fixed by either the dealer or SIGLENT? I bought mine from Saelig in October of 2017. I queried Saelig about the problem and they replied that:

Siglent is aware of the problem that you mentioned and are putting together a solution with the factory on this. The problem is not on all of the SSA3000X spectrum analyzers so part of the process is putting together a test to verify the unit has a problem and then getting everything in place to take care of the units that have problems. As soon as the details come in Siglent I will let everyone know how they will proceed going forward.

From reading about the issue here, I was under the impression that the problem had been identified by Siglent. Is that not the case?

This is the official position:
https://www.eevblog.com/forum/testgear/siglent-ssa3032x-input-related-spurious-99091/msg1393911/#msg1393911

I have one unit in stock affected and Siglent are sending me new shielding for installation.
Some dealers won't have the means to install them, therefore it's best left to an authorized service center.
Contact Saelig again or Siglent America in Ohio.

North American Headquarters
SIGLENT Technologies America, Inc
6557 Cochran Rd Solon, Ohio 44139
Tel: 440-398-5800
Toll Free:877-515-5551
Fax: 440-399-1211
info@siglent.com
www.siglentamerica.com

Edit
Siglent will ask for the SN#.

[TheSteve]

What about a re-calibration? There is no way anyone can change the RF shields and still say the cal is good.