Siglent SSA3000X and SSA3000X-Plus Spectrum Analyzers

[rf-loop]

One of the improvements in 7.07 was to enable the saving of jpeg file types, not only bmp as was in 7.05

This is shame they select jpeg as alternative for BMP. Better than nothing but..

Jpeg is ok for normal photographs. But not at all good for this kind of detailed technical images.  It destroy colors (when thin lines) and produce lot of artefacts and pixelization. Poorest format for this purpose. So, least I need continue BMP - PNG conversion using example very fast and simple irfan view.

Here is example about difference. I have rised gamma from 1.0 to 1.5 so that difference is more visible.
And what best, png  file size is less (with this kind of images) around 15k to around 25k as can see with most of my images. And quality is lot of better than jpg "scrambling".

[tautech]

I've asked for more file types for a little while, SSA3kX is the first of any Siglent AFAIK to offer jpeg and we hope there will be more in the future. Png? Maybe. Discussion with Tech support has been around the time to convert each of the different types and the effect it has on the Save file computation times.

Jpeg takes a few more seconds to compute than BMP and with the jpeg format I have missed files by removing a USB pen drive too quickly before it has fully loaded. Annoying, yes but with the SSA3kX it is very quick to use the File management UI to check the file is present and does actually contain data.
I need to learn to more patient. 

[rf-loop]

I've asked for more file types for a little while, SSA3kX is the first of any Siglent AFAIK to offer jpeg and we hope there will be more in the future. Png? Maybe. Discussion with Tech support has been around the time to convert each of the different types and the effect it has on the Save file computation times.

Jpeg takes a few more seconds to compute than BMP and with the jpeg format I have missed files by removing a USB pen drive too quickly before it has fully loaded. Annoying, yes but with the SSA3kX it is very quick to use the File management UI to check the file is present and does actually contain data.
I need to learn to more patient. 

It can nearly say that jpeg in this kind of use is mistake. It is widely known. It is very ok for photograph but not at all ok for technical graphics. PNG is widely known and supported. Of course fully lossless packaging is best but if use example GIF (one of most widely supported format and old also) then need change SSA colors to match 256 colors (what is not at all problem, only it need do)

When designer put jpg to this kind of purpose in equipment it can nearly say it is design error.
PNG is perhaps best possble what support also wide color pallette but for technical purposes GIF is even more better because it is true lossless. Still it produce small file size. But, color pallette is reduced and if we convert higher color depth image to GIF there may come very fun results, so it is best if original is also 256 colors. In teccnical equipment display like Spectrum analyzer imho 256 color depth is not at all problem.

Please Siglent:
Give us freedom to select whole display color map by user. I do not like these "childrens toy" colors.
This is just one table in UI.
Take example old ancient R&S Spectrum and look how they have done it (at this time display was color crt). So simple, so very simple.
User select example one trace, After then he turn knob and trace color change. After user find ok color he leave it and go to other things for adjust. 2 minutes and you have adjusted whole display just as you want, simple. Next time you turn it on it remember it. Only if do deep reset (not preset) set it back to factory state. Or something like it.
Only way to do it is fully freedom to user. Manufacturer made "skins" are not ok, mostly they are just terrible.

All peoples color taste is not same, also culture may affect this. Then, peoples eyes are different. Some times it is nice to "harmonize" all equipments so that when I look long time red display and then I move my eyes to blue display it is  just - terrible. (it may some times take long time before eye adapt and "negative" color disappear.
Then some times we need work dark. How hell these equipments can use in dark, example in observatory. Only way is buy lot of color and gray filters, or do hardware modifications.

Please Siglent:
If you really want keep JPEG there it is ok, no need delete but please give us least PNG.  (GIF only if also native color depth is reduced to 256)

[hendorog]

PNG is lossless.
http://www.howtogeek.com/203979/is-the-png-format-lossless-since-it-has-a-compression-parameter/

There was a patent dispute over GIF IIRC. I just googled it and the patents have expired FWIW. PNG would be better anyway.

[rf-loop]

PNG is lossless.

Yes. You are exactly right. Thanks.

Somehow I was thinking it is very nearly as lossless. But never checked it exactly.
Before I write my opinion I fast check using my png and not at all check same image bmp because this I know is direct bitmap.
When I zoom my png it show some small artifacts/softness but very lot of less than same image jpg.
Jus after your msg I recheck. It is irfanView and/or my computer what destroy image (when I zoom image). It is fun, when I change zoomed image window size  in irfanView it return just clean). No any difference between original BMP and then PNG and used maximum compression. (also my laptop monitor do not use native reolution because I have reduced it due to my old dated eyes.)

[KE5FX]

When designer put jpg to this kind of purpose in equipment it can nearly say it is design error.

+1.  .JPG is for photographs, not high-contrast line and vector imagery.  .PNG is the appropriate format.

[Bicurico]

Hello,

I come from the TV world and I am into TV, CATV, SAT, IPTV & RADIO monitoring.

I own several TV analyzers and field meters, most with spectrum analyzer, which in comparison to a "real" spectrum analyzer are pretty basic.

However, to educate myself, I am really getting interested in owning a SAA3021X. Just because. Others collect stamps, I like to monitor the ether with spectrum analyzers and I like to learn and there is no better way to do it, than owning the required equipment.

Anyway, since I have no real experience with this kind of product, I would like to clarify the following issue, which is bugging me (especially because I grilled one simple spectrum analyzer due to this - see my other post "Presenting my new software for the SMA Spectrum Analyzer 138MHz-4.4GHz" - https://www.eevblog.com/forum/testgear/presenting-my-new-software-for-the-sma-spectrum-analyzer-138mhz-4-4ghz/ - but I was able to repair it):

The RF input connector of the SAA3021X is specified as follows: "To avoid damage to the instrument, for the signal input from the RF input terminal, the DC voltage component and the maximum continuous power of the AC (RF) signal component can not exceed 50 V and +30 dBm respectively."

Would it therefore be safe to connect the LNB Loopthrough cable from a satellite receiver (which would be responsible for feeding the DiSEqC-motor, LNB, etc.)? This cable would carry the RF signal from the LNB, together with 14/18V and 0/22kHz Signal.

How could I be sure that the "maximum continuous power of the AC (RF) signal component" will be less than +30 dBm?

The same goes for:
- CATV: sometimes active elements like amplifiers will feed current into the coax cable for additional remote powered amplifiers
- Antenna: what if I connect a huge antenna which will produce a really strong signal?

Thanks!

[tautech]

As you've discovered damage to the input is to be avoided at all costs.

Unfortunately even a basic DC block is not offered in the list of accessories from Siglent and it's something I'm currently investigating alternative sources for. Some cheap ones are available from eBay with dubious specs. 

A good general recommendation of accessories needed are listed in this previous post:
https://www.eevblog.com/forum/testgear/siglent-ssa3000x-spectrum-analyzers/msg884523/#msg884523


Just like any test equipment the base unit is only the start of other pieces needed in order to complete many tasks. 

[nctnico]

When unsure about a signal it is best to check the signal first by feeding the signal into the spectrum analyser through a 40dB (or more) attenuator and perhaps also an extra DC blocker.

@Tautech: I just tested some cheap DC blockers from Ebay and they work quite well.
This one is the blue line (yellow is no DC blocker): www.ebay.com/itm/141638175057
However it seems the Siglent SSA3000X series has DC blocking capacitors build in so there is no need for an external one for DC levels up to +/-50V.

[tautech]

@Tautech: I just tested some cheap DC blockers from Ebay and they work quite well.
This one is the blue line (yellow is no DC blocker): www.ebay.com/itm/141638175057

Yep, spied that one and found it for less from other sources.
But if you check the specs it's true DC blocking voltage is only 50V. 

However it seems the Siglent SSA3000X series has DC blocking capacitors build in so there is no need for an external one for DC levels up to +/-50V.

Quite, but with a rated 50V input how much more isolation is wise to use?
IMO if a significant amount more is not added there's still a risk that a SA newbie like me might do damage.
I'd rather be safe rather than sorry with a $3K+ SA .

I'm looking for something reasonable and a bit higher spec to offer as an option to customers rather than have unnecessary equipment failures as a result of misuse.

Anyone suggest a source of a suitable DC block, a bit better spec'ed?

[hendorog]

Quite, but with a rated 50V input how much more isolation is wise to use?
IMO if a significant amount more is not added there's still a risk that a SA newbie like me might do damage.
I'd rather be safe rather than sorry with a $3K+ SA .

I'm looking for something reasonable and a bit higher spec to offer as an option to customers rather than have unnecessary equipment failures as a result of misuse.

Anyone suggest a source of a suitable DC block, a bit better spec'ed?

Maybe 50V is the standard -  e.g. the Signalhound accessory pack - they use Minicircuits DC blocks which all seem to be rated to 50V.

If someone blows both that _and_ the built in 50V input protection - well it reminds me a bit of this old joke
http://www.waynesthisandthat.com/response.htm

[tautech]

Quite, but with a rated 50V input how much more isolation is wise to use?
IMO if a significant amount more is not added there's still a risk that a SA newbie like me might do damage.
I'd rather be safe rather than sorry with a $3K+ SA .

I'm looking for something reasonable and a bit higher spec to offer as an option to customers rather than have unnecessary equipment failures as a result of misuse.

Anyone suggest a source of a suitable DC block, a bit better spec'ed?

Maybe 50V is the standard ......

Maybe its is, but for some this might be considered an essential accessory too:
http://www.aliexpress.com/store/product/N-M-F-DC-VARIABLE-POWER-block-DC-2-5GHz-200W-1000V-RF-Coaxial-high-voltage/1247757_32628218601.html

Edit
comprehension

[hendorog]

Quite, but with a rated 50V input how much more isolation is wise to use?
IMO if a significant amount more is not added there's still a risk that a SA newbie like me might do damage.
I'd rather be safe rather than sorry with a $3K+ SA .

I'm looking for something reasonable and a bit higher spec to offer as an option to customers rather than have unnecessary equipment failures as a result of misuse.

Anyone suggest a source of a suitable DC block, a bit better spec'ed?

Maybe 50V is the standard ......

Maybe its is, but stuff for some this might be considered an essential accessory too:
http://www.aliexpress.com/store/product/N-M-F-DC-VARIABLE-POWER-block-DC-2-5GHz-200W-1000V-RF-Coaxial-high-voltage/1247757_32628218601.html

Thats an impressive beast. Just curious is it just projection from accidents when probing around or is there a type of application where they might need that type of protection - kinda sounds like you have something in mind?

[tautech]

Quite, but with a rated 50V input how much more isolation is wise to use?
IMO if a significant amount more is not added there's still a risk that a SA newbie like me might do damage.
I'd rather be safe rather than sorry with a $3K+ SA .

I'm looking for something reasonable and a bit higher spec to offer as an option to customers rather than have unnecessary equipment failures as a result of misuse.

Anyone suggest a source of a suitable DC block, a bit better spec'ed?

Maybe 50V is the standard ......

Maybe its is, but for some this might be considered an essential accessory too:
http://www.aliexpress.com/store/product/N-M-F-DC-VARIABLE-POWER-block-DC-2-5GHz-200W-1000V-RF-Coaxial-high-voltage/1247757_32628218601.html

Thats an impressive beast. Just curious is it just projection from accidents when probing around or is there a type of application where they might need that type of protection - kinda sounds like you have something in mind?

Not really, there's stuff out there for 1kW and more. Have a hunt on eBay, you'll see. 

No I haven't got a personal need, just looking for something appropriate to offer customers for additional input protection.

But should customers need to go looking at HV RF, those that might should have the nouse to know what they're doing.....I hope.

[hendorog]

Ah I see kW not kV- i.e. you want to block DC and an RF power limiter. That makes sense, I'm sure people have transmitted 100W into their SA before.

Someone will know more about that. I just use a 20dB attenuator _and_ an external DC block on the front end when I need to be careful, but thats only good to maybe a couple of Watts and 50V. Of course something like that big limiter would be too heavy for the SMA connector on my Signalhound, would be like the tail wagging the dog

[rf-loop]

Why typically 50V dc.

Look typical cheap SMD capacitors specifications.

Note also that for 9kHz in 50ohm system this capasitance need be quite high.

Many DC blocks looks like they have nF class capacitors there instead of around uF class.
Example one in ebay tell that center pin DC block capacitor is 1nF
Now look this corner frequency just for simple estimation without more complex math.

But if do not really need 9kHz and starting from MHz or so is ok  it may give much more protection in practice what can read from DC value. Rising from 50V to 200V feeels like some improvement but... think also if freq range bottom is more high, it may also give more protection in practice because it is high pass filter before all internals and more low frequency things are attenuated.

Pity these eBay sellers do not give enough data or data is copied and not exactly from what they are selling..
Also it need think if need DC/High pass block both, center and shield and not only center.

[nctnico]

@Tautech: I just tested some cheap DC blockers from Ebay and they work quite well.
This one is the blue line (yellow is no DC blocker): www.ebay.com/itm/141638175057

Yep, spied that one and found it for less from other sources.
But if you check the specs it's true DC blocking voltage is only 50V. 

The breakdown voltage is specified at 1kV. See below.

However it seems the Siglent SSA3000X series has DC blocking capacitors build in so there is no need for an external one for DC levels up to +/-50V.

Quite, but with a rated 50V input how much more isolation is wise to use?
IMO if a significant amount more is not added there's still a risk that a SA newbie like me might do damage.
I'd rather be safe rather than sorry with a $3K+ SA .

If you look at Dave's SSA3000X teardown you'll see the MLCC capacitors for the DC blocking near the input. Even though they are rated for 50V MLCC capacitors in general have a very high dielectric breakdown voltage (over 1kV). This PDF has more information on it: http://www.koaspeer.com/pdfs/MLCC.pdf 
Note the PDF is not about safety but it shows MLCC capacitors can deal with significant overvoltages based on emperical data.

HOWEVER(!) when using a high voltage DC block the capacitor inside still needs to be charged to level out the DC bias. This energy will go into the SA's input so IMHO it is better to use an attenuator as well when dealing with potentially high DC levels.

Also DC blocks don't work very well at low frequencies. I also measured that (see below). Yellow is a 0dBm signal with a 20dB attenuator (to protect the input against the generator misbehaving), the blue line is the SMA Ebay DC block I listed earlier.

[joeqsmith]

Sensitive test equipment input protection ideas would have made for a great sticky thread.   Rare I want to look at anything exceeding 20bB but I keep an assortment of attenuators, blocks, GDTs and diode clamps.  Combo depends what the signal is.  Looking at a gas grill ignitor is going to be different than looking at a 10W 100MHz signal.     Link from when I was increasing the BW of my current probe.  You can see a small load I made with a tap. 

https://www.eevblog.com/forum/projects/home-made-pa-and-load/msg640180/#msg640180

[tautech]

Why typically 50V dc.

Look typical cheap SMD capacitors specifications.

Note also that for 9kHz in 50ohm system this capasitance need be quite high.

Many DC blocks looks like they have nF class capacitors there instead of around uF class.
Example one in ebay tell that center pin DC block capacitor is 1nF
Now look this corner frequency just for simple estimation without more complex math.

But if do not really need 9kHz and starting from MHz or so is ok  it may give much more protection in practice what can read from DC value. Rising from 50V to 200V feeels like some improvement but... think also if freq range bottom is more high, it may also give more protection in practice because it is high pass filter before all internals and more low frequency things are attenuated.

Pity these eBay sellers do not give enough data or data is copied and not exactly from what they are selling..
Also it need think if need DC/High pass block both, center and shield and not only center.

It's all starting to make a little more sense now. (to me)
So if we look at the datasheet spec:

Amplitude and Level
Measurement range DANL to +10 dBm, 100 kHz~1 MHz, preamplifier off
                                DANL to +20 dBm, 1 MHz~3.2 GHz, preamplifier off
Reference level -100 dBm to +30 dBm, 1 dB steps
Preamplifier 20 dB (nom.), 9 kHz~3.2 GHz
Input attenuation 0~51 dB, 1 dB steps
Maximum input DC voltage +/- 50 VDC
Maximum series RF power 33 dBm, 3 minutes, input attenuation >20 dB

Then apply each rating in turn to our intended use there seems to be quite adequate front end protection for normal use.

 Opinions to the contrary?

[Bicurico]

"Sensitive test equipment input protection ideas would have made for a great sticky thread. "

I am glad, my question was not as stupid as I feared...

Again, coming from TV/CATV/SAT field meters, I never had to worry about connecting the coaxial cable, but it would worry me to connect a unknown source to an expensive equipment like a spectrum analyzer.

Would it be wise to open a separate thread on this?

Regards

[hendorog]

"Sensitive test equipment input protection ideas would have made for a great sticky thread. "

I am glad, my question was not as stupid as I feared...

Again, coming from TV/CATV/SAT field meters, I never had to worry about connecting the coaxial cable, but it would worry me to connect a unknown source to an expensive equipment like a spectrum analyzer.

Would it be wise to open a separate thread on this?

Regards

Seperate thread is probably a good idea.

In the meantime, what about static charge which can build up on an antenna coax?

I know you should discharge the static before connecting the coax to anything sensitive like a SA or analyzer. Anyone know more about that?

[Bicurico]

Antennas, SAT and CATV should all be grounded, hence there should be no static at all.

I will open a separate thread.

Regards

[hendorog]

Antennas, SAT and CATV should all be grounded, hence there should be no static at all.

I will open a separate thread.

Regards

I'm not sure its that simple for all of the antennas out there. I thought the coax is acting like a capacitor and building up charge - grounding one leg of a capacitor won't discharge it.

[Bicurico]

I was referring to TV antennas and SAT/CATV distribution! These are definitly grounded - in Germany by law as a protection against lightening.

The signal is transported by the inner core (aluminium or copper) and the outer mesh (normally steel) is used for grounding the antenna and acting as a shield (hence the name) against interferences coming inside the cable (i.e. Dect phones) or against emissions from the cable core, causing interferences. In the US, it is mandatory to perform tests against signal leakage in CATV networks, to prevent interference i.e. in aeronautics communication.

I would not expect any static here, but you might end up with some voltage from DC@RF devices like SAT-receivers, amplifiers, etc.).

Not so sure about best practices in amateur radio antennas...

Regards

[hendorog]

I was referring to TV antennas and SAT/CATV distribution! These are definitly grounded - in Germany by law as a protection against lightening.

OK understood. Yes I was thinking of amateur radio antennas mainly.
 
These are the things I try(I'm being honest!) to remember before connecting:
* DC voltage
* RF Power
* Static/ESD