EEVblog
Vložit
- čas přidán 5. 07. 2018
- Teardown and look at the new $1395 Siglent SVA1015X 1.5GHz Spectrum and Vector Network Analyser
Well, $2000 when you include the actual VNA option :-/
SSA3021X Spectrum Analyser Teardown: • EEVblog #892 - Siglent...
Forum: www.eevblog.com/forum/blog/eev...
EEVblog Main Web Site: www.eevblog.com
The 2nd EEVblog Channel: / eevblog2
Support the EEVblog through Patreon!
/ eevblog
Stuff I recommend:
kit.com/EEVblog/
Donate With Bitcoin & Other Crypto Currencies!
www.eevblog.com/crypto-currency/
T-Shirts: teespring.com/stores/eevblog
Likecoin - Coins for Likes: likecoin.pro/@eevblog/dil9/hcq3
💗 Likecoin - Coins for Likes: likecoin.pro/@eevblog/dil9/hcq3 - Věda a technologie
I'm not sure how you can release a VNA with that name on the sticker and make the VNA optional extra, marketing innovation?
It's like releasing a Digital Storage Oscilloscope and making the storage optional extra.
Won't take long before someone jailbreaks it.
yup...it will be an easy firmware update...
Well, you can't be sure that its an "easy firmware upgrade". They could have different daughterboards on the models.
of course they could...and of course you would need to compare the 2 models before hand...but its not out of the question for stuff like this
jort93z - No, they don't. You can buy just buy the software upgrade that enables VNA. Also, look at the teardown, the VNA stuff adds pittance, which is why they are able to include it. Siglent actually rejected the idea of a model (SSA1015X) without the VNA capability.
I think the sharp edges on the shieldings are a secondary "Waranty void if removed" device. If somebody cut themeself there would be blood on them and easy checkable if it's the blood of the owner.
Huh? i've never been so in over my head watching a video and trying to understand it. So much to learn. Thanks Dave!
Thanks Dave. Long time lowballer on Patreon bit you've helped me in my professional life volumes. Keep up the great content and telling it like it is. BNE fanboy.
The prices on those options are insane. This machine would never get near my workshop for that reason. Excellent review!
The VNA capability is wonderful for anyone doing filter design and impedance matching/feed networks. S11 and S21 is all you need.
The processing on the daughtercard is a really spiffy idea! Solid points, regarding reducing the size of the multilayer board area, and potentially standardizing the processing across products!
Any chance on getting hands on a rigol 7000 o-scope? Also, I want the batt bump enabled version.
I find it really odd that in a bit of $1k+ test gear there is corner cutting on the filter cap but yet on a $500 projector I took apart that failed every single cap on the PSU is Rubycon
The picture quality of this video is amazing.
I notice at 40:30 that D46, D47, D48 are not installed in the tracking gen, I suspect that they would of been protection diodes for reverse DC input, maybe they effected the output in some way and were not really needed because of the capacitive coupling from the port anyway.
My friend just bought a new build house where interior flooring wasn't included in the price. This video reminds me of that.
After watching this I feel like my BS in EE I got last month is just not enough. Absolutely interesting video though, loved it.
What you miss is material used for those filters. In "highish" version is there some ceramic substrate used for those filters (like Rogers).
Not sure if you mentioned it, but is the SSA3021X model using a PCB with a rogers RF material top layer and the new SVA1015X is getting away with regular FR-4?
How much extra is the main power supply?
LOL
Hi Dave, the crappy Leylon 400V Cap only has to worry about 100Hz ripple
When refurbishing some repairable SMPS I have not found any failed DCBus 400V caps.
It is good to see better quality caps in the HF switchmode part of the supply were ESR can be problem particularly if they are high density value caps or low value/high voltage types.
Yeah, even the shittiest primary input capacitors don't seem to fail in the SMPS's I've repaired.
The options policy killed this, sadly. But I really am a sucker for looking under the RF shielding, just delightful!
It's always a joy to tear down RF stuff.
Pity about the de-modulation not being enabled at 53:00, I wanted to see how that worked!
Me too. I just assumed it had all the options enabled
EEVblog - I’m sure if you ask they will tell you the unlock code.
Many "HMC" devices went obsolete after Analog purchased Hittite. Some of them only temporarily, appearing back in production with a new part number, some permanently. Some devices became "obsolete" overnight. We had a design ready for review by our international customers, only to be told they can't accept a new design with obsolete components.
It seems like this thing has a ton of optional features. I don't know if they were all mentioned in the video. Is there a list of all the optional features and what they cost (what would the total cost be for everything)? At around $300 per feature, it adds up quick..
Love the saw-tooth directional couplers. They switched to a shitty laminate, compared to the previous PCB. I wonder if those obsolete Hittite parts are counterfeit? I used the HMC-488 years ago.
Looks like the ribbon cable for the display was a bit skewed in the plug on the driver board.
53:36 Hahaha a weekend of Sundays would be one Sunday, so a short amount of time. I think you invented a nice new version of the phrase!! My dad used to say "not in a bloody month of Sundays" all the time :)
what happens if you were to try to analyse a signal below the minimum frequency?
A big part of deciding to pick up the Siglent SDS1204X-E scope was that so many "options" were standard -- signal decoding among those. Bummer that Siglent have returned to making so many things option$. In a SA, I need coverage through 2.8 GHz. I have an old HP 8594E SA with the 85630A scalar network analyzer that does a very nice job (it's portable -- it has handles). Let's see what Siglent (or Rigol) do in a 3 GHz model with TG and VNA. You can get okay 3 GHz cal kits (open, short, 50 Ohm load) on eBay for around $200, look for OSL cal kit.
Didn't get how they measure S11. Where is the coupling device?
Maybe touch anywhere else on the screen to "get rid" of the menu.
Can anybody tell me if the ‘N’ connectors on the Rigol 832 and the Siglent SSA 3032X spectrum analyzers are the same distance apart? I want to buy the Siglent, but their directional bridge is only good up to 2GHz, while the Rigol one is good for the whole 3.2GHz. I’m going to try to buy the Rigol directional bridge, to use with the Siglent, if it will fit without using extra cables. Thanks!
Hey now, Ne nice to the watch crystal. I know a DMM that originally had the watch crystal flapping in the breeze. :-)After it was pointed out, the PCB didn’t have a pad to solder so the manufacturer decided on hot glue to hold it down. Just took apart a smart meter hub that did the block gold traces since it had Bluetooth, zigbee, and WiFi, just no shielding.
That 3M stuff is high temperature double stick tape, not the cheap double sided tape like in some. It must be sticking down a cable I’m guessing so it doesn’t work loose?
If there is coupler (or directional bridge) that will reduce noise floor. That will explain lower noise floor then spectrum analyzer.
does it have multiple attenuation levels on the 1015 vs a single option on the 3000?
I like that they are using a BR cell instead of a CR cell. BR cells are more expensive and don't have the current capacity of CR cells but they do have a substantially lower rate of self-discharge.
Under 2K for a VNA /SA combination is a winner, shame it's upper limit is 1.5GHz though.
Yeah, bit of a shame.
I wonder if the high cost for the VNA feature (and the fact that its optional) has to do with patents, i.e. someone out there holds patents on the tech involved and either charges a large royalty (hence the high cost) or insists on a high price point as a condition of the license (because they dont want Siglent to undercut whatever their own product costs by too big a margin...)
If there are patents involved, I bet they are Cisco ones.
Can you do a review about analog arts scopes ?
Have you done a video on the Siglent 1200 series Oscilloscope?
Any idea why most spectrum analyzers don't cover audio frequencies? This one the BW starts at 9 kHz and that is typical.
i need one of these for my RF adventures!
the base material of the distributed element filters locks much different in both instruments -
could that be the reason for the larger structures ? like, more expensive ceramics in the 3GHz analyzer as opposed to cheaper ceramics or kind of epoxy or similar in the lower bandwith sva ... ?
Isn't the larger size simply because they're lower frequency? If I were to guess, I'd say that's why they've converted some of the filters to discrete components as well, because they'd start being impractically large as DEFs.
you are right, I was shooting too fast :-) ... ...
different materials could be used, because losses in a (supposedly) lower priced material are frequency dependend and less contributing at half the maximum frequency
Can you comment of the relative value of the very lowend nano-vna versus this sort of VNA? For a ham I am not seeing the big justification for the 10+ X price.
I'm wondering what a quote from a dealer for this instrument fully loaded would look like? Last time I ordered instruments with add-ons, some of the add-ons were bundled together and the price of the bundle was significantly less than the prices listed on the manufacturer's website. The bundle offered also included some features I wouldn't have otherwise purchased.
I'm still going to agree with everybody, if it says VNA on the box, it should come off-the-shelf with a VNA in the box.
Dave you've got a typo in your description "SSA3021X Specrtum Analyser Teardown".
LOL there is a not soldered ceramic filtercap in the Section between IF Output and 12bit ADC --- C112 is open :D
In the topdown pictures its a little hard to see but still noticeable (better distinguishable on the angled view found on the Daves Flickers page)
Good to see that even big brands still have such easy escapes with AOI inspection.
I'm considering this puppy for my lab and ham adventures. As of 2019.09.23 Siglent started including the VNA option for free which is fantastic. TEquipment currently has it listed for $1635.62. It doesn't seem like there's much in the way of competition right now so I may have to bite! 73 de KN6MEC
I know I'm a bit late, but the HMC213B is not obsolete, only the HMC213A is. For some reason Analog Devices does not have a recomendation for the HMC213B on the product page of the HMC213A.
Useful video
20:39 Good question - why would they shove 3M VHB tape down there? Don't get me wrong, I love it - use it at work and around the house for lots of stuff. And even some things that don't require me to peel back the other side (making it essentially one sided super tape..) which this appears to be - as neither an adhesives or electronics expert; only a hobbyist, I'm sort of curious. When I use it, I'm using it mostly for bodge jobs, so perhaps I have a jaded perspective on it's practical industrial and electronics uses.. :)
where do you find these? must be a dumpster near a factory
So this stops just too low to work with the 2.4GHz ISM band (RC toys, Bluetooth, WiFi etc.) let alone the ~5GHz bands used for modern WiFi,cordless phones etc.
But it might be able to analyze wired Ethernet, TV broadcasts etc.
Jesper Henriksen And that'll be just for 2.4GHz and GPS, I guess?
That 3M is VHB tape, a type of goo like tape from them, which acts as a double sided tape, good adherence in general
Replacing the Bowties with SMD components probably means they were throwing away a lot of PCBs that didn't meet the standards, or the material for manufacturing to the tolerance was significantly more expensive.
Vector Network Analyser: $609
Demodulation Analysis (ASK/FSK): $629
Distance to Fault: $289
EMI: $449
AM/FM Modulation options: €319.00 = $375
Advanced Measurement Kit: €364.00 = $428
So... $2779 in total for all upgrades!
Doesn't include the hardware necessary.
That sort of milling is also found inside LNB's if you ever took a LNB apart, I just hope that you can put it back together without any problems, you can do the same analysis by doing a IFFT then to get the magnitude by doing sqrt((i*i)+(q*q)) and the phase by atan2(i,q);, I think it would be wise to just use 4096 bins and not to go above this as it gets progressively slower.
Do people hack these? Send worth it for $600 saving?
I'm sure someone will hack it.
I would be somewhat hesitant to buy this spectrum analyzer for radio amateur (ham) use, as it does not cover e.g. the 40 m and lower bands. If I read the specifications sheet correctly, the tracking generator starts at 5 MHz and the network analysis option even higher at 10 MHz. Tracking generators in other low cost products from Siglent and Rigol start as low as 100 kHz, although they do not have the network analysis option. However, radio amateurs may find these products more useful if combined with a separate ham-oriented analyser such as the Funkamateur FA-VA5.
The Tracking Generator on the SSA3021X is listed as going down to 100KHz, while the generator in this bottoms out at 10MHz.
It's another point to consider, if you're trying to choose between these models
Actually, the Tracking Generator on the SVA1015X starts at 5 MHz for the spectrum analyzer. The VNA option does not go below 10 MHz.
Love to get your take on the new NanoVNA compared with the Siglent VNA.
those gold plated boards are "Schmick "
This is some nice piece of engineering.
This dude didn't notice the metallisation on the side of the tracking generator board? It's always nice when this dude tells you something without actually not knowing what he is talking about.
SO does Siglent provide any form of calibration kit for their VNA? That's normally a pretty spendy of add-on for most VNA's....
I mentioned that. Yeah they do but it's optional
Thanks! Love your channel BTW.
It is only a one port VNA. The other port is scalar not vector?
would like you to test the cheap portable VNA found on ebay for a couple of $, just to see if they are worth for the hobbyist... like you did test the cheap component tester identifier, you could do the same for the cheap VNAs...
I like that they are starting to mimick an UI of the Rhoden-Shwartz equipement. RS uses Windows afaik and some native code, and Sigilent and Rigol use Linux, but they managed to make it similar, and more importantly very usable and nice. I wonder if they use Qt framework maybe.
I can't really imagine that many useful cases for a VNA that can only do up to 1.5GHz, especially given that almost everything is 2.4GHz and above nowadays.
Kontra ever heard of amateur radio? All the really interesting stuff is from 0-50mhz.
To be useable on the microwave bands, you'd need far higher frequency support for the harmonics
Can you do everything with the buttons or is there anything you can only do with the touchscreen?
I don't think you could do everything with the touch screen, but not 100% sure.
More, except power off
If it was only 400 mhz higher in spec i would grab one, so close, maybe next year
Wow, I was excited until I heard the price. I will just rent one when I need it at that price point.
Are these hackable for some options?
If so, I'll get one
We saw a Chicken Caesar Salad on a menu and it was an additional $3 to have chicken on it. So the Chicken Caesar doesn't come with chicken...
Plus $4 for Caesar dressing, plus $2 per crouton, plus $7 to rent a bowl to eat it out of ($125 partially-refundable bowl deposit required), but free unlimited* salt and pepper (limit 1 ground peppercorn or 1 tsp salt per customer).
You should've tried the "marker" buttons while in VNA mode.. Awesome video
The Marker function does as you'd expect, also delta marker works. But that's it as far as analysis goes AFAIK.
It will only be good enough for pre compliance work if your UUT only clocks up to 200 MHz, otherwise you need to go 5 times the max clock speed
At 53:03 you see that the tracking generator and DMA are extra paid options. In the device in this video they were included as temporary licences. The optional extras Dave listed in this video add up to almost $2,700 for a device that retails just under $1400. I don't remember if that included the tracking generator. If they included the TG and VNA options as stock for just under $1400 it would be a possible killer of the Rigol DSA-815TG.
The tracking gen IS included in the base model.
The hardware is included in the base model. The licence to use it is extra. If you look at the screen shot at 53:03 you see that the model you were using in the video only had a 128 hour licence for the use of the TG and DMA. After the time expired you would no longer be able to use either feature without paying for the options.
is this... test gear teardown in 4k?! Oh my god.
Hi Dave! Please make a review of NanoVna
I'd like to see this compared with a very similar unit, the R&S FPC1500. I'm aware of the significant price differences and general limitations between the two, however this introduction definitely peaks my interest to see which one would fulfill my "home lab" for my graduate studies.
Whilst I'm quite impressed by Siglent's significant improvements over recent years, my experience with them has been "fair" compared with the usual "high price brands."
EDIT: "The Signal Path" ep. 121 has an in-depth review of the FPC1500.
I haven't seen that one yet. Bit pricey for 1GHz with S11 only VNA, but it's expandable. But it IS an R&S
Hmmm. I just looked over at the EEVBlog website and haven't seen a forum posting yet. But yes, the FPC1500 is actually much higher price as expected.
Btw, I tried to look, at whether the Siglent's Smith Chart will work with S21, which their website does say and the menu option during your Smith chart appears as an option, but the user manual isn't very clear whether it is or not. The FPC1500 data sheet doesn't appear to allow S21 Smith Chart displays but does appear to measure S21 magnitude.
Just a few thoughts before I carry this topic over to the EEVBlog page.
$1300 with 20k in options? Lost interest.
Until someone figures out how to hack it.
Wholly Mindless it's just pirateable software options, so pretty awesome once people figure out how to crack it
It's like dlc in real life
How do you know they are piratable software options? the processors are on a seperate boards. They possibly use different daughterboards on different models.
jort93z - As per comment above, the hardware is the same, the VNA is a software option.
Shame that they used only 2 screws (instead of intended 4) to fix the N-connector. That is the only mechanical support and you have always a lot mechanical stress on the peripheral connectors, especially for those main IN/OUT connections... But otherwise seems very nice apparatus 🙂
I saw the thumbnail on my phone, and thought you’d ported Stardew Valley to an oscilloscope.
It's like buying a car with, No wheels, no engine, no transmission, no interior, and no sheet metal, all optional extras.
How much do they charge for a full wave rectifier power supply board and the screen?
A $600 upgrade each too?
I still think the KC901V is the best value performer in this field.
Scaling is limited to 1 dB/. Even the basic DSA815 goes to 0.1 dB/. My 1990s vintage HP VNA goes down to 0.001 dB! VNA option is limited to 10 MHz or higher! Shouldn't these have been mentioned in the review?
With the price of these things, not using an IPS LCD is really quite sad...
TN panels are cheap, but this kit is not. Gimme an IPS screen please! I want good viewing angles!
Personally I like S-parameters with dB rather than a Smith chart as it's more intuitive for me to read the amount of power being reflected or transmitted at a frequency of interest (and the phase in some cases). Smith chart seems a bit old school for me.
Yeah, i's just a different way of looking at the same thing. Some prefer one to the other.
derre98 the dB magnitude is for gain and return loss, so might as well use a scalar analyzer. With the smith chart, you can actually see what needs to be done for impedance matching, though you must have a good calibration with well defined reference planes.
S-parameters include the phase as well and I use phase information so I need a vector network analyzer. I just most of the time fit some data to S-parameters to characterize some samples over wide range of frequencies rather than doing simple impedance matching so never gotten into the Smith chart much. Same information, different representation.
1.5 GHz is very small. VNA should be atleast upto 6 GHz.
51:00 EEVblog presents the world’s first one-wire Quine.
I hope someone figurers out how to hack all of the software upgrades soon. I would love to have one, but I will not pay an extra $2000 for license keys to use the hardware I buy.
would be nice if someone started to make "open hardware + freesoftware + all fancy gnu license" scopes and spectrum analyzers for hobbiest
Great video Dave but SCREW siglent for the nickel and dime BUTT REAMING for options.. I'll keep my older HP SA and VNA which I can repair wo BUTT REAMING !!
No way... Just discovered this a few hours ago...
Same
Not a bad looking unit. Definitely a lot of "moist" HAM's around the world. :-)
Since you mentioned Rubycon capacitors. Did you know there is a class action lawsuit here in the US against them, Nippon Chemi-Con and several other companies. Apparently they have been price fixing prices since 2002. Screwing us for 16 years!!! Got my notice in the mail the other day. www.capacitorsindirectcase.com/
Yeah, they'll just pay the fine which is way under what they would have gained from the price fixing.
EEVblog
And no future business from me. I'll never buy any of their capacitors again.
Calibration kits are not free to make regardless of how low-end they are assuming they're N-type and still within the range of being usable for general purpose stuff up to 1.5 GHz. Open, short and thru are cheap enough, only so-so machined blocks of brass in the cheaper end of the spectrum, but a decent 50 ohm load will always cost you something. If you're looking at a $2000 instrument, they're not going to throw in a $350 cal kit. That would absolutely destroy the bottom line and plenty enough people would complaint about the quality anyway. RF engineers and HAMs get the level of calibration standards they require. Some need $20,000 kits per connector type and some are fine with getting a simple set of $300 standards. There's a whole spectrum of different price points available for calibration standards.
Teardown starts at 11:25.
Distance to fault $289 - NanoVna with 4.3 screen < $150 . BAM
I think those smith charts would look far better if you calibrated the thing...
I'm suffering a cognitive dissonance. You turned it on before you tore it apart.
What the?
;^)==
You got that right!
Maybe it can be modified to add a battery
The thumbnail looked like were playing zelda on that oscilloscope