Author Topic: Why are spectrum analyzers so expensive?  (Read 7840 times)

0 Members and 1 Guest are viewing this topic.

Offline codingwithethanolTopic starter

  • Regular Contributor
  • *
  • Posts: 73
  • Country: us
  • always tilted
Why are spectrum analyzers so expensive?
« on: June 09, 2019, 09:22:11 am »
I've recently been making myself a mini lab so I can move past arduino and do some actual engineering. I wanted to get a spectrum analyzer to view frequency response of components, but I looked at the prices of the cheapest analyzers and had a stroke. What makes these things so expensive, are they made out of platinum and diamonds? Not saying that I don't understand why the super serious business pro models are expensive, but where are the budget hobby tier models?
its your boi, dj ethanol
 

Offline bd139

  • Super Contributor
  • ***
  • Posts: 23099
  • Country: gb
Re: Why are spectrum analyzers so expensive?
« Reply #1 on: June 09, 2019, 09:28:35 am »
Because they’re very complicated, the bits cost a lot of money and the engineering effort to design one is pretty extensive.

Not joking but they are pretty cheap now really. Just the SA plug in for my tek scope I had recently (7L12) with inflation would cost $25000 today. Add another $20000 for the scope mainframe and a couple of basic plugins for it so you get a scope too. So $45k down.

That Rigol or Siglent looks like a bargain now.

Or you can buy the 25 year old $45k one which is vastly inferior for $200 and fix it which is what I did.
 

Offline codingwithethanolTopic starter

  • Regular Contributor
  • *
  • Posts: 73
  • Country: us
  • always tilted
Re: Why are spectrum analyzers so expensive?
« Reply #2 on: June 09, 2019, 09:34:31 am »
WOW

Also, speaking of plugins, I don't have a 7000 series tek to plug that into, but i was also wondering why there isn't just a black box with some BNC ports you can connect to your oscilloscope to have SA functionality on a budget?
its your boi, dj ethanol
 

Offline codingwithethanolTopic starter

  • Regular Contributor
  • *
  • Posts: 73
  • Country: us
  • always tilted
Re: Why are spectrum analyzers so expensive?
« Reply #3 on: June 09, 2019, 09:56:09 am »
http://prntscr.com/nzhwh4

HRNNNNNNG
When I say budget im talking $100-$200. I've checked a few DIY SA projects but they are all far beyond my expertise to build

its your boi, dj ethanol
 

Offline bd139

  • Super Contributor
  • ***
  • Posts: 23099
  • Country: gb
Re: Why are spectrum analyzers so expensive?
« Reply #4 on: June 09, 2019, 10:04:07 am »
WOW

Also, speaking of plugins, I don't have a 7000 series tek to plug that into, but i was also wondering why there isn't just a black box with some BNC ports you can connect to your oscilloscope to have SA functionality on a budget?

Yes Thurlby TSA250 and TSA1000 spectrum analysers. they are pretty crap though. They come in under your budget.

It's possible to get a good deal but you have to lurk and there's a risk element.
 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 22436
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: Why are spectrum analyzers so expensive?
« Reply #5 on: June 09, 2019, 10:15:46 am »
There are handhelds, and RTL-SDR, and so on.  Haven't spotted any yet? ???

The good old ones are either well sought after (e.g., classic HP 8590 and such), high spec and quite practical (pretty much anything newer: faster sweep, PLL locked or DDS, FFT...), or both.  $1000 is a good starting point for a proper benchtop instrument that you want to do real work with.  (Coincidentally, the DSA700 is around $700, not a bad deal considering.)

As for why -- keep in mind, second hand market prices fluctuate with demand, so they aren't necessarily representative of the effort needed to build one of these.

The old ones are made from radio parts, exactly as given: a high frequency oscillator into a mixer, IF1 filter, another oscillator and mixer down to an easier-to-manage IF2, then detector and plotting.
- There's no preamp, or not usually; that would impair dynamic range.  Mixer also adds ~6dB mixer loss, but specs aren't usually optimized for noise figure.
- IF1 is quite high, above the input bandwidth -- to avoid images.  It's filtered to avoid images on the next stage.  It's converted down to IF2 where things are easier to handle.
- IF2 (or maybe there's an IF3, easier to manage images/filtering this way) is where most of the actual filtering is done (i.e., receiver bandwidth RBW).  These are usually separate, carefully tuned filter chains, selected by relay or whatever.
- Detector is linear or log, and peak or average reading.  Or something inbetween, like the quasi-peak used for EMC work.  More selectable signal paths.
- Final output might be plotting sweep and detector output on a screen, or a sampler, and plotting that on a digital readout, or transmitting via data port.

There may also be selectable IF stages, when wide input ranges are needed.  When you see teardowns of newer specs, with all those shiny planar filter circuits, you're most likely seeing some of this at play.

There's also the DSA (Digital Signal Analyzer), basically an ADC married to an FFT.  Classic ones like HP 3562A only do ~100kHz, but modern ones (like Rigol's DSA700) do RF in real time (what a difference a half dozen orders of magnitude in computation power can do, right?).

Which, on that note -- a DSO with FFT functionality isn't much different.  It won't have the spectrum-related features that a DSA does, but if all you want is to see some peaks, it's a good deal.

A DSA is basically a DSO with FFT on top, whereas a DSO is a DSA with waveform capture (triggering) on top.  Not too too different, but different enough to be separate products.

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 

Offline codingwithethanolTopic starter

  • Regular Contributor
  • *
  • Posts: 73
  • Country: us
  • always tilted
Re: Why are spectrum analyzers so expensive?
« Reply #6 on: June 09, 2019, 10:19:17 am »
Awesome recommendation but there arent any listings on ebay, so im gonna stay posted and wait for someone to put one up for sale. I had googled it and found 2 older listings that had gone for 60 and 250, so im hopeful
its your boi, dj ethanol
 

Offline exe

  • Supporter
  • ****
  • Posts: 2647
  • Country: nl
  • self-educated hobbyist
Re: Why are spectrum analyzers so expensive?
« Reply #7 on: June 09, 2019, 10:23:19 am »
Another "cheap" option: https://www.sdr-kits.net/

I myself going to try https://www.sdrplay.com/rsp2pro/ ... eventually.
 

Offline bd139

  • Super Contributor
  • ***
  • Posts: 23099
  • Country: gb
Re: Why are spectrum analyzers so expensive?
« Reply #8 on: June 09, 2019, 10:26:03 am »
Which, on that note -- a DSO with FFT functionality isn't much different.  It won't have the spectrum-related features that a DSA does, but if all you want is to see some peaks, it's a good deal.

Yeah they're not too terrible at that actually. Not a bad solution.

Here's one I did with a receiver I was building recently. Pre IF filter mixer products, on DS1054Z:



After crystal IF filter.



Obviously not great for measurements or particularly high frequencies but it does help design considerably.

I will buy a DSA815TG at some point however. Next time I need more than HF band SA.
 

Offline Psi

  • Super Contributor
  • ***
  • Posts: 10385
  • Country: nz
Re: Why are spectrum analyzers so expensive?
« Reply #9 on: June 09, 2019, 10:27:37 am »
Is there anything like the HackRF One but in a spectrum analyzer style box with a LCD/touchscreen ?

I'd love to get a basic spectrum analyzer with up to 6ghz for under $1k.
I'm not messing around with something that needs a PC, it has to be portable, ready-to-go in a box.

I think there is a market for something cheap that can get the 5.8ghz band covered.
A lot of people don't really need the advanced spectrum analyzers functions that you go without in a SDR style solution.
Since the hackRF One is $299 it should be totally doable to have something in a box with a screen for under 1k
« Last Edit: June 09, 2019, 10:33:31 am by Psi »
Greek letter 'Psi' (not Pounds per Square Inch)
 

Offline codingwithethanolTopic starter

  • Regular Contributor
  • *
  • Posts: 73
  • Country: us
  • always tilted
Re: Why are spectrum analyzers so expensive?
« Reply #10 on: June 09, 2019, 10:44:28 am »
@bd139 I have a budget tek right now but i was thinking about getting a little rigol down the line so thats good to know
@blueskull appreciate the input but those are still wallet biters
@T3sl4co1l yeah ive seen those but theyre mostly useful for hams, im trying to measure frequency response of filters and other circuits
its your boi, dj ethanol
 

Offline codingwithethanolTopic starter

  • Regular Contributor
  • *
  • Posts: 73
  • Country: us
  • always tilted
Re: Why are spectrum analyzers so expensive?
« Reply #11 on: June 09, 2019, 10:47:50 am »
@exe that looks interesting but im wary of all PC based equip due to trash software support
@Psi does that have any affiliation with the GS youtube channel? Interesting coincidence if not
its your boi, dj ethanol
 

Offline Bicurico

  • Super Contributor
  • ***
  • Posts: 1816
  • Country: pt
    • VMA's Satellite Blog
Re: Why are spectrum analyzers so expensive?
« Reply #12 on: June 09, 2019, 10:48:29 am »
Spectrum Analyzers are expensive for several reasons:

1) There is not a big market for them, so all engineering, research and development costs are split up between a fairly small amount of customers. This has changed recently with companies like Rigol and Siglent, mass-producing them, which lead to incredible affordable prices.

2) Spectrum Analyzers use expensive components. The core component of older swept spectrum analyzers (SSA) is the YIG, which is contains small Yttrium spheres for frequency down-mixing. If you search this forum, you will find posts from people trying to repair them and it is amazing technology - almost black magic! I still wonder how this can possible work at all. Then, modern spectrum analyzers offer FFT modes (Fast Fourrier Transformation). Instead of sweeping the frequency range, they digitize millions of samples and convert these time-domain samples into frequency domain. This requires a lot of processing power! It also allows a super smooth spectrum and waterfall display. Look at the 6-figure price tag Rohde&Schwarz spectrum analyzers and you get an idea...

3) High Frequency stops being "reasonable" at around 1GHz. Once you get above that value and especially above 2GHz, things start to behave oddly and you really need to take care of every detail: connectors, cables, everything needs to be "special" to not interfer with the signal you are analyzing. This becomes extremely expensive!

Regarding your choices:

50-200 Euro: You can purchase chinese clones of the NWT series. These work surprisingly well and I developed an alternative software for these devices. Take a look at my blog about this: https://vma-satellite.blogspot.com/

500-2000 Euro: You can consider second hand spectrum analyzers. These were normally very expensive machines, which are now sold at ridiculously low prices. But beware that you get no support, calibration and often no spare parts. What I have done is buy several broken units and then repair them, using some for spare parts. This way I ended up with 2x R&S CRTU, 2x R&S CMU200 and 1x HP 8594E. All in working order. But I have another 2x HP8594E (broken PSU) and 1x R&S CRTU (missing some boards) and spare parts. This takes space, time and money to get them in working order. And you still have no valid calibration, if that is important for you.

1500 Euro: Get yourself the best performance/price ration currently on the market and buy a Siglent SSA3021X. New, calibrated with warranty. And it can be hacked for increased bandwith (3.2GHz) and all options.

Regards,
Vitor
 
The following users thanked this post: plurn

Offline Bicurico

  • Super Contributor
  • ***
  • Posts: 1816
  • Country: pt
    • VMA's Satellite Blog
Re: Why are spectrum analyzers so expensive?
« Reply #13 on: June 09, 2019, 10:57:49 am »
Forgot to add some thoughts on SDR:

Yes, you can buy a HackRF One and run this amazing software: https://github.com/pavsa/hackrf-spectrum-analyzer. But you will soon find out that functionality/measurements are limited and that the software/device crashes a lot.

Then you have the ADALM Pluto: I cannot recommend this as a spectrum analyzer replacement, at all.

SDR: Yes, with SDR# (sdrsharp) you can do spectrum analysis and even demodulate most analog and some digital signals! BUT: In my opinion this is in no way a replacement for a real spectrum analyzer. These devices do not support a swept spectrum analysis, so that your frequency range is limited (4MHz to 40MHz, depending on the signal). If you want to look i.e. at the CATV band from 50-850MHz, you won't be able to do that, unless you buy the SDR# specific AirSpy, which costs a lot more than a regular RTL2832 dongle.

Plus, most tools are Linux only and you can literally spend hours to get a particular driver/software up and running. GNU Radio is great and you can learn so much using it - especially if you are interested in HF engineering and have a Python/C++ programming background. Sounds complicated? It is!

In my honest and modest opinion, if all you want is a spectrum analyzer, then either go cheap with SMA/NWT devices or buy a SSA3021X or spend some time reading about which are the best second hand devices to buy. I think there was a whole thread here about second hand spectrum analyzers.

Good models are HP/Agilent 859x series (beware of the PSUs - they are a nightmare to repair: I have 3 broken ones and repair seems impossible) and Rohde&Schwarz CMU200/CRTU - these are actually Protocol/Communication Tester, but offer a fully working spectrum analyzer, too.

Regards,
Vitor

Offline codingwithethanolTopic starter

  • Regular Contributor
  • *
  • Posts: 73
  • Country: us
  • always tilted
Re: Why are spectrum analyzers so expensive?
« Reply #14 on: June 09, 2019, 10:58:53 am »
Why is Yttrium required???? I had a hunch these things ran on fairy dust but HUH?
its your boi, dj ethanol
 

Offline codingwithethanolTopic starter

  • Regular Contributor
  • *
  • Posts: 73
  • Country: us
  • always tilted
Re: Why are spectrum analyzers so expensive?
« Reply #15 on: June 09, 2019, 11:08:05 am »
Anyways the reason I started this thread is because im working on student budget here. I had gotten a decent tek and psu and function generator for <$200 and couldnt justify shelling out more than that for ONE piece of equipment. Im taking all the suggestions made into consideration, also I found this thing:
 https://www.ebay.com/itm/U-V-UHF-VHF-Dual-Band-Spectrum-Analyzer-w-Tracking-Source-136-173MHz-400-470MHz/292520566371?hash=item441b959e63:g:gBoAAOSwQFVapi2a
But it looks like it'd shit itself and blow up if I so much as looked at it wrong. Also I found a Hameg SA at auction just now but it was a little over the $200 mark so I didn't pull the trigger.
its your boi, dj ethanol
 

Offline HB9EVI

  • Frequent Contributor
  • **
  • Posts: 722
  • Country: ch
Re: Why are spectrum analyzers so expensive?
« Reply #16 on: June 09, 2019, 11:28:52 am »
Most things were already pointed out. I think, analyzers nowadays are quite affordable;
first it was the Rigol DSA815TG, today it's the Siglent with better technical data than the Rigol for entry class.

After having twice time tough luck with 2 older storage Tek scopes breaking down irrepairably, I decided to not buy anymore old measurement devices, so I went for the Rigol which does an ok job here; if you know about its limits you can fairly work with it and I'm using it rather often - it's certainly much better than all those FFT functions modern DSOs offer, what can imo not be seen as 'measurement'.

If you personally see a need for a SA, new or second/third hand go for one with a tracking generator; it's a must-have for many measurements
 
The following users thanked this post: codingwithethanol

Offline Bicurico

  • Super Contributor
  • ***
  • Posts: 1816
  • Country: pt
    • VMA's Satellite Blog
Re: Why are spectrum analyzers so expensive?
« Reply #17 on: June 09, 2019, 11:56:46 am »
I would not recommed wasting money in this device.

For your budget you can only buy a SMT/NWT device, which requires a PC.

All options with a display is money thrown away.

Regards
Vitor
 
The following users thanked this post: codingwithethanol

Offline Theboel

  • Frequent Contributor
  • **
  • Posts: 278
  • Country: id
Re: Why are spectrum analyzers so expensive?
« Reply #18 on: June 09, 2019, 12:14:53 pm »
I think You don't give information about how high You want to go 100MHz, 2GHz ?

based from Your post :
"I've recently been making myself a mini lab so I can move past arduino and do some actual engineering. I wanted to get a spectrum analyzer to view frequency response of components, but I looked at the prices of the cheapest analyzers and had a stroke. What makes these things so expensive, are they made out of platinum and diamonds? Not saying that I don't understand why the super serious business pro models are expensive, but where are the budget hobby tier models?"

You can try to use Arduino for made Your own spectrum analyzer, something like this perhaps
 
The following users thanked this post: Mechatrommer, codingwithethanol

Offline Mechatrommer

  • Super Contributor
  • ***
  • Posts: 11714
  • Country: my
  • reassessing directives...
Re: Why are spectrum analyzers so expensive?
« Reply #19 on: June 09, 2019, 03:07:05 pm »
Which, on that note -- a DSO with FFT functionality isn't much different.  It won't have the spectrum-related features that a DSA does, but if all you want is to see some peaks, it's a good deal.
Yeah they're not too terrible at that actually. Not a bad solution.
with PC aided SW, one can do much more... GSps and vertical bit resolution are the only limitation.
Nature: Evolution and the Illusion of Randomness (Stephen L. Talbott): Its now indisputable that... organisms “expertise” contextualizes its genome, and its nonsense to say that these powers are under the control of the genome being contextualized - Barbara McClintock
 

Offline bd139

  • Super Contributor
  • ***
  • Posts: 23099
  • Country: gb
Re: Why are spectrum analyzers so expensive?
« Reply #20 on: June 09, 2019, 03:14:13 pm »
Might have a play with that. The software FFT on the Ds1054Z is a bit poo for some things to the point for low frequencies I use a sound card and Spectogram for it instead.
 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 22436
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: Why are spectrum analyzers so expensive?
« Reply #21 on: June 09, 2019, 03:50:07 pm »
Why is Yttrium required???? I had a hunch these things ran on fairy dust but HUH?

"Physics.  It works, bitches." :-DD

That was an incomplete description, but yttrium is indeed involved -- the module is a YTO (YIG Tuned Oscillator).  The active ingredient is a YIG (yttrium iron garnet) sphere.  YIG is a ferrite (i.e., a compound of some metal oxide with Fe2O3, or to put it another way, iron(III) ferrate), specifically one taking the garnet crystal structure.  (Whereas most ferrites are in the spinel group*.)

*Chemistry aside:
The mineral spinel has the formula MgO.Al2O3, but the Al can be substituted with Fe giving a ferrite, and the Mg with a variety of metals (Fe, Mn, Zn, Ni, etc.), making this a huge family of continuous mixtures of similar atoms.  Soft ferrites are (Mn,Zn) and (Ni,Zn) ferrite.  The mineral magnetite is actually iron ferrite, FeO.Fe2O3!

In terms of electrical properties, YIG has lower saturation flux density (because there's less active magnetic iron, and more diluting or opposing atoms), lower relative permeability, higher resistivity, lower losses, and notable paramagnetism.

It's the last property we're interested in.  When subject to a static magnetic field, unpaired electrons (the part responsible for bulk magnetism in the first place) precess at a frequency proportional to that field.  (The physics is similar to the precession of a spinning top, where the precession rate depends on the force trying to tip it over, i.e., gravity in that case, ambient magnetic field in this case.)  The precession in turn manifests as electromagnetic energy (photons; this is a quantum process too, after all), which is to say, light waves.  Or at typical field strengths, radio waves more specifically.

So the really interesting property is this: when an inductive loop is coupled to a YIG crystal, and the crystal is placed in a magnetic field, there is a blip in the frequency response of that loop, which varies with magnetic field intensity.  The blip looks much like the impedance response of a quartz crystal, having a series and a parallel resonance, and you can tune in either one.  Or, as far as I know, it's usually done with two perpendicular inductive loops, so that they are only coupled through the YIG crystal's resonance (getting a bandpass characteristic), and serve as the input and output terminals of a microwave-frequency resonator.  Add an amplifier and buffer, and you've got a YTO module!

The tunable range doesn't go all the way to zero, of course; that would be kind of horrible to design (those inductive loops won't work very well at DC!).  Usually it's, like, say, 3-6GHz or such.  Which is perfectly fine because as mentioned before, LO1 will be somewhere at high frequency

A few other practical considerations include magnetic design (the module housing is usually steel, cut in a "pot core" sort of shape to give good shielding to external fields, and better field intensity (and uniformity) around the crystal itself), thermal design (the electromagnets are normally wired differentially, so as to maintain the same total heat dissipation while varying the total field), and so on.

So, because the YTO is tuned by magnetic field, it can be tuned electronically.  Fantastic, huh? :)

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 
The following users thanked this post: blueskull, Gregg, genghisnico13, bd139, DC1MC, codingwithethanol

Online ebastler

  • Super Contributor
  • ***
  • Posts: 7375
  • Country: de
Re: Why are spectrum analyzers so expensive?
« Reply #22 on: June 09, 2019, 04:14:37 pm »
I think You don't give information about how high You want to go 100MHz, 2GHz ?

Yep, without any information about the required frequency range, this discussion and any recomendations are flapping in the breeze.
@codingwithethanol: What signals do you want to look at?
  • In the audio range, a decent sound card and free FFT software are a good bet. You can even get a combined sweep generator and Bode plotter for your sound card.
  • In the 100 MHz range (order of magnitude), the FFT function on entry-level DSOs may do what you need -- or, again, the Bode plot function some scopes offer.
  • If you want to look at GHz frequencies, things get expensive...
 
The following users thanked this post: codingwithethanol

Offline RoGeorge

  • Super Contributor
  • ***
  • Posts: 7012
  • Country: ro
Re: Why are spectrum analyzers so expensive?
« Reply #23 on: June 09, 2019, 05:01:33 pm »
Why is Yttrium required???? I had a hunch these things ran on fairy dust but HUH?

If you want to know more, take a look inside a YIG oscillator (starting from minute 33:28):


Offline Bicurico

  • Super Contributor
  • ***
  • Posts: 1816
  • Country: pt
    • VMA's Satellite Blog
Re: Why are spectrum analyzers so expensive?
« Reply #24 on: June 09, 2019, 05:30:36 pm »
Thanks T3sl4co1l and RoGeorge:

YIG/YTO are magic, no matter how great the explanation on how they work!

:)

Regards,
Vitor


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf