Educate myself on spectrum analyzers

[JPortici]

Good day folks,
At work we are starting to tamper with 433/860 MHz transceivers and we don't have a spectrum analyzer.
I also have never really used one, not even at uni as crazy as it may sound, so now that we (I) have to choose the next lab tool i face myself with a serious lack of knowledge on the matter.

I have read several threads and documents all over the web but i wasn't really able to find what i was looking for, or maybe i don't know what to look for

I understand the basics, I undestand how the instrument work but i'm not sure what else i can do or what other info i can gather from the spectrum besides "good, there's the carrier"

So, Question number one: Where can i find resources to educate myself? Talking about books, videos, ...
Number two: what else do i need besides the SA to look at radio waves? A piece of wire directly on the BNC? An Antenna? which kind?

I also have some difficult in choosing what to pick. Budget should be around 2-3k€.
Beginner in this field, i obviously look at Rigol, usually works fairly good without costing too much, but the 815 goes to 1.5 GHz. We might want to play with gps in the future so 2 GHz minimum.
The new siglent is pretty nice on paper, but it's new, it's siglent and i have been following the topic.
I wouldn't buy used gear on this one.

Ant that's that..
Well for now, while i still look around and learn i think i can get one of those 15€ dvb dongles and use them with rtl-sdr, at least i should be able to know that there is the signal (or not)
If sombody can give me a hint? Much Appreciated

[Psi]

You can get the 1.5GHz Rigol with tracking gen for US$1,495

There is a Rohde & Schwarz in your price range too, around US$2,800

[Smith]

Look for some diferent companies that sell something you might consider based on specs/price. Ask them to come over and show them what you want to do, and ask them for what they think might be a good solution for you. Tell them you still consider other brands too, just be honest about that. They can probably lend you some demo tools to try them for yourself. Most companies have no problems with this and are eager to help you out / take your money.

It not just handy for the tools, but they mostly come with some demo probes/leads etc. This is verry handy to test too, because probes and leads can become verry pricey too.

[rf-loop]

You tell you need spectrum analyzer in work. Serious work?  If you need do 860 MHz things seriously you need also access to its harmonics. With 1.5GHz spectrum you do not see even 860MHz 2rd harmonic!
Some kind of minimum is that you see 3rd but better if 5rd (and in some cases it is mandatory). So you can check spurs-harmonics are in accepted levels. You need least 3GHz spectrum analyzer for any serious work with 860MHz. With this budget you can get very good second hand professional class spectrums instead of buying new toy hobby class "spectrum".   If it is hobby and you are not responsible about anything with measured result data then you can play with toys or with limited features, eg without looking 3rd or 5rd harmonics levels.

[tggzzz]

At work we are starting to tamper with 433/860 MHz transceivers and we don't have a spectrum analyzer.

I understand the basics, I undestand how the instrument work but i'm not sure what else i can do or what other info i can gather from the spectrum besides "good, there's the carrier"

So, Question number one: Where can i find resources to educate myself? Talking about books, videos, ...
Number two: what else do i need besides the SA to look at radio waves? A piece of wire directly on the BNC? An Antenna? which kind?

The most important other thing you need is "objectives". If you don't know what you want/need to measure and why, then measuring anything is equally (in)effective.

You should start by determining what your equipment must, legally, not do. That will guide you in your dealings with the relevant local and international authorities, who are typically only interested in ensuring that you don't screw other parties.

If that is too expensive (in time or money) then you should purchase pre-qualified industrial modules and/or purchasing the expertise of those with experience in this topic. RF is a different world to analogue, digital, and software.

[pascal_sweden]

Is there some tutorial available online, with an overall list of what different types of measurements you can do with a spectrum analyzer, and maybe also a short description on what and how.

In this overall list, I would like to see the the following entries for each type of measurement.

Overall list with multiple types of measurements, for each type of measurement the following entries:
...
Title of measurement
Description of measurement
What to measure exactly
What is the center frequency of interest
What is the actual bandwidth which you need (e.g. harmonics, etc.)
How to measure with the spectrum analyzer
Which decision based on measurement
Which correction to make based on measurement
...

[Wuerstchenhund]

Good day folks,
At work we are starting to tamper with 433/860 MHz transceivers and we don't have a spectrum analyzer.
I also have never really used one, not even at uni as crazy as it may sound, so now that we (I) have to choose the next lab tool i face myself with a serious lack of knowledge on the matter.

I have read several threads and documents all over the web but i wasn't really able to find what i was looking for, or maybe i don't know what to look for

I understand the basics, I undestand how the instrument work but i'm not sure what else i can do or what other info i can gather from the spectrum besides "good, there's the carrier"

So, Question number one: Where can i find resources to educate myself? Talking about books, videos, ...

The best start is probably the book "Fundamentals of Spectrum Analysis" by C. Rauscher, one of the RF gurus at Rohde & Schwarz. I found a download of the 1st 6th Edition here:

http://qtwork.tudelft.nl/~schouten/linkload/spectrumanalysis-rs.pdf

http://www2.electron.frba.utn.edu.ar/~jcecconi/Bibliografia/05%20-%20Analizador%20de%20Espectro/Fundamentals_of_Spectrum_Analysis_Rauscher2008.pdf

That should give you all the basics.

Number two: what else do i need besides the SA to look at radio waves? A piece of wire directly on the BNC? An Antenna? which kind?

That depends on your specific problem, what the UUT is and so on. In general, you probably want a direct connection (i.e. not via antenna), and you want good quality cables, adapters and passives (like attenuators). For tx testing, you probably also want a dummy load (where the output power can go into) and a suitable splitter to couple out a smaller amount of power for measurements.

But again, the specifics depend on what exactly you want to do.

Also, don't underestimate the cost factor for such "accessories".

I also have some difficult in choosing what to pick. Budget should be around 2-3k€.
Beginner in this field, i obviously look at Rigol, usually works fairly good without costing too much, but the 815 goes to 1.5 GHz. We might want to play with gps in the future so 2 GHz minimum.
The new siglent is pretty nice on paper, but it's new, it's siglent and i have been following the topic.

Well, it's B-brand, at the end of the day if you pay bargain-bin prices then you get bargain-bin products.

I wouldn't buy used gear on this one.

Your loss, really, as your price range pretty much leaves you with new B-brand gear with average at best. For example, recently bought a 2nd hand Anritsu 13GHz spectrum analyzer, four years old and in pristine condition for less than £1200 delivered, offering a far better RF performance than anything you can get from the B-brands.

And if you find that your requirements are more modest, there are various mobile communications testers like the R&S CMU200 and CRTU-RU which have a built-in SA that goes to 2.7GHz and which still offers an RF performance comparable to the B-brand analyzers, at prices often far below 1kEUR.

There's also the Agilent E4406A VSA Transmitter Tester, which goes for litlle money these days and despite it's limitations (i.e. 10MHz span) could be a very cost-effective alternative, depending on what you want to do exactly.

There are lots of 2nd hand options within your budget.

Well for now, while i still look around and learn i think i can get one of those 15€ dvb dongles and use them with rtl-sdr, at least i should be able to know that there is the signal (or not)

One of the cheap SDRs is always a worthwhile investment, no matter if you end up buying an SA or not. It makes a decent spectrum monitor, plus because of its flexibility it's a great tool for further signal analysis and demodulation on a separate PC.

[nctnico]

I also have some difficult in choosing what to pick. Budget should be around 2-3k€.

I wouldn't buy used gear on this one.

If I where you I'd seriously consider buying used from an equipment dealer (with warranty ofcourse).

[Wuerstchenhund]

Is there some tutorial available online, with an overall list of what different types of measurements you can do with a spectrum analyzer, and maybe also a short description on what and how.

Well, there's the book I mentioned in my previous reply to this thread, which goes into details as to how an SA works and what information it can pull out from a signal. But you still need a basic understanding of signal theory.

Also, these days old-style Spectrum Analyzers are being replaced by Signal Analyzers which can do even more things than a common SA.

In this overall list, I would like to see the the following entries for each type of measurement.

Overall list with multiple types of measurements, for each type of measurement the following entries:
...
Title of measurement
Description of measurement
What to measure exactly
What is the center frequency of interest
What is the actual bandwidth which you need (e.g. harmonics, etc.)
How to measure with the spectrum analyzer
Which decision based on measurement
Which correction to make based on measurement
...

It doesn't work that way, what the best tool is depends very much on what your problem is, plus a range of other parameters that can't be put down into a simple table.

There's a reason EE is a degree education 

[pascal_sweden]

Yes, I understand. But some kind of a cookbook would be nice!

For oscilloscopes, they have this demo boards (Rigol, Siglent), with a bunch of signals, to demonstrate the different functions of your scope.

Would it be too expensive to have such a demo board for a spectrum analyzer?

Just like these DVB-T reception dongles are a cheap alternative for an actual spectrum analyzer, there might be a cheap counterpart goodie at the sender side

[nctnico]

A spectrum analyser is easier. Just attach a piece of wire and go hunting for FM radio stations.

[rf-loop]

A spectrum analyser is easier. Just attach a piece of wire and go hunting for FM radio stations.

And switch on AM demodulator isnstead of FM demodulator, listen AM.

[JPortici]

Thank you for your answers..

Yes, of course the first thing i should do is define what i have to do but given my lack of knowledge in the area i am not sure what my need are, hence this topic

Let's see.. This is what i have to do: we are designing a car antitheft device and we want to add a remote to unlock it. This remote will operate in the 433 MHz band which can be used without restriction in our market's area. To achieve communication we are using ready made transceiver modules that make two microcontrollers talk with each other. First requirement is to have at least a working prototype and i need to see it they are transmitting or not.

Then, in the future we may adapt the design for other markets, AKA reconfigure the transceiver to work in the 860MHz band.

For now i think i'll go with the cheap sdr transceivers to see if there is signal, until i find a real solution. Maybe it's enough? but i want a real tool to use. A better tool is always the best choice in the end.

About used instruments, not knowing spectrum analyzers makes me feel wary about getting a used one, as i don't know how to discern good from bad as with other equipment and most if not all salesman will sense your lack of knowledge and try to give you shit for cash (i know, i've been one for some time. in the end we all do that).
Yes it should absolutely come with warranty and calibration, but from what i've read older equipment is much more delicate. Also it's slow (which is annoying. Not a huge deal breaker but, you know..)
I guess i'll make some calls monday, but second hand equipment is getting very pricey around here

I almost forgot... Thanks for the PDF reading right now. very useful so far..

[tggzzz]

Let's see.. This is what i have to do: we are designing a car antitheft device and we want to add a remote to unlock it. This remote will operate in the 433 MHz band which can be used without restriction in our market's area. To achieve communication we are using ready made transceiver modules that make two microcontrollers talk with each other.

Sensible.

First requirement is to have at least a working prototype and i need to see it they are transmitting or not.

You can use a simple $15 SDR dongle for that, and, depending on the modulation scheme, it might also let you decode what is being transmitted.

Then, in the future we may adapt the design for other markets, AKA reconfigure the transceiver to work in the 860MHz band.

By "reconfigure the transciever" I hope you mean "use a different ready-made transciever".

[Wuerstchenhund]

About used instruments, not knowing spectrum analyzers makes me feel wary about getting a used one, as i don't know how to discern good from bad as with other equipment

Well, I know a forum where you could ask for help 

and most if not all salesman will sense your lack of knowledge and try to give you shit for cash (i know, i've been one for some time. in the end we all do that).

Which is even more important you know what you want and are well prepared before entering any sales negotiations.

Yes it should absolutely come with warranty and calibration, but from what i've read older equipment is much more delicate. Also it's slow (which is annoying. Not a huge deal breaker but, you know..)

Yes, it is delicate and slow, at least if you buy a boat anchor from the late '70s. But if you stay within the 2000-or-newer time frame then you'll find plenty of fast and non-delicate instruments that are mostly self-calibrating and need very little manual adjustment (if at all).

And getting a 2nd hand brand name SA calibrated is a no-brainer, while most calibration facilities will struggle to calibrate something like a Rigol or Siglent SA simply because of the lack of proper maintenance documentation. Plus many won't touch Chinese B-brands for other reasons anyways.

I guess i'll make some calls monday, but second hand equipment is getting very pricey around here

It will be, if you go to some of the many used T&M gear resellers that want ludicrous prices for 2nd hand gear. If you do a bit of leg work, research options and keep calm then you can find a lot of great deals on ebay, even from business sellers which offer warranty.

But since you said you're using ready-made RF modules anyways you won't need the best RF performance to verify that your modules work as intended, so one of the cheap communications testers like the R&S CMU200 or CRTU should be perfectly fine, and they often cost little money.

I almost forgot... Thanks for the PDF reading right now. very useful so far..

You're welcome! It's still one of the best books about spectrum analyzers I've read.

[steverino]

adding to Wuerstchenhund's reading, google the following

Fundamentals of Spectrum Analysis filetype:pdf

provides a good list, including an updated pdf to the R&S document.

[Wuerstchenhund]

adding to Wuerstchenhund's reading, google the following

Fundamentals of Spectrum Analysis filetype:pdf

provides a good list, including an updated pdf to the R&S document.

The one in the link is the 1st Edition from 2001, the book has seen several more since then. Here's the 6th Edition from 2008:

http://www2.electron.frba.utn.edu.ar/~jcecconi/Bibliografia/05%20-%20Analizador%20de%20Espectro/Fundamentals_of_Spectrum_Analysis_Rauscher2008.pdf

There's also R&S's Spectrum Analyzer Primer, which seems to be a heavily compressed version of the book:

http://mwrf.com/site-files/mwrf.com/files/uploads/2013/05/SpecAnFundamentals3.pdf

[G0HZU]

HP/Agilent provide some good info too. HP released a lot of papers about spectrum analysers back in the 1970s and these were all variants of application note AN150.

Over the years the AN150 series has been updated and I think it's well worth reading. I studied these notes a LOT when I was younger and just starting out in RF design.

Here's a recent version although you can also search for AN150-x where x is a number as this leads to some of the older/classic versions of AN150 that deal with various aspects of spectrum analysis in more detail.

http://cp.literature.agilent.com/litweb/pdf/5952-0292.pdf

However, you really can't beat having a decent tutor to show the basics and also to show you the common traps/mistakes people make with a spectrum analyser

[pascal_sweden]

Thanks for the share about AN150.

Could you also give a recommendation on a specific older AN150-x edition with the most deep technical coverage? Or did some versions go deeper in one topic, and other versions deeper in another topic?

Would be cool to find out the specific version that went most deepest of all AN150 editions

[Wuerstchenhund]

Thanks for the share about AN150.

Could you also give a recommendation on a specific older AN150-x edition with the most deep technical coverage? Or did some versions go deeper in one topic, and other versions deeper in another topic?

Would be cool to find out the specific version that went most deepest of all AN150 editions

AN = Application Note

ANs are documents that quickly discuss a specific topic. Sometimes they barely scratch the surface, in other cases they go a a bit deeper, but if you're looking for something that pretty much starts from Adam and Eve then ANs are the wrong type of document.

If you want to dig deep then read the book I posted the download link to.

[G0HZU]

ANs are documents that quickly discuss a specific topic. Sometimes they barely scratch the surface, in other cases they go a a bit deeper, but if you're looking for something that pretty much starts from Adam and Eve then ANs are the wrong type of document.

The AN150 series of app notes from HP pretty much defined how people went about making 'better' measurements with a spectrum analyser for at least a generation. I can only assume that you have never read them.

The whole point of these application notes was to improve the awareness of the user so they can make a better measurement with a spectrum analyser. I think HP/Agilent are far and away the best at understanding and communicating with the customer and I think this is why these app notes were released. They are often 50 pages long and deal with various aspects of analyser usage in each app note.

The content in the R&S book you linked to will have been influenced by these classic application notes, either directly or indirectly.

[pascal_sweden]

I downloaded the sixth edition of "Fundamentals of Spectrum Analysis", as recommended by Wuerstchenhund. Nice book!

Note: They sell the hardcopy of this book in the Rohde&Schwartz online bookstore:
http://www.books.rohde-schwarz.com/go/rohdeschwarz/en/home/store.xhtml

But the hardcopy which they sell seems to be a much older edition (4th edition).
Rather weird that they don't update the bookstore with the latest edition

I also found the following tutorials from Tektronix
http://www.tek.com/learning/spectrum-analyzer-tutorial

[Wuerstchenhund]

ANs are documents that quickly discuss a specific topic. Sometimes they barely scratch the surface, in other cases they go a a bit deeper, but if you're looking for something that pretty much starts from Adam and Eve then ANs are the wrong type of document.

The AN150 series of app notes from HP pretty much defined how people went about making 'better' measurements with a spectrum analyser for at least a generation. I can only assume that you have never read them.

Actually, I did read the AN150 Series and have used them myself as reference since the '80s. And I did not say they were not helpful, which they most certainly were. That wasn't the point.

The whole point of these application notes was to improve the awareness of the user so they can make a better measurement with a spectrum analyser. I think HP/Agilent are far and away the best at understanding and communicating with the customer and I think this is why these app notes were released. They are often 50 pages long and deal with various aspects of analyser usage in each app note.

That is true, at least to some extend, but you're completely missing the point, which is that even the 90 pages in the Agilent AN150 aren't really a lot for such a complex topic, and I really would have expected you being able to appreciate that a 200+ page document about a certain topic is more likely to go into more details than a 90 page or even 50 page app note.

I certainly didn't suggest to not read the AN150s, so no need to become so defensive.

The content in the R&S book you linked to will have been influenced by these classic application notes, either directly or indirectly.

That may well be the case, but again, no-one argued otherwise.

[JPortici]

No, i meant "reconfigure" as in this specific case we are using the mrf49xa transceiver from microchip and the same part will work in three different bands. (433, 860, 915) The balun needs readjustments of course

So, to recap, for now the sdr dongle will do, while i look around for a good deal on a calibrated used instrument.

As for bandwidth, I didn't consider higher order harmonics because the plan is to only use ready made transceivers (and i assume they are proven designs that work as they should. Bad speculation? i think they would be banished by the community if they didn't work, right?) so i though that 2 GHz would be sufficient, but i certainly won't say no to three.

Just to be fair, i know that chinese b-brands usually are what they are in terms of quality and documentation, which are both of extreme importance, but i wouldn't automatiacally rule them out.. look at siglent: siglent designs and rebrands for many others and those products are still well built, good quality instruments.. maybe not the best, but not as bad as you would think if they didn't have the shiny name.
Just like when you get a mercedes, open the hood and find out that it has a renault engine

[tggzzz]

No, i meant "reconfigure" as in this specific case we are using the mrf49xa transceiver from microchip and the same part will work in three different bands. (433, 860, 915) The balun needs readjustments of course

Make sure you understand what is necessary to certify your equipment so that it can used legally in destination countries. In the UK the general principle is that it is legal to sell/buy items even if they cannot be used legally.

As for bandwidth, I didn't consider higher order harmonics because the plan is to only use ready made transceivers (and i assume they are proven designs that work as they should. Bad speculation? i think they would be banished by the community if they didn't work, right?) so i though that 2 GHz would be sufficient, but i certainly won't say no to three.

But you are changing the balun. Are you certain the balun cannot affect the harmonics? Don't forget that any non-linearity will generate harmonics. Any magentic material will be non-linear.

Quick sanity check: what are the IP2 and IP3 specifications, why are they important, and what will be their effect in your circuit?