Newbie working in sub 1Ghz, equipment and reference knowledge.

[Robotec]

Good morning everyone,

I'm a complete stranger to the RF world but due the wonderful decisions of management now i'm in charge of a redesign a PCB working in 868 - 915 Mhz due to the lack of components.
In a first phase it will be a change of front end from a CC11190 to a SE2435L which shouldn't be too hard since the later in 50 ohm terminated, but the real problems will start when the transceiver core will be changed and all the fine matching will start.

Thats why im trying to read everything possible about the RF basics ,impedance adapting,etc , but i would like to know is there is read you consider a MUST, also what instrumentation is considered a minimum, spectrum analyzer ,etc , I've seen that the ADALM-Pluto is regarded a a good performance/cost option.(somehow the first version was designed without any rf instrumentation, very likely to be a copy of some example module.)

The device is based in narrowband channels of 25 Khz with a very low baud rate so I don't know how of a hi -end instrumentation would be necessary of if even with my oscilloscope DS1054Z could work(doubt it).

Any tips or help is welcome.

[Uunoctium]

tl;dr

It's not a bargain*, but the DG8SAQ VNWA 2 Port (up to 1.3 GHz) Vector Network Analyzer is an very useful instrument (based on USB- audio codecs). Includes also an simple spectrum analyzer and is very easy expandable to an 4-Port (forward-reflected)**.

There are many YT tutorials out - best are from Knut Poulsen.

If you buy an used one, confirm that you get the license key with the device!

* ~400$
** Philips/Sievers coax-switch PM 7551 + small electronic stuff

[Robotec]

tl;dr

It's not a bargain*, but the DG8SAQ VNWA 2 Port (up to 1.3 GHz) Vector Network Analyzer is an very useful instrument (based on USB- audio codecs). Includes also an simple spectrum analyzer and is very easy expandable to an 4-Port (forward-reflected)**.

There are many YT tutorials out - best are from Knut Poulsen.

If you buy an used one, confirm that you get the license key with the device!

* ~400$
** Philips/Sievers coax-switch PM 7551 + small electronic stuff

Thank you, but after speaking with management  something like the Sigilent SSA3021X could be on the table, looks far easier to make it work and by what I'm seeing to troubleshoot the system it will be a must.

It includes a tracking generator so based in my limited knowledge that should be enough to debug the output signal of the system and if necessary improve the filters , but the board only has a output SMA connector so i don't see how a tracking gen or VNA could be connected into the system.

[rfclown]

While a VNA is an great tool, if I could only have one thing, it would be a spectrum analyzer with a tracking generator. With a tracking generator you can make gain measurements (magnitude of S21), and if you add a directional coupler (can get one on eBay for not too much like this https://www.ebay.com/itm/133922573005?hash=item1f2e684acd:g:D6AAAOSwJrxhgZh4) you can measure return loss (magnitude of S11). If you don't have a spectrum analyzer, you are basically in the dark. ADALM-PLUTO is a cool thing (I have them, and use them), but they are a project. I have HP analyzers, VNAs, etc. Put me on an island to do some RF work with one instrument, and I'd choose a spectrum analyzer with a tracking generator.

[Robotec]

Thank you this was the kind of information I was looking for.

[switchabl]

Yes, what he said, if you can get just one thing, get a spectrum analyzer. Both for debugging and for pre-compliance testing. It would be helpful to maybe cover at least your fifth harmonic (so >4.6GHz). If it turns out your antenna filter provides insufficient harmonic suppression, you probably want to fix that early and not after failing compliance testing ($$$). For a one-off renting may be an option.

[coppercone2]

just don't think you will hit every problem with the pre compliance test. When it comes to the chamber you will be surprised by what can happen.

Some manufacturers don't bother at all trying to find the problems in the beginning because they just learn that its too precise/chaotic to try to track down those problems. I have a feeling the bosses learned that a simple system that fails hard is easier to fix then something that was soouped up and fails mildly.

AKA don't give people the expectation that your precompliance test is gonna save boat loads of money and be a magic bullet.

The utility of these tests depends entirely on your company expectations.

[switchabl]

AKA don't give people the expectation that your precompliance test is gonna save boat loads of money and be a magic bullet.

Yes, I didn't want to leave that impression. Just that, if you are designing a radio transceiver (for the first time in this case), not being able to check the output of your actual antenna port up to a couple harmonics really feels like tempting fate to me.