Kit to help with measuring microwave radio system characteristics

[RobBarter]

Having just built my first 23cm contest system I've been spending a lot of time measuring noise figure (NF), adjusting, measuring etc.  Been using a BBGen+ noise source (very nice and simple to use) and a HP8561E spectrum analyser.  Measuring every stage from the antenna in, to output of circulator & bpf, output of preamp, output of transverter and finally the output of the Elecraft K3.

Now I've got the bug for this (blame my contest team G3M and G5LK/P for this) I was wondering if any of the Radio Comms Analysers I'm seeing on EBay would be of help here in getting consistent results in a single setup?   I've noticed the R&S CMU 200, R&S CRTU, Anritsu MT8820A  etc. etc.)   Or am I barking up completely the wrong tree?

[mjs]

With these you basically get a signal generator and spectrum analyzer that you can use for TX measurements and with 10/20dB attenuator for LNA/RX measurements. You can't effectively measure NF, but other stuff like frequency, spectrum, IP3 etc. you can do. CRTU can measure power up to 50W.

CMU200 has also audio generator/analyzer (at least as option), so you can start from baseband, with CRTU you get SA from 10MHz and SG from 100kHz.

If you wanna try CRTU, I've got a few to sel for cheap :-)

[RobBarter]

Thanks mjs

Looks like I'll be sticking with the noise source and a spectrum analyser and doing it by hand.

[tggzzz]

Thanks mjs

Looks like I'll be sticking with the noise source and a spectrum analyser and doing it by hand.

If you use a noise source, consider the effect of non-linearities in the UUT and spectrum analyser front-end: there will, of course, be intermod products.

Naturally it is desirable to ensure those products don't fall on the frequency at which measurements are being made. But since the noise source contains all frequencies, some of the products will be on top of the measurement frequency.

[rastro]

Thanks mjs

Looks like I'll be sticking with the noise source and a spectrum analyser and doing it by hand.

If you use a noise source, consider the effect of non-linearities in the UUT and spectrum analyser front-end: there will, of course, be intermod products.

Naturally it is desirable to ensure those products don't fall on the frequency at which measurements are being made. But since the noise source contains all frequencies, some of the products will be on top of the measurement frequency.

A signal generatior would probably be better.  It seems if you where going to use a noise source you could consider using lower power levels to minimize the contribution of non-linearities/intermod products of UUT.  These contribute significantly more as power is increased toward the UUT's max input levels; so its better to keep the UUT well within its linear range.  Of course you would need to know beforehand some of the operational parameters of the UUT.  This approach would not eliminate intermod products issues but should reduce its effect.

[rastro]

You might want to also consider a cheap RF generator (35MHz-4GHz) module there are many of these, with user interface, offered on eBay for about $50.00.

Just search "ADF4351 RF Sweep Signal Source Generator Module 35M-4.4G+STM32 TFT LCD " 

Probably not the cleanest signal but you should be able to get a good relative comparison going directly to your SA and then through your UUT.

Just a thought...

[tggzzz]

You might want to also consider a cheap RF generator (35MHz-4GHz) module there are many of these, with user interface, offered on eBay for about $50.00.

Just search "ADF4351 RF Sweep Signal Source Generator Module 35M-4.4G+STM32 TFT LCD " 

Probably not the cleanest signal but you should be able to get a good relative comparison going directly to your SA and then through your UUT.

Just a thought...

Any frequency below 2.2GHz comes from a digital divider, with the expected harmonics.

[RobBarter]

I have an HP ESG-D4000A signal generator (250KHz - 4000MHz) so good on that side.

[mjs]

Indeed, with noise generator you may run into non-linearities. You can actually use noise generator with some filters to measure amplifier linearity: https://www.lintech.com/PDF/npr_wp.pdf