Note: groups.io will be down for maintenance this evening, starting at 11pm Pacific Time (6am Tuesday 8/3/2020 UTC), for up to two hours.
I too was wondering how theS22 is +- 50Ohms.
On the other hand I am wondering why was an MMIC used in the first place. I can get equally good results if I directly use the 9851 output. A match with a 4:1 balun or in my case UNUN does the impedance match.
Any idea why an MMIC was used? I just adds to harmonic and other distortion.
I think some more explanation is called for.
I guess you might remember that many moons ago I pasted some pictures of SNA I built from scratch. I was considering posting it to the group but wanted to improve is somewhat and be ready to answer questions regarding the HW/SW. I could not get around to doing it due to some bit of illness. ( It was all assembly so hard to remember the details) During this time I decided to learn some embedded C and as an exercise to rewrite the SNA SW in C. I also did not like the slow serial connect to the Graphic LCD ( GLCD!) etc. S0 I decided to use a 3.2 inch GLCD. This one has a 16 bit parallel data interface so the old 32 pin IC would not do and I switched to a more modern C8051F020 from Silicon Labs.
I was looking at the output of the AD9851 and decided that I don't really need the MMIC. I have a few AD9851 boards with the filter built in and decided that a single ended signal take off would do.
I have used a 200 Ohm R on the output of the AD9851 board removed the filter and used a 4:1 Ruthroff UNUN The output is over the top. Enough to do any old testing of networks.
The output level is calculated as follows. ( excluding the attenuation of any built in LP filter. I have removed mine)
The Peak I out is 10mA. The 200 ohms at the output and the 50 ohms reflected as 200 ohms at the output of the SNA = total 100 Ohms. The 10 mA x 100 ohms gives a peak of 1000 mV and 0.707 V rms output which reduces to .707/2 =353 mV rms at the SNA terminals ( UNUN). This equals +4dBm
Enough for my needs. The out impedance is 50 Ohms. I intend to build a few small external filters for use in different applications. Sweep scan is variable from 480 Hz to 50 MHz. I think the fastest scan is 300mSec. Slowest is several seconds.
I will try upload the pictures and schematic and SW as soon as I am happy with the details. It is written in C except a few fast routines for the GLCD in Assembly. I have used no library routines as all is home brew. and the SW is as efficient as I could make it. This is an 8051 U-controller so some people might not be familiar but it is the best uC in the world. ( In my view, that is)
I think the real DIY guys should find it interesting.