TNP #18 - Repair & Analysis of a W-Band (75-110GHz) WR-10 12dBm Power Amplifier Module

In this episode Shahriar takes a look at a non-functional W-Band (75-110GHz) power amplifier module with WR-10 waveguide ports. A quick inspection reveals a broken DC connector which can be easily repaired. Since a W-Band VNA or source is not available, the amplifier's excess noise is instead measured using an external mixer and a Keysight MXA-B. The noise power spectral density shows the the amplifier is likely functional with the expected noise and gain.

[trevorvanbremen]

FINALLY!!!

TSP has encountered a scenario I can readily relate to...
As amazing as this might seem, just like TSP mentions @ 4:20, I _ALSO_ "don't happen to have a 75GHz-110GHz source that I can turn on right now"
The fact that I could barely scrape together a 50Hz source to turn on (yes, I really mean 50Hz and not 50GHz!) is another matter entirely!!!
I guess I could 'cheat' and say that I do have ready access to a high power (around 1kW) 2.4GHz 'source' that I can turn on at will.
It's even got a little light inside, a turntable and a fairly rudimentary timer. It was built by Panasonic.
It comes in REALLY handy when I want to reheat last nights dinner!

Footnote:
Despite the above, I still DROOL at the little tidbits of info within a TSP video that my tiny brain can actually comprehend!
I tend to concentrate my efforts towards the 'DC' end of the spectrum and leave the 'daylight' end of the spectrum to TSP et al

[ruhnet]

Nice when the repair is something on the DC side rather than mm-wave side hehe

[rfengr]

I’m glad It worked, but I also wanted to see it pulled apart!

[vincei]

Hmm. Pretty sure Pooch is a 75-120 Ghz signal source in the lab. I call foul on the termination preconditions. You'll need to finish the 75Ghz VNA build to convince us :)

[hugov]

Would have been good to see the setup on the bench attached to the spectrum analyser, and not just the instrument's screen. I'm curious how the mixers, waveguides, etc all fit together. I would guess for most viewers, you could skip it for lower frequency setups (SMA coax etc) but for mm wave, those practical matters are rather important.

[ecospider]

Good job saving that component. Great video thanks.

[auke]

how was the spectrum analyzer connected to the amplifier ??

[urlkrueger]

Aww Gee, I was hoping we would be able to see an emerald green glow inside the waveguide.
As time passes usable frequencies are getting higher and higher and components are getting smaller and smaller, or more appropriately physics-al.
At 100 GHz we are only a factor of 3 away from 300 GHz which some consider the lower end of Far Infrared radiation which I guess could be defined as the boundary between electronics and optics. What magic awaits us in that last 200 GHz?

[christianweagle]

"In a parallel universe somewhere" is as good a theory as anything. I have this idea that all these little round things should actually be made out of hex stock so that they stand a chance of not rolling into the next time zone when dropped.

[hosseinkhalilirad]

Hi, what applications are these kind of amplifiers with this frequency band are used for ?

[antoineroquentin]

Starting from which frequency do you expect a lab to be interference free? I work below 2 GHz and find interference everywhere

[robertdixon]

Nice work on the repair.
What is the external mixer?

[PapasDino]

I've never understood why noise sources use BNC connectors for power since the chance for mis-connecting something could be high if one isn't careful; maybe it's an industry understanding that no one would ever make that mistake. As a hobbyist I'd never mix power and signal connector types. Thanks for sharing!

[durantjames]

Thanks for sharing! How to calculate the noise？I can't find the relevant formula。

[fbnx]

Did you already release the video on the 110GHz network analyzer?

[timothykeller]

Nice video. I think the characteristic impedance of WR-10 waveguide would be closer to 500 ohms. I’m not sure what the impedance of the horn looks like though…

[huawatuam]

Video request: Maybe you can make a video explaining how the RCWL-0516 works and if it is really using doppler or another effect.

[craigs]

Place a couple of aluminum plate reflectors between the horn and the input port, see if you can get it to oscillate. Regenerative 100GHz receiver???

[stevefriedl]

Is the E36313A power supply new to your bench?

[paulmoir]

I'm sure you're aware, but in case you're not, all workshops naturally develop a small black hole. If you're ever on the other side of it, would you mind picking up the #8-32 tap I dropped? It has been about 20 years.