cheap THz experiments?

[jjoonathan]

Suhner clicky wrench is still MVP

[Joel_Dunsmore]

Speaking of soft wrenches,

NASA designed one, should be too hard to modify adding a open end 5/16.  Here's the thingiverse version: https://www.thingiverse.com/thing:608397

[Joel_Dunsmore]

Re: Marki MMIQ, are we talking about harmonic mixing on the MMIQ-4067L?

I realize now I might have been speaking slightly out-of-school.  Sometimes Christopher Marki (main mixer designer at Marki) sends me a proto or two to test out.   It's was like the MMIQ-4067 just smaller, with wee connectors and with a bigger second number.  So I was testing it fundamentally mixed.  But almost all balanced mixers operate well in 3-rd harmonic LO drive.  We typically figure 12-15 dB lower conversion loss, and that is how the Gilbert-Cell mixers in the PNA work above 27 GHz, in fact. 

We've been using the MMIQ-4065 for several experiments recently, expect some papers to come out on those experiments later this year.  I haven't tried it in 3rd harmonic LO mode, but next time I get in the office I think I'll spin up an experiment

[jjoonathan]

Nice! PNA mixers are something I can point to if I need to sell someone on the idea.

Yeah, when Marki Microwave released their 125GHz quadrupler I figured they were tooling up to fill out their 1.0mm portfolio. Or 1.35mm portfolio -- as far as I'm concerned, those can't arrive soon enough. 1.0mm just takes things a bit far. New wrenches, no "male thru" connectors, and a constant lingering awareness of how if a mechanical accident happens in your setup the 1.0mm connectors will ensure that the damage gets funneled towards the most expensive components. If you need the mode-free operation range, you need it, but I really like the sales pitch of 1.35mm. If only it had more market penetration... or any at all, for that matter.

Good news on the option front, though: if MMQ-40125 is any indication, Marki has decided to stay with .009" pins instead of .005" pins, so the Southwest connectors can be swapped. My entire setup hasn't grown into the 1.0mm range yet, so I'm very much in this business -- I'm sure opinions will vary, but mine is that swapping a Southwest connector beats a 1.0mm gooseneck from a between-series connector, at least if you only plan to do it once in a blue moon.

I'll keep an eye out for those papers!

[0culus]

For giggles: https://www.keysight.com/us/en/product/85059B/1-mm-calibration-kit.html

I wonder, is there a 1mm ecal or are we in mechanical cal only territory? And how long does your $35,000 cal kit last?

[jjoonathan]

Ooh free shipping!

[0culus]

Ooh free shipping!

I guess that's the least they can do. 

[jjoonathan]

Looks like RCal goes to 110: https://www.keysight.com/us/en/assets/3120-1426/technical-overviews/U9361-RCal-Receiver-Calibrator.pdf

EDIT: Ah, but it only calibrates receivers, not VNAs. I could have sworn I heard something about receivers being used in newer VNA ecal architectures, but maybe I just let my imagination run wild after hearing the term. Certainly by 70GHz it seems that ecals start to look a lot less like a switched SOLT set and more like a "well, they're linearly independent enough" set: https://ieeexplore.ieee.org/document/5165942 . It's not limited to ecals, of course, but in the mechanical calibration world the gigahertz giveth conveniently sized air lines, offset shorts, and dielectric loading, even as they taketh away ideal opens, so I could understand if overall calibration strategy pivoted in that direction at those frequencies.

I suspect material has been written on the subject... and I have a suspicion about who wrote it, too but that's a search for another day .

[0culus]

Looks like RCal goes to 110: https://www.keysight.com/us/en/assets/3120-1426/technical-overviews/U9361-RCal-Receiver-Calibrator.pdf

Is that the same thing as a VNA ecal though? My understand was this isn't a traditional SOLT style calibration, and is intended to be used with signal analyzers and other "regular" receivers to calibrate out the cable connecting you to whatever your experiment is?

[jjoonathan]

Yep, you're right! I think I found out where my intuition went off the rails: I watched a R&S video about new power sensors that used receivers, not ecals. Mystery solved!

[Joel_Dunsmore]

Looks like RCal goes to 110: https://www.keysight.com/us/en/assets/3120-1426/technical-overviews/U9361-RCal-Receiver-Calibrator.pdf

R-cal is kind of like the Phase-Reference calibrator (such as U9391G ) in that it provides a very well characterized set of of multi-tones (mag and phase are both known) that can be used to calibrate across a span of a receiver such as UXA or UXR.   The U9391 produces a set of harmonic tones (typically every 10 MHz up to 67 GHz) so power is quite low per tone, the R-cal can produce a narrow-band (e.g. 4 GHz compared to 67 GHz) focused at some frequency so the power per tone is much high.  It calibrates the instantaneous BW across its range, so you can remove mag and phase response.  Theoretically one could use it to calibrate a PNA (which uses receivers instead of samplers, just like a UXA) response ... just a matter of a bit of software.

[0culus]

Looks like RCal goes to 110: https://www.keysight.com/us/en/assets/3120-1426/technical-overviews/U9361-RCal-Receiver-Calibrator.pdf

R-cal is kind of like the Phase-Reference calibrator (such as U9391G ) in that it provides a very well characterized set of of multi-tones (mag and phase are both known) that can be used to calibrate across a span of a receiver such as UXA or UXR.   The U9391 produces a set of harmonic tones (typically every 10 MHz up to 67 GHz) so power is quite low per tone, the R-cal can produce a narrow-band (e.g. 4 GHz compared to 67 GHz) focused at some frequency so the power per tone is much high.  It calibrates the instantaneous BW across its range, so you can remove mag and phase response.  Theoretically one could use it to calibrate a PNA (which uses receivers instead of samplers, just like a UXA) response ... just a matter of a bit of software.

Very interesting. Joel, if corporate would let you, you should make a youtube channel like Shahriar. I would subscribe in a heartbeat to see you talk about this stuff and provide demonstrations. If anything it's positive advertising to show off flagship gear... 

[cdev]

Here is how to get rid of dust. Brand new leaf blowers and HEPA filtration, and a spiral airflow. (go in one direction so the air describes a circular path in a room.) Use the leaf blowers to dislodge the dust and large size HEPA filtered air cleaners to suck it up. Wear a P-100 mask yourself while you do this, and of course, use HEPA vacuums to clean everthing too. By doing all this you can create an environment with very minimal dust, when you are through. Good for microscopy, etc.

[Ultrapurple]

Harking back to the original direction of this thread, the OP may want to raise their head in the Thermal Imaging section. Single-digit THz is considered a bit low-frequency there, but you'll find a bunch of people with a different spin on things. And a lot of experience.

[Bud]

Here is how to get rid of dust. Brand new leaf blowers and HEPA filtration, and a spiral airflow. (go in one direction so the air describes a circular path in a room.) Use the leaf blowers to dislodge the dust and large size HEPA filtered air cleaners to suck it up. Wear a P-100 mask yourself while you do this, and of course, use HEPA vacuums to clean everthing too. By doing all this you can create an environment with very minimal dust, when you are through. Good for microscopy, etc.

You are ef-king kidding me. Go try a leaf blower in the garage first and tell us how it went.

[coppercone2]

The optical approach to THz is unfamiliar with me but I need to think of some topics.

[doktor pyta]

'Low-cost THz heterodyne detection by miniature neon indicator lamp glow discharge detector'
https://www.spiedigitallibrary.org/conference-proceedings-of-spie/8188/81880P/Low-cost-THz-heterodyne-detection-by-miniature-neon-indicator-lamp/10.1117/12.898035.short?SSO=1

[coppercone2]

sounds like you can increase sensitivity by using a parabolic reflector or channeling the energy otherwise. If you use a antenna with a waveguide it should act as some sort of filter.

Does anyone know like what power level you can expect to receive energy in this band from space?

But I got the paper on the neon thing and it looks like you need a 300GHz source.. so maybe the detector is cheap but you still need something very expensive. Maybe neon and a photomultiplier if it glows slightly or something? I think some kind of diode is more practical for the "cheap" part of this. They don't show 'direct detection' at the higher frequencies, they only show it for the lower frequency. I assume that means they had some kind of problems.

[jjoonathan]

Yeah my impression was that detectors were the easy part: I bet you could get a bolometer to work literally from DC to daylight. DC sees a resistor, RF sees a termination, IR sees a blackbody, optical sees a tiny black strip of material. Energy conservation says that if the power isn't reflected or stored or re-emitted, it's coming out as heat, and if you can detect that heat, there you go.

The flip side of that intuition is "RF techniques stop working before optical techniques start working, and everything is super lossy." Diodes become tiny and lossy before SHG and THG crystals start doing their thing. Transistors become tiny and lossy before losses drop enough to make a laser gain medium that actually has, you know, gain.

I'm far from an expert, but every year that I see a flurry of papers claiming to bridge the THz gap, I know that the gap is not yet bridged. Once it is, papers will stop claiming to have bridged it. In the meantime, you just have to pick your poison by taking an RF or optical technique and driving it far into a regime where it doesn't work very well.

[coppercone2]

I kind of wonder if maybe rather then neons passing them through some kind of larger field might produce some effects, but it would require a glass shop to even begin to experiment with it. Like large gas devices not optimized for light emission or circuit use or anything like that. Density fluctuations measured with lasers (totally random thought).

[TheUnnamedNewbie]

Also a millimeter wave and THz guy here. I've mostly focused on D-band, but I've also done measurements at 330 GHz, and assisted colleagues in the design (and mostly measurement) at 660 GHz and 1.1 THz.

I think spectrum analysis and signal generation should be doable, provided you can already generate a few GHz at decent power. If you drive a diode with say 10 GHz at 10 dBm, even if the conversion loss is like 50 dB, you get -40 dBm at the 10th harmonic, which is detectable (with another diode in a similar setup) by a good spectrum analyzer? Could start at a few tens of GHz this way and see where it goes.

Those who know me might see this one coming, but consider doing experiments with dielectric waveguides as method of getting signal from a to b. A 'cheap' piece of PE rod can work just fine for this to carry signal a few meters with only 10 or 20 dB loss.

[coppercone2]

another option for the high GHz band might be to use automotive radar.

I see the modules are not that expensive.. but I am not sure how much trim adjustment you can do on them.

[0culus]

another option for the high GHz band might be to use automotive radar.

I see the modules are not that expensive.. but I am not sure how much trim adjustment you can do on them.

I think this is (was?) not uncommon for some of the advancements made for the highest-frequency amateur QSOs.

[Paul B]

hi can you give me any information about the multiplayer with an output at 125 GHz as i am looking for a sig gen from 120 to 140 Ghz
Paul

[coppercone2]