Hack Chat Transcript, Part 2

A event log for Weird World of Microwaves Hack Chat

Let's get small

lutetiumLutetium 12/18/2019 at 21:090 Comments

James Finch12:34 PM
Rainbow Photonics AG is who I am thinking of.

David Troetschel12:35 PM
@James Finch oh, this looks cool!

James Finch12:35 PM
W-Band is interesting.

The Signal Path12:35 PM
@David Troetschel (Power Factor (loss Tangent) @15°C: 0.0001 - 0.0004) That is pretty good. The issue is more than just the cost of the raw material. Lithography resolution, layers, etc. all have to be taken into account.

James Finch12:36 PM
Amazed me observing the transition where the circuits turned into optical train like systems designs.

David Troetschel12:37 PM
for sure; just an idea

The Signal Path12:37 PM
@David Troetschel I'll investigate. Thanks!

A reminder that I'll post a transcript right after the chat, so everyone can refer back to links and such

David Troetschel12:38 PM
You can usually split a single sheet into about 20 but I have no idea how crazy you would have to be to try and smoosh em back together after laying traces haha

home12:38 PM

The Signal Path12:38 PM
Any suggestions/requests for videos/ect.?

Ethan Waldo12:39 PM
@The Signal Path I've seen a lot of research in creating digital logic in optical ranges, are you aware of anything being produced in high mm range?

The Signal Path12:39 PM
One of the challenges with my channel is that the topics can get fairly advanced, as a result the audience size is always going to be fairly small.

ottingerg12:39 PM
... I really like the sampling osciloscope project from Ted Yapo ... would be nice if you could feature this one in your videos

BSA Secure12:39 PM
Yes :) How to re-purpose existing devices to put your hands on microwave stuff while still using RTL-SDR / HackRFOne

BSA Secure12:40 PM
that is price baseline for most of folks

James Finch12:40 PM
@The Signal Path Maybe detailing circuit designs correlating the range of electronic design as well as optical like designs.

The Signal Path12:40 PM
@Ethan Waldo Ultra-high speed digital communication is not really digital, it is mixed signal. Wireline links can be made to beyond 100Gbps.

The Signal Path12:40 PM
@BSA Secure Ok. I will keep that in mind.

michal77712:41 PM
Can you make video about the cheapest chinese microwave test equipmen (spectrum analysets, VNAs)?

James Finch12:41 PM
@The Signal Path Might even be interesting to demonstrate making a cost effective spectrum analyzer interface into an oscilloscope. Might be a cost effective way to get more out of a scope and maybe in analog ways too first off.

The Signal Path12:42 PM
@michal777 It is amazing how much far the Chinese vendors have come in this regards. I will take a look again and see what I can find.

Ethan Waldo12:42 PM
But that almost always uses some specialized transceiver chip as opposed to nuts and bolts digital logic (i.e. NAND)

home12:42 PM
@michal777 - I agree there must be some good stuff out there for us cheapskates.

The Signal Path12:42 PM
@Ethan Waldo Yes, of course. These are not traditional logic gate circuits.

James Finch12:42 PM
@The Signal Path Would be sweet and totally awesome to be able to detect from the Gamma down to DC... though that's a little excessive of a goal for a smaller package design... though would be interesting to see what the present state of the smallest package can look like.

Jacob joined  the room.12:43 PM

BarceP12:44 PM
thank you for the advice, i'll keep it in mind, I don't know if it is possible, but i think it would be interesting an overview or analysis of an IC or IC fabrication, i aIso loved the 24Ghz Doppler radar module, because of the nice analysis of a "simpler" circuit type.

Ethan Waldo12:44 PM produces stuff in the 1Ghz max range, and then there's one other that is ridiculously expensive that I can't find right now.

James Finch12:44 PM
@The Signal Path Seems as soon as the chips come out and there is an evaluation board available... I'm guessing there are Chinese out there trying to engineer the cheapest feasible modules to put on the market with the bare minimum interface capabilities.

BSA Secure12:44 PM
Hughes, ViaSat, other radioline producers allows people to get a hands on ready to re purpose equipment to send and receive 20Ghz+ stuff. Challenge is how to estimate performance of the Ka-Band (~19-30 Ghz) equipment that is widely available. Like impact of the RTL-SDR on the community.

James Finch12:45 PM
@The Signal Path The Microwave Hearing Effect or Voice to Skull technologies would be interesting to watch demonstrated. Even acoustic hailing with microwaves would be interesting to watch.

The Signal Path12:46 PM
@Ethan Waldo This link from PotatoSemi made me winch: " High Frequency Noise Cancellation Technology."

Frank Buss12:47 PM
@The Signal Path I've read that the transistors on an Intel chip are switching at 300 GHz, why is my CPU still running only at a few GHz? I guess at these frequencies the distances on the die matters and synchronization is a problem?

James Finch12:47 PM
@The Signal Path Some are considering downconverting cost effectively. I'm wondering what the most cost effective evaluation board with the lowest noise floor, highest sampling rate and highest bandwidth is since is nice to see more of the spectrum real time. Seems the price has to be coming down with the new 5G chips out.

The Signal Path12:48 PM
@Frank Buss The fT & fMAX of FinFet devices in the 12nm range can easily be 300GHz. This does not mean you can make a processor at that frequency.

Frank Buss12:49 PM
would it be possible with async logic instead of a central clock? I think there was some research about this

Ethan Waldo12:49 PM
@Frank Buss That's how a lot of the original room-sized computers were designed

The Signal Path12:50 PM
@Frank Buss The issue is complexity and what is to be achieved. An ARM core or x86 core requires a certain architecture for compilers to work with.

The Signal Path12:50 PM
@Frank Buss There is also a huge thermal issue to overcome.

Frank Buss12:51 PM
makes sense, so this would need new CPU architectures as well, new compilers etc., sounds interesting :-)

The Signal Path12:51 PM
@James Finch I am sure in the next few years equipment up to 40GHz will come down in price even more.

The Signal Path12:52 PM
@Frank Buss That is once possibility, but there are a lot of other considerations. In the past 5 years, we have gone to parallelism, rather than faster serial rates for CPUs. And software has made that transition as well.

salec12:52 PM
@Frank Buss iirc the problem is that working regime in which that transistors achieve such high frequencies are not switching, completely off, completely on, and that translates into too much power dissipation

James Finch12:53 PM
I wonder if there is a mirroring of tech like limiting CPU's main stream more-so with around 2.4GHz since the earliest research was using and if that has anything to do with the weird world or remote sensing?

The Signal Path12:54 PM
@James Finch I don't believe there is any connection. At least I don't know of any.

michal77712:54 PM
A question: How to understand the idea of impedance matching and mismatch in case of highly nonlinear circuits (mixers, doublers)? It seems easy in case of passive, linear networks but when the waves are for example chopped by diodes and they get any non-sinusoidal shape then... I've no idea how to relate that to the simple theory.

James Finch12:54 PM
Or is there something to do with the manufacturing capabilities since the infrastructure doesn't need to be re-tooled so much for the next smallest size or different method system.

The Signal Path12:56 PM
@michal777 You have to be careful about the terminology here. There is small signal impedance and large-signal impedance. Just because a waveform is square, it doesn't mean it is large-signal matching. The square signal has a lot more frequency content, so matching would have to be more broadband if you want to preserve the shape of the waveform.

James Finch12:56 PM
@The Signal Path I've not read anything either... though wondered with the bus frequencies, clock speed and being at those more common easy to access ranges if there was something TEMPEST going on since those intel types demand a backdoor or way in from my course work.

The Signal Path12:56 PM
@michal777 When things become non-linear, then you have a whole other domain of signal matching.

RoGeorge12:58 PM
@The Signal Path I would like to see some videos about basic microstrip structures

- main theoretical ideas

- practical implementations, examples

- building them (DIY techniques)

- measuring them and see how it all worked (or not) according to theory

Ethan Waldo12:59 PM
+1 on that 4th item

The Signal Path12:59 PM
@RoGeorge Yes, that has been on my list. I have talked about some of those topics as part of various videos. But there is no dedicated video yet.

James Finch12:59 PM
Seems the volume of backhaul systems devices isn't so large scale and I'm guessing that doesn't aid the price point for the main stream consumer market.

RoGeorge1:00 PM
yes, especially the measurement part, thatk you

The Signal Path1:00 PM
@RoGeorge Sure, we can measure SP or TDR to verify.

James Finch1:01 PM
@RoGeorge Good call... even the range of "strip" and "line" designs overview would be interesting.

James Finch1:03 PM
With properties measured of course.

The Signal Path1:03 PM
@James Finch I do have a PCB milling machine, so we can go from design, simulation, manufacturing and measuring in one video.

Ethan Waldo1:03 PM
That would be fantastic

Well that was fast - we're already through an hour. We'll let Shahriar go if he has to, but of course everyone is free to stay around and continue the discussion. I just want to say thanks to Shahriar for coming on today and closing the 2019 Hack Chat series in great style. I really appreciate it!

JMGarofalo1:04 PM
@The Signal Path Thank you!

Ken Kaiser1:04 PM
@The Signal Path Thanks

The Signal Path1:04 PM
You are very welcome and thank you all. Especially to my Patreon supporters who keep the lab going and allow me to purchase the instruments/components I need for the videos.

The Signal Path1:04 PM
Without the Patreon supporters, it would not be possible at all.

BSA Secure1:05 PM
Re-purposing 20$ radar to measure (estimate) performance ...or just to know if signal appeared in the XXGhz band woud be interesting. E.g practical removing of filters to extend detection rage

Ethan Waldo1:05 PM
Yes, thank you for sharing your wealth of knowledge and experience with mere mortals.

The Signal Path1:05 PM
Cheers all, if you like to find the videos and subscribe you can go here:

James Finch1:05 PM
@The Signal Path Excellent, thanks for your time and consideration. Excellent topics looking forward to seeing and more of in general.

The Signal Path1:06 PM
Bye everyone! :)

We'll be back in January, so watch for announcements on those Chats. And of course, if you've got anyone you want to hear from on the Hack CHat, or if you've got something cool to discuss, drop me a line - danmaloney@hackaday,com

Thanks everyone! Transcript coming up soon.

Ethan Waldo1:06 PM
Thank you too Dan

James Finch1:07 PM
Thanks Dan,

RoGeorge1:07 PM
Bye & see you soon in the videos, thank you for sharing!