RF Product Design HackChat

Dan Maloney says:13 minutes ago

Hey Jenny

13 minutes ago

thanks for coming :)

Jenny List says:13 minutes ago

(Well, it is here, anyway)

Les Hall says:13 minutes ago

oh Jenny is in europe i presume then?

Dan Maloney says:13 minutes ago

Blimey, yes

Jenny List says:13 minutes ago

ANy time

Les Hall says:13 minutes ago

hah, ok then!

Jenny List says:13 minutes ago

Yes, I'm in the UK.

11 minutes ago

so we have some questions: https://docs.google.com/spreadsheets/d/1kzALEbqoawQ6IWc-n5EUY5quY5-t4HgAFg57wlvSK74/edit#gid=0

Jenny List says:11 minutes ago

Our radio waves wear tuxedos, y'know

10 minutes ago

Is the Raspberry Pi receiver board the giveaway?

Jenny List says:10 minutes ago

That's right.

Jenny List says:10 minutes ago

OK, just addressing questions...

steverobillard says:8 minutes ago

looking forward to this one my first electronics project was a radio telescope

Jack Laidlaw says:7 minutes ago

Hey folks.

Shulie Tornel says:7 minutes ago

Here's a good question from the doc: What are good modern resources for learning RF? (assuming already know digital & analog electronics)

M.daSilva says:6 minutes ago

Hey guys!

steverobillard says:6 minutes ago

when it comes to RF I like the ARRL handbook

Tim says:6 minutes ago

Hey gang!

Jenny List says:5 minutes ago

"How do i best do low power digital commms?"

It depends as always what you want to do. If it's just a simple serieal link between small devices, any one of numbers of off the shelf UHF modules. Usually 435 MHz, 868 MHz. (Are those Euro frequencies, US might be different)

If you want to go further afield, I'd suggest a LoRaWAN board. These can make miles with a decent antenna, are pretty cheap, and easy enough to program. Upcoming I'll be doing a demo for a HaD feature. Google "The Things Network" for an example of an open Internet backend for LoRaWAN.

Shulie Tornel says:5 minutes ago

HI Jack, M.daSilva, Tim! Welcome!

skot says:4 minutes ago

Is the Microchip RN2903 the best module out there to get started with LoRaWAN?

4 minutes ago

how much power does a LoRa board pput out for transmission?

Shulie Tornel says:4 minutes ago

GoogleDoc for questions (and RaspPi Receiver Board Giveaway) https://docs.google.com/spreadsheets/d/1kzALEbqoawQ6IWc-n5EUY5quY5-t4HgAFg57wlvSK74/edit

Adam Vadala-Roth says:3 minutes ago

hi Jenny!

Jenny List says:2 minutes ago

"What are good modern resources for learning RF?"

That's a tough one, because of course I learned it 30 years ago :)

I would suggest looking at publications from the ARRL, the RSGB, or other national amateur radio societies. The RSGB's Radio Communication Handbook, is that still a thing? I am pretty sure the ARRL do a similar book.

But of course the best way to learn in anything like this is to get hacking. Search the WWW, try a few projects, modify them, stick them togehter, the usual.

Jenny List says:2 minutes ago

Evening Adam!

Jacob Christ says:a minute ago

Is there a plug in that will do text to speech for this chat?

Adam Vadala-Roth says:a minute ago

:) added a question

a minute ago

no, sorry

a few seconds ago

here's another question from the sheet: What's your suggestion for a fun project I can lead with my local radio club to introduce HF? We are mostly VHF/UHF users but looking to explore HF. Thank you, Jenny!

Jenny List says:a few seconds ago

As to individual LoRaWAN modules, I'm afraid that's beyond my knowledge. However it's a certainty that among Hackaday's users will be people who know each module intimately.

Radu Motisan says:3 minutes ago

hi Jenny, Sophie & all

Jacob Christ says:3 minutes ago

Is there a published REST interface for this chat room so I can build a text to speech ap?

3 minutes ago

hey Radu!!

Dan Maloney says:2 minutes ago

Andreas Spiess has some good videos lately on LoRaWAN range tests and the like, FWIW

Radu Motisan says:a minute ago

@skot RN2903 is an excellent choice to start with LoraWAN. I used it in 3 projects

Jack Laidlaw says:a minute ago

Right here's a question I have, I want to buy first oscilloscope do I go for a second hand analogue or one of those DSO plug into the laptop things?

Adam Vadala-Roth says:a few seconds ago

go second hand digital, the 100MHZ scope in my profile I got at a yardsale for 1 dollar

Jacob Christ says:a few seconds ago

@Jack Laidlaw Get a stand alone scope. I have 4 scopes and the one that never gets used is the one that uses needs a computer.

Adam Vadala-Roth says:a few seconds ago

I have this one http://hackaday.com/2014/04/26/playing-tetris-on-an-oscilloscope/

Jenny List says:a few seconds ago

"Suggestion for radio club as an HF project" How about the little QRP transciever kits from China? The "Pixie", it's a long established design. They cost not a lot, and while the instructions can be patchy there is plenty of accumulated knowledge online.

Jenny List says:a minute ago

Apologies, this is rather slow here.

M.daSilva says:a few seconds ago

If I can add as a small HF project, there's the BITX40 board, a small 40m SSB transciever for about 60 bucks

M.daSilva says:a few seconds ago

(hence my question ^^ )

Jack Laidlaw says:a few seconds ago

right so consensus is standalone & digital 100mhz should put me in good standing for quirte a while?

Jenny List says:a few seconds ago

Scope choice? If an old CRT comes your way for nowt, grab it. But if you're buying new, I'd say get a scope rather than a USB device. Prsonal choice, in 10 years time will your usb device have software support?

Michal Fapso says:a few seconds ago

Hi Jenny, when I design a PCB where I use an RF module, communicating with the MCU at 2MHz over SPI, how much harm can I do to the RF performance when I am new to PCB design and don't follow any RF-specific guidelines for PCB traces? MCU: Atmega328p, RF module: NRF24L01+ SMD. Thank you for any hints

Jenny List says:2 minutes ago

Keep traces short and straight, Google RF design guidelines, use your PCB CAD package's design tool to calculate 50 ohm impedance stripline for RF traces.

Adam Vadala-Roth says:2 minutes ago

michael ossman's video is a great resource on this subject

Jenny List says:a minute ago

"Do I really need to pay extra $$ for controlled impedance PCBs for a short 50 ohm microstrip? (BLE)" Probably not, unless you need every ounce of performance. You'll get away with FR4.

Michal Fapso says:a few seconds ago

Thank you Jenny and Adam. And how much important is the metal casing seen on some RF modules?

Jack Laidlaw says:a few seconds ago

Jenny you make a really good point. drivers etc for old devices are a bummer. I tried asking universities if I could buy (well donate) them cash for a donation of an old scope but EU regulations on waste don't allow that. :(

Jenny List says:a minute ago

"How can I convert a BitX40 SSB kit to 20m? Still trying to figure out the IF calculations (uses 12MHz IF, not sure if I can keep it that way): VFO at 2MHz or 26?" Gut feeling says always keep VFO frequency as low as possible for stability.

Tim says:a few seconds ago

@Michal FapsoIt's also effective to us metal shielding over areas that are prone to unintentional radation

John Aboud says:a few seconds ago

Thanks, Jenny for the suggestion on the Pixie. and M.daSilva for the BITX40. Interesting that both the Pixie (I just Googled it) and the BITX are 40m projects. I had heard about the BITX from the HamNation podcast.

Jenny List says:a few seconds ago

@Jack, take to hanging around the skips :) (dumpsters)

M.daSilva says:3 minutes ago

Thanks Jenny! I'm using a Si DDS, I'll try and see how it works out :)

skot says:3 minutes ago

Okay, thanks for the microstrip tip. Is there a rule of thumb for trace width for cheapo 4-layer PCBs? (I'm thinking seeed studio)

Jenny List says:3 minutes ago

If DDS, then less of a problem. In fact higher means better image rejection. When you said VFO I pictured traditional VFO.

Jack Laidlaw says:2 minutes ago

@Jenny List I'll give that one a go lol, EEVBLog Dave seems to pick up a lot of stuff that way.

Michal Fapso says:2 minutes ago

Thank you Tim

Jenny List says:2 minutes ago

Can't give width for trace, depends on variables. Your CAD package should have a calculator, if not there are ones online I'm sure.

M.daSilva says:a minute ago

I'll give it a go!

loxodes says:3 minutes ago

@skot I recommend OSH Park 4 layer for cheap RF boards, they use FR408 which is a little better characterized and lower loss than standard FR4

Jenny List says:a minute ago

"How does one convert a discrete PA/LNA match section into one with a balun? Like what is the process I should go through to figure it out if there is one ?" Apologies, probably beyond this engineer and this hack chat session without knowing more about the device in question. Sorry.

skot says:a few seconds ago

Seeed gives a dielectric separation distance range of 0.075 - 5.0mil, which basically makes microstrip calculators useless.

Adam Vadala-Roth says:a few seconds ago

its all good Jenny, I just thought I'd ask LOL the dsicrete implementation on the S1G project I have on here is just really expensive to buy all the inductors

Adam Vadala-Roth says:a few seconds ago

you need to buy 10 at a time for 1 buck each

Adam Vadala-Roth says:a few seconds ago

And there are 6 unique parts

Adam Vadala-Roth says:a minute ago

so looking to reimplement it with a balun

Jenny List says:a minute ago

"What one thing do you wish you had known when you started working with RF" Accountants make loads more money :) On serious note, can't instantly think of anything, as have constantly learned by osmosis.

Jenny List says:a few seconds ago

@Adam, apologies :)

steverobillard says:a few seconds ago

@LOL I learned by doing radio astronomy projects and back then the ARRL Handbook was one of the best resources (this was before the first web browser)

Jenny List says:a few seconds ago

"Any insight into 3D printing your own oscillators? Also where is a good place to buy oscillators/resonators? Thanks!" Not entirely sure what this question means.

Gummy Rawn Jawb says:a few seconds ago

I read you there have been developments in piezeocermaics and polymers

Jenny List says:a few seconds ago

(buying oscs/crystals) The usual suppliers for me, mouser/digikey/newark/farnell/cpc/rs etcetc

Gummy Rawn Jawb says:a few seconds ago

to replace the traditional crystal used in oscillators

Gummy Rawn Jawb says:a few seconds ago

I was just wondering if you had any info on that

M.daSilva says:a few seconds ago

Hey Jenny, about crystals: I'll need to match 20MHz crystal for a bandpass filter, never done it before, do you have any pointers?

Jenny List says:a few seconds ago

AH, in that case then I'm afraid it's outside my knowledge. Could be very exciting though!

Jenny List says:a minute ago

Buy a heap of crystals, test each one in an oscillator with a frequency counter. Annoying, tedious, expensive. Look for surplus packs of crystals.

M.daSilva says:a few seconds ago

Thank you! :)

Jenny List says:4 minutes ago

"What options can you recommend to send a video feed from a drone on a distance of 5 - 10km ?" I'm afraid yet again this is outside my knowledge. I'd guess you'd need modules of a watt or more for that, though I don't know the multirotor biz well enough to advise. There's a bit of a fuss atm though about them because some are sold with illegal power/frequency. Search HaD for recent ARRL article on the subject. Take care what you buy.

Radu Motisan says:3 minutes ago

Thank you!

Jenny List says:3 minutes ago

"For a 2.4GHz Wifi modules (ESP8266) with a ceramic antenna, how to get the signal out of a metallic enclosure? Will a few holes placed close to the antenna do any good?" I'd always say your best bet would be to have the bit of PCB with the antenna on it protrude outside the box. But try holes, see if it works. Suspect it'll be disappointing though.

Jenny List says:a minute ago

"I recently read about NFMI for data communications, does anyone have experience with that or can suggest any IC's and antennas to use? NFMI stands for near field magnetic induction" Sadly I don't know enough here to comment as yet. Yet, I say, because a fellow OxHack member is working in that arena and I hope to pick their brains.

Tim says:a minute ago

Thanks Jenny! I had found that NXP has some cool stuff, they call it MiGLO, but all the info on it seems to be hidden behind NDAs

Jenny List says:a few seconds ago

"If you were to build a local network of sensors, why would you choose to connect them via LoRA (for example) rather than to the wifi (for example)" Lower power, for a start. Longer range, as *much* lower bandwidth and lower frequency.

Radu Motisan says:a few seconds ago

Thank you! I expect the holes to lower the signal considerably. Still might be the only doable solution for that thing. Are there any tools that can help doing some debug on the path of the radio waves ? (in this case the 2.4GHz)

steverobillard says:2 minutes ago

@radu I would look at the wavelength and make holes that match much like determining the proper length for antennas

Jenny List says:a minute ago

@Radu, try it. There will be software almost certainly, but sadly I don't know which package.

Jenny List says:a minute ago

@steverobillard, good answer.

Radu Motisan says:a minute ago

@steverobillard thanks!

M.daSilva says:a few seconds ago

BTW, do you have the specs on the Raspberry Pi receiver board? I'm curious :D

Les Hall says:a few seconds ago

ty all for the chat!

steverobillard says:2 minutes ago

any link to the give away board it might make a good toic for the raspberrypi.stackexchamnge.com blog

Jenny List says:2 minutes ago

"Is a wireless electrode like reading possible by measuring a difference in an RF signal focused on a specific muscle? What would it take to measure the frequency of a nerve impulse via RF, in other words." That's an interesting idea. Have to say I'm an electronic engineer not a medic though. I'd be worried pumping too much RF into muscles, it's not always a good thing to have in your body.

Les Hall says:2 minutes ago

i read mostly everything, thought some of it went "woosh" over my head

Gummy Rawn Jawb says:a minute ago

It's possible though you think?

Jenny List says:a few seconds ago

Raspberry Pi radio kit: https://shop.languagespy.com/collections/frontpage/products/pi-hf-direct-conversion-radio-receiver-for-the-raspberry-pi

Jenny List says:a few seconds ago

Did I catch all the questions?

loxodes says:4 minutes ago

Radu, if you have the space for an external antenna it might be possible to desolder the chip antenna and solder on a pigtail coax -> bulkhead SMA to get through the shield. A cheap source for bulkhead SMA connectors would be salvaging one from a U.fl to SMA cable.

Jenny List says:2 minutes ago

It should be possible to sense changes in impedance through human flesh, yes. Perhaps if you measured impedance over a series of points you could compute the likely underlying structure. Suspect there are already imagers that use these kind of effects, though as I said I'm no medic.

Jenny List says:a few seconds ago

Re SEEED microstrip calculations, sometimes you have to take a view and fudge the figures and see if what you end up with works.

Greg Bushta says:a minute ago

Very interesting chat. Now I have to solder up the pixie that arrived two days ago.

skot says:a minute ago

thanks @Jenny List!

Jenny List says:a minute ago

Any time! :)

Jenny List says:a minute ago

Good luck Greg! :)

Tim says:a minute ago

Thanks @Jenny List!

a minute ago

We'll pick the board winners and get them out late next week

a few seconds ago

transcript here: https://hackaday.io/event/19718-rf-product-design-hackchat/log/53464-edited-transcript-for-rf-hackchat

M.daSilva says:a few seconds ago

Thanks! I'll get started on my Bitx assembly and Minima design!73 de EI/ON6RF ;)

Radu Motisan says:a few seconds ago

@loxodes that's a great idea, the only issue is that the enclosure is fully metallic, and I need to keep its surfaces flat

Shulie Tornel says:a few seconds ago

Thank you Jenny

Greg Bushta says:a few seconds ago

Thanks Jenny, been studying the ARRL Tech manual to pass the test. Soldering is taking a back burner.