Michael Ossmann: Simple RF Circuit Design
Вставка
- Опубліковано 16 тра 2024
- This workshop on Simple RF Circuit Design was presented by Michael Ossmann at the 2015 Hackaday Superconference. It sold out almost immediately and for good reason. He has designed numerous popular tools like the the HackRF One and YARD Stick One. Michael's depth of knowledge and experience make him a leader in a field that is often called a dark art. There is no reason to fear RF design. Follow his recommendations and remove some of the mystery from the topic.
Read the accompanying Hackaday article:
hackaday.com/?p=196883
Visit Michael Ossmann's Website:
greatscottgadgets.com/ - Наука та технологія
As an experienced RF design engineer, I recommend this and agree with the key points. The wavelength guidance is however out by about a factor of 10. To reduce the importance of the signal path length, it needs to be less than lambda/4. Then on a PCB the dielectric slows the velocity by about 2.
Glad someone noticed that other than me. Also speed of propagation of a signal on a micro strip is not near the speed of light.
please i want to make fm receiver circuit without earth ground wire
If every component is perfectly matched to the system characteristic impedance, then the length of the transmission lines will have no effect on the impedance match. Only the loss in the dielectric will change with transmission line length. But if your impedance's are mismatched, then having a lambda/4 transmission line will effect your system performance.
There is a "Rule of a thumb" in the microwave community that you should't worry about transmission lines if the signal lengths are smaller than 10% of the wavelength in the dielectric.
Great I found a right person. Sir you are a RF design engineer, I'm sure you'll help me, rather guide me. I'm a beginner.
Sir actually want to know the working of the 433 MHz RF transmitter receiver module circuit diagram,and the components used in the circuit diagram. I want to know about the topics covered while designing this RF transmitter receiver module circuit diagram
@@mohamedelghmrawy7850 buy an FM radio from ebay remove the case and pretend you made it yourself
I love how Ossmann says in the beginning of the video "my qualifications are.. I've made some rf circuits," right on man. Learning and teaching doesn't need to be confined to the classroom or a corporate position! Ossmann's delivery is perfect for a layman like myself. I want an Ubertooth and/or YARD stick. My students would enjoy playing around with these!
I love your attitude. With most of the internet and people everywhere calling it extremely complicated and difficult etc, your attitude is really helpful for amateurs and self learners.
That was one of the best RF design lessons I've ever had, all summed up in 1 hour! Great work Michael Ossmann! Thank you for putting this together and sharing with everyone. Really great notes came out of this presentation and I look forward to implementing them into my own designs going forward.
What a great talk! I really like the emphasis on just being smart, reading the documentation, re-using things that have already been figured out. People are often very quick to deem something "too complicated" before even really trying. Not just with RF but with all electronics.
Thank you, Michael!
It's amazing just how much _useful information_ you can gather from this presentation.
Keep up the good work!
Follow the rules and design recommendations until you know enough to break the rules. Great presentation.
In RF, keep your lead length as short as possible.
In D.C, add filter caps as often as necessary.
Grounding everything properly, is a must.
Always keep the supply rail away from the RF.
Cross-coupling is not desired.
I’m an experienced RF engineer and I watched the whole video. You have some very good and intuitive ways of coming up with RF solutions. All the best.
You saved me a lot of time. I was struggling with calculations for 50 ohm characteristic impedance with a 2-layer board. The main point and trick I found that saved me is moving to 4-layer board because it adjusts the dielectric thickness between the top copper and the inner copper to a very convenient value of around 0.2mm for most fab houses. Thanks.
Fantastic presentation. I wouldn't have dreamed of trying to design an RF board until i heard this presentation, now I might give it a shot. Thanks!
delivery of complex information in a simple way is a talent which a very few people have
wow this was one of great presentation on RF circuit design
A great video demystifying practical RF circuit design. A minor correction 1-dB point (P-1dB) is 10-dB lower than third-order intermodulation (P-IP3) point. Thanking HACKADAY for posting this video.
The best explanation i've ever encountered on PCB design. Finally something that i can understand
Thanks for uplaoding, UA-cam is so amazing as a resource because of channels like yours!
I used to work in production making thin film resistor and microwave/rf networks but have to say this video explained alot of the reasons why things were laid out
Awesome!
small world lol love u samy
Samy Kamkar thank you for your contributions
Samy Kamkar samy is my hero
samy is my hero
Love you samy dude
Thanks for the very helpful talk! Lots of great advice!
After I finished my undergrad I applied at a laser manufacturer and got thrown into a entry level RF engineering position with no training whatsoever from the company. I learned so much on my own and was able to learn everything you've talked about on my own and I just wished I had seen this so much earlier. Their biggest issue was being cheap and constantly switching components for cheaper ones and manufacturers causing each batch of boards to be different from the next, it was extremely frustrating and I'd be pressed to make them work
23222😅😊😊😅😊😊😊😊😊😊😊😅😅😊 1:42 😅😅 1:43 😊😅😊😊 1:43 😊😮😊😊😅😊😅😊😊😅😊😊😊😊
Really enjoy how Mike takes a difficult subject and makes it accessible! For example, the recent fomu provides an open source (verilog) usb stack (see the crowd supply's 'TEARDOWN 2019: Making USB Accessible Developing Ultra low cost, Open USB Tools')
Thank you Michael, I've owed one of your HackRF One's since it was pretty much first offered and love it. You are absolutely an expert engineer on this, far more than most university-educated engineers.
@jshowa o You apparently don't understand why you want to use more than one or two layers. When you deal with RF, you especially need separate power and ground planes.
@jshowa o by the way, stitching is a bad idea when your using RF frequencies. You guys should take more notes from Michael.
@@fjs1111 No people should take notes from current experts in the field like Rick Hartley and a multitude of others not from a 7 year old video of something that (luckily) worked at the time! This stack-up may have worked for him but it ignores modern understanding of high frequency PCB design, signal integrity and return paths / fields.
@@ingmarm8858 I've been developing embedded systems for quite some time. When it comes to RF, Michael has it nailed. If you're not familiar with his work, take a look at some of his open source OpenBTS and cellular network contributions. He's not a 7 year old... I don't even understand why I took the time to write this other than to defend him
Really great info clearly gleaned from the trenches! A few extra points that I learned the hard way:
I can't stress enough the need to carefully read the datasheets but also the application notes that the manufacturers publish. As many of these transceivers have programmable PLL's/DDS's check to see if the manufacturer has written app notes specifically for your country's regulations and follow them to the letter if you're able. I got caught out by the PLL crystal I was using producing a harmonic on the output that bust the spurious emissions limits. I was able to change the crystal frequency (in hardware) and correspondingly the PLL divider (in software) to get close to the same output frequency but that nugget of info was buried deep inside one of several app notes although the datasheet headlines promised compliance...just not at max legal output power apparently.
The single chip filter-baluns can perform worse than discrete front ends but if you can make use of them then do. Only really important if you're trying to get approvals at maximum legal power but again check app notes comparing the two if possible.
Newer devices, although potentially better performing, generally have fewer app notes covering actual test results - buyer beware!
Decades ago, I was all about power...that magic RF deity.
Now, I am all about RX front end sensitivity and noise figure levels.
Taking even amateur radio RX to new lows in sensitivity without degrading selectivity is where I focus my efforts now.
Awesome talk. Thanks Taco!
Regarding the over 90-degree trace angles. back in the 1980s when I learn PCB design with red and blue tape on a drawing board (!!) it was the PCB manufacturers that told us not to use 90-degree or more acute angles to avoid ferric chloride build up in tight corners that could dissolve into your trace. Wide angles helped prevent the acid eating too much trace copper around corners. And it looks better too :)
Using acute angle pcb corners created reflections in your transmission lines
Great video , thanks for sharing
Thanks a lot of things dear Ossmann
excellent video about RF circuit design
Michael Great show I learned a lot. I viewed your demonstration in Nov 2020 and found that when you were on screen and pointing to a display the pointed at area was not visible to me; the same when the time the Screen has the display only and you are pointing off stage to a component part the pointer is not visible therefore very confusing. This is because the filming of the demonstration when you are making it, does not follow where the pointer is pointing to.
a very good look at the topic for beginners
Great presentation.
Thanks for sharing!
Good presentation thank you so much
His repeated advice to refer to the manufacturer’s datasheets is very useful. Manufacturers have done the hard work for you and provide you with proven designs. This is done to make their product easier to use and when their product is easier to use, you’re more likely to implement it into your design instead of something from one of their competitors. If you need to deviate or add to the design, you at least have a proven starting point. This also works in reverse, as well. Many times, over many years, I have been diagnosing circuits (especially, consumer grade equipment) where I didn’t have a schematic. I will often look up the datasheet for the IC in the section of the circuit which is malfunctioning. More times than not, the manufacturer copied the circuit right from the datasheet and you now have a schematic for that section.
Great talk, thankyou
Just what I needed! Thanks.
Ossmann = Awesome man
This guy is by far the best self-taught RF engineer I have seen. Does anyone have any other recommendations close to his level?
Thanks for the video! PLEASE tell me what the intro song is :D
Brilliant!
does the same design and layers and stack work for flex pcb also ? 4 layers?
Amazing !
What the h is that cylindrical single color coded smd on TV tuners and IF transformers which not found anywhere else in other field of Electronics?
I learned a lot
thanks a lot!
Test equipment, if it's an intentional radiator of RF signals, is not exempt. The exemption refers to digital devices, not radio transmitters.
1 hour video is more valuable than 5 years engineering college
Isn't that usually the case?
yeah
I'm not sure I'd go that far. It was a very informative video though.
This is a summary of ways to bypass the right way of doing RF circuit design. With this, you could never innovate in the field.
Breadth vs depth man. It's a thing.
That's bullcrap. There is so much not covered in this video...
Congratulations Michael for all that . I have a question, which circuit design software are you using for designing the specific circuit? And what is the preferable circuit design software for RF applications ??
Thank you very much.....
Mike uses KiCAD!
nice job
What is the name of the intro song?
very great video
Thanks!!
really awesome speach~!
Very nice!
25:40 everywhere you calculate 50 Ω, shouldn’t the differential also be 50 Ω, but with another trace width?
great demo...
Great explanation!
73!
MIle Kokotov, Z33T
What is he saying in 32:01? ballain or something like that, I can not figure it out, I am not native english speaker. Thank you so much!
Balun. A transformer, used to convert RF signals from balanced to unbalanced. The name comes from the words BALanced and UNbalanced.
en.wikipedia.org/wiki/Balun
Thank you :)
Does anyone notice that Mike didn't pour ground on the top layer?
I haven't seen any reference design without ground pour on TOP/Bottom layer.
I know that it it's a very dense board, ground pour maybe cause more trouble.
But the 'ground' also constitutes a port of RF circuitry, no? Anyone RF expert shed some light here?
I am no RF expert, but I'd like to know.
By the way, thanks Mike, for this great tutorial!
What are those small silver colored inductors and capacitors at 35:45 called? I have not seen them before. I have usually these cylindrical or paper caps and wire wound inductors
there are through-hole devices and smt(surface mount technolog)devices. in this videos, these are smt capacitors and smt inductors
Very nice presentation. At 42:48 though we can see a bad practice with the decoupling capacitor. The decoupling capacitor should be as close as possible to the pad of the IC. When you need to connect the supply pad to a via, you place the via further away, not in between the capacitor and the pad.
The funny thing is: it’s not necessarily needed… the via can be in between. But I would also place it as close as possible, unless there is a good routing reason for it.
Placing it farther away can add an unnecessary stub depending on what's being accomplished
Will my spark gap transmitter hooked up to a tesla coil pass FCC certification? It works on all frequencies and uses 1000 watts.
Yes, that should be fine
lol xD
As long as it's "test equipment" !
Tell them your first name is Nikola that will help.
You might wish to chat with professor Van De Graf at some point. **chuckle**
Those Johanson baluns are worth every penny!!!!
This video can really hack your day ;)
I learned a lot from this than=you!
Thxsomuch.
@10:00 I totally agree. Choose a chip that does it for you.
I am a programmer, but no electronics expert (at all). For a project i chose an nRF24L01+. I "just" had to poke some values into the chip's registers, and it would rock. What a giant difference that was from earlier radio-projects i did, as a total RF noob, with "dumb" radios circuits. pffff capacitors, resistors, filters, ..
I dove into datasheets before knowing what to write to the IC, of course.. but i guess that is not even needed for your Arduino project (meaning: there will probably exist ready-to-use libraries for your RF-IC of choice).. so you do not need to write your own code for interfacing with the RF-chip.
My point: If you know not much about RF, but you want to add some radio to your project: Choose such transceiver-ICs as mentioned @9:40
many thanks 👽
Just curious. If you have a DC blocking cap, why is it I can still measure the Vcc?
Parasitic voltage is usually quite low, rarely enough to affect a receiver, unless the supply voltage is quite large.
Nothing is absolute in electronics(mostly).
There are also hysteresis voltages that are present in a D.C field as well.
Watching this from 2022. "There is a huge array of RFICs available today" kinda makes me cry
?
excellent
designing RF circuitry is not my main major; as I am more classified with power electronics (dealing with high current AC And DC) but my B.Sc. was in telecommunications and electronics as I must ask; since the beginning of your lecture you said transceivers what are they and what is their major role in the design I mean if we omit them from the design what will happen a mismatch?
Transceiver is a device that is both a TRANSmitter and reCEIVER. He is saying that if your design requires a transmitter and a receiver, don't design them both separately, use a chip that includes both functions and is easy to use. Great simplification of your design.
Thanks Taco
Meet the man !!
It would be so awesome to see the slides at the same time while the speaker is talking. Maybe someday technology will make it possible.
There are some videos like how you mentioned...
Please post chapters too
I never heard of a 4-layer PCB. Thanks! If you can't use 50 ohm for Z matching, use a R sub box or pot and an SWR meter. It would be more accurate to use a scope for Z matching, and since Z = square root of R squared - the difference of X sub L and X sub C, you can use a regular calculator. People, you can also learn dB via watching my UA-cam video. On UA-cam, type in Frank Reiser M.S., and then type in decibels.
Can you send the output of your transceiver to a power amplifier (possibly a op amp with PNP transistor to avoid clipping) and not worry about additional Z matching?
I didn't know that there were dedicated chip antennas. I want to build my own 2 m ham radio. I sold my old 2 m and 70 cm Icom, and a lot of electronics equipment, years ago when I got freaked out over how to build and repair SMD. However, I just learned how to deal with this, and now I am back in business. I am fixing all types of electronics.
Thanks. Great information; very practical!
Frank
Frank Reiser Video Audio Service
I am into high power rocketry and need a a data down link that can run off a Li-ion battery (3.7V). The data is a 3.3V TX TTL signal generated by the flight computer ( Adafruit Feather MO Adalogger). I am wondering if you could help me design a simple RF circuit. Maybe I should mention altitudes be over 50,000FT.
What did you end up going with? What frequency? Antenna?
I realize that the newest comments in this section are 3 years old but I'm designing a PCB which has an MCU and a sim800l module and a GPS receiver module. I think it's considered an RF circuit?
The ADF7242 is badass!
yeah arduino approach for RF design, that helps
All good points but - seriously in error on the "pop quiz" . . . internal impedance matching is not the only criteria in selecting components. The BGA777 has a lower noise figure, more gain (and selectable), and lower power consumption. It "lacks" internal matching (which is not a flaw) for flexibility. And a note on datasheets: When a chip maker lists "various" under recommended components, it means "various" companies manufacture that recommended component. An important point also - the datasheets give you a >basic< "recommended" layout, circuit, etc. It's up to the engineer to get his calculator out and build it to his own specifications, performance, etc. Datasheets commonly have errors also. The wise engineer will do his own calculations and verify his design thru measurements.
😍 Mic.🙏
Great video!
The bypass capacitor is on the wrong side of the power via at 43:11... the power should have to pass by the capacitor before it goes into the IC pin. It's not using the capacitor at all for high frequency operation if the pin goes directly to a via.
No it’s not haha.
51:45 I have replaced those switches on my hackrf. Do-able but yeah, hateful is accurate.
So basically he's saying: Use four layers and common sense. Like the 50 ohm rule: common sense. Use premade part instead of inventing your own; common sense. Still good to know as engineers often come up with the most convoluted ways to solve simple problems.
Loved it ..!!!
Michael, I can't get over the fact that you sound like Jon Lajoie.
4 layers....well...guess im not building that ADL5902 RF power sensor optimized for scattered field measurements anytime soon.
Unless you're living on third world wages, 4 layer PCBs aren't expensive. There are plenty of places you can order them online. OSHPark ( oshpark.com/ ) was mentioned in the video, and there are cheaper suppliers, such as DirtyPCBs ( dirtypcbs.com/store/pcbs ).
Super
It's all true. It just lacks the fun of refactoring designs until you've created something novel no one else has. I can spend 3 bucks on an RDA chip and make the equivalent of a pi controlled baofeng, or I can add a few transistors to the other stuff already in the pi and do the same thing without the RDA chip. One is definitely easier, the other way more fun.
what about counterweight of antenna?
Counterpoise?
Is this the brother of William Ossmann? or are the two entirely unrelated?
Watching master pa shout out
Can anyone will help me, I want to know, or you can say want to learn about RF transmitter receiver module circuit,so that I can design my own transmitter receiver module circuit for my RC control toys,so what are the stuff I've to study, in other words how do I start learning step by step. Hope someone will guide me
I would request u to start from aurdino based RF transceiver like nrf24 or lora modules. They are good for beginners to start designing RF circuits for toys and for other projects. U can find lots of videos about aurdino based remote controller. It would be a lot easier to begin with. And learn about the basics of RF like frequency,wave length etc.. and programming for the RF without that it would be just nightmare for u to design RF circuits!.
@@gokulkannan1845 Sir I'm so grateful to u, thanks 🙏💕 for the reply. Sir actually want to know the working of the 433 MHz RF transmitter receiver module circuit diagram,and the components used in the circuit diagram. I want to know about the topics covered while designing this RF transmitter receiver module circuit diagram
@@kamleshchavan7451 the circuit requires 433mhz transmitter and a 433mhz receiver and a aurdino nano. I request u to learn about aurdino boards working principle and about the GPIO pins in it. And I would like to know about the application of Ur project for clear idea.
@@gokulkannan1845 sir I want to make RC car
@@kamleshchavan7451 then it's better for u to go with nrf24 module with 2.4ghz frequency. For RC car, u has to use 2 aurdino nano boards and nrf24 transmitter and receiver. The nrf24 also has the same range as 433mhz transmitter . U are using aurdino based control system or any other custom boards?
hey new idea old subject that is if you have knowledge of computer programming and how to build them. led panels controlled by bios chips with voltage sensors placed at the end of every led to control and prevent over voltages of the led solar panel system. a bios can be augmented to allow the pass through of electricity only when needed just like the operation of a microprocessor. in this way only a portion or even a large portion can be used to create effective solar energy. this is possible because the bios program will only allow energy in when it is needed. this will prevent over voltaging of household circuits. a simple switch sensor will prevent too much power from reaching the circuit by saying it is not ready to use. this can be tested with very bright flashlights to prove how many chips will be needed in a circuit and should be test by someone. good luck.
verdim şukunu deluğanlu (y)
14:49
justin roiland??
I become sad and funny at the same time when you missed your favourite design hhh
Mate, every one of your products is out of stock in the UK.
The DC 50 Ohm value you're talking about isn't what you measure across the entire circuit. The impedance at any given point on the transmission line is defined as Z(d) = V(d)/I(d). This means that the electromagnetic wave peak voltage and current define the impedance at a given point. So although it's a good rule of thumb, this is why a one hour video is NOT better than 4 years of engineering college.
He is clearly talking about AC impedance, why else would the thickness and composition of dielectric figure into his calculation?
Cuz Mr Academy is only 3 years into his 4 year degree :)
Your comment isn't helping your case. Z = V/I also known as R = V/I or ohms law. Impedance includes AC reactance as well as the DC resistance that your explanation included
he never said DC .... be fair there is more to it then ''50 ohms''...fortunately for many applications you don't have to exactly know.