EEVblog  

Could I pick your brains Winter? I would love to build some high frequency antennas and boosters ranging from 890Mhz - to 30Ghz range. For personal point-to-point data transmissions towers. We need high data throughput for free! I want to build a wireless mesh, by the people for the people. Could you recommend any of your favourite designs with dimensions and materials? It seems like some serious precision work and without the correct test tools. Currently not able to for pay RF chamber testing time but I'm always willing to learn from my mistakes, as this is just a hobby of mine. As I understand it, to avoid disturbance to other services on particular spectrums government authorities allocates/ "sells out" them to large corporate companies. Last time I checked the air we breath is free :) Thank you very much in advance! -Trev

Tradie Trev The air you breath is free, but when you start interfering with the GPS signal of the airplane trying to find an airport in whiteout conditions, you pose a serious safety threat. Dont be stupid. If you interfere with someone's cell phone, its annoying, when you interfere with stuff like GPS, you can get someone killed. That said you can create mesh networks from off the shelf equipment. Creating your own high frequency, high efficiency antennas and TX/RX equipment is not a small trivial task. It will require expensive equipment and a tremendous amount of knowledge. You may want to brush up on your Maxwell's equations.

Tradie Trev EMC testing and Antenna design are very different areas of expertise. I use antennas in my work but we do not design them. Also, working up at 30 GHz frequencies requires incredibly high precision in terms of board design/layout and component selection. Creating antennas and circuitry at frequencies that high is way beyond my, and most people knowledge. It really takes an RF design expert to make something that operates at 30 GHz. Best of luck & stay safe.

theonlyari Hadn't planned on killing anyone or disturbing other vital services, I'm not that evil :) It's quite amazing the amount of engineering that's gone into mobile phone towers, I know it possible to build such high through-put networks as many country towns rely on their microwave links as a means of communication.

10 years later: what is good bang for the buck solution for EMC pre-compoliance tool today? :) Today talked with R&S representative and 5-6k euro for FPC1500 + EMC module i little steep for one man startup company 😅 at least for sub-GHz RF product prototyping… thanks in advance! :)

+rovfitta ya I never really understood why they didn't teach it, we barely even touched spectrum analysers!

" the University of applied Sciences in Amsterdam a rigorous EMC course is compulsory, for ee students.

Hey Ronald, I'm currently at the same stage you were 6 years ago. Could you please share your experience on how you went through pre-compliance in the end? Our product is a 5V USB powered microcontroller device, wondering how I can test that for pre-compliance having a spectrum analyzer. Many thanks.

The forum appears to be unavailable at this moment.

Old tektronix Oscope like TDSxxx/C/D serie do offer FFT's so spectrum analysis.

3 types of measurements are common with this testing: Peak, Average, and Quasi-Peak Peak is exactly what it sounds like - the highest level measured at that frequency. It gives no indication of how strong or persistent the signal really is only the maximum value measured during the sweep. Average is also pretty self explanatory - it is a value that reflects the average level of the signal at that frequency. If the signal at a certain frequency is very common you will have a higher average than if that signal were intermittent, or infrequent. Quasi-Peak is harder to explain and somewhat like an average but calculated differently. Taken from an Agilent Technologies website: (quote) "Most radiated and conducted limits in electromagnetic compatibility (EMC) testing are based on quasi-peak detection mode. Quasi-peak detectors weigh signals according to their repetition rate, which is a way of measuring their "annoyance factor." They do this by having a charge rate much faster than the discharge rate. Therefore as the repetition rate increases, the quasi-peak detector does not have enough time to discharge as much, resulting in a higher voltage output (response on spectrum analyzer). For continuous wave (CW) signals, the peak and the quasi-peak response are the same. The quasi-peak detector also responds to different amplitude signals in a linear fashion. High amplitude low repetition rate signals could produce the same output as low amplitude high repetition rate signal. Quasi-peak detector readings will always be less than or equal to the peak detection. Because quasi-peak readings are much slower, (by 2 or 3 orders of magnitude compared with peak) it is very common to scan initially with the peak detection first, and then if this is marginal or fails, switch and run the quasi- peak measurement against the limits. For further information, please refer to Application Note 1328 "Cookbook for EMC Precompliance Measurements" literature part number is 5968-3661E." (end quote) Most, if not all limit lines that are defined by IEEE, FCC & CISPR will be Average and Quasi-Peak limits. Average and Quasi-Peak information is more telling of the true activity at a frequency than just the peak measurement. As far as I know Peak measurements are not compared to any limits for Pass/Fail criteria.

@@twomorestars app note does not google :-)

Goddammit

details, only TINY details XD (sorry for my nazi-grammar)

You can make your own LISN with higher voltage parts, or just change low-voltage parts in that device.

The title of the video is EMC Precompliance Conduction Emissions Testing. It's supposed to be a quick and dirty check you make before you go to the actual test house to have the unit officially tested. Hence the term "precompliance".

Chuck Sampson Yes, definitely. So wrong setup and assumptions shall cost a lot of money, time and effort.

I have.

It's the opposite of what you said.