EEVblog Electronics Community Forum

Electronics => RF, Microwave, Ham Radio => Topic started by: MarkBRR on September 18, 2017, 06:06:54 pm

Title: Frequency specific RF peak detection
Post by: MarkBRR on September 18, 2017, 06:06:54 pm
Hi guys!
I'm thinking about doing racing drone a track timer based on VTX signal strengths. Current DIY solutions uses one RF reciever module (https://www.banggood.com/FPV-5_8G-Wireless-Audio-Video-Receiving-Module-RX5808-p-84775.html) per drone racing.
Is there a circuit or product that can do peak detection of multiple specific frequencies?

the VTX (video transmitters) are specified as followed:
5.3-5.9GHz - full table avaible here (https://oscarliang.com/5-8ghz-frequency-bands-b-e-f-different-brand/)
PAL/NTSC video signals carried over FM
25-600mW (most common is 25mW and 200mW)

The drone receivers are populating the 2.4GHz frequencies.

I've been looking into solutions like the ADL5910, but it seems like there has to be a smarter solution than to add RF filters for every frequency. Can you provide me with some guidance?

Best Regards
Mark
Title: Re: Frequency specific RF peak detection
Post by: randsl on September 19, 2017, 04:26:28 am
Can you elaborate your problem a little bit more please...

Do you need to measure RSSI within 5.3 - 5.9 GHz frequency?
What's the channel spacing for the 5.3 - 5.9 GHz range?

How much is the FM bandwidth? (I'm not very familiar with video transmission)


Title: Re: Frequency specific RF peak detection
Post by: MarkBRR on September 19, 2017, 10:03:20 am
Sure!

It is basically RSSI between 5.3GHz - 5.9GHz indeed.
Channel spacing in non-consistent but if the "Race" band i used, then 37MHz.
I believe the FM bandwidth is 6MHz but i lack a spectrum analyzer to validate this.
Title: Re: Frequency specific RF peak detection
Post by: dmills on September 19, 2017, 12:09:48 pm
For PAL, the baseband video is 6MHz, but the FM bandwidth will be very much larger because FM is a mess of bessel function sidebands (In fact bandwidth is poorly defined for FM).

You could probably do something like use one of the Linear tech microwave mixer ICs to mix down to something like 70MHz, then use a IF filter there to limit the bandwidth to maybe 20MHz, then feed a log amp.

Make the LO come from a programmable PLL and step rapidly between the required frequencies, digitising the log detector output at each step.... 

I am not actually sure that RSSI is the way to go here however, at 5GHz it is going to suffer severe multipath, possibly something based on the Doppler of the received colour burst would be better, but that would require a real recever per channel.

Regards, Dan.
Title: Re: Frequency specific RF peak detection
Post by: randsl on September 20, 2017, 06:29:39 am

The attached concept might work with your requirement.

I've used a similar setup to identify different frequencies which have been spreaded over about 500 MHz bandwidth.
I had a frequency counter instead of env. det.  and I used a IF frequency about 5 MHz. and it worked well.

Not sure whether it'll work well for measuring RSSI, just thought that it might work for your requirement.

Regarding the RF synthesizer, you can use one like this: http://www.ebay.com.au/itm/ADF5355-Phase-locked-Loop-RF-Output-54M-to-13-6G-Development-Board-VCO-Durable/232459959470?_trkparms=aid%3D555017%26algo%3DPL.CASSINI%26ao%3D1%26asc%3D20160706105120%26meid%3D78d52b927bc04647bd7dfdf250c64983%26pid%3D100508%26rk%3D1%26rkt%3D1%26&_trksid=p2045573.c100508.m3226 (http://www.ebay.com.au/itm/ADF5355-Phase-locked-Loop-RF-Output-54M-to-13-6G-Development-Board-VCO-Durable/232459959470?_trkparms=aid%3D555017%26algo%3DPL.CASSINI%26ao%3D1%26asc%3D20160706105120%26meid%3D78d52b927bc04647bd7dfdf250c64983%26pid%3D100508%26rk%3D1%26rkt%3D1%26&_trksid=p2045573.c100508.m3226) and it has a SPI interface which can be hooked up with a micro. without much hassle and you can easily achieve channel spacing even down to 10 kHz and frequency adjustment takes ~3 -5 ms per step...
 
You may need a amp + filter to drive the mixer if you use this one as the output has 2nd harmonic components and output is -5 dBm.

Title: Re: Frequency specific RF peak detection
Post by: randsl on September 20, 2017, 06:34:14 am
For PAL, the baseband video is 6MHz, but the FM bandwidth will be very much larger because FM is a mess of bessel function sidebands (In fact bandwidth is poorly defined for FM).

You could probably do something like use one of the Linear tech microwave mixer ICs to mix down to something like 70MHz, then use a IF filter there to limit the bandwidth to maybe 20MHz, then feed a log amp.

Make the LO come from a programmable PLL and step rapidly between the required frequencies, digitising the log detector output at each step.... 

I am not actually sure that RSSI is the way to go here however, at 5GHz it is going to suffer severe multipath, possibly something based on the Doppler of the received colour burst would be better, but that would require a real recever per channel.

Regards, Dan.

ooops, I didn't read your reply before posting my one... Just noticed that you've mentioned the same thing...
Title: Re: Frequency specific RF peak detection
Post by: MarkBRR on September 21, 2017, 10:52:42 am
Thank you very much for your replies. I'll be investigating the downmixing=> IF filter => log amp circuit for estimated costs and performance :)
Title: Re: Frequency specific RF peak detection
Post by: MasterT on September 21, 2017, 01:36:12 pm
Use the  same receiver linked in original post, since its analog video all you have to do is to switch receiving channel a few times in a second and estimate RSSI. Arduino could do it easily.
Title: Re: Frequency specific RF peak detection
Post by: MarkBRR on September 22, 2017, 09:11:33 am
Channel switching takes 25ms. With 8 racing drones the time resolution will be 0.2s - it is a bit too slow. A couple of modules may be efficient.