Any opinions on this article? https://archive.org/details/ModernVHFUHFFrontEndDesign I think I basically understand how amplifiers, filters and mixers work, including how intermodulation products arise and how noise gets amplified by subsequent stages; however I've been struggling to get an intuitive grip on things like noise figures and 2nd/3rd order intercept points. I've now scanned through this article and I think that if I sit down one evening and work through it properly, things might just start to go "click" for me. To those with more experience: does it look sound? Is it still relevant today, 33 years after it was written? Mike
"Modern VHF/UHF Front-End Design" (1985)
Started by ●October 27, 2018
Reply by ●October 27, 20182018-10-27
MikeP wrote>Any opinions on this article? > > https://archive.org/details/ModernVHFUHFFrontEndDesign > >I think I basically understand how amplifiers, filters and mixers work, >including how intermodulation products arise and how noise gets amplified >by subsequent stages; however I've been struggling to get an intuitive >grip on things like noise figures and 2nd/3rd order intercept points. > >I've now scanned through this article and I think that if I sit down one >evening and work through it properly, things might just start to go >"click" for me. > >To those with more experience: does it look sound? Is it still relevant >today, 33 years after it was written? > >MikeNice article, makes noise power very clear. That will always stay valid. Reading further down the pages.. 2 meter transceivers are better now I'd think. We now live in the days of RTL-SDR USB sticks and software defined radio... My RTL-SDR stick has no problem receiving GPS signals at 1.5 GHz, but outdoors only with a GPS mouse antenna. Large scale integration... Even the 10 GHz LNB PLL and mixer in one chip: http://www.panteltje.com/pub/octagon_twin_LNB_OTLSO_inside_RT320M_PLL_IMG_6538.JPG So.. where will it go? hard to DIY, more dependency on special chips. war, large EMP, all your Baofengs and LNBs, and back to tubes? I used to have a metal box for EMC protection with some spare RF transistors in it. Now I have one to protect my FLASH USB sticks and cards. Imagine archaeologists digging up what's left shortly after trumpy canceled the nuke treaty, ... Look, in the 21st century they already made GHz chips? How did they do it?... Like now we look at the pyramids. Or it will be just ants and mosquitos that survived.
Reply by ●October 27, 20182018-10-27
On Sat, 27 Oct 2018 08:15:56 +0000, Jan Panteltje wrote:> We now live in the days of RTL-SDR USB sticks and software defined > radio... > > My RTL-SDR stick has no problem receiving GPS signals at 1.5 GHz, > but outdoors only with a GPS mouse antenna.I have a LimeSDR Mini. It can receive everything... and so it has trouble receiving anything, because it has no ability to _not_ receive stuff. You can't do the whole lot digitally. The front end still matters, if only as a way of keeping unwanted stuff out. Mike
Reply by ●October 27, 20182018-10-27
On Sat, 27 Oct 2018 04:36:12 -0000 (UTC), MikeP <mjp993@googlemail.com> wrote:>Any opinions on this article? > > https://archive.org/details/ModernVHFUHFFrontEndDesignIt is still a very valid article which essentially appeared in RSGB VHF/UHF manuals for decades. However, the RF-environment has changed during the last decades. - For ground based communication, in which part of the antenna lobe sees the "hot" (300 K) earth, the noise figure or noise temperature requirements are modest as in the article. - In satellite (and EME) communication, the signal source is against the "cold" (above 4 K) sky, getting the ultimate preamplifier performance makes sense - Current semiconductors have a decent noise figures even at slightly higher collector/drain currents, making it possible to adjust between noise figure and strong signal handling. - These days, the spectrum is full of strong (nearby) signals up to at least 6 GHz. If a preamp has, say a 10 GHz bandwidth, it will amplify all this crud, causing potentially thousands of intermodulation products. - For best performance, the input spectrum should be band limited to the band of interest. Unfortunately a bandpass filter has also some passband attenuation, which adds directly to the noise figure. A bandpass filter with very high unloaded Qu (such as silver coated 1/4 wave resonator) will have a low attenuation. - Mixer intermodulation performance has been a problem, forcing a simple low loss filter between antenna and first preamplifier and a sharp high loss bandpass filter between the preamp and the mixer. - In SDRs with direct (or sub)sampling seems to tolerate big signals better. However, if there are mixers in front of the SDR, there are still these strong signal issues. To get the full performance of SDR, front end filtering is still needed, see FunCube Dongle vs. FunCube Dongle Pro+.
Reply by ●October 27, 20182018-10-27
MikeP wrote>On Sat, 27 Oct 2018 08:15:56 +0000, Jan Panteltje wrote: >> We now live in the days of RTL-SDR USB sticks and software defined >> radio... >> >> My RTL-SDR stick has no problem receiving GPS signals at 1.5 GHz, >> but outdoors only with a GPS mouse antenna. > >I have a LimeSDR Mini. > >It can receive everything... and so it has trouble receiving anything, >because it has no ability to _not_ receive stuff. > >You can't do the whole lot digitally. The front end still matters, >if only as a way of keeping unwanted stuff out. > >MikeStrange, I use it for 2 meters and 70 cm, wrote a simple spectrum analyzer for it, band is further clean: http://panteltje.com/panteltje/xpsa/index.html Antenna type matters a lot for selectivity, a simple rod or wire will pick up strong unwanted signals. I use slim-jims for AIS (ship traffic) and 70 cm. Those are basically a short on other frequencies. Have several of these sticks, main one has a temperature controlled oscillator: https://www.ebay.com/itm/272411458376 I designed and build several receivers from the sixties upwards, best one with cascode JFET pre-amp stage, low noise, good selectivity, for short wave and 10 meters. Tuners for VHF TV.. dual gate MOSFET. So now I play with that LNB stuff for 10.4 GHz... Not much I can do to improve that, only to make it more stable. I am sort of out in the country, small village, so far away from big transmitters, OTOH some years ago I called my cross the road neighbor on CB and told him I have a LED connected to my GPA antenna and it lit up every time he pressed 'transmit', asked him how much power he was running... eeeh So YMMV RF bands are polluted completely due to all the switchers around. I build a down converter for in my drone to send it up with a slim-jim hanging under it, to get real antenna height, thin coax cable with 100 kHz few hundred volts to power the drone and for the down converted signal. Antenna height matters. Have not done much testing with it yet, 'they' declared the area a no fly zone after that, close to mil airport. Oh well. I am not into hundred element yagis and that sort of stuff. I Eshail 2 sat is working, and I can RX it, then maybe start on the 2.4 GHz upload, will need a bigger dish, already have the power amp.
Reply by ●October 27, 20182018-10-27
upsidedown@downunder.com wrote>It is still a very valid article which essentially appeared in RSGB >VHF/UHF manuals for decades. > >However, the RF-environment has changed during the last decades. > >- For ground based communication, in which part of the antenna lobe >sees the "hot" (300 K) earth, the noise figure or noise temperature >requirements are modest as in the article. > >- In satellite (and EME) communication, the signal source is against >the "cold" (above 4 K) sky, getting the ultimate preamplifier >performance makes sense > >- Current semiconductors have a decent noise figures even at slightly >higher collector/drain currents, making it possible to adjust between >noise figure and strong signal handling. > >- These days, the spectrum is full of strong (nearby) signals up to at >least 6 GHz. If a preamp has, say a 10 GHz bandwidth, it will amplify >all this crud, causing potentially thousands of intermodulation >products. > >- For best performance, the input spectrum should be band limited to >the band of interest. Unfortunately a bandpass filter has also some >passband attenuation, which adds directly to the noise figure. A >bandpass filter with very high unloaded Qu (such as silver coated 1/4 >wave resonator) will have a low attenuation. > >- Mixer intermodulation performance has been a problem, forcing a >simple low loss filter between antenna and first preamplifier and a >sharp high loss bandpass filter between the preamp and the mixer. > >- In SDRs with direct (or sub)sampling seems to tolerate big signals >better. However, if there are mixers in front of the SDR, there are >still these strong signal issues. > >To get the full performance of SDR, front end filtering is still >needed, see FunCube Dongle vs. FunCube Dongle Pro+.I still have a superconducting fitler for the GSM bands, bought on ebay comes from an cellphone tower. Now that has high Q. http://panteltje.com/pub/super_filter/super_filter_front_plate_img_2576.jpg>
Reply by ●October 27, 20182018-10-27
On Sat, 27 Oct 2018 10:37:00 +0000, Jan Panteltje wrote:> I still have a superconducting fitler for the GSM bands, > bought on ebay comes from an cellphone tower. > Now that has high Q.Did it cool itself, or did they have to send a guy out to top up the liquid nitrogen now and again? ;) Mike
Reply by ●October 27, 20182018-10-27
On Sat, 27 Oct 2018 10:28:36 +0000, Jan Panteltje wrote:> Strange, I use it for 2 meters and 70 cm, > wrote a simple spectrum analyzer for it, > band is further clean: > http://panteltje.com/panteltje/xpsa/index.htmlI tried to do that with my LimeSDR thing. It didn't work very well, and I don't honestly understand why - I suspect that (a) the input and output ports aren't really 50+j0 ohms, and (b) it has some kind of AGC that I can't switch off in software.> Antenna type matters a lot for selectivity, > a simple rod or wire will pick up strong unwanted signals. > I use slim-jims for AIS (ship traffic) and 70 cm.I've had some great success building antennas.> Not much I can do to improve that, only to make it more stable. > I am sort of out in the country, small village, so far away from big > transmitters,Sadly things are different for me. There are pagers, DAB radio broadcasts, and cellphone signals, all being chucked out at enormous strength from the top of the water tower at the end of my road. I prefer to see it as a challenge, rather than a setback.> Have not done much testing with it yet, 'they' declared the area a no > fly zone after that, > close to mil airport.At my place I have an official restriction, courtesy of the local planning department, that I'm not allowed to build an airport. LOL! It's because I'm too close to the local USAF bases. Mike
Reply by ●October 27, 20182018-10-27
On Sat, 27 Oct 2018 10:28:36 GMT, Jan Panteltje <pNaOnStPeAlMtje@yahoo.com> wrote:>MikeP wrote > >>On Sat, 27 Oct 2018 08:15:56 +0000, Jan Panteltje wrote: >>> We now live in the days of RTL-SDR USB sticks and software defined >>> radio... >>> >>> My RTL-SDR stick has no problem receiving GPS signals at 1.5 GHz, >>> but outdoors only with a GPS mouse antenna. >> >>I have a LimeSDR Mini. >> >>It can receive everything... and so it has trouble receiving anything, >>because it has no ability to _not_ receive stuff. >> >>You can't do the whole lot digitally. The front end still matters, >>if only as a way of keeping unwanted stuff out. >> >>Mike > >Strange, I use it for 2 meters and 70 cm, >wrote a simple spectrum analyzer for it, >band is further clean: > http://panteltje.com/panteltje/xpsa/index.htmlOn that spectral display, what does your 0 dB represent (approximately) compared to -174 dBm/Hz (300 K antenna temperature).? Using a good SSB receiver with good front end selectivity on VHF, a lot of variable band noise can be heard (not limited by receiver front end noise). On UHF, the band noise level is lower, so only front end noise is often heard, creating a horizontal "0 dB" spectral analyzer display. Your VHF "0 dB" line must be much stronger than -174 dBm/Hz.
Reply by ●October 27, 20182018-10-27
On Sat, 27 Oct 2018 10:37:00 GMT, Jan Panteltje <pNaOnStPeAlMtje@yahoo.com> wrote:>upsidedown@downunder.com wrote >>It is still a very valid article which essentially appeared in RSGB >>VHF/UHF manuals for decades. >> >>However, the RF-environment has changed during the last decades. >> >>- For ground based communication, in which part of the antenna lobe >>sees the "hot" (300 K) earth, the noise figure or noise temperature >>requirements are modest as in the article. >> >>- In satellite (and EME) communication, the signal source is against >>the "cold" (above 4 K) sky, getting the ultimate preamplifier >>performance makes sense >> >>- Current semiconductors have a decent noise figures even at slightly >>higher collector/drain currents, making it possible to adjust between >>noise figure and strong signal handling. >> >>- These days, the spectrum is full of strong (nearby) signals up to at >>least 6 GHz. If a preamp has, say a 10 GHz bandwidth, it will amplify >>all this crud, causing potentially thousands of intermodulation >>products. >> >>- For best performance, the input spectrum should be band limited to >>the band of interest. Unfortunately a bandpass filter has also some >>passband attenuation, which adds directly to the noise figure. A >>bandpass filter with very high unloaded Qu (such as silver coated 1/4 >>wave resonator) will have a low attenuation. >> >>- Mixer intermodulation performance has been a problem, forcing a >>simple low loss filter between antenna and first preamplifier and a >>sharp high loss bandpass filter between the preamp and the mixer. >> >>- In SDRs with direct (or sub)sampling seems to tolerate big signals >>better. However, if there are mixers in front of the SDR, there are >>still these strong signal issues. >> >>To get the full performance of SDR, front end filtering is still >>needed, see FunCube Dongle vs. FunCube Dongle Pro+. > >I still have a superconducting fitler for the GSM bands, >bought on ebay comes from an cellphone tower. >Now that has high Q. > http://panteltje.com/pub/super_filter/super_filter_front_plate_img_2576.jpgHow much liquid nitrogen (or even liquid helium) does it use daily ? I might believe that some radioastronomy telescope site might be using superconducting band stop GSM filters, but why use such filters on cellular towers ?
