Receiver Performance - Radio Design 101 Epilogue 1

In this Epilogue to the Radio Design 101 series, we look at receiver performance - using our own design as well as commercial radios to demonstrate design challenges. Strong interfering signals are shown to be a large problem and various techniques for dealing with them are illustrated.

[@SuperFredAZ]

I designed TV front ends 40 years ago, we built tracking filters with ,3 tuned circuits for low band and high band VHF using varactors plus a tracking varactor tuned oscillator. We always used the oscillator at ~45 MHz higher than the rf frequency. Nice to see someone covering this topic.

[@MegawattKS]

Awesome. Those definitely helped. We lost a lot of front-end protection when things got more integrated on-chip. I'm still amazed though at how well the single-chip receiver in the ATS25 radio works. Judicious use of AGC seems to help quite a bit - more than I would have expected !

[@SuperFredAZ]

Yes, we used AGC on both the front-end (called a tuner back then) and the IF, we used about 45 dB of gain reduction on the RF and ~ 60 dB in the IF, of course tv signals were basically AM and the output need to be constant for the video circuits@@MegawattKS

[@xDR1TeK]

I'm interested in circuit design and architecture. Both are equally important. We can draw block diagrams all day, we can do math all day and I am happy, but need to see how ideas find their way into circuit design. Books don't cover that much. Need more circuits and math and block diagrams. If people want to play, it's fine, but the pros want to see different orientations and articulation in every idea possible without going themselves on this path. That's why we come to KZbin. I love what you have done. It's absolutely educational and best have seen so far.

[@MegawattKS]

Thanks for the kind feedback !

[@acestudioscouk-Ace-G0ACE]

This subject is really above my knowledge grade and initially, I thought above my ability grade... but with your excellent clarity of teaching, I'm beginning to understand and absorb the points made. The beauty of KZbin videos is that you can watch them repeatedly till you have learnt what you need. Thanks so much for making this series.

[@MegawattKS]

Sorry for the delay in replying. My house AC system went out, so I've been getting behind on the channel. Thanks so much for the comments. It is good to hear, and you make an excellent point about being able to re-watch stuff, or backup and pause. KZbin makes video similar to reading books in that sense I think (although indexing and highlighting options aren't the same of course).

[@acestudioscouk-Ace-G0ACE]

@@MegawattKS Thanks for the reply. I hope your AC is now fixed and you are cool in all senses of the word.

[@dri50]

Looking forward to your future episodes on Receiver Sensitivity especially SINAD measurements.

[@MegawattKS]

Thanks for the feedback / comments. I'll try to include SINAD in that part (as well as BER vs Eb/No stuff for the digital modulations). FWIW, in the class, we used a surplus HP8903E Distortion Analyzer to measure SINAD sensitivity (together with an HP signal generator modulated with an audio tone at the FM standard deviation).

[@asdf56790]

Wow! I am just getting into RF design and your videos are just absolutely amazing!

[@MegawattKS]

Thanks for the complement. I'm glad they're helpful. Enjoy the RF journey :-)

[@sullivanzheng9586]

The last 3 min of your video is a real cliff-hanger and can't wait to see following videos. Actually I have started searching for digitally tuned capacitor ICs right now. My guess on why your circuit totally failed after removal of collector resistors (my wild guess, might be wrong) -- RF amp compression can have different causes and I guess in this situation, the input signal to IF amp is so high (0 dBm, which translates to quite high voltage, considering the high input impedance) has saturated your BJT. Simply by introducing more "gain"(actually small signal gain) will not effectively translate to real gain given this large signal input (which can not be modeled using small signal assumption), but could possibly make things worse (saturate the BJT even worse). My intuition is: maybe we need smaller gain for IF amp to keep that amp close to its linear (small signal area) by different biasing setup. Then filter it again, and add one more stage of IF amp. But still, I feel pre-filtering is an attractive, ultimate solution since you reduce/remove the interference of loud stations from the very beginning. After all, complicated intermodulation in LNA could be a trouble anyway, even if IF amp compression problem is perfectly solved. That is why I started searching for digitally controlled variable capacitor ICs. My intuition/guess might be totally wrong. I am curiously and eagerly looking forward to your next vid.

[@MegawattKS]

You're definitely on-target with the issue of intermod. That, combined with overload in the LNA or mixer can make big problems. But the reason the RX failed after the mod is not overload. Overload is OK in some situations like FM IF subsystems (so that any signal level at the input can reach a constant final output level desired by the demod), but only _after_ the IF filter, which it is in this case. I don't want to give away any spoilers, but a hint is contained in the old RF Murphy's Law adage: "amplifiers oscillate and oscillators don't" :-)

[@sullivanzheng9586]

@@MegawattKS Read several refs on FM demod and now I understand for most FM demodulator, amp overloading after IF filter is not a problem - and sometimes even desired. Oscillation - ha I had experience in audio amplifiers - If the VCC/VDD is not well AC grounded it will happen sooner or later. Thanks for the very helpful reply. Looking forward to your next video.

[@zidaniabdallah3741]

it will be very intresting if you schow as a schematic of a commercial receiver and an explanation from it motherboard and some measurements and a study of the stages of the radio receiver and the components used and theire values according to signals... I've been looking for that for many years ago so thanks Cordially

[@MegawattKS]

Agreed. I tried to include a little of that in parts of the videos in the Radio Design 101 series. It's hard of course since everyone's background is different, so what a person already knows varies. The Radio Design 101 series (and the course it supports) tries to take the learner through from the beginning with minimal math, starting with the basics and covering all the parts. Here's the full semester course set of videos that preceded this Epilog 1 video. kzbin.info/aero/PL9Ox3wpnB0kqekAyz6blg4YdvoEMoJNJY

[@rickharold7884]

Cool

[@zidaniabdallah3741]

I tried to identify the schematic that you had been schown us in your videos but I dont found any one of them and it was so hard to understanding what's going on so I need your help

[@MegawattKS]

Please see the videos in this playlist. Hopefully these will provide what you're looking for :-) kzbin.info/aero/PL9Ox3wpnB0kqekAyz6blg4YdvoEMoJNJY

[@y_x2]

In your schematic (around 27:00) the base of Q2 is connected directly to V1, is it an error?

[@MegawattKS]

Good question. It's actually OK (and even intended) in this circuit. While Q2b would normally be biased lower in voltage (and bypassed), this direct connection saves a resistor and a capacitor. It's OK in the DC bias sense because the collector quiescent voltage is V1, so with the base at V1 also, the collector can dip as low as about 0.6V below V1 before leaving the active region. The intent was to actually use the limiting that takes place when it tries to go further, since this is meant to be a limiting IF amp :-) So the negative swing a the collector can't go below about 4.4V. The positive swing is limited by Q3 base-collector becoming forward biased. (Actually, as the second "Epilogue" episode will show, the latter is actually a problem with this design. When Q3 goes into saturation (due to base-collector forward biasing during limiting), the capacitance on the node goes way up and the frequency shifts. More on that when the next Epilogue episode comes out :-)

[@lanceklein7957]

Wow! Amazing job! Increase your stats = P r o m o s m!