Typically in this kind of application, a very few other important parameters to stimulate and test is the 3rd order intercept point (IP3), blocking margin, and because it is the first stage in your RF chain, Noise Figure. The first 2 are important because you don’t only have a single signal entering the amp, but signals varying hugely in amplitude and frequency. Love your videos.
For those out there that wish to build their own amplifier, I highly recommend the "Norton" design. Using a 2n5109 the NF should be ~1db across the HF spectrum, the amplifier is very stable and broadband, Using 43 or 61 material will move the response of the amp. Gain ~10-13db, depends on the amount of turns used in the binocular transformer. I've been using these amps for decades and currently use a single amp, for my SDR receiver. Don't forget to use FILTERS! in front and out of the amp!
The Norton amps, sometimes also called lossless feedback amps are great for RF-front ends, especially for their low Noise-figure. However, their PCB-layout is very critical, as they tend to oscillate at very high frequencies, and are also prone to BE-breakdown if a large signal hits them, so they can get tricky to get working just right.
If you are using this as a remote antenna amplifier, it might be worthwhile including diode clamps to ground and the supply to limit voltage excursions induced by lightning or atmospheric charging of the antenna.
Regarding input protection, I've seen a common implementation with parallel diodes from the signal line to ground - each conducting in a different direction - this way you limit the input signal to +/- 0.7V; letting it rise to the supply voltage might be too much.
Hey Mr. FesZ, Great video again. In radio, we would call that a pre-amplifier or a receive amplifier. both are used interchangeably and you can find circuit diagrams with that name.
For anyone interested, the main reason an inductor of a specific value is used in order to produce high amplifications at specific frequencies. This is called inductive degeneration. It is based on the fact that transistors, and any other electrical component will have some parasitic capacitance that on low level layout can be calculated, simulated, and even measured. The inductor is used to cancel out the impedance effects of the capacitor, since at the center frequency, these two impedances C & L expressed in the Laplace transform will cancel each other, producing infinite gain. This is not possible in the real world due to other impedances in the circuits, of course, but with additional resistance introduced by us (or the manufacturer), like the resistors in the emitter of the transistors (source/emitter degeneration), will in turn produce defined gains. I've simplified a bit the explanation since I just learned this last semester in my Master's in Electrical Engineering.
Awsome video thank you. It would be great if you could make a video explaining and guiding us torugh the proces that you used for the component calculations. So that one can actualy understand the math behind it you used and be abel to applye it to thiere own designs
Some typical 680 nhy chokes have self resonance of 1.5 MHz, which implies ~ 16 pf shunt capacitor across the choke. I am not sure why you left this out, it is certainly easy enough to calculate and would give more accurate results.
I've built once a few amplifiers with different SOT-89 gain blocks, but every time I experienced way higher harmonics than the data sheet suggested. I measured the source, it came through a SAW filter, with minuscule harmonic content. The output was properly attenuated too, so the culprit wasn't the overdriven input of my spectrum analyzer. The only thing helped was raising the supply voltage to the absolute maximum value or even above - with appropriately huge current draw - to get those harmonics to fall within the datasheet range. Any ideas what I was doing wrong?
I enjoy RF things. However in my college we don’t have much subject about it. Do you recommend any book to start with. (My background is EE, i am currently trying to design MOS op amp following Razavi analog book)
I recently bought a few basic wideband LNA boards for SDR. I would really like to have the ability to adjust the gain, via a trim pot. Is it possible to add one to these LNA boards?
Not that I know of.. One approach I've seen though, is the usage of a switched or stepped attenuator. So use an amplifier with more gain than you need, and then have an attenuator either before or after, and based on the exact attenuation that is set, the total gain can be adjusted; I guess in the same idea the amplifier can be switched - so have multiple amplifiers either inserted or removed from the chain.
Please do make a vid about filters in the 50 to 500 MHz range. My experience is that the frequency range is too low for HF trickery but too high for lumped inductors as they have to be in the nH range.
Had a discussion with an oldtimer youtuber, about the supply inductors. It was for final power amplifier of a HF transmitter. Classic. Where i pointed that the supply inductor is not a "choke". The idea of choke being like the higher inductance, the better. I pointed that the supply inductor cannot be big, therefore not a choke, as it is a matter of optimising, not just increasing inductance. Now, in your case here, if you make those two supply inductors thousand times bigger, in the assumtion there is no resistive component in the inductors, will it change the amplifier regime?
The problem with making inductors bigger is that almost invariably, their capacitance increases as well, also lowering its self-resonant frequency - i.e. the range over which it still looks like an inductor, and then starts looking more like a capacitor. That parasitic capacitance essentially shunts to ground, so it will reduce the signal level at higher frequencies. When you're simulating RF circuits, you really need to include all the parasitic elements of your device models in the simulation. The good manufacturers will have device models (typically S parameters) that can be used in the simulations. The higher the freqency you go, the more you need to consider parasitic effects.
@@Cynthia_Cantrell true, but what i was talking about, has nothing to do with the imperfections. Even if reactances are perfect, is still problematic. That is why i talk about it. I know even professionals, many of them, did not learn this well. I have a lot of experience of finding missinterpretations and bad recommendations. Sometimes am exasperated. And i remember how i was fooled in the past. In fact you never supply anything through an inductor, supposedly meant to block high frequencies. It is never an inductor that does the filtering. It is always an inductor plus a capacitor. Always! If you try to filter only with the inductor, it will not work. Filtering is always a matter of more than one device. Further, this omission will go together with some bad use of the superposition principle. To which i reply "what superposition? you can't have the superposition you think of, you can only have superposition that is achievable"
@@SergiuCosminViorel I got my EE degree in 1989, and have designed circuits operating up to 2GHz, including the front end for a measurement board for DirectTV (the satellite TV company). It worked on the first spin - better than they expected. I'm not sure what you're trying to prove here... I see a lot of handwaving.
@@Cynthia_Cantrell playing silly? if you did not get it, read again, carefully. do not contradict, it makes you stupid. i described very well, better than you will ever describe. and i too have a degree in technological engineering. and am a teacher. read again, maybe you'll get it!
@@Cynthia_Cantrell you do not sound like a good professional. read again, carefully, maybe you get something! Am not criticizing the amplifier. Am talking about something colateral to the amplifier, to any amplifier, am talking about the problematic of supplying through inductors. Why this? Because there is general confusion in this topic. What can bring the problem to the amplifier itself, is when a storage and transmission inductor, that is in some projects named "choke", has its role missunderstood. You cannot just throw in some high inductance, the amplifier may not work as intended.
I was wondering the same. Just guessing, but L3 would serve to put a DC bias on the cable between the two capacitors (C9 and C3). Since the DC bias is then on both sides of C3, the capacitance effects of C3 would be eliminated for the signal but still block DC current flow from cable to BFU520 base. If this wasn’t done, the reactance of C9 and C3 would form an AC voltage divider and there would be higher signal loss from antenna to BFU520.🤔
The idea was to show a method to send supply voltage to the block that is placed in the signal chain before the main amplifier - maybe you wish to add an extra amplifier stage or something. The signal source is V2+C9 and the line in-between is powered trough L3.
Hello ..could you please share the BFU520 Model file which you are using in your simulation .. it would be nice .. great video .. thanks a lot in advance
I used the model available on the manufacturers website - you will find it under "Design Resources" www.nxp.com/products/radio-frequency/rf-discrete-components-low-power/rf-wideband-transistors/2-ghz-rf-wideband-transistors/npn-wideband-silicon-rf-transistor:BFU520
The idea was to have the opportunity to add extra circuits down the line, if needed. I did not anything in the simulation, but I put in the inductor to confirm it has no adverse effect on the initial amplifier.
Exactly, the emitter resistor is there for DC bias setting only - for AC gain, I don't want to have that element. The AC gain will still be however limited by Re' - the equivalent emitter resistance built into the transistor - which is dependent on the emitter current.
My experience: Gain blocks are good for a quick solution with however compromises especially in frequency range and noise. Better quality is almost always achieved with discrete components.
@@glasslinger No, my experience is current. Signal to noise ratio was always better with discrete components in my antenna amplifier designs. In addition, price-wise I would not have a chance to get into production with the fancy ICs. Example: A 4 band automotive antenna amplifier PCB at 100 k pcs populated and tested should not exceed 70 € cents.
The "gain block" is always restricted to RF only? Are there any individual BJTs, single or combination, that could function up to the gain block's 5 GHz upper end? (I'm assuming that BJTs in an IC will always perform better that individual ones.)
I'm not sure if there is an upper frequency limit - on the AD gain block page, the component with the highest operating frequency goes up to 18GHz; anyway, in general an IC will have better performance at very high frequency compared to a discrete transistor since its easier to control and minimize parasitics when everything is on a small die. I guess transistor implementations can only be better at relatively low frequencies.
Audio gain blocks were quite common in Hi Fi receivers of the 70's and 80's. Simple, low noise, low power - one or two transistor class A circuits worked well as preamps or line level amps for tone controls etc. Many op amps of that time had crossover distortion and were expensive. Recently, I designed a gain = 6 ,one transistor, 5 MHz, 9v preamp for my scope to make up for the 15dB loss of a bridged T notch filter. Why use an opamp that is bigger and cost more with more external parts? KISS (keep it simple stupid) still applies...
This was our test in high school electronics class: Take a transistor with all data erased off it; determine what it is; bias it for some value of gain (4X+).
![Gain block RF Amplifiers - Building and Testing [2/2]](http://i.ytimg.com/vi/9m9lW0dX3n4/mqdefault.jpg)
![Gain block RF Amplifiers – Building and Testing [2/2]](/img/tr.png)
Typically in this kind of application, a very few other important parameters to stimulate and test is the 3rd order intercept point (IP3), blocking margin, and because it is the first stage in your RF chain, Noise Figure. The first 2 are important because you don’t only have a single signal entering the amp, but signals varying hugely in amplitude and frequency. Love your videos.
You have developed to be my absolute favourite YT channel right now. thanks!
For those out there that wish to build their own amplifier, I highly recommend the "Norton" design. Using a 2n5109 the NF should be ~1db across the HF spectrum, the amplifier is very stable and broadband, Using 43 or 61 material will move the response of the amp. Gain ~10-13db, depends on the amount of turns used in the binocular transformer. I've been using these amps for decades and currently use a single amp, for my SDR receiver. Don't forget to use FILTERS! in front and out of the amp!
The Norton amps, sometimes also called lossless feedback amps are great for RF-front ends, especially for their low Noise-figure. However, their PCB-layout is very critical, as they tend to oscillate at very high frequencies, and are also prone to BE-breakdown if a large signal hits them, so they can get tricky to get working just right.
Very educational video, keep going deep in the knowledge!
I like your red lighting. It is good for both of our mitochondrial health.
If you are using this as a remote antenna amplifier, it might be worthwhile including diode clamps to ground and the supply to limit voltage excursions induced by lightning or atmospheric charging of the antenna.
Regarding input protection, I've seen a common implementation with parallel diodes from the signal line to ground - each conducting in a different direction - this way you limit the input signal to +/- 0.7V; letting it rise to the supply voltage might be too much.
How did you achieve matching to 50 Ohms?
Pls continue making such videos.
Hey Mr. FesZ, Great video again. In radio, we would call that a pre-amplifier or a receive amplifier. both are used interchangeably and you can find circuit diagrams with that name.
For anyone interested, the main reason an inductor of a specific value is used in order to produce high amplifications at specific frequencies. This is called inductive degeneration. It is based on the fact that transistors, and any other electrical component will have some parasitic capacitance that on low level layout can be calculated, simulated, and even measured. The inductor is used to cancel out the impedance effects of the capacitor, since at the center frequency, these two impedances C & L expressed in the Laplace transform will cancel each other, producing infinite gain. This is not possible in the real world due to other impedances in the circuits, of course, but with additional resistance introduced by us (or the manufacturer), like the resistors in the emitter of the transistors (source/emitter degeneration), will in turn produce defined gains.
I've simplified a bit the explanation since I just learned this last semester in my Master's in Electrical Engineering.
if im correct, i belive 1% harmonic distoriton would mean the first harmonic is 20dB below the fundamental, i could be wrong though.
Awsome video thank you. It would be great if you could make a video explaining and guiding us torugh the proces that you used for the component calculations. So that one can actualy understand the math behind it you used and be abel to applye it to thiere own designs
Some typical 680 nhy chokes have self resonance of 1.5 MHz, which implies ~ 16 pf shunt capacitor across the choke. I am not sure why you left this out, it is certainly easy enough to calculate and would give more accurate results.
Great Video Content, Really Great And Easy To Understand, Wish You Could Make Video For HF Bias Tee Frequency Amps
Hello,
I just discovered your channel, and it is already one of my favorites
thanks a lot for sharing
from which country are you from please ?
Really nice!!!! Thank you
SHould have included some parasitics in the circuit and think about stability. BFU520 is a FAST transistor and will take any reason to oscillate.
Great video ❤
Awesome video. It will be even greater if you include how you did the calculations for the biasing
I mainly covered biasing in a previous series - maybe this will be of use: ua-cam.com/video/B3Oi2XIIRME/v-deo.html
Great video
I've built once a few amplifiers with different SOT-89 gain blocks, but every time I experienced way higher harmonics than the data sheet suggested. I measured the source, it came through a SAW filter, with minuscule harmonic content. The output was properly attenuated too, so the culprit wasn't the overdriven input of my spectrum analyzer. The only thing helped was raising the supply voltage to the absolute maximum value or even above - with appropriately huge current draw - to get those harmonics to fall within the datasheet range.
Any ideas what I was doing wrong?
I enjoy RF things. However in my college we don’t have much subject about it. Do you recommend any book to start with.
(My background is EE, i am currently trying to design MOS op amp following Razavi analog book)
RF circuit design, by Bowick is a nice one
I might get in trouble with this. I was just youtubing trying to find a use for an MPSA29 Darlington.
Thanks, FesZ 👍
interesting, thank you
Would love a video explaining Noise Figure... :-)
I recently bought a few basic wideband LNA boards for SDR. I would really like to have the ability to adjust the gain, via a trim pot. Is it possible to add one to these LNA boards?
Not that I know of.. One approach I've seen though, is the usage of a switched or stepped attenuator. So use an amplifier with more gain than you need, and then have an attenuator either before or after, and based on the exact attenuation that is set, the total gain can be adjusted; I guess in the same idea the amplifier can be switched - so have multiple amplifiers either inserted or removed from the chain.
Thank you!
Please do make a vid about filters in the 50 to 500 MHz range.
My experience is that the frequency range is too low for HF trickery but too high for lumped inductors as they have to be in the nH range.
do you have a vid about how to write a mosfet spice model from a datasheet?
Had a discussion with an oldtimer youtuber, about the supply inductors. It was for final power amplifier of a HF transmitter. Classic. Where i pointed that the supply inductor is not a "choke". The idea of choke being like the higher inductance, the better. I pointed that the supply inductor cannot be big, therefore not a choke, as it is a matter of optimising, not just increasing inductance.
Now, in your case here, if you make those two supply inductors thousand times bigger, in the assumtion there is no resistive component in the inductors, will it change the amplifier regime?
The problem with making inductors bigger is that almost invariably, their capacitance increases as well, also lowering its self-resonant frequency - i.e. the range over which it still looks like an inductor, and then starts looking more like a capacitor. That parasitic capacitance essentially shunts to ground, so it will reduce the signal level at higher frequencies.
When you're simulating RF circuits, you really need to include all the parasitic elements of your device models in the simulation. The good manufacturers will have device models (typically S parameters) that can be used in the simulations. The higher the freqency you go, the more you need to consider parasitic effects.
@@Cynthia_Cantrell true, but what i was talking about, has nothing to do with the imperfections. Even if reactances are perfect, is still problematic. That is why i talk about it. I know even professionals, many of them, did not learn this well.
I have a lot of experience of finding missinterpretations and bad recommendations. Sometimes am exasperated. And i remember how i was fooled in the past.
In fact you never supply anything through an inductor, supposedly meant to block high frequencies. It is never an inductor that does the filtering. It is always an inductor plus a capacitor. Always! If you try to filter only with the inductor, it will not work. Filtering is always a matter of more than one device. Further, this omission will go together with some bad use of the superposition principle. To which i reply "what superposition? you can't have the superposition you think of, you can only have superposition that is achievable"
@@SergiuCosminViorel I got my EE degree in 1989, and have designed circuits operating up to 2GHz, including the front end for a measurement board for DirectTV (the satellite TV company). It worked on the first spin - better than they expected.
I'm not sure what you're trying to prove here... I see a lot of handwaving.
@@Cynthia_Cantrell playing silly?
if you did not get it, read again, carefully.
do not contradict, it makes you stupid.
i described very well, better than you will ever describe. and i too have a degree in technological engineering. and am a teacher.
read again, maybe you'll get it!
@@Cynthia_Cantrell you do not sound like a good professional. read again, carefully, maybe you get something!
Am not criticizing the amplifier. Am talking about something colateral to the amplifier, to any amplifier, am talking about the problematic of supplying through inductors. Why this? Because there is general confusion in this topic.
What can bring the problem to the amplifier itself, is when a storage and transmission inductor, that is in some projects named "choke", has its role missunderstood.
You cannot just throw in some high inductance, the amplifier may not work as intended.
Can you explain the function of L3 please?
I was wondering the same. Just guessing, but L3 would serve to put a DC bias on the cable between the two capacitors (C9 and C3). Since the DC bias is then on both sides of C3, the capacitance effects of C3 would be eliminated for the signal but still block DC current flow from cable to BFU520 base. If this wasn’t done, the reactance of C9 and C3 would form an AC voltage divider and there would be higher signal loss from antenna to BFU520.🤔
The idea was to show a method to send supply voltage to the block that is placed in the signal chain before the main amplifier - maybe you wish to add an extra amplifier stage or something. The signal source is V2+C9 and the line in-between is powered trough L3.
I wonder what simulator software is being used.
LT spiceVII
Hello ..could you please share the BFU520 Model file which you are using in your simulation .. it would be nice .. great video .. thanks a lot in advance
I used the model available on the manufacturers website - you will find it under "Design Resources" www.nxp.com/products/radio-frequency/rf-discrete-components-low-power/rf-wideband-transistors/2-ghz-rf-wideband-transistors/npn-wideband-silicon-rf-transistor:BFU520
What is simulation software used for circuit design?
I use LTspice in all my videos
12:27 - why do we need connection with L3 inductor, establishing DC level between C9 and C3 ? section between C9 and C3 looks not used anywhere ...
The idea was to have the opportunity to add extra circuits down the line, if needed. I did not anything in the simulation, but I put in the inductor to confirm it has no adverse effect on the initial amplifier.
Hello sir. What is purpose of C1? At high freq it just shorts Re pushing ac gain to max.
Exactly, the emitter resistor is there for DC bias setting only - for AC gain, I don't want to have that element. The AC gain will still be however limited by Re' - the equivalent emitter resistance built into the transistor - which is dependent on the emitter current.
Transistor !👍🤖🦾👽
My experience: Gain blocks are good for a quick solution with however compromises especially in frequency range and noise. Better quality is almost always achieved with discrete components.
You must be using early designs of the gain block IC's. The later designs are AMAZING! Flat gain over bandwidths only dreamed of with discretes.
@@glasslinger No, my experience is current. Signal to noise ratio was always better with discrete components in my antenna amplifier designs. In addition, price-wise I would not have a chance to get into production with the fancy ICs. Example: A 4 band automotive antenna amplifier PCB at 100 k pcs populated and tested should not exceed 70 € cents.
Why power drone jammers are so expensive? (for legal civilian use in war zone)
The "gain block" is always restricted to RF only?
Are there any individual BJTs, single or combination, that could function up to the gain block's 5 GHz upper end?
(I'm assuming that BJTs in an IC will always perform better that individual ones.)
I'm not sure if there is an upper frequency limit - on the AD gain block page, the component with the highest operating frequency goes up to 18GHz; anyway, in general an IC will have better performance at very high frequency compared to a discrete transistor since its easier to control and minimize parasitics when everything is on a small die. I guess transistor implementations can only be better at relatively low frequencies.
Audio gain blocks were quite common in Hi Fi receivers of the 70's and 80's. Simple, low noise, low power - one or two transistor class A circuits worked well as preamps or line level amps for tone controls etc. Many op amps of that time had crossover distortion and were expensive.
Recently, I designed a gain = 6 ,one transistor, 5 MHz, 9v preamp for my scope to make up for the 15dB loss of a bridged T notch filter. Why use an opamp that is bigger and cost more with more external parts? KISS (keep it simple stupid) still applies...
@@FesZElectronics Thanks
@@jim9930 Thanks
@@jim9930With an op amp the DC response is preserved, which may be desirable depending on the application.
This was our test in high school electronics class: Take a transistor with all data erased off it; determine what it is; bias it for some value of gain (4X+).
Incompetents pure