It's so in details and it's so technically in-depth in its description that I've no idea what he's talking about. It's nice to watch anyway... One day, I will understand what you are trying to say...
Damn.... your videos are addicting, interesting, and straight to the point. I also like how you record and document your trial-and-error process on some videos. Excellent videos, dude!
i build a VERY basic one recently. Basically a metal detector works(in my mind) by having a coppercoil connected to some OPamps. Ferric metal creates an inductive load and opamps are sensitive enough to pick up on it. Mine isnt very sophisticated and i just made it with a mains transformer and a LM358 connected to some (rather pretty) LED ladder.
Joe Toe me too. But each video you watch you understand 1 percent more than the last. If you watch 20 videos, then start the playlist over again, you would understand more and more each cycle.
It's really very well produced but to him I think all the concepts are far too basic for an explanation, for us it's quite hard to follow, even though it's interesting.
You do realize that some of your diagram drawing edits for the sake of animation are so good that at least half the viewers aren't aware of them? Oh, and the diagrams are perfect as well :D mad respect on both counts :D Cheers from Slovenia
You only need 2 times the maximum frequency (call it f) for the sampling frequency. The imperfections you get have higher frequency components that were not present on the sampled signal, so the reconstructed signal will be exactly the sampled signal, if you pass it through a low-pass filter with a cutoff frequency equal to f. In the real world, you want to sample at a slightly higher frequency than 2f, because filters are not perfect. This is the reason for 44100Hz sampling frequency of CD, you get 22050Hz as your maximum frequency, but your low-pass filter is set to about 20000Hz, to remove the sampling artifacts.
This is true for audio applications, where phase is not critical. In other applications, such as oscilloscopes, having a sampling frequency only slightly above the Nyquist minimum will necessitate a very steep filter, which will invariably result in hefty phase shift. In these applications, you're better off with a sampling frequency around 5× the highest measured frequency (or more), and a shallower filter.
I don't see how this can be the case. The point of the nyquist limit is it's the minimum frequency you need to sample at to reproduce the frequency of the sampled signal, but as he says, you'll get the right frequency but you won't have a remotely accurate wave shape: It will just turn everything into a triangle wave. You will also have potentially horrible aliasing distortion.
Awesome, as always! Thanks a lot. I'm and Electronics Engineer and you refresh my knowledge in few minutes. Even far better explained than my professors at college.
About the nyquist shannon theorem: if you use a low pass filter on the output, the double frequency of sample rate would be sufficient to recreate the sine. This is what the theorem stands upon. Its the backbone of digital audio. It has to do with fourier transform. Check Technology Connections video on the subject about nyquist shannon theorem.
Everytime i surprise when i see the Nyquist ratio, In order to reconstruct your signal perfectly you have to choose your sampling frequency greater then your signal (maximum frequency of the bandwith of your signal) otherwise there will be aliasing and that cause loss of information. That is actually fantastic.
Strictly speaking, the Nyquist-Shannon sampling theorem does not state that the sampling rate must be higher than twice the highest frequency of the signal, but higher than the signal's bandwidth. So if you sample at 20 kHz and have a signal with frequencies ranging from 90 kHz to 100 kHz, you can still perfectly reconstruct the signal, since if your digital version of the signal contains a frequency of x kHz, you know the original signal must have been at (x + 90) kHz (the signal is "aliased" to below 10 kHz). This is called undersampling and is frequently done in ultrasonic positioning systems, where your signal can be bandpass-filtered in hardware, before being sampled at a sample rate much lower than the signal's frequency, which decreases the computational cost of analysing the signal.
It would be great if a project was picked and this type of explanation was used showing all the used components and what happens if you used to strong or weak a unit. Thanks for the video.
I've experimented with AD conversion using an Arduino in the past: my goal was to sample an audio signal, filter the lows, mids and highs, get their amplitudes, and appropriately analogWrite() the R, G and B lines of an LED strip. Then Signal Theory hit me, with a *heavy* stick called FFT, and the computational cost of such filtering. Luckily, I also bumped into the wonders of analog electronics, and eventually built a low, band, high pass filter using OPAMPS. Good times.
Really very informative. Wish they were longer by a couple of minutes and explain the quantization error in the ADS's real quick. Keep up the good work please!!
When it comes to ADC I personally love the theory over the practice of it... I mean, how beautiful and round is the entire concept of converting real world information to bits?. Now, the practive of it (sampling, aliasing, noise, etc) is dirty!
very well presented.. I'm an electronics engineer and can say that you did an amazing job as compared to the text book or a lecture on this. keep up the good work ! will be your patron soon :)
Looks to me like you've misunderstood the sampling theorem, by the drawing and argument you made :) Oscillating between 1 and -1, that is the fastest frequency you can reproduce. That frequency should be reflected in your system by the sampling-rate. In the case of human hearing, we can detect up to around half of 20kHz, which is reflected in the common audio sampling-rate 44.1kHz, allowing us to reproduce a maximum frequency of ~22kHz. I have no idea where you got the 10-times rule you're mentioning, but sampling at twice the maximum of required frequency range is quite enough.
I have no clue what you are talking about...worrying when this is called 'basic'!! Nevertheless I still watch everyone of your vids lol! If nothing else you make me want to learn which I guess is the whole point. Keep it up!
GREAT SCOTT! Yet another good video :) Nyquest Shannon also applies to those USB desktop audio converters for recording your own music or voice at home. Stepping up to a 24bit 48kHz sampling, A very noticeable difference when recording a piano or guitar compared to the basic 16 bit card that is in your PC. Bravo and well done explanation.
you can get a nice signal out of 2 samples per period but in post-procesing, when you already have future samples - use some sort of spline (cause we can't get the ideal shannon interpolation formula). You can also get a good result if you delay the output by 4 or 5 samples and do a spline over them. (similar to matlabs interp1 with spline function). but the more samples you can get the better
Good gravy. You may have just set some sort of record on how to teach the basics of ADC circuits. That was very short and yet effectively conveyed how ADC work.
You are not stuck with the number of bits of an ADC. If you need higher resolution, you can oversample and average, or use noise shaping and filtering to remove some quantization noise (equivalent to extra resolution) from a frequency band of interest below the Nyquist frequency.
The Serial output makes the Arduino run slower. That's why you have 9 kHz. Normally it takes 100 microseconds, so more something like 10 kHz to read analog inputs.
There's a reason that people only use 2x the desired max frequency. It's not just theoretical; most audio uses ~44kHz sampling rate, where the highest humans can hear is about 20kHz
I gt an adc0808 the other day. Its a nifty chip. It has an 8 channel mux on the input so you can connect 8 analog signals to it. Of course you can only convert one channel at a time.
hey Scott i want to start working on a special project. A DIY multi channel mixer. about 32 channels. kind of an ambitious project, but i have everything figured out, except the equalizer. i tried to watch some videos but they were not helpful at all, it would be awesome if you could maybe do a video on how to make a 3 band parametric eq? i am personally going to try to be using some rotary pots, but i dont know how to link them to a rang of a frequency, so yea. hell even a whole thing on making mixer would kinda of a fun project. i tried looking at videos and none of them explain it all that well, and your style realy works for me!
You should have used serial plotter at the beginning to see what we get when the prescaler is 128 and 16, to see the difference, then we could see what we get on oscilloscope and on arduino
why does it have to be 666? just saying... BTW, dang your handwriting is amazing. not a lot of people have that skill anymore, due to computers being more popular.
I think you have gotten Nyquist Shannon wrong. Any bandwidth limited signal can be accurately represented with the sample rate that is twice the frequency. From Wiki : (The exact error that you and everyone makes) Intuitively we expect that when one reduces a continuous function to a discrete sequence and interpolates back to a continuous function, the fidelity of the result depends on the density (or sample rate) of the original samples. BUT...(Again from Wiki) The sampling theorem introduces the concept of a sample rate that is sufficient for perfect fidelity for the class of functions that are band-limited to a given bandwidth, such that no actual information is lost in the sampling process. It expresses the sufficient sample rate in terms of the bandwidth for the class of functions. The theorem also leads to a formula for *perfectly reconstructing* the original continuous-time function from the samples. (No loss of data, lossless).
What do you think about creating an adjustable switching bench power supply? Should be able to at least use the entire primary side of a PC power supply, since it boosts voltage to over 300 volts. You could either make a new secondary side, or just upgrade the secondary side with higher voltage rated caps and whatever other components need upgraded. The power monitor chip is probably just powered off the 5 volt stand-by transformer. Should be as simple as hooking up your secondary side's output to a variable voltage divider so the the voltage monitor chip always see's 12 volts; and, when you dived the voltage more the actual output goes up. i guess to get less than 12, you could use a simple linear regulator.
greatscott! what would be a good way to learn more in depth about the components and boards themselves and their inner workings? i dont know where to start, but i think a mechanical and physic understanding of all these parts would greatly help me understand all the things you say that i currently dont
Hey @GreatScott I really enjoy watching your electronics basics videos! Can you make a video about how to develop a project from a Breadboard on to a strip board, more specifically how to construct a circuit on a strip board? Thanks!
You know. I see a lot of info on ADCs and DACs out there. I even had classes on how they work in Physics Engineering. But now I'd like to select ADCs and DACs for a specific project and I'm having an issue no one addresses anywhere: I'd like to select a pair of ADC and DAC that match. Surely that's a thing electronics designers have to do all the time. What considerations have to be made? What I mean by match is: have the same protocol communication characteristics and play well in the same system, connected to one MCU or FPGA. I'm not even referring to resolution or other specific characteristics.
I’m working on a reverse engineer of the Digital J11 CPU from a PDP 11/70. I’m wanting to add it to a diy synth I making. I think it will top out at around 12 MHz. I’m past the point where I really need an Oscope.
I have a few remarks. First of all, I would like to comment that your handwriting is absolutely immaculate. I wish I wrote as beautifully as that. Secondly, I believe your sketch, the first graph you drew (Frequency wave = Frequency Sample), must be incorrect. If the frequencies match, then there should be one dot per period right? In the sketch it's still one dot per half period.
Why was the signal being reconstructed with straight lines? My DSP class taught we use the lowest frequency sinusoid and that is why we worry about aliasing.
Greatscott, are breadboards bad because of their capacitance? Are they good for just random tinkering? Or can they keep up with precise timing chips and power supply circuits?
the capacitance of a breadboard is in the pF range, even lower. I've used them for signals in the 100kHz range and never had a problem. You shouldnt be dealing with microwave circuits unless you're a professional. Also a breadboard can handle several amps of current, but if you're doing anything with higher than 1A, I suggest you just use regular wire to connect things.
Capacitance gets annoying when dealing with high frequency signals. You can get a lot of crosstalk between stripes, and the signal tends to be rounded (aka the higher frequencies gets cut off). That said, I already used signals around 8MHz on a breadboard, but it looked more like a sine wave than the square one it was supposed to be (although a sine wave was perfectly fine for the application). Regarding timings, if you make sure to keep coherent lengths for the signal paths you'll be fine.
Only knew AC/DC but that seems to be quite interesting too!
back in black? highway to hell?
xD made my day even in 2020
It's so in details and it's so technically in-depth in its description that I've no idea what he's talking about. It's nice to watch anyway... One day, I will understand what you are trying to say...
Damn.... your videos are addicting, interesting, and straight to the point. I also like how you record and document your trial-and-error process on some videos. Excellent videos, dude!
Thanks mate :-)
Leo Takacs // Scam Baiting 100% Agree
GreatScott! Can You Please Explain How Metal Detectors Work? Cause Im Not Able To Find Nice Videos Anywhere Else On YT
AAYUSH AGRAWAL I remember julian illet had a video on them.
i build a VERY basic one recently. Basically a metal detector works(in my mind) by having a coppercoil connected to some OPamps. Ferric metal creates an inductive load and opamps are sensitive enough to pick up on it. Mine isnt very sophisticated and i just made it with a mains transformer and a LM358 connected to some (rather pretty) LED ladder.
I understand 10% of these videos... but I still watch them
Joe Toe me too. But each video you watch you understand 1 percent more than the last. If you watch 20 videos, then start the playlist over again, you would understand more and more each cycle.
lol me too😁😁
It's really very well produced but to him I think all the concepts are far too basic for an explanation, for us it's quite hard to follow, even though it's interesting.
What's the good old saying..... "Throw enough sh&t at a wall some will stick"
Same. Gotta watch it in .75 ha
You do realize that some of your diagram drawing edits for the sake of animation are so good that at least half the viewers aren't aware of them? Oh, and the diagrams are perfect as well :D mad respect on both counts :D
Cheers from Slovenia
You only need 2 times the maximum frequency (call it f) for the sampling frequency. The imperfections you get have higher frequency components that were not present on the sampled signal, so the reconstructed signal will be exactly the sampled signal, if you pass it through a low-pass filter with a cutoff frequency equal to f. In the real world, you want to sample at a slightly higher frequency than 2f, because filters are not perfect. This is the reason for 44100Hz sampling frequency of CD, you get 22050Hz as your maximum frequency, but your low-pass filter is set to about 20000Hz, to remove the sampling artifacts.
Good knowledge!!
This is true for audio applications, where phase is not critical. In other applications, such as oscilloscopes, having a sampling frequency only slightly above the Nyquist minimum will necessitate a very steep filter, which will invariably result in hefty phase shift. In these applications, you're better off with a sampling frequency around 5× the highest measured frequency (or more), and a shallower filter.
Which is what any decent and recent audio ADC does, by means of oversampling
I don't see how this can be the case.
The point of the nyquist limit is it's the minimum frequency you need to sample at to reproduce the frequency of the sampled signal, but as he says, you'll get the right frequency but you won't have a remotely accurate wave shape: It will just turn everything into a triangle wave. You will also have potentially horrible aliasing distortion.
@@vapourmile the interpolation in audio applications isn't necessarily linear. So good sine functions are created from few samples
Ohne dass ich das Video anschaue weiß ich schon, dass du das perfekt erklären wirst. Du bist eif der Beste.
your explanations are so easy and to the point that I can easily digest your understandings more easily than our professor's.
I never really liked electrical circuits or electronics before finding this channel. You are awesome!
Thanks mate :-) Always a pleasure to show people how awesome electronics can be.
You have great penmanship.
Awesome, as always! Thanks a lot. I'm and Electronics Engineer and you refresh my knowledge in few minutes. Even far better explained than my professors at college.
About the nyquist shannon theorem: if you use a low pass filter on the output, the double frequency of sample rate would be sufficient to recreate the sine. This is what the theorem stands upon. Its the backbone of digital audio. It has to do with fourier transform. Check Technology Connections video on the subject about nyquist shannon theorem.
the handwriting and drawings are so satisfying
Everytime i surprise when i see the Nyquist ratio, In order to reconstruct your signal perfectly you have to choose your sampling frequency greater then your signal (maximum frequency of the bandwith of your signal) otherwise there will be aliasing and that cause loss of information. That is actually fantastic.
Good stuff. Don't use straight lines to reconstruct samples, use low frequency sine waves.
i love how neat your schematics are
Strictly speaking, the Nyquist-Shannon sampling theorem does not state that the sampling rate must be higher than twice the highest frequency of the signal, but higher than the signal's bandwidth. So if you sample at 20 kHz and have a signal with frequencies ranging from 90 kHz to 100 kHz, you can still perfectly reconstruct the signal, since if your digital version of the signal contains a frequency of x kHz, you know the original signal must have been at (x + 90) kHz (the signal is "aliased" to below 10 kHz). This is called undersampling and is frequently done in ultrasonic positioning systems, where your signal can be bandpass-filtered in hardware, before being sampled at a sample rate much lower than the signal's frequency, which decreases the computational cost of analysing the signal.
Thanks. Nyquist-Shannon is super interesting.
Does anyone else think about this but this guy has amazing hand writing!
Ok this is amazingly comprehensive and informative
I am writing an exam on mixed analog and digital circuits this week. This video was a good revision on flash and sar adcs!
What can I say to you Scott, thank you every day!!!
It would be great if a project was picked and this type of explanation was used showing all the used components and what happens if you used to strong or weak a unit. Thanks for the video.
I've experimented with AD conversion using an Arduino in the past: my goal was to sample an audio signal, filter the lows, mids and highs, get their amplitudes, and appropriately analogWrite() the R, G and B lines of an LED strip.
Then Signal Theory hit me, with a *heavy* stick called FFT, and the computational cost of such filtering. Luckily, I also bumped into the wonders of analog electronics, and eventually built a low, band, high pass filter using OPAMPS. Good times.
Very informative. You are skilled in both your pedagogy and video editing, excellent work.
I really like this series please make more.
Really very informative. Wish they were longer by a couple of minutes and explain the quantization error in the ADS's real quick. Keep up the good work please!!
I love watching your videos even though I don't understand most of it 😂
same lol i feel like im watching chinese
i always wait for the videos the are quite helpful to me
When it comes to ADC I personally love the theory over the practice of it... I mean, how beautiful and round is the entire concept of converting real world information to bits?. Now, the practive of it (sampling, aliasing, noise, etc) is dirty!
very well presented.. I'm an electronics engineer and can say that you did an amazing job as compared to the text book or a lecture on this. keep up the good work ! will be your patron soon :)
Looks to me like you've misunderstood the sampling theorem, by the drawing and argument you made :) Oscillating between 1 and -1, that is the fastest frequency you can reproduce. That frequency should be reflected in your system by the sampling-rate. In the case of human hearing, we can detect up to around half of 20kHz, which is reflected in the common audio sampling-rate 44.1kHz, allowing us to reproduce a maximum frequency of ~22kHz. I have no idea where you got the 10-times rule you're mentioning, but sampling at twice the maximum of required frequency range is quite enough.
10 times is good if you actually want the curve to be right. If you only care about the frequency > 2 times is enough. I don't think he got it wrong.
Next level demonstration
Love it ❤️❤️❤️❤️
you are the best great scott
Next: #10 DAC (thanks, very interesting to learn about the lowest level details of such components)
Better explainations than my teacher !
I have no clue what you are talking about...worrying when this is called 'basic'!! Nevertheless I still watch everyone of your vids lol! If nothing else you make me want to learn which I guess is the whole point. Keep it up!
Thanks for share your knowledge. This playlist is awesome. I will waiting for video #28.
this video was very fany, congratulations your videos inspire me for mi projects
I should be studying geometry, but this is more interesting :3
actually geometry is very usefull in signal processing :D
Your handwriting is beautiful
that's more like electronic advanced than basic. hope to understand that in the future
hey great job, understood SAR method, I had my doubts but your video made it clear, thanks
GREAT SCOTT! Yet another good video :) Nyquest Shannon also applies to those USB desktop audio converters for recording your own music or voice at home. Stepping up to a 24bit 48kHz sampling, A very noticeable difference when recording a piano or guitar compared to the basic 16 bit card that is in your PC. Bravo and well done explanation.
I have no idea what you just said, but I believe you.
you can get a nice signal out of 2 samples per period but in post-procesing, when you already have future samples - use some sort of spline (cause we can't get the ideal shannon interpolation formula). You can also get a good result if you delay the output by 4 or 5 samples and do a spline over them. (similar to matlabs interp1 with spline function). but the more samples you can get the better
can you make a video where you show us how „Tesla Coils“ work? could be interesting.
Good gravy. You may have just set some sort of record on how to teach the basics of ADC circuits. That was very short and yet effectively conveyed how ADC work.
I don't know what are you talking about but it looks awesome hehehe. I will try to figure it out in the future, well-done bro :D
You are not stuck with the number of bits of an ADC. If you need higher resolution, you can oversample and average, or use noise shaping and filtering to remove some quantization noise (equivalent to extra resolution) from a frequency band of interest below the Nyquist frequency.
The Serial output makes the Arduino run slower.
That's why you have 9 kHz. Normally it takes 100 microseconds, so more something like 10 kHz to read analog inputs.
Awesome job! Love your videos, Love your teaching skills, You are just awesome!!
Awesome video and love your work
love your tutorial videos! keep up the great work.
idk why i watch because i dont get anything but i like his voice
There's a reason that people only use 2x the desired max frequency. It's not just theoretical; most audio uses ~44kHz sampling rate, where the highest humans can hear is about 20kHz
Nice drawings, what kind of pen do you use?
Project Paul he is using a black stabilo fineliner
Thanks.
Your question reminds me of those people thinking that cameras take pictures. By that logic, pens write poetry.
ADC is inside in computer's soundcard, it's used for audio recording.
I gt an adc0808 the other day. Its a nifty chip. It has an 8 channel mux on the input so you can connect 8 analog signals to it. Of course you can only convert one channel at a time.
hey Scott i want to start working on a special project. A DIY multi channel mixer. about 32 channels. kind of an ambitious project, but i have everything figured out, except the equalizer. i tried to watch some videos but they were not helpful at all, it would be awesome if you could maybe do a video on how to make a 3 band parametric eq? i am personally going to try to be using some rotary pots, but i dont know how to link them to a rang of a frequency, so yea. hell even a whole thing on making mixer would kinda of a fun project. i tried looking at videos and none of them explain it all that well, and your style realy works for me!
I liked the new intro!
neatest left hander ever, or second after Flanders
Jee man, I love watching you write and draw! That is some seriously good drafting skills!!!
Great.... now my brain is melting down xD. Tolles Video Scott :3
Good video.
Funny: ADC was one part of my "Messtechnik" - exam yesterday. :)
I wish you were my electrical engineering professors.
Thank you very much for such awesome content all the time :D
You have such lovely penmanship :).
If you have a 12-bit SAR converter operating at 1MHz, what will be the maximum sampling frequency to use in hertz?
Thank you I got some grip on this topic.🙏🙏🙏❣️🥰🥰
You should have used serial plotter at the beginning to see what we get when the prescaler is 128 and 16, to see the difference, then we could see what we get on oscilloscope and on arduino
How on Earth this is a BASIC electronic , for me it is very advance , but weirdly i am keep watching, good vid
Great insight into ADC'S.
What amazing video, keep going on!!
why does it have to be 666? just saying...
BTW, dang your handwriting is amazing. not a lot of people have that skill anymore, due to computers being more popular.
Juho L Because science is the work of the devil.
My handwriting was bullshit even before i started to use computers .-. I just didn't learn it properly and well...
Damián Cupo my handwriting was fantastic in cursive but I had to switch to print and now it's shit again
I always write in print, my cursive is just... not easy to te eye, why you had to switch to print?
Damián Cupo I grew up right as many schools in my area stopped teaching cursive. I learned it but many of my peers didn't and can't read it
Try using voltage references that match bit resolution, I.e, 2.048V or 4.096V.
Nice video GS
I think you have gotten Nyquist Shannon wrong. Any bandwidth limited signal can be accurately represented with the sample rate that is twice the frequency.
From Wiki : (The exact error that you and everyone makes)
Intuitively we expect that when one reduces a continuous function to a discrete sequence and interpolates back to a continuous function, the fidelity of the result depends on the density (or sample rate) of the original samples.
BUT...(Again from Wiki)
The sampling theorem introduces the concept of a sample rate that is sufficient for perfect fidelity for the class of functions that are band-limited to a given bandwidth, such that no actual information is lost in the sampling process. It expresses the sufficient sample rate in terms of the bandwidth for the class of functions. The theorem also leads to a formula for *perfectly reconstructing* the original continuous-time function from the samples. (No loss of data, lossless).
What do you think about creating an adjustable switching bench power supply? Should be able to at least use the entire primary side of a PC power supply, since it boosts voltage to over 300 volts. You could either make a new secondary side, or just upgrade the secondary side with higher voltage rated caps and whatever other components need upgraded. The power monitor chip is probably just powered off the 5 volt stand-by transformer. Should be as simple as hooking up your secondary side's output to a variable voltage divider so the the voltage monitor chip always see's 12 volts; and, when you dived the voltage more the actual output goes up. i guess to get less than 12, you could use a simple linear regulator.
That was very informative !
greatscott! what would be a good way to learn more in depth about the components and boards themselves and their inner workings? i dont know where to start, but i think a mechanical and physic understanding of all these parts would greatly help me understand all the things you say that i currently dont
make a usb midi keyboard using arduino nano. your tutorial are very easy to learn and great... like you
Hey @GreatScott
I really enjoy watching your electronics basics videos! Can you make a video about how to develop a project from a Breadboard on to a strip board, more specifically how to construct a circuit on a strip board? Thanks!
Scott du bist n geiler Leo, 4 Vorlesungen in einem Video erklärt!
You know. I see a lot of info on ADCs and DACs out there. I even had classes on how they work in Physics Engineering. But now I'd like to select ADCs and DACs for a specific project and I'm having an issue no one addresses anywhere:
I'd like to select a pair of ADC and DAC that match. Surely that's a thing electronics designers have to do all the time. What considerations have to be made?
What I mean by match is: have the same protocol communication characteristics and play well in the same system, connected to one MCU or FPGA. I'm not even referring to resolution or other specific characteristics.
I’m working on a reverse engineer of the Digital J11 CPU from a PDP 11/70. I’m wanting to add it to a diy synth I making. I think it will top out at around 12 MHz. I’m past the point where I really need an Oscope.
0:22 devil confirmed? XD
Hi! The video's image says "Eletronic Basics", you are missing the C. Just a small catch. Keep up your great videos. Bye!
I need to learn some basics basics and then maybe go back to check this tutorial.
I have a few remarks. First of all, I would like to comment that your handwriting is absolutely immaculate. I wish I wrote as beautifully as that. Secondly, I believe your sketch, the first graph you drew (Frequency wave = Frequency Sample), must be incorrect. If the frequencies match, then there should be one dot per period right? In the sketch it's still one dot per half period.
Great video !
I actually understood this video! Am I smart now?
You, Marius and My Playhouse should do a group video.
Ooooh the old intro ❤️.❤️ :3 !!!
Why was the signal being reconstructed with straight lines? My DSP class taught we use the lowest frequency sinusoid and that is why we worry about aliasing.
good introduction 👍😀
holy hell, i have alot to learn!
Greatscott, are breadboards bad because of their capacitance? Are they good for just random tinkering? Or can they keep up with precise timing chips and power supply circuits?
the capacitance of a breadboard is in the pF range, even lower.
I've used them for signals in the 100kHz range and never had a problem. You shouldnt be dealing with microwave circuits unless you're a professional. Also a breadboard can handle several amps of current, but if you're doing anything with higher than 1A, I suggest you just use regular wire to connect things.
Capacitance gets annoying when dealing with high frequency signals. You can get a lot of crosstalk between stripes, and the signal tends to be rounded (aka the higher frequencies gets cut off).
That said, I already used signals around 8MHz on a breadboard, but it looked more like a sine wave than the square one it was supposed to be (although a sine wave was perfectly fine for the application).
Regarding timings, if you make sure to keep coherent lengths for the signal paths you'll be fine.
Speaking of sampling, I'd like to sample the audio at 1:42.
A truly LOL moment.
I'm surprised he didn't mention the mcp3008 ADC. It's a decent chip, and works with the Raspberry Pi ☺
Nice Video! Could you please also talk a bit about oversampling? Thanks