Dave! This is an old video, but I watched it when I was in middle school, now I'm in my third year of EE! You have no idea how much of an inspiration you've been.

I get bombarded with comments on an almost daily basis on how crap and annoying my voice is etc. There are so many people out there who think you have to sound and talk in a certain way that they expect. You can't be your natural self, unless it fits in with their narrow view. Screw them. As humorous as they are, it's gotten to the point were I'm contemplating just blocking and deleting every such pointless hatred filled comment.

Genius. Exactly what I'm after to try to help get a good intuitive understanding of filters. Thanks Dave!

Dave, i just saw this this very useful video. Thanks for the upload. Good job!

Wow, this will come in so handy - I recently specifically build myself a mini lab to learn more about audio range filters and was a bit baffled that there are no affordable audio frequency range spectrum analyzers. Now I will know how to solve that problem.

Thank you so much. This was a huge help in my Electronic Devices II Class. We used this for a variety of High-Pass and Low-Pass Filters. It really saved us time rather then setting each individual frequency and then plotting manually by hand or in excel. Our instructor is actually making labs to use you technique for the following semesters. I'd love to see more videos on using the features of the equipment, ie; oscilloscope, wave generator etc in analyzing networks and different elements. I'm hooked on your Blog! Thank you again!

Dave, you're my bloody hero!

Just dropping by to say that this was very very helpful and saved me many a time. Thanks!

This is a neat trick. I've been doing it a while to set up audio systems for client builds. It works great for tweaking filters for surround systems to have a perfect overlap between the tweeters, midbass, and subwoofers.

Block 'em Dave. Keep your videos coming. I find them very easy to follow and very educational. Thanks for taking the time to share your knowledge.

Very impressive. EEV is the best! Thank you for sharing this!

I've seen a lot of your videos before, wish I had seen this one earlier. I'm building a 10-Band EQ circuit with gyrators instead of inductors with no idea how to test it. I've built another circuit with just one LC filter for bass. I will try this method to test its correctly working. Thank you for this video!

Two big thumbs up for you Dave!!!

It would have been nice to see the generator sweep set to a log scale on the bandgap filter experiment. I was kind of expecting it. KEEP UP THE GOOD WORK!

Mind blown. Looking at my scope in a totally different way now. Thanks mate.

I use this trick mainly on my CRO , lovely trick Dave.

Very useful vid, I've got some projects in mind that require messing with various audio filters (which I've not played with before) so this should make my life easier! :D Many thanks.

BO - DEE That is a really interesting trick. I think it is near useless for any qualitative measurement but it does make you think about the way you can extended the reach of your instruments. A good engineer always thinks about the boundaries and you are certainly a good engineer. I loved this video, Dave. Keep it up.

Thanks so much for sharing your great trick!

thanks from México. You saved my soul

Thanks Dave, big thumbs up x2

Best advice I ever had!

I have used this to see the response of all of the IF/Audio stages with an RF input sweep, since it does not matter if the frequencies are different.

Thats a great video. I'm going to be playing with this Monday.

Wow, thank you! Was wondering if it would be possible to do this, and seems it is!

That's a cool trick! I usually do it using the XY mode on the oscilloscope

You're a genius Mr. Jones.

Very nice! You can select the scope to show the Y scale in log. This way is easier to analyze the graph. But yeah, the 8-bit ADC may be too low for this... Next step, show the phase plot :D

2 thoughts: #1) use math math=abs(ch.1). #2) use infinite persistence - you’ll get rid of the low frequency sweep problem.

awesome work

sweet info and a great ltspice test example project... thanks...:)

great video... hope you do more like this because most of us ( i mean, "me" ) don't know how to use the full potential of our scope... thank's a lot...

nice:) this the kind of lecture we all want go on bro,

Thank you so much, fantastic.

Great video! Thanks.

Great way to save $10K if you need to design/debug some complicated filtering and see what's going on. Thanks. Brilliant ! Any insights on how to do the same for getting a phase plot (perhaps another video ?).

remarkably innovative

I tried that some years ago to find the frequencies of my tesla coil. Thou I used X-Y mode and the ramping output on the function gen. Also, connecting a frequency counter I could easily find the exact number without calculating it from the analog scope.

I really wish you had looked at one feature of the signal generator - marker frequency. The trigger pulse indicates the beginning of a sweep with a rising edge. Normally, the falling edge is simply placed half way between rising edges. The trick here is that when the marker is enabled, the falling edge of the trigger occurs when the generator output frequency matches the marker frequency set by the user.

Hi Dave the way I have done this is using a noise source, and FFT on the scope

Huh, clever idea. Had to try it. I don't have fancy digital signal generator, but tested by hooking two analog ones; feeding triangle wave to VCF. Works....but tedious. And frequency range is only 2MHz max. But better than nothing...

Thanks! This might be usefull for my Bachelor thesis.

Great video thanks.

Watching this in 2021 and its still useful

Brilliant!

When using the band pass filter, why does the plot show an initial gradient that's steeper in the first half of the rise compared to what's seen in the second half of the rise? I'd have expected it to look more like a bell curve, without that little hump on the left. And why does it not have the same little hump on the descent?

That's a great trick - thanks a lot ! Now I wonder if it is also possible to retrieve the group delay as well ?!

It's fun to play with, enjoy. Try a peak detector as well.

I mentioned that. You'd have to convert the AC signal into DC.

Nice video. At the days I was dreaming about having oscilloscope I new next thing after scope would be a sweeper. You say: you only need a multimeter & a function generator but you mean probably a multimeter and a SWEEP function generator (We are talking to the young players). Not all function generators have that function even not all spectrum analyzers on the market have that TG (tracking generator) which are coming often as an option (+$$$).

Possibly with the aid of a frontend (synth chips perhaps) you could do the log/lin input power conversion.

if your oscilloscope has a log-math function, you should even be able to scale the y-axis logharitmicly or am i missing something here?

Bodeplot is Frequency-AND Phaseresponse. You just did the frequency part. But still nice. There are people call me "Pünktlischiisser" (Nitpicker).

I kind of like the filled in effect though, it looks groovy.

When I got my waveform generator this is literally the second thing I did right out of the box... Screwed with sync-out and trigger-in of the scope. Didn't try in logarithmic mode though! Guess I got to go back and try it.

Some scopes (Tek I think) have an envelope detect mode that could be used.

somehow the wave on the display of the scope makes me incredibly uncomfortable, can't tell why...

This is what i never thought off ... I had to measure / test some filters while building phase shifters and compensation filters... Thank you 😊👍

Nice trick to know

really amazing, now I only need a function generator and a scope

Using Syscomp CGM mini. They're like $50 now and ok in audio ranges. Can dump readings to CSV and do math on it with Excel or R and compare plots with reasonable ease. Getting better data will cost a fortune. This is quite near though.

Made my day..... You get wings in Heaven!

You don't absolutely need a Spectrum Analyzer with a tracking generator to generate a frequency response plot. If the spectrum analyzer has peak tracking, you can use a slow sweeping (or manual sweeping) signal generator. You will need to set the resolution bandwidth pretty high, and the sweep rate should be very slow. It may take a minute or two to sweep a few hundred MHz, but you can do it.

Bloody genius dave! Not gonna pay up 10k for a fancy spectrum analyser, no sir ree bob!

That's pretty neat Dave. The only negative (or "delta" in project manager speke) is that technically a Bode Plot includes both Gain and Phase. You're only showing phase. However, as I'm sure you'll agree, this detracts in no way from the value of your creative energies.

Ideally, would the signal generator's amplitude be adjusted, to maintain a constant amplitude, as the frequency is adjusted?

Funky use of the scope. Interesting video however I will have to watch it a few more times as I can't stop ogling at that MSO. I love Agilent "one day" so before I get arrested thank you for making this video. What a winner!

I love that agilent scope, we used those in a lab i had for a microprocessor class. Im pretty sure those things costs like 6000 $ though :S

Using your oscilloscope *and* a rather expensive waveform generator

fucken aweeseome. this actually has the potential to be way more high res than an FFT in certain respects because you can be more specific with ya frequencies in question where-as an FFT is limited to buckets.

"Poor mans bode plotter"... 40k oscilloscope... Yeah right. :P Anyway, nice video Dave. I'll have to try this when I (hopefully) get my function gen for Christmas.

I might shed some some light on this matter. Hendrik Wade Bode, the inventor of the Bode plot, was an American with Dutch roots and ditto name. I'm from the Netherlands, so know how his name originally was pronounced; It's "Boduh". A 'bode' was actually a job in the early days. It's someons who delivers internal mail and messages in offices and such. Wether the pronunciation is still used as the mainstream term I can't say.

Essa função é muito interessante p/ testes em amplificadores!

Venable dot biz has a few pieces of equipment that measures frequency response, phase response, etc. I'm not sure how much they run for, though.

Once you've got the plot you want turn display persistence on to infinite ;)

Dave, why didn't you use your scope in X/Y mode and use the sweep output of your generator for the Y axis?

Dave, I have a new DS1104Z and DG1022Z. Using the sweep, sync and XY setup I am trying to plot a basic audio freq response curve. Once working I could use it to test audio amps. Per your video's approach I was hoping that setting the scope to peak acquisition mode I would get just the top line displayed. No luck. The only Volts by Freq curve I can display is with FFT and it is not very good. The XY arrangement doesn't display a Volts by Freq plot. I can't figure out what I'm doing wrong. Any ideas?

thanx boss.....

Nice, but with an analog oscilloscope with z-Input and with a signal generator which generates markers (as Z-Modulation), it is much more comfortable to show the bode plot at a scope and to analyze the frequency response.

A crystal radio style envelope detector would help in this case I think.

Strange how it's in the Macquarie dictionary then!

Check out the Wikipedia page for Hendrik Bode (sorry, it's really Boh-dee). Incredible range of contributions: WW2 anti aircraft fire control systems, co-authored a paper with Claude Shannon, robotics, etc. He's responsible for the systems that shot down Wernher von Braun's V1 rockets and later worked with him in the US on space technology. Good video, thanks.

Another way is to take the FFT of the impulse response. But you would need a fast pulse generator.

Isn't it important to also note that the function generator should have a 50ohm attenuator on its output so the sweep doesn't cause the function generator itself to change voltage because of different impedance in the circuit over frequency?

Thrifty! Think you could get a similar result in XY mode if you put your function generator in VCO mode with the control voltage (frequency) on the X-axis?

Personally, I like the variety among "English" speakers. I don't expect everyone to sound the same - that would be boring.

Yes, I've heard that too. But I was taught the way I pronounce it and found that the most common. There was a time before the Internet and video when languages would grow and develop fairly independently. Now everyone expects people to talk like Americans!

Got to try it. How the sweeping waveform change to almost to bode plot after feeding into the device under test(filter) escape me. I was expecting to see a sweeping signal as before but with magnitude change.

Details... details...!

What are the nine little dots under each of the BNC connectors on your scope?

Dave I have tried this using the Sync output on my Siglent SDG 1020 gen on a Siglent SDS 1052DL scope but cant get it to lock. Can you help?

See the article Hendrik_Wade_Bode on Wikipedia. His co-workers pronounced it Bo-dee and the actual correct Dutch is Bo-dah.

Cool!

Dave, just a question and not trying to be rude. Did you get lost at any point through Peter's walk around, as a standard electronics hobbiest i found it overwhelming at parts.

Maybe you could generate a log signal with the other channel of the function generator and generate an XY plot on the scope. Then, you would have even better, you would have your log axis.

i still use a mic input to my daw, then look at the signal with a plugin

or can't you also get something resembling a proper bode plot etc by using the math function? dividing the filter output by the original output in channel 2? or something like that?

For active probes (i.e. expensive stuff)

Nice to know, but a cheaper Method is to use the analog discovery Form digilent

great video .. i was bale to get this.. however at high freq, if you zoom in the waveform is distorted due to low sampling rate (1s time span, but 20k) .. so lot of aliasing can happen. let me know if you think there is some solution to this..