EEVblog

[@Schwuuuuup]

The Video was worth watching for the hint towards the Memory Setting for FFT on my Rigol Scope alone... Can't wait for the SDR one

[@skeggjoldgunnr3167]

The SDR spectrum analyzer with this near field probe is something I have been thinking about for a couple years now...I think some failing components act like a little radio station. I will use this for circuit board diagnostics in conjunction with thermal cam + nano-ohm differences along a circuit net to locate culprit. I've even bought a new cheap plastic boring toshiba laptop specifically for this application.

[@InteraliaTony]

Also looking forward to the SDR report.

[@Rodrickfung]

We had EMI / EMC testing for our homemade BLDC Motor Controller last week. Unfortunately, we failed the test. The result was nasty between 30MHz and 200MHz which is a low frequency band, but clean and beautiful in high frequency band which ranges from 200MHz and 1GHz. We figured out the bundle of noises mostly come from the high power switching nodes as the noise level significantly goes down when I put the ferrite bead at those nodes. I have intention to alternate the configuration in the BLDC board e.g. Gate Resistances, Switching Frequency, The cable's length blar blar blar to see any differences beforehead the second-test and development of the final prototype. It gave me surprise as I found this video useful from EEVblog ! Kinda fascinating to give me a full picture from EMI theory to pre-test stage before in the lab! I literally learnt alot from the video and plan for the internal test now! I've made the order for the coming DIY EMC probes! Although it's limited in near field test, at least, it's pramagtic and afforadable way for me to have a brief understanding to see what's going on and to "predict" the optimal parameters for mitigrating the issues! Thank you for the lecture!

[@gwc1410]

YES, do the software defined radio spectrum analyzer video.

[@PilotPlater]

Thanks for showing off all the different scopes. That was pretty slick, thanks Dave!

[@lincolnworsham1085]

What worked for us is to take the EUT to the test house for a "prescan" and then use the near field to debug problems. With the probes i'm just looking for x dB drop to bring the peek under the limit line.

[@sparqqling]

A quick and cheap pre-scan in the lab, find the source with the probe back at the office. Make a fix, test it with the probe and go into certification.

[@HighestRank]

‘peek’, what’s that?

[@lincolnworsham1085]

@@HighestRank The dyslexic version of peak....

[@UpcycleElectronics]

I do not need a new scope. I do not need a new scope. I do not need a new scope... Doesn't work I still want a new scope.

[@KeysightLabs]

You need a new scope.

@@KeysightLabs That's some real good superliminal marketing!

A brand new scope magically appears on my desk right now ... A brand new scope magically appears on my desk right now ... ;-P ... doesn't work either :-O

[@KeysightLabs]

@Mate You should buy all our things.

[@jafinch78]

You know you want a new scope. I totally bought an oscilloscope since was more cost effective than a spectrum analyzer. Then I bought like 4 more literally. Then decided I want to study the analog oscilloscopes and see how to update those to tap into the signals to make a spectrum analyzer, invested in 3 and was about given the other three (two free and the other like $10... technically the others I bought were around $25) and haven't gotten much further. Project Scope Creep (pun intended)! You know you want a new scope and a real time signal analyzer.

[@iamdarkyoshi]

The DS1054Z is honestly doing quite well in this considering its age and price point

[@EEVblog]

Yeah, it does get the job done, thanks to that software update giving 16k points

[@tablatronix]

I will have to rewatch the fft episode of all the scopes, although I think its a bit old now. Very unsatisfied with my fft, ready for a new scope

[@eliotmansfield]

didn’t even realise my 1054 could do that.

[@GadgetUK164]

Wow, that's fantastic! I would have never thought a scope with FFT could be used that way.

[@DanielLopez-up6os]

The RTL-SDR radios are really awesome, using "Q branch direct sampling", on some of them allow you to go down to the hundreds of KHz, without needing any sort of extra circuitry, albeit "only" at 3.2 Million Samples per second.

[@stevetobias4890]

I have ordered the necessary parts to make my own analyser like you made in the other video. Am looking forward to playing with it.

[@kevincozens6837]

At 8:07 I like how Dave refers to the 200MHz 4 channel scope as a "low end scope". :) Dave should use the amplifiers when he tests the SDR.

[@bastelwastel8551]

That low end crap scope can stand no chance against my analog 2 chan 60MHz hameg.

[@skeggjoldgunnr3167]

@@bastelwastel8551 I know those analogue scopes are great. Please make video showing us how!

[@rowifi]

Useful concepts, but only useful if you can subsequently take the product to a calibrated chamber, or if you already know what frequencies are above the limits. What can look like bad emission on a near field probe may often not radiate far at all, and vice versa.

[@EEVblog]

Yes, but still can be a very useful tool, and dirt cheap.

[@christopherj3367]

loved it, fantastic scopes.

[@GiGaSzS]

The Keysight looks promising :) I have tried Rigol 1054 and found out that it has many issues including occasional serial signal loss (displays as nothing is connected!). A temporally fix is to click AUTO. Not to mention annoying slow controls.

[@jafinch78]

Great presentation! The use of an oscilloscope as a spectrum analyzer is what got me into this test equipment addiction mess. One of my project plans is to see if I can hack the TDS-520 frankenscope into having more options for FFT. I still need to determine if the memory upgrade is really worth is since the sampling rate isn't so hot. Really need to upgrade the sampling rate.

[@hoggif]

A preamp is a great idea. I have a set of home-made probes but I never though about preamps to increase signal.

[@God-CDXX]

man you have some nice scopes

[@God-CDXX]

PS first

[@seanet1310]

One good thing the scope is great on is low frequency noise if you are having issues even if you have a nice SpecAn handy. Most cheap to mid range SpecAn and EMI Recievers only go down to 9khz. If you play in the military space or select parts of the commercial world you may end up with sub 150khz conducted emissions or even radiated emissions in the case of mil-std and def-stan.

[@mroek]

The Siglent does have a 1M point FFT, but the FFT analyzes what's on the screen (down to the nearest power of two). And when you sample at 1 GS/s and have a horizontal setting of 1 uS/div, you're getting just 14k points to work from (and the FFT then analyzes 8192 points). At lower sweep speeds, you can get the FFT to analyze the full 1M points no problem. But it will be quite slow at updating....

[@EEVblog]

Ah, makes sense, thanks. Obviously I didn't play around with it enough.

[@mroek]

@@EEVblog I think you'll find it is quite capable, and also remember that you can use the history mode (also works for segmented memory capture, of course) to do FFT on each and every frame that has been captured. And also, if you want to do FFT on a smaller part of the waveform, you can zoom in using the horizontal control, and position (using the horizontal position) the sweep to where you want. The FFT then analyzes just what you see on the screen. This is for example useful if you've captured something like a string of DTMF digits on a long timebase, and you want to see what each digit is. Just zoom (by adjusting the horizontal, not actually using the zoom function) into the part of the signal you want, and use the FFT to see which two frequencies are there (and then look that up to see which digit it was).

[@mohammad-mahditaghipour4307]

Useful info.Thanks. Also please do a video on the E-field probe as well.

[@Funkenflug96]

The Siglent has such a low resolution because you are using just 8k points as it is displaying on the top left. To use the full 1M point you have to adjust the timebase. Of course you have to make sure that you enabled enough memory under acquire. I have used the FFT on the SDS1204X-E quite a bit and i am really impressed by its performance

[@EEVblog]

Yep, goofed that

[@23RaySan]

When your SDR Dongle (rtl sdr i guess?) arrives, give the software qspectrumanalyzer a bowl

[@springer9406]

If the clock is a square wave, you should see the fundamental tone as the highest, and then each odd one decreasing in power as you go up in frequency. But you don't see that with either your scope or spectrum analyzer. I'm not sure why....???

[@JackZimmermann]

Whoa! That Rohde & Schwarz RTB oscilloscope. Thinking instead of buying six-piston calliper brakes for my sports car, getting a new R&S scope instead. So if I crash and die, it's partly your fault Dave! ;) Great video, as always!

[@무적쏭]

정말 큰 도움을 받았다고 생각했는데, eev였다니! ㅋㅋ

[@1kuhny]

It'll be interesting to see some more RF stuff from you. Hoping you do a lot of expierments with the SDR. Lots of cool things to do with it.

[@richosthoff7212]

GREAT SCOTT, I could sure use an oscilloscope!

[@BloodAsp]

I hope you do a bunch with your SDR, I got one to play with.

[@strangersound]

That Keysight is a beauty. The gray color scheme is way easier on the eyes than that bromide off white thing that seems to never go away. ;) Great video, Dave. This was more than the title suggests. Scope use, comparisons, and more. :)

[@EEVblog]

Thanks. Yeah, ended up longer than I wanted, I kept talking...

[@velinr]

The OWON scope. At 12:40, the vertical units are Vrms, not dB. This is the reason for the crap measurement...

[@foxabilo]

Dave we need to design a mount for 3d printers like the Creality, Prusa, A8, etc etc for that probe so we can 3d map the EMI values, that would be a killer app, the app being able to go to the same place in space to 0.1mm repeatedly , heat maps over different spins. Just need a mount on the print bed to keep the pcb in place... I see big projects from this

[@andymouse]

this is a great idea …...would love to see it

[@Valenorious]

Actually that would be a 4d map. 2d for horizontal plane, another 2d for frequency and amplitude. My hat goes of for you if you can manage to visualize that. Now, if you would like to image a specific frequency then that might work as a heatmap. But then you won't see those frequency comb paterns

[@foxabilo]

​@@Valenorious i.imgur.com/Ji8sndf.png just an idea, that would be 15 or so Z axis zones as the stacks, the Y is represented as the coloured intensity graphs go back , and the X in this instance is in 3 zones, i am sure with a bit of manipulation it would be fairly trivial to display a 64 by 64 grid of such plots which would then show the Y as depth and the amplitude as colours and frequency as the position inside each box, you could zoom in and out of such data sets to gain more perfective and perhaps overlay a 20% transparent image of the PCB for orientation

[@Valenorious]

@@foxabilo I can't quite fathom what you are trying to describe. But it is past bedtime here, so that may be just me. I would avoid plotting several similar graphs next to each other. That will be information overload on your visual pattern recognition neurons.

[@Elektronik-1]

With the latest firmware, fft on the Siglent is very fast...

[@uberpixzels]

With 16384 points the update speed is very fast (>12fps) on a SDS 1204X-E. It starts to get slower with 64k points but is still very usable (I'd estimate around 5-7fps). In terms of frequency, it seems that 500MHz is the limit. The black-padded level/power scale on both sides of the window takes up valuable horizontal space that could otherwise be used to show the signal. An option to hide one or both sides would be useful. This is with the V6.1.26 firmware from 2018-09-26 from www.siglent.com I also tried using the scope with FFT over the web interface; still a usable framerate. I see an HTML5 VNC client in use for mirroring the display. If the raw data was accessible in the browser, or say as a UDP stream, it could make for some very interesting development indeed. I'm willing to wager that there is enough raw processing power to do that just fine.

[@JAKOB1977]

A shame you never got a look on the old Micsig scope from that same era. as I recall that one had like 110k FFT points, but it was extremely fast, if I made a sweep on one of these MAX2870 synth chip Sweep board with 2.8" inch toucscreen that can go from 23.5Mhz to 6.5Ghz and set the sweep from 24Mhz to 500Mhz with step sizes at 1Mhz at 40ms pr sweep.. the sweep was suppresingly fluent, though with some bouncing harmonics at the lower sub100Mhz and significant lower amplitude up around 500Mhz, though its a 400 buck 100Mhz scope.. but you need to dial it in first on the waveform and then you can remove that channels and let math take center.. the FFT can display up to like 13Ghz, but in that fast timespan you have so few FFT points, and the BW is a fraction.

[@bobkozlarekwa2sqq59]

I’ve been following your KZbin channel for quite a while, and I’ve learned a lot! Many thanks. I have a repair situation that I’m hoping you can provide some suggestion for. I use a Flex 6500 for my HF transceiver (ham radio). I have a 2M transverter and amp connected to it. Both devices are keyed by the 6500 using an open collector connection the 6500 provides (TX3), providing a switched ground. I discovered that the TX3 outputs was stuck in a constant “on” position. The open collector transistor is obviously shorted. Sounds like an easy repair, right? Flex has a hard and fast policy not to assist anyone in self repair, nor have they released any schematics. Their only response is, send it in. So, I’m faced with finding the surface mount transistor and replacing it. Do to the density of the two PCB’s it’s impossible to do it visually. My idea was to inject a low-level RF signal into the TX3 output connection. I’m thinking of making a very small loop, maybe ¼” connected to my spectrum analyzer and trying to sniff out the connection. Have you ever been faced with such a situation? What I’m unsure about is because the transistor is shorted, won’t the RF be all over the radio on the ground plane? Keep up the good work!

[@peterellis2764]

Fab vid as always Dave :) SDR here we come!! Ive been designing amplifiers lately and I only have a pc based oscope. I does me for what i need but a spectrum sdr now that would help me massively Cant wait Thanks Dave

[@leppie]

You did not use the audiophoole power cables! Dammit Dave

[@TheBmannumber1]

I've got a small question: how good are those

[@JackZimmermann]

Most of the ones I've tried actually are. They use amplifiers from Mini-Circuits and other high-quality providers.

[@EEVblog]

First you have to define "low noise". They use standard amplifier chips so you can check the datasheet

[@yaidontknowwhattoput]

You should try exporting the raw and fft data to a .bin file and do the fft with your computer then compare the two.

[@idahosagebrush5662]

Check out the SDR Play receivers for making a Poor Man's Spectrum Analyzer. 10 Mhz visible bandwidth, front end filters, built-in preamp, etc. and the software is very well done. Well worth the extra price which is very reasonable considering how well they work compared to a cheap "SDR Dongle".

[@DAVIDGREGORYKERR]

What about running the FFT code on a computer with the input coming in from the sound-card with the H-Field probe plugged into the Line input of the sound-card, try to stick to 32768bins and no more.

[@tonysfun]

What is the adapter called that connects the oscilloscope to the RF thing? ( BNC to DIN 4.x? what size is it? Or is it BNC to SMA? Are there several sizes or one size is used on most RF input/outputs?) I do not know the terminology of RF connectors; where can I find it so I can order me some cables and adapters. How did you make the probe? What is under the big "bump" hiding? How thick should the insolation bee? Does liquid-tape OK to use to create the insolation? What material is used to make the sensor/probe itself? I never worked with RF so I need way more info than provided here, but thanks for the video. And thanks other guys for answering my questions!

[@JackHalkabar]

Please do a video designing a PCB probe like the professional one!

[@EEVblog]

I plan to.

[@markkrutzmann6862]

@EEVblog you might even include an amp in there, as an all in one solution.

[@etmoimeme]

Hi Dave, do you know how to verify the claim (when I was student, I heared that...) that blowing on a solder joint when mounting a hf part (let's say a transistor) can ruin its functionnality ? One of my professor said the porous joint would affect the hf thing. I have no idea if that's true, but I suppose a lot of other things could affect the high speed signals before a poor solder joint. What do you think of that?

[@EEVblog]

That would have to do with the "plastic region" of the solder when cooling I'm guessing. As to it changing the HF properties, maybe, if it's a really horrible dry joint? But sounds like more myth than reality

[@HighestRank]

Which may be why even in cheap Mexico, clean rooms would use a hooded vacuum to desmog solder bath areas instead of pressurized desiccant- here I could be giving away corporate secrets plus charged with espionage for my insensitivity.

[@redoverdrivetheunstoppable4637]

Dave: "let's try another scope" Me: "let's try another battery" but i have to try that, now "we" have the proof of working, sweet to see the 200mV\div showing something, sometimes having a probe like that around saves you a bit of headscratches for circuits that just need to "go"... and to "go dangerously" in voltages... or to avoid hooking up leads to a tuned circuit etc (but i know it's totally different from a connected probe)

[@jjoeygold]

DAVE a quantitative RF peak value will tell you if you will pass in the EMC test house or not.

[@EEVblog]

Yes, but test houses usually don't test for near H-field emissions, they are far field EM fields, quite different.

[@jjoeygold]

@@EEVblog yes but they do calibrate the test site with a comb generator. You can measure the comb generator in the near or far field and compare results with the DUT.

[@catalinalb1722]

Hi Dave, you also said about a PCB version. Do you have any updates on that? Thank you

[@JGnLAU8OAWF6]

Question is, could you (at least in theory) export whole memory of DS1054Z and do better FFT on a PC?

[@EEVblog]

Yes, people have done that.

[@jjoeygold]

DAVE why don't you measure a RF comb generator with calibrated RF peak and then home brew one!

[@bennguyen1313]

Did Dave ever do an EMI test using an RTL-SDR + $10 EMC Probe?

[@Cb80001]

Hey how can I measure magnetic field of high frequency with Beehive Electronics 100B EMC Probe?

[@electronic7979]

Very good 👍

[@forresterickson6225]

Could you add a link back to the video on building the probe? You have enough videos it is hard to find others with out a link or an EXACT NAME.

[@forresterickson6225]

I found it at: kzbin.info/www/bejne/aKncZHujZsSNp6s

[@ki4dbk]

:O This is totally cool!!!

[@sjcwoor]

hey Dave. any chance of a Teledyne LeCroy LabMaster 10-100Zi teardown???

[@Chris_the_Muso]

"fast FFT" = fast fast fourier transform LOL

[@johnfrancisdoe1563]

Chris Robinson Well Fast implementation of the fast Fourier transform algorithm. FFT itself is is ISTR an O(n log n) step operation, but the question is how many of those individual multiplications you end up doing per second. O(n log n) means that If 1024 sample FFT takes x arithmetic steps, 1048576 samples takes about 2048 times x steps. And that's per screen update. It would be interesting to know how this demo would go with a PC-based USB 3 oscilloscope (USB 2 is limited to about 10 Msps, while USB 3 might pull 200 Msps). Doing the FFT in the DSP/FPGA doesn't save bus bandwidth (1 FFT point per sample processed), so might as well use that multimedia math hardware already in the PC.

[@SwingFish]

"she just got Serious!" .. haha really great stuff! .. thanks alot

[@GoldSrc_]

He said "shit" not "she" :v

[@SwingFish]

@@GoldSrc_ oups .. yeah . well still funny as hell ;)

[@GoldSrc_]

@@SwingFish Forgot to mention, that it was probably a reference to the movie back to the future.

[@Muck-qy2oo]

I always thought that spectrum analyzers are much more efficient and precise than scope FFTs. How can it be so exact here? I'm astound of the similarity.

[@JAKOB1977]

They are, and and also optimised with SA feaures, internal attenuation, and filters all that jazz that are crucial, waterfall, and control over resolution bandwidth (RBW) build in LNA tons of markers and deltas and the list just goes on. I just purchased myself a spanking new 12Ghz full mental spectrum analyser that can reach 12.072Ghz with H-mixing and I had to shll out a tad over 100 of the big ones. Created by Erik & manufactered by Hugen.

[@Happyoldfashionfamily]

Off subject. Is there any decent online learning courses that dont cost alot and dont bribe you with crappy soldering iron and tools to sign up? I have a good amount of good quality tools for the job. Just need better understanding of how to diagnose.

[@vidasvv]

So how how do you "fix" it once you locate the source of the radiation ?

[@davidl8523]

shrink the loop, as he said in the video

[@xortan666]

Don't know why but I built almost the same probe and it does nothing, can't figure it out.

[@theminer3746]

Do you recommend keysight oscilloscope for beginner? I’ve watched your oscilloscope recommendation video and you didn’t mention keysight. I’m planning on getting an oscilloscope but the keysight one seems a bit over budget for me. But if you insist I may increase my budget.

[@velinr]

The price seems to be very important factor for beginers. Keysight/Agilent/HP is a very reputable manufacturer. They make nice scopes. But the price... Ah well... I mean - if you can afford it, go for it.

[@EEVblog]

The Keysight is my preferred scope for everyday use, it's just nicer to use and super responsive. But doesn't have the bang-per-buck of the Siglent and Rigols

[@dhpbear2]

Are most of these scopes Australian brands?

[@TheDefpom]

I fell I am lacking in the scope department...

[@thenaimis]

Not a single Tektronix scope... That OWON scope, though, the knobs look like some sort of potty training toy.

[@karlmartell9279]

Poor mans probe needs rich mans scope. Find out, what's wrong in the equation.

[@XtianApi]

Lol. My cheap scope does it too. Just slower update rate

[@jort93z]

I am quite excited for the SDR one. I am trying to justify getting an SDR to play around with for a while, and if you have sucess with it i will definitly get one.

[@eliotmansfield]

jort93z I bought a £20 rtl, followed by an airspy, then became licensed and now have a yaesu 991 - so i blame the humble rtl for getting me into radio.

[@bertblankenstein3738]

Dave Cad. Was already thinking it before he said it.

[@redtails]

3:12 4CHAN

[@huseyin.goktas]

Is there any way to get spectrum info on a scope which does not have fft.

[@EEVblog]

If you can export the sampled data you can run an FFT on a PC.

[@JAKOB1977]

I feel so priviliged from the way Dave says "4 channel" 3:11 Thats right bb, my unit also got 4 channels... daddy went deep, when acquiring labgear,

[@peppem94marsala]

SDR usb dongle have about 2MHz bandwidth!

[@EEVblog]

Yes, but you can sweep them across the range.

[@elbrujo3253]

What fume extractor do you use

[@MrAbrandao]

Better cover the probes with "liquid electrical tape"

[@ki4dbk]

Do you have this content on Odysee? I could not find it there.

[@MrAwyork]

Were did you get that fancy scope on the left?

[@EEVblog]

The manufacturers send them to me

[@MrAwyork]

@@EEVblog Looks like one of ones in your give away. They look very nice :D

[@danielcampanoflorido854]

By radio spectrum analyser do you mean measure radiated emissions?, About the usb that is at the beginning of the video

[@EEVblog]

See my previous video on the $10 EMC probe

[@dvlachy]

Hi guys. I created a 3D PRINTED CASE for the amplifier mentioned in Dave's video. For interested people: www.thingiverse.com/thing:3490143

[@jasoneyes01]

Bobby Desler is my uncle

[@charliebMZ]

You cant use the Owon scope correctly!

[@kapioskapiopoylos7338]

poor litle owon, how did they make such a bad fft function?

[@troyw5832]

☺😊😀😁😂5 Smile's Should off had a H u/T, E v/m reader and the probe attached to a meter reader/probe via cable to the scope boy well glad that's cheap $10na😥 even a cheap one is around $30 dollars for a low amount reader, boy a lot off equipment has high levels yet a old hydrogen light on ac zip and a gauss reader $130min as H/E only work with a magnet when it's live or moving.

[@eddiesomesan]

Hi Dave. Maybe you could take some ideas regarding the RTL-SDR dongle from here: kzbin.info/www/bejne/omWuZ3eBnNupbdk between 2:41:50 and 3:03:00. It`s a short presentation I`ve made last year about "seeing" a filter characteristics using a RTL dongle and a wideband noise source. Sorry, it`s not in english but maybe the pictures would inspire you. It`s not a professional presentation, only some experiments i`ve made as a ham radio operator searching for cheaper solutions. Some other software might be available now for that.

[@petertech6135]

I come from Vietnam, TEXBOX company

[@gamerpaddy]

40 bucks for a rtl2832u dongle. Lol dave you just got scammed. Even the original improved rtl-sdr dongle is half that price. The Chinese ones without tcxo are like 5 bucks

[@WacKEDmaN]

oh man...youre buying a FAKE V3 RTL SDR dongle...its probably not going to be a V3 board...only buy V3 RTL-SDRs from RTL-SDR.Com...they designed the V3...and a bunch have ripped them off...

[@EEVblog]

Who do I tell?

[@WacKEDmaN]

@@EEVblog they already know theres a bunch of knock offs all over the place...the problem is that they say is a v3 dongle, but its usually a V2 with a bit tacked on...performance is nowhere near the same as the V3..eg missing things like TCXO..or claim 0.5ppm or missing the bias-t function... V3s have some nice filtering aswell that removes alot of the usb line and power noise...i guess we'll find out when it arrives!..might be a chance for another scam video! LOL check here... www.rtl-sdr.com/cloned-sdrplay-and-airspy-units-now-appearing-on-aliexpress-ebay/

[@WacKEDmaN]

the problem with RTL-SDRs for this is that they only have 2.8Ms/s... there are some spectrum programs but they slowly scan across the selected window at about 1sec per 2.8Mhz chunk.. how about sending me a Keysight DSO1204X so i can do some tests against my SDRs! :D

[@firedeveloper]

Clean the lab :D

[@EEVblog]

Yes mum.

[@trickyrat483]

@@EEVblog Mum? Nah, worse than that - it's the wife! :)

[@Milwaukeelors]

Marketing bull shit ..your doing it too ...no one throws away that many scopes in the rubbish..how much did they pay you to talk about their scopes..Rhode and schwarts rule above all other scopes.