Breadboards - Trash or Treasure? | S-Parameter Bandwidth Test + Wave Winners
HTML-код
- Опубликовано: 5 июл 2024
- Is the frequency limit of a solderless breadboard really 10 MHz? The VNA will tell us!
Sign up now for the Wave giveaway: bit.ly/YTWave2020
Click to subscribe! ► bit.ly/KLabs_sub ◄
Wave 2020 overview: bit.ly/Wave2020Blog
Helpful Links:
How fast of a signal can a breadboard handle? The answer is not super straightforward, but a simple S21 S-parameter test with the vector network analyzer will give us a lot of insights!
We test a single row, a single row with jumpers, two rows jumper wired together, and multiple rows chained together with a TON of jumpers.
Keysight Bench Facebook page:
/ keysightbench
Keysight RF Facebook page:
/ keysightrf
EEs Talk Tech Electrical Engineering podcast:
www.eestalktech.com
/ keysightpodcasts
Check out our blog:
bit.ly/KeysTechBlogs
Twitter: @DanielBogdanoff:
/ danielbogdanoff
Is this breadboard as good as garbage or can you actually do real prototyping with one? In a moment we’re going to break out the good ‘ol network analyzer and find out.
Conventional internet wisdom tells us that you shouldn’t use a breadboard for signals over 10 MHz. Is that true? Where did that frequency limit come from? I couldn’t find any answers I liked, so today we’re going to actually quantify the performance of a breadboard.
Breadboards get a bad rap, but they are fantastically helpful for hands-on work - especially if you’re playing around with a new component or prototyping a simple build. If you’re like me, though, I’m always worried the schematic won’t transfer well to an actual PCB.
Let’s see if 10 MHz is actually the maximum a breadboard can handle.
We’ll use our FieldFox, which goes up to 28 GHz - so it’s safe to say we’re over-equipped for this measurement.
To connect it to the breadboard, we need cabling and fixtures. This fixture is a bit sketchy, but breadboard work is suspect anyways so it should be good enough. And now you have a super geeky pair of earrings.
I’ll then put the FieldFox into Network Analyzer mode and set it up for an S21 measurement.
For an S21 measurement, we’re going to output a frequency sweep at a known power into port 1, run the signal through the device, and see what we get at port 2.
It’s called an S21 measurement because it tells us what happens on port 2 as a result of what happens on port 1.
Before you can make any decent measurement with a network analyzer, you have to do a calibration. We’re going to do a quick and dirty cal, but VNA cal is a whole deep dive topic that we don’t have time for today.
So, here’s what it looks like before calibration, and here’s what it looks after we cal with a simple coax cable. When we add in our fixture we see that it looks different again. this is the effect of these fixtures on the measurement.
This highlights the importance of both calibration and using good fixtures and cabling. Because we only want to see the breadboard’s parameters, let’s run another cal.
Now it’s finally time to measure the breadboard - we’re going to start with just one row and nothing else connected.
Anyways, the general consensus is that solderless breadboards work ok for medium-speed, medium-impedance-level, and medium-precision circuits. We only looked at bandwidth today, but there are a ton of other factors that come into play here beyond just the bandwidth. For example, how’s the noise? How close can you actually get your bypass capacitors, what about EMI and crosstalk? How repeatable is it?
I also think just about anyone who’s ever used electronics has a horror story from using these things, I’d love to hear yours in the comments!
#Breadboard #BreadboardFrequency #BreadboardElectronics #BreadboardTutorial #RFEngineering #VNA #VectorNetworkAnalyzer #SParameters #ScatteringParamters #S21Measurement #RFElectronics #Wave #KeysightWave
#electronics #electricalengineering 
Наука
So I liked this video a lot - very cool to see a Fieldfox actually quantify the breadboard's losses like that. But I disagree with your conclusion saying that breadboards are fine for medium speed and medium power circuits based on the fact that S21 is extremely low at 10MHz. The problem with breadboarding isn't always signal attenuation. For example try prototyping a 500kHz, 2A SEPIC converter on a breadboard. That is low frequency, medium power. There would be almost no attentuation in magnitude of voltage or current. And yet, the performance will be abysmal and the feedback loop may not even operate correctly. The problem isn't attenuation, it's the added inductances and capacitances which would lead to phase shifts, unwanted ringing, and sometimes full blown resonance at certain nodes in the circuit. Even a slight amount of ringing on a FET's gate can lead to chaos. I am sure you know this already but I think beginners watching the video may get the wrong idea of what a breadboard can handle.
This is an excellent point! I agree 100%
0:30 Bonus points for that К155ИД1 just casually sitting there
After literally melting my breadboard I can quite safely suggest not to build a ghetto (not microcontrolled) BLDC power driver circuit if you are not replicating a known schematic. Very fun, though!
I used a breadboard to prototype a slayer exciter and a blown transistor melted some holes in mine...
Did you say... a 6 FIGA HERTZ at 6:24?? :D Welcome to the club!
dont think so...
Man, I flub soooooo many words it's not even funny
maybe he said Fiescal 😂
maybe he said Fiescal 😂
@@KeysightLabs Dude, you do great. I even watch some of the ones I don't really care about, and I also usually hate everyone.
When they make a breadboard that takes BGAs, sign me up.
Seriously, I'd be interested in seeing S21 measurements between *unconnected* adjacent columns - i.e. crosstalk. Plug in your fixture's G-S-S-G on 4 adjacent columns then use a very short, low inductance jumper to tie the grounds together. I suspect you'll get something around -20 dB, maybe even closer to -15.
I remember back when I was in my teens, doing a simple LM386 amplifier circuit and wondering why my output level was so low. I could distinctly hear my music, it was just quiet. Then I realized I had the headphone jack wire one column away from my amplifier's output with no DC connection - just coupling through the parasitic C between columns! Gave me a new appreciation for parasitics.
Also maybe use a nicer fixture like www.zebax.com/doc/ZXACC/ZX00SMAH.pdf? Seeing a SMA going into a breadboard would be pretty funny, but might actually give more accurate measurements too.
I like those ideas a LOT.
@@KeysightLabs Then maybe you could follow up on that by disassembling a breadboard, measuring dimensions of the contact strips, and modeling them in ADS and seeing how close you can get the field solver model to match the measured S2P. I might actually try to do a 2D thick-metal model in Sonnet later on and see how that looks.
Maybe also try doing some eye patterns of common low/mid speed protocols, like SPI from 100 Kbps out to tens or hundreds of Mbps?
You can get yourself a solderless BGA breakout board. Something like this :
e2e.ti.com/support/microcontrollers/hercules/f/312/t/462170
I mostly make testbilds just soldering the components together "dead bug style", this is in Danish called "fuglerede" (translates to "Birds Nest")
@MichaelKingsfordGray "I used that method in the 1960s.
Nothing new."
Yeah, you tell him who's the boss man, keep up it up!
3:50 I think this must be a bad contact (i.e. capacitance in series). -50dB at 580kHz seems impossible.
Yeah, that and breadboards have bad contacts like swiss cheese has holes.
@@jjoonathan7178 It depends on the brand. One of Ben Eater's videos talks about breadboards
It’s like a Mythbusters episode. Great info!
I came here from Wave giveaway site, but damn, I'll stay because of the content! Love the lack of formality and practical projects mixed with a bit of madness!
Congratulations to the winners! And to Daniel for your new fashion line, I think Olivia Palermo finally has some true competition.
Wave sign up link: bit.ly/YTWave2020
Ok ok, I know. There are a LOT of variables when it comes to breadboards and I'm sure many of you are here to comment with other tests I should run. AWESOME! I'd love to come back to this topic and do a deeper dive with it and have a longer video later this year. Put your thoughts and worst breadboard horror story below!
**** [Add comment pointing to main CTA and pin it]: check!
My main gripe with breadboards are the bad connection problems. I usually spin a small PCB to prototype or solder up the circuit on perfboard.
Lookin good! Grats to the winners! Dig the earrings.
Fashion forward! Test Equipment, snging, ditties, fashion.. this channel is going places!
Hilarious and educational, keep up the good work.
I use breadboards a lot! I am a computer science student with a high intrest in electronics (RF and EMC in particular). I use bread boards a lot for prototyping microcontrollers, like a ATmega328P, to test te circuit before making it on a perf board. Really handy to modify low frequency circuits!
I'm so happy that ElectroBoom introduced me to Keysight's channel :-D
Love the video!
nice earnings Daniel.
i use breadboard for every prototype or tinkering and things get a lot messy.
1:15 after double 'E' s talk tech, keysight now proudly presents double 'E' s talk techo
LOL!
Great video, thanks :)
Great breadboard tips. Congrats to the winners, again!
I haven't used a solderless Breadboard in many years, as there aren't enough circuit nodes to be useful.
A technique I do still use is the Veroboard with the square pads, and Vero "Speedwire" which uses enamelled wire point-to-point soldered connections, routed with plastic combs.
This will give a surprisingly high board density and performance, approaching that of a basic PCB. Perfectly fine for typical microcontroller speeds.
With adapter-boards that breakout SMT-chip pinouts to a 0.1" pitch, prototyping with square-pad Veroboard is still a viable technique.
This is really interesting. Even being just out of college I was suspect of that 10MHz limit. I had built a 100MHz FM transmitter with a breadboard that was actually able to pretty clearly transmit.
Keysight better put the earrings in the catalog.
Thanks again for the informative and entertaining videos.
Hi, You can get these perma proto boards that allow you to transfer your bread board circuit layout exactly like you built it, they are available at adafruit in various bread board sizes. I don't usually have a problem with bread boards and they are good for an experimental circuit you want to try, but, maybe not make permanent on perf or pcb. Best, C.
Tier 1 prize goes to: Chris "Hack"
Yaghiyah Brenner My heart skipped a beat for a second 😅
@@HoppyStar Congrats mate, hope you have fun with all that gear
Thanks guys, I will...
Enjoyed your video. I found it after doing a search on VNA Breadboards and was surprised when yours came up.
I have been working on software for the low cost NanoVNA to allow measuring power rails impedance. The breadboard makes it easy to build up different decoupling networks to test the software. Eventually, I will move onto testing with a PCB. The NanoVNA is nice as it only costs about $50 and can provide some decent data for the RF hobbyist.
I've seen various numbers tossed around for breadboards speeds. For fun someone on EEVBlog started a contest where the idea was to construct an oscillator using a breadboard. As you can imagine, the rules had to be clarified as the speeds went up. The definition of wires vs components became very important. In the end, I was limited by my test equipment.
Are you still planning a part II?
Those are some good looking earrings! Looking stylish Daniel!
I like to use them but I never consider to use them with RF signals.Congratulations to the winners!
Would the maximum frequency be related to the mstl characteristics of the parallel rows?
I like retro table in the background, 8-bit Atari was my first computer
I once assembled simple LED clock prototype on the breadboard using 8bit PIC at 4MHz.
The time deviation per day was huge -> more than 30 minutes.
So when I transferred the design on PCB I had to remove all compensations I made in the code.
Up vote for earring demo with EDM music. Awesome.
That pair of electro earrings scene it's kinda cool.
Prizes come second, KNOWLEDGE IS IMPORTANT! (Of course, you still need some equipment to apply them, but it comes second)
WOO HOOO!!! I am so excited to be the tier 1 winner! I've already replied to the email with the documents... Just let me know what's next! I'm really looking forward to using that DSOX6004A oscilloscope...
Congrats!
@@KeysightLabs Thanks! I'm really just blown away by this. Been entering since 2018 and honestly didn't think I'd have a chance...
A really interesting topic in the upcoming video. There is a project for a device that i take the designed of it from the Japanese company Hioki and when applying the design to the bread board, I found a difference with the frequencies of some sections of the oscillator circuit.
So, in some cases, it is not possible to use the bread board, because it can create parasitic capacities in the tracks, and also the quality of the bread board can affect the results. In the circuit that you test it
So in the end tkanks for posting god bless
Sure do need a REAL Oscilloscope in my life.
I used the breadboards up to 8MHz and never had a problem (by choosing the right topology). In my opinion, for high frequency circuits, the conductors should be as short as possible. It's always good to know the limits of applying something. It was interesting. Thanks.
I use my breadboards for experimenting but I also use one for soldering pin headers nicely onto my Arduino Nano's for example that way I make sure that the pins are all in line and standing straight up so I know that my Arduino will always fit perfectly into all my breadboards, but when you do that you should dedicate an old breadboard to your pin header soldering because after a while you'll see the plastic starts to deform and or melt in some places, you're definitely gonna destroy it over time but it's a handy way for soldering pin headers in a consistent way that will fit onto every other breadboard... well, most of them.
That's a good idea
I'm just a disabled putz that loves to play with electronics to keep my brain off of the pain. I use breadboards a lot. I can't afford to make anything permanent with PCBs but it's fun to build things. Thx for the vid. 👍👍
i use breadboards and they mostly work fine but i tried a precise chronometer and it delayed the time by 1 ms every 23 seconds but when i used a regular pcb it didn't have the same issue
Yeah!! That's really a usefull analysis, I've always wanted to do this measurement at my place. But with this equipment it is much much easier and quicker! :)
The inevitable bad contact, either due to bad cable, or bad point in the breadboard, has always been the main problem in my breadboard builds. Quality varies wildly from (eBay) sellers, and really good ones are hard to find.
Breadboards are very useful!
What you have not tested is the capacitive crosstalk between neighboring rows - in a student lab in the second year we build a CE amplifier on these breadboards with a 2N3904 which most of the time ends up with base and collector in neighboring rows. This creates a parasitic capacitance in the order of 10 pF between base and collector which - due to the Miller effect - reduces the achievable bandwith of the amplifier to below 1 MHz as compared to a simulated bandwith of 30-50 MHz. Luckily, since our students would not be able to measure such a high bandwith with our lab equipment!
Someone recommended an S21 on a neighboring row, I'll have to try it when I get the FieldFox back
I like to use them but I never consider to use them with RF signals. Hope to be lucky next time.
Good Video
3:55 that low cut filter is kind of weird, that should not happen on a connection, what about DC? So apparently you have a loose contact in somewhere in there...
Maybe you should have started with some of that old school Radio Shack solid Black/White wire spool which was used for inductance coils for a receiving antenna.
When I was in high school in the 1970s we were taught to use brass escutcheon pins nailed into a piece of wood for soldered bread boarding.
Bread Boards have they're place to be sure. Most RF work is not one of them. Just plain bad Ju Ju. Better to go DEAD BUG build. Or even Point to point.
the inter capacitance between rows are ~2pf.
Breadboards are wonderful to use cause I'm always constantly changing things. But go into switch mode power supplies and logic circuits and you'll have problems, probably, depending on what you're using. In all cases, always look up that data sheet! Also try EMI breadboard testing sometime if it hasn't been done yet. Nice vid overall. Didn't know you could do up to 100+ megacycles with that mess lol.
I was making a prototype of a simple radio signal amplivier on a breadboard and it worked horible. I coldn't figuire out what was wrong.😅 Then i just soldered it all together on a pcb and it worked all right.👌
I dont know everything you explained... but why I am so interesting with this video and this channel ? -_-
I think i must learn something in here .. maybe..
I use them for low frequancies like sound waves app.
At 3:23 the 1:1 patched ports don't look like they would HP filter the signal .... at least DC & LF signals should come through if directly patched.
Maybe one jumper wire was broken / has had a bad contact and hence the coupling was only capacitive (hence the HP behaviour) ... @Keysight Labs : What do you think are the reasons here?
I'm not sure that I could live without Pomona mini-hooks! However, I know that I can live without them hanging from my ears. :-) Nice video.
@Keysight Labs I do hope that the solder work in your professional gear is better than we can see at 0:58...
I didn't realize when I asked that this was a giveaway video haha
Breadboards are great for quick and dirty stuff, iterating and hotswapping parts, but they'll drive you crazy if you try and do high precision anything. If things start acting funny when you wiggle random stuff slightly - solder it and save your sanity.
I like to use as few jumpers as possible, it makes for a fun layout challenge of it's own.
If you need to add jumpers to increase available fanout on more than one or two nodes you need to move on to something better, even perfboard gives you much more routing options.
Now I can blame the breadboard for my fails!
@Keysight Labs You know those bandwidth and response measurements you just showed, about them......
....still can't use them for my custom autonomous UAV antennas without a FieldFox or other VNA!!!!
NanoVNA v2?
Build it on a breadboard, does it work?
#1 No, you've probably screwed up the interconnects or your design is wrong.
#2 It's an RF circuit and you ought to know better than to risk a breadboard
#3 Breadboards are so 1980s but *very useful* providing you understand their limitations.
Thanks for an excellent video dude.
This is a good summary! Especially #1
I didn't enter the giveaway but I did love the fireworks at the end.
I think the 10 MHz are still a very good rule of thumb - on the safe side even for most transients and crappy boards. if you have anything higher frequency you can not really get any useful numbers for a final solid design from it - if you need some extra filtering or smoothing you need to measure that on the actual layout as the breadboard will have too much of an impact there.
(But i have seen them used together with a Cyclone III to hock up some extra stuff to the DDR )
Once I assembled a circuit on a breadboard and it did not work whatever I tried. I even contacted a man from Japan who designed this circuit, he did not help either. Then I decided that circuit has some flaws and it is an opportunity to design my own version of it. So I spent a lot of time reading about the topic and finally built my own version of the same device (floppy emulator for Apple II computer). And when this project come to an end I decided to give another chance to breadboard circuit. I checked every connection and found that one leg of microcontroller had no contact with adjacent wire in the same row. So, sometimes a bad breadboard can give you a chance to learn and build something by yourself.
1:09 Perfect time for someone to walk in on you watching this. Hah
I love breadboard prototyping because whenever I prototype on veroboard, it doesn't work out. I don't know why maybe additional exposed copper pads? Any advice anyone?
Most of the times, probably 95%, it's wrong connections. Even when you are sure everything is exactly as it was on the breadboard, something(s) is wrong.
The remaining 5% can be anything... From flux residue on high impedance traces, to crosstalk,...
@@anonimo1633 thanks for the info.
How can we be sure that our names are in the name picker for this giveaway?
One of my classmates, accidentally connected full AC voltage to our pic microcontroller and some LEDs on a breadboard. one capacitor made a loud bang.
"Accidentally". 🤔
"Life is like a box of breadboards" ... of course it is... it's a better analogy than a box of chocolates... but what about a box of Chocolate Breadboards!? Hmm... To answer your question: I use breadboards at home for prototyping stuff, mostly Arduino or Rasberry Pi based. At work the hardware engineers cook up a PCB or make a LPKF prototype, rarely do we use a breadboards. Good to know the limits -Thanks a bunch for posting this video.
I play around testing new ideas with them but nothing serious. Love teaching people electronics with simple led circuits on them. Gets them hooked quickly on electronics hobby
Maybe thats the reason why my measurements fails. Thanks for the video!
I'm using breadboards, prototype PCB's with dots and homemade PCB's.. All depends on what i need to do. For example, when i need some heavy load, etc, i may use combination of breadboard and prototype PCB, in that case, transistors and heavy part are soldered on prototype PCB, but low level logic are on breadboard. In some cases i create projects in parts - some module at project starting are assembled on breadboard, after testing, PCB are created and soldered and parts from breadboard are disassembled to make place for next project module.. After that PCB's finalized part are connected to breadboard for next module testing..
I don't use cheap breadboards and have lots of good quality board's.. This eliminate debugging for bad connection reason..
Anyway, breadboards have current limits as it is with any type of connector's and, usually, i take it in account..
By experience - breadboards can handle 20MHz without any problem.. Long time ago, i tested FM stereo modulator and mini bare bones (with TO92 case transistors) FM transmitter on breadboard - at the end it worked, but i need to put some other capacitors, not as planned, because with calculated values i can't get transmitter in the needed frequency range.. This situation was awful for debugging.. :D
One time when i assembled new project, i accidentally shorted power lines on breadboard with jumper wires. Luckily, i left fuse in 12V power wire from previous project.. I understand my mistake when i see how fuse blows up like small bulb.. :D
So what you are saying is that instead of buying expensive RF filters in my circuits, I should replace them with tuned pieces of bread board and wires? ^^'
Do it. And report back
Btw. How do Keysight make prototypes?
Depends. We have lots of great simulation tools, and do a few board revs as needed. Exclusively SMD, though.
We learned a lot about Daniel on this one! LOL
Congratulations for the winners. I hope I can win a oscilloscope this time, to Sri Lanka :)
One more week to go, I'm still hoping for that scope :)
К115ИД1 wow, vintage tech :D
Just about everything I testbuild with is surface-mount so I'll typically just run jumpers from some SMD breakout boards.
i love them
I love breadboards. When my circuit doesn't work properly, it's always a mistake I've made. However, I can't stand jumper wires. They are usually small gauge wire and have horrible pin/socket ends.
I just realized this was 2020 keysight wave event and video -_-.
I think the 10MHz "rule of thumb" comes more from the issues due to cross-talk, unwanted coupling due to the parasitic capacitance and poorly defined impedances. Also breadboards are notorious for poor contacts which could give you a lot of bizarre problems at higher frequencies (DC circuit is broken but RF still gets through due to capacitive coupling and such).
I am building HAM radio receivers (1-30MHz) and I can't imagine getting one perform satisfactorily on a breadboard - it would be super "touchy", that would drive me crazy. But I am not worried too much about running digital circuits with microcontrollers with fast edges on a breadboard - those aren't that sensitive for it to cause problems in most common cases.
DC circuit is broken but RF gets through? Where have I heard this? Oh right 3:45 a strange low cut...
Need an oscilloscope, never could afford one.. just a bitty tiny entry one 😍👍🚀
Wow
My worst story with these is that one USB power supply was so uncapable that the internal resistance and connection resistances were enough to make the ESP32 brown out
What I want to know is who has ever seen the MB102 with "gold plating" as claimed on every $2 breadboard box from China? I think the terminal plating quality is the primary factor in breadboards getting a reputation for bad/unreliable connections. Some are plated too thick and it makes them difficult to insert pins, while others are too thin and corrode easily. Some are shipped pre-corroded for good measure.
I had a couple of projects where I cut up breadboards to make my own connectors. Even the ones that look good on the bottom can have corrosion issues in the pin contact region. You really can't tell unless you pull all of the terminals and look. I just wish someone sold a gold plated contact version at a fair price that is feasible for retro computer shenanigans @ 8-12 breadboards. I'm not forking over $30-$50 each in that kind of quantity. I'll just etch a board from scratch before I'll pay $500 for a temp breadboard setup.
I use solderless breadboards all the time at 20Mhz, and sometimes at 80Mhz, and I've never had any horror stories. Love it.
You need to put a LOAD or a C.I.; only coneccting cables with out LOAD is tricky.
Just don't expect the component values you use to get things working on a breadboard to work if you go to dead-bug or protoboard.
The graphic pop at 7:01 is wrong as it said Tier 2 prize and not Tier 1.
Shucks, good catch! We'll fix it for next Friday.
My horrible story about breadboards is at the college where the teacher had no idea about breadboards (row and column continuity). We have done many experiments after experiments using op-amps, resistors, 555 timers but they failed. I went home and research about breadboard and I was triggered!
I mean how a teacher of an electronics degree don't know about breadboards 🤨
I don't mind that much about that situation because "everything is useful in moderation, a good servant or a bad master"
I use it for every prototype
EARRINGS 🙌👏
Rule of thumb for breadboards? Avoid 830 tie point MB-102 breadboards. Pretty much every Taiwanese or American ones are good to go. KH, EIC, WBU are some of the good series that are reliable.
My experience so far is up to 20mhz they still work fine
Life IS like a box of breadboards :)