Seriously props to the fact that this video was that long and you literally kept me interested the entire time! (LITERALLY ABOUT BUTTONS....though I love tech and wiring things could be part of it :P )
Something to remember when using power rated switches for logic level control. Power rated contacts ( the round ones ) rely on an arc during switching to keep the contacts clean from oxidation as well as cleaning off general " goop " left over from from assembly. Logic level contacts use gold plated knife edge contacts at 90* to keep the contact area small. I've had this bounce issue with hydraulic pressure switches linked to a CNC controller. Pulling the switch out of circuit then using a resistor and power supply then cycling the switch to create arcs would clean the contacts and bring it back into operation. As for the old time wall switch, the spring is more for a quick make / break in order to minimize arcing and eventual failure of the switch. Another issue I've run across when using twisted pair cabling. Don't use both wires in a pair for different actions. The twist is basically a linear transformer and can lead to spikes / ghost operation especially with AC .
With how extreme and fast-paced mr beast videos have become I feel like there's so many stories that happen and get documented but no one ever gets to see
@@JoeGallantChurchComms With all the 3rd party fallout videos I watch it feels like Mr. Beast and his core staff are using their power to ask way too much of people in too short of a time, so that they can save money on a project that would cost any other entity 2 - 3 times as long, and cost 2 - 3 times as much.
The reason is because since most people use TikTok, their attention span is usually very short which is the reason why Mrbeast tries to make every second of the video entertaining
During the days I spent in that cube I was immensely impressed and overwhelmed by the amount of work and attention to detail from everyone. Thank you for helping to make this an unforgettable experience. Sincerely, #41
@@goosenotmaverick1156 People of the ages 1 to 100 were staying and sleeping in those 100 boxes and competed for a prize of $500.000. They did challenges and used these buttons to vote for challenges or vote each other out. In the end after over a week in there only age 52 and age 40 were left. The video has 230 million views.
@@goosenotmaverick1156 it's a good video if you watch it, but the tl;dw he put people from every age from 1 to 100 in small little boxes and whoever stayed in their box the longest won a lot of money. There were also challenges that the people could vote to participate in which is what the buttons and LEDs were for.
@@RobinClower I ended up watching it when I couldn't sleep late last night. I'll admit, it was entertaining. Can we all just collectively hate 54? That guy was an ass lol
@@ElusiveXSetups but you see... content... it's not as interesting I there isn't a time pressure, it kinda just becomes a educational video at that point
Wow... MrBeast's deadlines are insane though, they call upon creators with 1-2 weeks notice to complete these crazy assignments. Too many hiccups with the other workers messing up the wires and stuff. I hope they compensated you well. That's super stressful.
Thing here is to respect yourself and don't take crap projects like that unless you have a solution at hand. Coz mr b surely doesn't respect anyone who he gives this assignment to. Want to bend over backward and crash and burn - go ahead.
They didn’t compensate Will Ossmann properly. At least I remember it like that. I don‘t get how they do it this last minute. I doubt they has the idea and started building the boxes in the same time frame.
Yeah, I guess UA-camr labor is counted differently when you can put 'Mr. Beast' in your title... but still. That's an insult to the creator I think if they weren't compensated well up front. @@DailyShit.
This is my experience with anything hardware. Coming from a programming world, it's really a shock just how much can and will go wrong with electronics.
but same like software. You can minimize the potential of problems by following best practice on every step. But no one actually do that because of limited time 😆
Nothing goes wrong with electronics, if you know what you're doing! I'm an embedded hardware/firmware/software dev. My software-engineering colleagues always claim that the hardware is at fault, the hardware-engineering colleagues always claim the software is at fault. Whilst everyone just lacks the right knowledge to really debug simple issues outside their field of expertise...
As a senior network & system engineer myself I can personally say you did a fantastic job with the mess of constraints! I watched the Mr. beast video a while ago but seeing this now is just awesome
Hot take: When using potatoes for CPUs, take lots of breaks and don't expect synchronicity on the first trial run. Great use of dad by the way. He brings the "ol' school" element.
@@retsamyar I am exaggerating a little. There is obviously a lot of extra report writing, diagramming, etc a class has to do too especially in the first semester. Sometimes make custom PCBs as well. Additionally, most of the time students will be taking other classes alongside being in senior design class.
i swear this and william osmans video are probably 2 of my most rewatched videos and funnily enough they are both about the nightmare of working on a mrbeast video lol
I have worked as intern with company that did stuff like this, and had opportunity to work on sets like this. I can fully understand the stress and hotfixing everything last minute, even watching this got my stress level up like flashbacks, great video!
As a programmer, that’s an absolutely epic collab! You, Network Chuck, and MrBeast?! I mean, I never thought I’d see you together, but this is just epic.
@@SunnyCress hacksmith. They accidentally broke a safe that was meant to be cut open with a lightsaber just before the video was supposed to be recorded
@@YourLocalWarGorilla they accidentally tugged on a carpet under the safe a bit too hard when sliding around an ice rink, and the gallium part of the safe fell onto the ice, causing it to break.
The challenge was a really good MB video. Who would have thought a highly technical behind the scenes video from JG would be so well viewed as well. Congrats man. You and the "button and lights" team made the experience great.
There's a bit of irony around the interrupt problem. For debouncing mechanical contacts polling is almost always the better solution. I would have recommended running a separate task that simply polled the switches on a regular interval (100Hz or so) and then fired off events for your main code to catch.
Nice that you were able to make it work! I'd have done it completely differently though, using the sorts of techniques you see in industrial automation that ensure high reliability. No need for a whole programmable computer in each cell; each would've had only an industrial modbus remote IO expansion module, mounted onto a strip of DIN rail, together with the power supply, relay modules, and a terminal block for every external connection that needs to be made after the pre-assembled rail is screwed into place. All wiring within the module would've been 18AWG silicone jacketed wire terminated with crimped ferrules, and both sides of the cable to the button box would've been keystone jacks with only factory-terminated CAT cabling between them.
This continues to be my absolute favorite tech channel! I’m personally not a fan of Mr beast videos but seeing the behind the scenes for a shoot like this makes me have a profound amount of respect for all the people involved in making these productions happen. The absolute pure passion and dedication from you and the whole crew involved is infectious and exactly why I keep coming back! Incredible work!
You guys did phenomenal work, definitely appreciating more of the bts of that video now that I've seen the stressful nights and the amount of manpower that went into making everything come together. Mad respect.
My Dad got me interested in electronics and IT very early on, I've always loved working on a project with him, especially puzzling over the best (and the most practical!) way of doing something.
My dad went in the oppisite direction as me. He is a farrier(Horseshoer). So our interests did not intersect often. Thats awesome your dad gets to help with these projects. Thanks for sharing. @@JeffGeerling
I've been doing events for years, and there's nothing as satisfying as working until 4am building a hack knowing patrons are in site in 4 hours, or listening to talent pad a live event while you're on stage under a table unplugging your carefully managed cables and running a hail mary one. Great stuff Jeff
Hehe I haven't done too much live production work but you can tell a good emcee when they can seamlessly delay for you and the audience is none the wiser!
jeff im glad to see you guys are thinking deeply about the fire hazard associated with this project, but i cant help but feel like such aggressive timelines on mr beast videos is going to result in disaster at some point. especially considering how big and potentially dangerous their videos can be. clearly yall did a fine job, but under this kind of pressure and crunch: the odds of mistakes go up very quickly. a bit concerning knowing they occasionally do videos with explosives, but even then its likely the danger in the details of things you wouldn't expect. just a thought
Seems like something that could really be helped by queuing up more projects and running them on rotation. That way each event has had double, triple or quadruple the time for work but he doesn't have to slow his schedule if he doesn't want to.
@@LanceThumpingthe problem with that would be having to pay people double, triple, or quadrouple the pay/having to hire double, triple or quadrouple amount of crew workers. That's not cheap when you have so big production team already. But yeah agreed on the crunch. Someone needs to put this problem into spotlight. I can only imagine how big of a tragedy could happen in a mr beast video with all of those people cluncked up together hastly made electronics and propably no fire extinguishers in the rooms
@@LanceThumping They already work on multiple projects at the same time. Not sure why such the short time frame, maybe they were using the warehouse for another project before this, maybe just to save money.
@@taisato2091 The crunch cost them a lot of money in components and rework labor. Their component choice was extremely limited by lead-times-they had to pay for next day shipping, change their specs to match hardware, and probably lost a lot of 100+ part bundle deals. Not to mention the amount of re-work labor or replacing ruined components due to ESD and conflicting construction schedules and OT. This also would increase the technological quality and safety with more R&D time. PLCs, shielded cabling, metal enclosures for ESD protection, better cables etc.
As soon as you ACTUALLY showed the blue walls, I immediately said "oh no" out loud. Immediately I was like "look, I know ElectroBOOM did that video with Linus, but that's a LOT of static." Brutal! But hey, what a challenge!
Great Video Jeff...it was exciting from beginning to end. Glad your medical issues are better and you were able to get everything working. Im not a follower of Mr. BEAST but I may watch that. I appreciate all you do, love watching your videos and learn a lot from them. Thanks!
this gave me a lot of chuckles and reminded me of many problems i have had debugging electronics on a much smaller scale so i can imagine compounding the issues at the level you were doing. bravo sir.
Being that I have worked in the Production/Events business for 40 years, I feel every second of your pain. Including the mild emotional trauma of leaving it all behind. Congratulations!!
I've honestly found the little jumper wires quite unreliable even for my small projects. They break extremely easily. Also, I would 100% watch the 12 hrs of footage of troubleshooting, building, etc. LOL
De-bouncing is necessary for just about every digital circuit with a button as an input, dealing with it is one of the first things you learn when learning to design interactive systems using FPGAs.
For me it was a 8 bit 8080 development board from the 80s that only had a hexadecimal keyboard for programming and no way to store your program, with a class final test to simulate a traffic light, using button instead of vehicle detection, all in a 150 byte, the debounce subroutine was a pain with memory limitations.
@@AstroStrongBoxPutting a pulldown resistor is better tho, it reduces bouncing well (when working with inverted logic *5V -Off, 0V -High*) and it's easy to integrate.
@@sebastienmonette6659 totally agree but hardware is hard to fix in scale and last second. Software will get you 90% there especially if absolute speed is not super important. If all buttons have the same software latency it’s still fair ;)
Between this and the Osman Squid video/stories, I feel like an infinite amount of stress and problems would go away if people were given an extra pre-week (before the 2week crunch) to prototype and test a tech solution at a peaceful schedule before it had to be done at scale (one box/room or one squib etc)
as someone that doesnt watch mr beast, this is my favorite mr beast video. also that button bounce issue is really interesting, i had never heard of that.
Amazing video. I was able to predict the cause of every issue and it shows how much engineering goes into even the smallest projects. It really helps to pay a little extra and not cut corners especially when the client can move goal posts at anytime. I was really impressed you stuck to dupont to the very end and didn’t use any project boxes.
Best episode ever Jeff. If anything, this should be worth a white paper on how diverse a situation, small form computers can be used, and the fact it was done in a simultaneous prototype/production scenario is pretty dang amazing and telling to the capability. The whole experience is impressive, /hats off my friend
People see the end result to massive events like this but have no idea what really goes into making them. This video is amazing in making people more appreciative to the amount of work it take to create something like this.
I worked on a project exactly like this. Watching you learn everything I did in this video was comforting and anxiety-inducing. I was building a large interactive arcade setup about the size of a football field. There were 22 intel NUCs, a bunch of rpi3, and over 100 esp32 custom boards with neopixels. our electronics team at the time was somewhere between 3-5 people. We had every exact issue you had, plus some. Was a truly wild experience.
Why was my comment deleted? Anyways, some suggestions for next time: - an operation like Mr Beast's will have an experienced light tech. You should have focused on what matters: the buttons. Let the light technician deal with the strips - send an ArtNet signal to his console and you'd be good - he would have mapped this stuff out in a few minutes with his console. - Don't buy cheap indoor LED strip, instead rent out some high quality outdoor DMX stuff especially for a one-off project. All that stuff will now become garbage or just pile up dust in a warehouse somewhere - Use proper Molex connectors and put all your electronics in the box. Less wire length = less susceptible to static charges - Should have ran a single ethernet cable + power from each box. Why buy cheap unshielded cable when again, event venues rent this stuff out all the time. Anyways I thought it was an entertaining video and I hope you got some experience out of it. The event business is a whole other world ;) Trust me, I've been there!
Your first bullet point was the exact thing a lighting tech pointed out on set :) Unfortunately they were only consulted for the overhead lighting, not in-room. Would've been great to have lighting be separate, and I could just manage it over the network. (They had a massive lighting console set up entirely independent.)
Cable relief is about #1 on the build list. Only fighting with ground/screening. It's common to use current loops for long unprotected cables because it's easy to inject a voltage spike but very much harder to inject a significant current. And a current loop has a quite low-impedive listening side. And it's easy to clamp any peak voltage. Good job that you got it working. But seems this was a task where you need some communications R&D engineer in your phone book.
Heh, that was my Dad, and he got involved *after* we had gotten the initial order in. Also was not able to be physically present once we started laying things in the floor (because of the slight delay with the rooms)... he would've probably helped us get through a few of those hurdles more quickly! As an RF engineer, he's seen it all (or, well, most things EMI/RF/interference!).
@@JeffGeerling One reason whe love twisted pair Ethernet is the 4 kV isolation in the magnetics of the network connector. A big reason why few people care about shielded network cables. Twisted pair reduces amount of bit errors in the transfer and the little transformer in the connector protects the electronics from big spikes. Then low impedance signals for the shorter cables. Many customers talks about "we can do this with a Raspberry Pi" about the hardware I work with. When ignoring voltage range, temperature range, EMI filters, clamping diodes etc of equipment for industrial or automotive use. 100% fault proof is hard. But a few tricks can take it quite far, allowing even hobbyists to manage quite robust designs.
EZ de-bounce is to simply add an RC low pass filter to each switch. Often satisfactory results can be obtained with 0R, only needing the capacitor, this will filter out the high frequency transitions typical of bounces.
What specifically amazes me about this whole process was the short time crunch. I would've assumed it'd take a week minimum for some of those parts to arrive in bulk
Heh, we wound up calling some suppliers direct and getting some things shipped straight from the origin, if we couldn't find stock somewhere in the US!
Damm, that bounce is insane. I've seen this multiple times in industrial applications. I could have used industrial controllers like PLC or networked io's.
Wow, what a great video Jeff!!! Very exciting all the way through especially with so much debugging. I felt second hand stress watching this, was almost like a thriller movie haha. I definitely could’ve watched all of the remaining footage without getting bored, seems like it was a blast!!!
Next time you need to de-bounce in a hurry just solder a capacitor, something in the 100nf to 1uf range across the button, with high impedance CMOS inputs that often does the job. The alternate way is just to program a hard pause after the button state change. Your polling issue could have been fixed with small sleep in micro-seconds in the main loop, that will cause the C runtime to yield back to the kernel and stop the process from CPU hogging and running hot.
What an exciting project! Congratulations on such a unique opportunity and many lessons learned. This video taught me something new about buttons. Also, it's cool to see Le Potato in use in the wild - this may be their biggest project yet.
This was incredible to see, the amount of stress you guys were under must've been insane. Although looking back there are probably a lot of things you could've done different, given the time constraints you guys made a miracle happen, seeing how much work you guys put in really made me appreciate all that goes into a video. Can you convince Mr.Beast to call you guys up ahead of time next time he has a massive project like this?🤣
This video gave me flashbacks on one of installations I've been in where certain LED boards or NUCs or the wires and networking between them were simply not working because of the odds when dealing with around a hundred of them all at once. When looking at electronics individually, it's easy to assume that they'll be fine but running hundreds of them together and all their little moving parts and considering the people that'd be messing with them - like frigging static on acrylic, you're likely to encounter a dud or a broken setup very easily and it's frustrating even with spares.
This is the first vid I've seen from you, i love it no only the behind the senes of a beast vid but what it takes to making the idea come to life ,A master at work.great video
Man, you should have called for this. I was the engineer for a 500 person trivia tournament about 7 years ago and had to set up response boxes for 50 tables and the setup was very similar (50 button boxes, 4 buttons but no LED strips). Every time we ran into big issues it was due to treating the project as a hack instead of a product. I made as much of the system modular as I could so modules could be independently tested and swapped as needed. We bought 3-gang electrical boxes and matching blank covers in bulk from Lowes which meant we had plenty of spares and could get more in an emergency. We used a fly press and a die to knock out the holes and the boxes had built in knockouts that we used for connectors. Inside the button boxes we used back lit arcade cabinet buttons and made harnesses with crimped spade connectors (no soldering) to 2 rj45 that connected to female terminals on the box. Boom, now all boxes are modular, easily assembled and tested and swapped (I made a test rig to test 5 boxes at a time). Then we ran cat5e (or cat6/e this was in 2016 so I can't remember off the top of my head) from 3 button boxes to one of 17 hubs that split out the rj45s to jumpers that fed into a pi zeros IO pins. The pi zero would read the inputs from each box, set their light's status, do some error checking, and send/receive data over TCP/IP to a central pi that would then host a website where you could see people's selections, reset the button states and check for errors on the hubs/button boxes. Even though everything was thoroughly tested before the start of the tourney, we had multiple spares of every module. If there was a failure with a box, we just had to disconnect it from the rj45 cable and swap in a new one (and troubleshoot the faulty box). If a pi went down we could swap it with another that was pre-flashed. If a harness was faulty, replace it with a spare. All of the wiring outside of the boxes was standard ethernet that could be quickly replaced and the hubs used POE so we didn't even need to run power. I wish I knew where the whole thing ended up.
I was planning on spade connectors but I thought the boxes would be set up by another group... when we had to do it, there wasn't any place nearby with enough spade connectors, so soldering it was! The setup at home, I re-built it using spade connectors and it was about the same speed as soldering, but infinitely easier to just do a single button swap.
I LOVED this video. Behind the scenes is some of my favorite content to watch. I hope they bring you back for some more adventures, if not for your electronics expertise, then for your story telling prowess! 🙂
That'd be my dream job... actually, I'd be there for free! Not for the Video or MrBeast but for the experience you went through. A super tight schedule, barely any sleep, fixing new problems every 5 minutes, debugging software that had be written in record time, praying everything will work, improvisation, the team work ... great job! (also on the editing part)
Being a dev, this whole video reminds me of why I love doing what I do. Go-lives are the most stressful thing in the world, but when everything works out in the end, it really is a great feeling! Although it must’ve been a stressful few days, it must’ve been incredibly fun!
Back in the 90s just out of HS, my boss had a need to have buttons in 100 apartments that when pressed would notify the security office. I was just learning about microcontrollers and said I could build it. I set the buttons up in a 10x10 matrix connected to the PIC uC and polled them by putting one row high and checking for high in each column.. Imagine my surprise when it worked fine on the bench but when I put long wires on it, one button press would trigger multiple apartments. That is the day I learned about capacitance. The capacitance of the long wires would lower both the leading and falling edge slew rates causing the button detection to shift to the next button, or cause multiple buttons to be triggered. Had to redesign the hardware to use high current drivers and slow down the poll rate. I could only imagine the EMI that system put out in that building lol! It was in service almost 20 years!
This video actually means so much to me. I love everything systems administration, and the thrill of those high-stress jobs, even thougu I've never been on any as high stress as this. But this video really means a lot to me because I just watched it to destress. I've had to take my little brother in to the hospital, and I was with him until a family member was able to tag team with me. Jeff, watching your stress for something positive helped me alleviate my stress for what's been going on. Thank you.
Believe it or not, I’ve worked in several projects of this magnitude and similar functionality. Last time I went with 125 ESP32 connecting to the server via wifi. In this case I would have put the ESP32 and relays inside the button box, connected to the LED strips. If one of them malfunctions: hotswap the whole box with one of the 20 spares. But I’ve done this 4-5 times, the first time I had similar problems. Congrats on pulling this off! :)
I was thinking Pi0w via wifi would've been decent. But ESP32 would probably make even more sense. You could probably get away with a half dozen APs and a lot less wiring. There are certainly more than one way to skin this cat.
I'm starting to hate this rushed aspect of Mr. Beast's videos. Every time he does a colab with someone it's just to make them work so hard on a super complex problem with the tightest schedule ever
Just of the top of my head I can think of Donut Media with their jet-powered super car, William Osman with the squid game kill mechanism, and Hacksmith Industries with light sabers (which in the end wasn't even part of the video). I'm sure there's more
Considering you've made this video I assume its safe to talk about - I worked on this set too, in a way. I am a CNC Operator at a chstom shop here in NC; about an hour away from Mr. Beast. I've spent many a weekend working 14-16 hour extra days cutting props for Mr. Beast sets. This set was actually pretty grueling to get done. Had to load, cut, unload, clean, reload, 500+ sheets of acrylic on the CNC and it was a ton of overtime getting it done in time for their timeframe. Kinda fun to see the whole place all put together!
Heh, I can only imagine! When I saw the sheets stacked on all those pallets, I immediately wondered who got stuck routing those things out-that's a LOT of cutting! Also had to be pretty precise. Nice work!
@@JeffGeerling It was a lot of work, and also probably the most boring thing I've made for them hahaha. Coolest thing was a giant sign made of milled MDF, painted, and carefully hand poured full of epoxy for the different colors. That one was at least really interesting haha. And same! Thanks! I've since finished watching your video and definitely great work! I've been dying to get back into playing with microcontrollers and hobby electronics again and got a bit of inspiration from seeing a few of your vids haha.
I think YOU'RE the beast. Way to get stuff done. There's not that many people that could consistently solve so many strange issues over successive days of long hours. Hope you are getting enough water, rest, vitamins and sunlight to counteract what you did to your body.
I honestly like watching these videos more than the actual video on Mr. Beast's channel. also I've heard so many good things about Micro Center but until they come to Seattle I'm just gonna have to take your word for it
Its insane how much actually goes through these Mr Beast Videos and its so awesome to see it all unfold. Didn't expect that setup to cause so much trouble when I watched th video.
Took me about five minutes to realize you're THAT geerlingguy! Hats off and thanks for the heaps of ansible resources, they came in handy a ton of times
I was recommended this video after a William Osman video, and I was extremely surprised to see "the Promenade" within the first minute or two. I'd recognize that hellscape anywhere.
This is such a good example on how simple problems become significantly harder to solve if you scale them up. This is why you need real experts to handle these kinds of situations, good job!
Cool! Most of your language was like Greek, however, I was “58” in the challenge! Great fun watching the challenge from the background!!! Peitz was my favorite in the vid 😍🤣😍
it is insane how many talented and hard working people it takes to make one dumb challenge video on a channel marketed towards children who can't put down an ipad
This is truly amazing how everyone overcame the trials & tribulations to deliver successfully. 💪👍 Wow, just wow!! The challenges faced by the teams in a span of 2 weeks are what most teams encounter on a 3yr build... and you left out the human-social issues.
As somebody who does special effects and engineering for theatre shows I can relate with the feeling he had about the crew and the stress you have for finishing shit like this before, during and after a show/shoot.
From my background as a project manager in IT - I'm wondering how did the two week time-scale get decided ? Bravo to you and the team for all you achieved in such a short space of time, but it seems an unreasonable amount of pressure to put you all under. I'm also concerned about the safety aspect - all those rooms have acrylic (flammable) walls and beds (also flammable) with hastily made electrical connections and not enough time to consider all the risks. I also hope you were sufficiently remunerated for your efforts on this - after all, I expect this was a money-making venture by Mr. Beast.
re: the timeline - I'm guessing (wasn't there until crunch time...) that they were investigating off-the-shelf button/quiz show solutions, but none were able to scale to the plans they had without months-long customization timelines. So they dug through UA-camrs until they found a couple of us insane enough to try pulling it off in that timeline!
@@gotmilkbuttJimmy is also learning on the go, no one has ever made videos like this at that scale and he has no technical knowledge, jeff could’ve said no like many others but he took on the challenge
With all the NASA and NASA-adjacent techies, Huntsville has a feasible claim to being a better location. Also gives Nashville geeks a better shot at a day trip.
I am in the Charlotte area and with the amount of electronics Micro Center has I will be happy to have one in Charlotte because it’s like every single project. I need at least one electronic thing that I don’t have.
Why did you go to such lengths with the button bounce thing? Wasn't simpler to just program it so that it only allows the user to press the button once per second or something like that? Meaning, after you receive a "button_pressed", you just ignore that signal for the following second.
That's pretty much how I ended up doing it (but with like 100ms of delay per press). Check out the blog post linked in the description for a little more on the actual code!
@@JeffGeerling also I wonder, how did you find out this is called "debounce" ? I find quite often during programming, or troubleshooting anything really, that you feel like surely you aren't the first person to encounter a problem, and you want to search for it, but even when rephrasing your problem a dozen different times you find nothing, so you come up with your own solution, and only way later realize the general consensus nomenclature was different but not straightforward and now that you know what it is called there turns out to be tons of information about that topic. How do you (or other people) handle this?
Wish I could make it for the opening! I have a baby on the way so I can't go but if I'm ever in the area I'll stop by, the new locations get all the cool stuff :D
When you inevitably make the 30 minute video essay on buttons, remember to give a shoutout to Technology Connections and his video on switches.
Heh... part of the inspiration ;)
Seriously props to the fact that this video was that long and you literally kept me interested the entire time! (LITERALLY ABOUT BUTTONS....though I love tech and wiring things could be part of it :P )
That's work for @TechnologyConnections haaaaa
And, through the magic of buying 300 of them ...
Something to remember when using power rated switches for logic level control.
Power rated contacts ( the round ones ) rely on an arc during switching to keep the contacts clean from oxidation as well as cleaning off general " goop " left over from from assembly. Logic level contacts use gold plated knife edge contacts at 90* to keep the contact area small.
I've had this bounce issue with hydraulic pressure switches linked to a CNC controller. Pulling the switch out of circuit then using a resistor and power supply then cycling the switch to create arcs would clean the contacts and bring it back into operation.
As for the old time wall switch, the spring is more for a quick make / break in order to minimize arcing and eventual failure of the switch.
Another issue I've run across when using twisted pair cabling. Don't use both wires in a pair for different actions. The twist is basically a linear transformer and can lead to spikes / ghost operation especially with AC .
Kudos to you and all the behind-the-scenes people for your work
With how extreme and fast-paced mr beast videos have become I feel like there's so many stories that happen and get documented but no one ever gets to see
Agreed; my Dad was commenting on how this whole thing would've been an 8-episode season on any TV network.
William Osman in Mr Beast Squid Game
Great as this is, there’s no reason it has to be done in such a short turnaround - that’s only bringing huge additional stress to everyone involved.
@@JoeGallantChurchComms With all the 3rd party fallout videos I watch it feels like Mr. Beast and his core staff are using their power to ask way too much of people in too short of a time, so that they can save money on a project that would cost any other entity 2 - 3 times as long, and cost 2 - 3 times as much.
The reason is because since most people use TikTok, their attention span is usually very short which is the reason why Mrbeast tries to make every second of the video entertaining
During the days I spent in that cube I was immensely impressed and overwhelmed by the amount of work and attention to detail from everyone. Thank you for helping to make this an unforgettable experience. Sincerely, #41
Call me stupid or lame but what was this challenge? I don't keep up with Mr beast.
@@goosenotmaverick1156 People of the ages 1 to 100 were staying and sleeping in those 100 boxes and competed for a prize of $500.000. They did challenges and used these buttons to vote for challenges or vote each other out. In the end after over a week in there only age 52 and age 40 were left. The video has 230 million views.
@@goosenotmaverick1156 it's a good video if you watch it, but the tl;dw he put people from every age from 1 to 100 in small little boxes and whoever stayed in their box the longest won a lot of money. There were also challenges that the people could vote to participate in which is what the buttons and LEDs were for.
@@RobinClower I ended up watching it when I couldn't sleep late last night. I'll admit, it was entertaining. Can we all just collectively hate 54? That guy was an ass lol
so what was the challenge?@@goosenotmaverick1156
My condolences for working with those timelines, I'll probably combust into flames from the stress.
Idk how Mr beast can stress people like this and with so quick deadlines.. you think you would have more time
@@ElusiveXSetups Im sure he pays well.
@@ElusiveXSetups Hopefully between the William Osman squid game and the 1-100 video they know to give more notice for future projects
My condolences for working with those timelines, I proudly combusted into flames from the stress
@@ElusiveXSetups but you see... content... it's not as interesting I there isn't a time pressure, it kinda just becomes a educational video at that point
Wow... MrBeast's deadlines are insane though, they call upon creators with 1-2 weeks notice to complete these crazy assignments. Too many hiccups with the other workers messing up the wires and stuff. I hope they compensated you well. That's super stressful.
Thing here is to respect yourself and don't take crap projects like that unless you have a solution at hand.
Coz mr b surely doesn't respect anyone who he gives this assignment to.
Want to bend over backward and crash and burn - go ahead.
They didn’t compensate Will Ossmann properly. At least I remember it like that.
I don‘t get how they do it this last minute. I doubt they has the idea and started building the boxes in the same time frame.
Yeah, I guess UA-camr labor is counted differently when you can put 'Mr. Beast' in your title... but still. That's an insult to the creator I think if they weren't compensated well up front. @@DailyShit.
@@DailyShit.I remeber watching William Osman video but there's no mention of them being not compensated. Where's the sauce?
@@Dvance I know who Jeff is but never heard of a mr beast?
This is my experience with anything hardware. Coming from a programming world, it's really a shock just how much can and will go wrong with electronics.
Hardware doesn't go wrong any more than software. Just different kinds of failures.
but same like software. You can minimize the potential of problems by following best practice on every step. But no one actually do that because of limited time 😆
Nothing goes wrong with electronics, if you know what you're doing!
I'm an embedded hardware/firmware/software dev. My software-engineering colleagues always claim that the hardware is at fault, the hardware-engineering colleagues always claim the software is at fault. Whilst everyone just lacks the right knowledge to really debug simple issues outside their field of expertise...
@@bzuidgeestwell, software is more likely to fail deterministically than hardware, making it easier to debug
I see what you did there… Very punny.
As a senior network & system engineer myself I can personally say you did a fantastic job with the mess of constraints! I watched the Mr. beast video a while ago but seeing this now is just awesome
I saw the opposite.
Hot take: When using potatoes for CPUs, take lots of breaks and don't expect synchronicity on the first trial run. Great use of dad by the way. He brings the "ol' school" element.
From what I seen, MrBeasts team seems very unprofessional for always giving short notice for some of the most important parts of their projects.
Helping with a Mr. Beast video looks like compressing a whole engineering senior design class into a mini mester. Great job!!!!
There's no better instruction than getting your hands dirty on a project, then reflecting on the hundreds of things you can do better next time!
@@JeffGeerling Agreed. Sweat & pain is the best teacher.
"mester"
if thats what one learns for a whole year they need a better school
@@retsamyar I am exaggerating a little. There is obviously a lot of extra report writing, diagramming, etc a class has to do too especially in the first semester. Sometimes make custom PCBs as well. Additionally, most of the time students will be taking other classes alongside being in senior design class.
i swear this and william osmans video are probably 2 of my most rewatched videos and funnily enough they are both about the nightmare of working on a mrbeast video lol
I have worked as intern with company that did stuff like this, and had opportunity to work on sets like this. I can fully understand the stress and hotfixing everything last minute, even watching this got my stress level up like flashbacks, great video!
In production, it's a day in the life!
@@JeffGeerling Fully agree, great, exhausting experience.
@adi265gsaplay quick question for you. How did you find an internship like that? Word of mouth? Indeed? Just curious, it sounds super fun!
@@ericwest3560 word of mouth is the best fit i think. It was my first serious job and thought me a lot.
What kind of company is this ? Sounds like a place I'd actually enjoy working at
As a programmer, that’s an absolutely epic collab! You, Network Chuck, and MrBeast?! I mean, I never thought I’d see you together, but this is just epic.
Amazing video! It's wild you still got it all done that fast.
with this I can officially bring my 'Times other youtubers have *almost* ruined a Mr Beast video' counter to 3.
Which isn't a lot, but odd it has happened three times.
There’s this guy and William osman who’s the third?
@@SunnyCress hacksmith. They accidentally broke a safe that was meant to be cut open with a lightsaber just before the video was supposed to be recorded
@@Dec_TGM Accidentally broke a safe?!!
@@YourLocalWarGorilla they accidentally tugged on a carpet under the safe a bit too hard when sliding around an ice rink, and the gallium part of the safe fell onto the ice, causing it to break.
Your Dad has great insight! He's right, you all provided a magical experience for all of the people in the challenge.
The challenge was a really good MB video. Who would have thought a highly technical behind the scenes video from JG would be so well viewed as well. Congrats man. You and the "button and lights" team made the experience great.
There's a bit of irony around the interrupt problem. For debouncing mechanical contacts polling is almost always the better solution. I would have recommended running a separate task that simply polled the switches on a regular interval (100Hz or so) and then fired off events for your main code to catch.
Agreed. The regular timing of the polling can make debounce over a few tens of milliseconds quite simple and with a low impact on other processes.
Yup
agreed
Nice that you were able to make it work! I'd have done it completely differently though, using the sorts of techniques you see in industrial automation that ensure high reliability.
No need for a whole programmable computer in each cell; each would've had only an industrial modbus remote IO expansion module, mounted onto a strip of DIN rail, together with the power supply, relay modules, and a terminal block for every external connection that needs to be made after the pre-assembled rail is screwed into place. All wiring within the module would've been 18AWG silicone jacketed wire terminated with crimped ferrules, and both sides of the cable to the button box would've been keystone jacks with only factory-terminated CAT cabling between them.
Thank the Lord 🙏
This continues to be my absolute favorite tech channel! I’m personally not a fan of Mr beast videos but seeing the behind the scenes for a shoot like this makes me have a profound amount of respect for all the people involved in making these productions happen. The absolute pure passion and dedication from you and the whole crew involved is infectious and exactly why I keep coming back! Incredible work!
You guys did phenomenal work, definitely appreciating more of the bts of that video now that I've seen the stressful nights and the amount of manpower that went into making everything come together. Mad respect.
I wish my dad would have been interested in Tech and engineering when I was growing up. Looks like you guys have a blast together.
My Dad got me interested in electronics and IT very early on, I've always loved working on a project with him, especially puzzling over the best (and the most practical!) way of doing something.
My dad went in the oppisite direction as me. He is a farrier(Horseshoer). So our interests did not intersect often. Thats awesome your dad gets to help with these projects. Thanks for sharing. @@JeffGeerling
Ha! I wish my kids would be, wanna switch? 😂
@@henderstechohhh, but making things with 🔥 and ⚒️ is awesome too!
I've been doing events for years, and there's nothing as satisfying as working until 4am building a hack knowing patrons are in site in 4 hours, or listening to talent pad a live event while you're on stage under a table unplugging your carefully managed cables and running a hail mary one. Great stuff Jeff
Hehe I haven't done too much live production work but you can tell a good emcee when they can seamlessly delay for you and the audience is none the wiser!
jeff im glad to see you guys are thinking deeply about the fire hazard associated with this project, but i cant help but feel like such aggressive timelines on mr beast videos is going to result in disaster at some point. especially considering how big and potentially dangerous their videos can be. clearly yall did a fine job, but under this kind of pressure and crunch: the odds of mistakes go up very quickly. a bit concerning knowing they occasionally do videos with explosives, but even then its likely the danger in the details of things you wouldn't expect. just a thought
Yeah, after watching a similar situation with William Osman helping with another project it seems bound for tragedy.
Seems like something that could really be helped by queuing up more projects and running them on rotation.
That way each event has had double, triple or quadruple the time for work but he doesn't have to slow his schedule if he doesn't want to.
@@LanceThumpingthe problem with that would be having to pay people double, triple, or quadrouple the pay/having to hire double, triple or quadrouple amount of crew workers. That's not cheap when you have so big production team already. But yeah agreed on the crunch. Someone needs to put this problem into spotlight. I can only imagine how big of a tragedy could happen in a mr beast video with all of those people cluncked up together hastly made electronics and propably no fire extinguishers in the rooms
@@LanceThumping They already work on multiple projects at the same time. Not sure why such the short time frame, maybe they were using the warehouse for another project before this, maybe just to save money.
@@taisato2091 The crunch cost them a lot of money in components and rework labor. Their component choice was extremely limited by lead-times-they had to pay for next day shipping, change their specs to match hardware, and probably lost a lot of 100+ part bundle deals. Not to mention the amount of re-work labor or replacing ruined components due to ESD and conflicting construction schedules and OT.
This also would increase the technological quality and safety with more R&D time. PLCs, shielded cabling, metal enclosures for ESD protection, better cables etc.
I'm thankful i live right between 2 Micro Centers. Love that store.
I'd love to see a video postmortem on this whole thing explaining what you'd do differently if you had to do it again.
Check out the blog post-it has some of that perspective in it!
@@JeffGeerling jeff is clearly done with this project. great work Jeff and team
As soon as you ACTUALLY showed the blue walls, I immediately said "oh no" out loud. Immediately I was like "look, I know ElectroBOOM did that video with Linus, but that's a LOT of static."
Brutal! But hey, what a challenge!
Great Video Jeff...it was exciting from beginning to end. Glad your medical issues are better and you were able to get everything working. Im not a follower of Mr. BEAST but I may watch that. I appreciate all you do, love watching your videos and learn a lot from them. Thanks!
this gave me a lot of chuckles and reminded me of many problems i have had debugging electronics on a much smaller scale so i can imagine compounding the issues at the level you were doing. bravo sir.
Being that I have worked in the Production/Events business for 40 years, I feel every second of your pain. Including the mild emotional trauma of leaving it all behind. Congratulations!!
I've honestly found the little jumper wires quite unreliable even for my small projects. They break extremely easily. Also, I would 100% watch the 12 hrs of footage of troubleshooting, building, etc. LOL
De-bouncing is necessary for just about every digital circuit with a button as an input, dealing with it is one of the first things you learn when learning to design interactive systems using FPGAs.
Bounce isn't even exclusive to buttons. Any kind of switching might have bounce. Though it goes by different names.
Yep and a software denounce delay of ~12 ms is usually the easiest solution.
For me it was a 8 bit 8080 development board from the 80s that only had a hexadecimal keyboard for programming and no way to store your program, with a class final test to simulate a traffic light, using button instead of vehicle detection, all in a 150 byte, the debounce subroutine was a pain with memory limitations.
@@AstroStrongBoxPutting a pulldown resistor is better tho, it reduces bouncing well (when working with inverted logic *5V -Off, 0V -High*) and it's easy to integrate.
@@sebastienmonette6659 totally agree but hardware is hard to fix in scale and last second. Software will get you 90% there especially if absolute speed is not super important. If all buttons have the same software latency it’s still fair ;)
Found out we are neighbors lmao via the microcenter so cheers.
What a fantastic insight into the technical challenges behind mass scale videos like that. Thanks for diligently documeting it!
Between this and the Osman Squid video/stories, I feel like an infinite amount of stress and problems would go away if people were given an extra pre-week (before the 2week crunch) to prototype and test a tech solution at a peaceful schedule before it had to be done at scale (one box/room or one squib etc)
as someone that doesnt watch mr beast, this is my favorite mr beast video. also that button bounce issue is really interesting, i had never heard of that.
It's a problem with any and all physical pushbuttons and switches.
@@nikkiofthevalley You mean model O's on release? XD. Gaming was fun with that POS for 3 mins before chucking it for a gpw many yrs ago >:D
Amazing video. I was able to predict the cause of every issue and it shows how much engineering goes into even the smallest projects. It really helps to pay a little extra and not cut corners especially when the client can move goal posts at anytime.
I was really impressed you stuck to dupont to the very end and didn’t use any project boxes.
Best episode ever Jeff. If anything, this should be worth a white paper on how diverse a situation, small form computers can be used, and the fact it was done in a simultaneous prototype/production scenario is pretty dang amazing and telling to the capability. The whole experience is impressive, /hats off my friend
This project doesn't really need 100 computers. It's the inefficient but quick (maybe) way to get it working fast
People see the end result to massive events like this but have no idea what really goes into making them. This video is amazing in making people more appreciative to the amount of work it take to create something like this.
“Creator youtuber talks about how they almost destroyed a mr beast video” has to be my new favorite type of content.
So nice of you to help out such a small UA-camr with his video! Great work!
I worked on a project exactly like this. Watching you learn everything I did in this video was comforting and anxiety-inducing. I was building a large interactive arcade setup about the size of a football field. There were 22 intel NUCs, a bunch of rpi3, and over 100 esp32 custom boards with neopixels. our electronics team at the time was somewhere between 3-5 people. We had every exact issue you had, plus some. Was a truly wild experience.
Ha, well I know I'm in good company then! All kinds of fun problems to be had when it's not all contained on a little breadboard on your desk...
Why was my comment deleted?
Anyways, some suggestions for next time:
- an operation like Mr Beast's will have an experienced light tech. You should have focused on what matters: the buttons. Let the light technician deal with the strips - send an ArtNet signal to his console and you'd be good - he would have mapped this stuff out in a few minutes with his console.
- Don't buy cheap indoor LED strip, instead rent out some high quality outdoor DMX stuff especially for a one-off project. All that stuff will now become garbage or just pile up dust in a warehouse somewhere
- Use proper Molex connectors and put all your electronics in the box. Less wire length = less susceptible to static charges
- Should have ran a single ethernet cable + power from each box. Why buy cheap unshielded cable when again, event venues rent this stuff out all the time.
Anyways I thought it was an entertaining video and I hope you got some experience out of it. The event business is a whole other world ;)
Trust me, I've been there!
Your first bullet point was the exact thing a lighting tech pointed out on set :)
Unfortunately they were only consulted for the overhead lighting, not in-room. Would've been great to have lighting be separate, and I could just manage it over the network. (They had a massive lighting console set up entirely independent.)
@@JeffGeerling glad to see I still got it then ;) Guess you'll know for the next one!
Cable relief is about #1 on the build list. Only fighting with ground/screening.
It's common to use current loops for long unprotected cables because it's easy to inject a voltage spike but very much harder to inject a significant current. And a current loop has a quite low-impedive listening side. And it's easy to clamp any peak voltage.
Good job that you got it working. But seems this was a task where you need some communications R&D engineer in your phone book.
Heh, that was my Dad, and he got involved *after* we had gotten the initial order in. Also was not able to be physically present once we started laying things in the floor (because of the slight delay with the rooms)... he would've probably helped us get through a few of those hurdles more quickly! As an RF engineer, he's seen it all (or, well, most things EMI/RF/interference!).
@@JeffGeerling One reason whe love twisted pair Ethernet is the 4 kV isolation in the magnetics of the network connector. A big reason why few people care about shielded network cables. Twisted pair reduces amount of bit errors in the transfer and the little transformer in the connector protects the electronics from big spikes.
Then low impedance signals for the shorter cables.
Many customers talks about "we can do this with a Raspberry Pi" about the hardware I work with. When ignoring voltage range, temperature range, EMI filters, clamping diodes etc of equipment for industrial or automotive use. 100% fault proof is hard. But a few tricks can take it quite far, allowing even hobbyists to manage quite robust designs.
14:44 "20-foot long giant static collecting antenna of death" ... fantastic
I respect your dedication, focus and slight display of madness learning the depth of bounce in buttons!
EZ de-bounce is to simply add an RC low pass filter to each switch. Often satisfactory results can be obtained with 0R, only needing the capacitor, this will filter out the high frequency transitions typical of bounces.
Awesome job! Video came out really great. The amount of work you did was insane lol. I can also add to my resume I was in a jeff geerling video!
It was great meeting you! Even if you're name wasn't Jimmy or James :D
What specifically amazes me about this whole process was the short time crunch. I would've assumed it'd take a week minimum for some of those parts to arrive in bulk
Heh, we wound up calling some suppliers direct and getting some things shipped straight from the origin, if we couldn't find stock somewhere in the US!
@@JeffGeerling Hats off to whoever was making the endless calls & rapid firing PO's. Who did this on the set? Yourself or ?
Damm, that bounce is insane. I've seen this multiple times in industrial applications. I could have used industrial controllers like PLC or networked io's.
youtube has been recomending this video to me non stop for 6 days and i finally decided to watch it, it was a great video
UA-cam's like "yeah... so you definitely want to watch this one"
Glad you made it :D
this video sound more like how the crew barely saved the video than "broke it"
gj on the whole team and hope you all got a lot of sleep after
Wow, what a great video Jeff!!! Very exciting all the way through especially with so much debugging. I felt second hand stress watching this, was almost like a thriller movie haha. I definitely could’ve watched all of the remaining footage without getting bored, seems like it was a blast!!!
Next time you need to de-bounce in a hurry just solder a capacitor, something in the 100nf to 1uf range across the button, with high impedance CMOS inputs that often does the job. The alternate way is just to program a hard pause after the button state change. Your polling issue could have been fixed with small sleep in micro-seconds in the main loop, that will cause the C runtime to yield back to the kernel and stop the process from CPU hogging and running hot.
Yeah on a super time crunch like jeffs I‘d take every shortcut and workaround
What an exciting project! Congratulations on such a unique opportunity and many lessons learned. This video taught me something new about buttons. Also, it's cool to see Le Potato in use in the wild - this may be their biggest project yet.
This was incredible to see, the amount of stress you guys were under must've been insane. Although looking back there are probably a lot of things you could've done different, given the time constraints you guys made a miracle happen, seeing how much work you guys put in really made me appreciate all that goes into a video. Can you convince Mr.Beast to call you guys up ahead of time next time he has a massive project like this?🤣
This video gave me flashbacks on one of installations I've been in where certain LED boards or NUCs or the wires and networking between them were simply not working because of the odds when dealing with around a hundred of them all at once.
When looking at electronics individually, it's easy to assume that they'll be fine but running hundreds of them together and all their little moving parts and considering the people that'd be messing with them - like frigging static on acrylic, you're likely to encounter a dud or a broken setup very easily and it's frustrating even with spares.
This is insane. So impressive that you pulled through!
This is the first vid I've seen from you, i love it no only the behind the senes of a beast vid but what it takes to making the idea come to life ,A master at work.great video
Man, you should have called for this. I was the engineer for a 500 person trivia tournament about 7 years ago and had to set up response boxes for 50 tables and the setup was very similar (50 button boxes, 4 buttons but no LED strips). Every time we ran into big issues it was due to treating the project as a hack instead of a product. I made as much of the system modular as I could so modules could be independently tested and swapped as needed.
We bought 3-gang electrical boxes and matching blank covers in bulk from Lowes which meant we had plenty of spares and could get more in an emergency. We used a fly press and a die to knock out the holes and the boxes had built in knockouts that we used for connectors. Inside the button boxes we used back lit arcade cabinet buttons and made harnesses with crimped spade connectors (no soldering) to 2 rj45 that connected to female terminals on the box. Boom, now all boxes are modular, easily assembled and tested and swapped (I made a test rig to test 5 boxes at a time).
Then we ran cat5e (or cat6/e this was in 2016 so I can't remember off the top of my head) from 3 button boxes to one of 17 hubs that split out the rj45s to jumpers that fed into a pi zeros IO pins. The pi zero would read the inputs from each box, set their light's status, do some error checking, and send/receive data over TCP/IP to a central pi that would then host a website where you could see people's selections, reset the button states and check for errors on the hubs/button boxes.
Even though everything was thoroughly tested before the start of the tourney, we had multiple spares of every module. If there was a failure with a box, we just had to disconnect it from the rj45 cable and swap in a new one (and troubleshoot the faulty box). If a pi went down we could swap it with another that was pre-flashed. If a harness was faulty, replace it with a spare. All of the wiring outside of the boxes was standard ethernet that could be quickly replaced and the hubs used POE so we didn't even need to run power. I wish I knew where the whole thing ended up.
I was planning on spade connectors but I thought the boxes would be set up by another group... when we had to do it, there wasn't any place nearby with enough spade connectors, so soldering it was!
The setup at home, I re-built it using spade connectors and it was about the same speed as soldering, but infinitely easier to just do a single button swap.
@jon1913 but how much time did you have to prepare all that? ;)
I LOVED this video. Behind the scenes is some of my favorite content to watch. I hope they bring you back for some more adventures, if not for your electronics expertise, then for your story telling prowess! 🙂
Had no idea you were in on this! Great job!
That'd be my dream job... actually, I'd be there for free! Not for the Video or MrBeast but for the experience you went through. A super tight schedule, barely any sleep, fixing new problems every 5 minutes, debugging software that had be written in record time, praying everything will work, improvisation, the team work ... great job! (also on the editing part)
You are joking right?
actually, no. I actually mean it. However, I can see how one might think I'd be joking with a comment like this
@@mattbrandt8207 I mean… you do you, but for that kind of stress, even if it is a fun experience, I would still want a bit of money for my efforts.
It is beautiful what you did Jeff. I think the take home here is how resilience and hard work can pay off. It really is inspiring to me.
Being a dev, this whole video reminds me of why I love doing what I do. Go-lives are the most stressful thing in the world, but when everything works out in the end, it really is a great feeling! Although it must’ve been a stressful few days, it must’ve been incredibly fun!
Back in the 90s just out of HS, my boss had a need to have buttons in 100 apartments that when pressed would notify the security office. I was just learning about microcontrollers and said I could build it. I set the buttons up in a 10x10 matrix connected to the PIC uC and polled them by putting one row high and checking for high in each column.. Imagine my surprise when it worked fine on the bench but when I put long wires on it, one button press would trigger multiple apartments. That is the day I learned about capacitance. The capacitance of the long wires would lower both the leading and falling edge slew rates causing the button detection to shift to the next button, or cause multiple buttons to be triggered. Had to redesign the hardware to use high current drivers and slow down the poll rate. I could only imagine the EMI that system put out in that building lol! It was in service almost 20 years!
Hehe, yeah... what's the saying about electrical vs rf engineers? Only an RF engineer actually knows when he builds a transmitter!
This video actually means so much to me. I love everything systems administration, and the thrill of those high-stress jobs, even thougu I've never been on any as high stress as this.
But this video really means a lot to me because I just watched it to destress. I've had to take my little brother in to the hospital, and I was with him until a family member was able to tag team with me. Jeff, watching your stress for something positive helped me alleviate my stress for what's been going on. Thank you.
I pray he gets better soon!
@@JeffGeerling thank you! He's on the mend.
Believe it or not, I’ve worked in several projects of this magnitude and similar functionality. Last time I went with 125 ESP32 connecting to the server via wifi. In this case I would have put the ESP32 and relays inside the button box, connected to the LED strips. If one of them malfunctions: hotswap the whole box with one of the 20 spares. But I’ve done this 4-5 times, the first time I had similar problems.
Congrats on pulling this off! :)
I was thinking Pi0w via wifi would've been decent. But ESP32 would probably make even more sense. You could probably get away with a half dozen APs and a lot less wiring. There are certainly more than one way to skin this cat.
At university I meshed ESP8266s to network to a root server. Definitely would've simplified the work here.
I'm starting to hate this rushed aspect of Mr. Beast's videos.
Every time he does a colab with someone it's just to make them work so hard on a super complex problem with the tightest schedule ever
I thought the same while watching the William Osman video when he colab with Mr beast
Just of the top of my head I can think of Donut Media with their jet-powered super car, William Osman with the squid game kill mechanism, and Hacksmith Industries with light sabers (which in the end wasn't even part of the video).
I'm sure there's more
Expensive but this is certainly a case where a mountain of Wago connectors would have been a massive time saver.
The lighting techs who were debugging the LEDs concurred. we couldn't find more than a handful of Wago connectors in town at that time though.
Woah hahaha I love the rabbit hole of debounce. I ran into bounce when creating buttons in javascript so it's bounce all the way down
Considering you've made this video I assume its safe to talk about - I worked on this set too, in a way. I am a CNC Operator at a chstom shop here in NC; about an hour away from Mr. Beast. I've spent many a weekend working 14-16 hour extra days cutting props for Mr. Beast sets. This set was actually pretty grueling to get done. Had to load, cut, unload, clean, reload, 500+ sheets of acrylic on the CNC and it was a ton of overtime getting it done in time for their timeframe. Kinda fun to see the whole place all put together!
Heh, I can only imagine! When I saw the sheets stacked on all those pallets, I immediately wondered who got stuck routing those things out-that's a LOT of cutting!
Also had to be pretty precise. Nice work!
@@JeffGeerling It was a lot of work, and also probably the most boring thing I've made for them hahaha. Coolest thing was a giant sign made of milled MDF, painted, and carefully hand poured full of epoxy for the different colors. That one was at least really interesting haha. And same! Thanks! I've since finished watching your video and definitely great work! I've been dying to get back into playing with microcontrollers and hobby electronics again and got a bit of inspiration from seeing a few of your vids haha.
Dang it Jeff! I want a Microcenter near me too! 😢😢
I think YOU'RE the beast. Way to get stuff done. There's not that many people that could consistently solve so many strange issues over successive days of long hours.
Hope you are getting enough water, rest, vitamins and sunlight to counteract what you did to your body.
Talk about Chris using pihole please
16:35
Yo, that kind of high stress quick thinking chaos, your job sounds so fun. Wish I had a job like that.
Money makes humans do very weird things...
I honestly like watching these videos more than the actual video on Mr. Beast's channel. also I've heard so many good things about Micro Center but until they come to Seattle I'm just gonna have to take your word for it
It sounds like you're not alone in wanting one in Seattle!
Microcenter in the UK please?
Its insane how much actually goes through these Mr Beast Videos and its so awesome to see it all unfold. Didn't expect that setup to cause so much trouble when I watched th video.
Just watched a short mentioning Mr Beast and this popped up 😅😅😅
Took me about five minutes to realize you're THAT geerlingguy! Hats off and thanks for the heaps of ansible resources, they came in handy a ton of times
Heh, no problem! Got to use some of those Ansible playbooks on this project :)
Ayyy that the micro center by my house
I was recommended this video after a William Osman video, and I was extremely surprised to see "the Promenade" within the first minute or two. I'd recognize that hellscape anywhere.
This is such a good example on how simple problems become significantly harder to solve if you scale them up. This is why you need real experts to handle these kinds of situations, good job!
Cool! Most of your language was like Greek, however, I was “58” in the challenge! Great fun watching the challenge from the background!!! Peitz was my favorite in the vid 😍🤣😍
One of the infamous 50s! I'm sorry you got knocked out on the Jenga round, that was a rough challenge!
I am very much here for this emerging genre on UA-cam of stressed out engineers trying desperately to make a Mr. Beast set work.
0:27 Literally that gave me ELECTRO BOOM vibe...!!!😂🤣
We all need to channel our inner BOOM every now and then!
@@JeffGeerling indeed 😂🤣
16:23 I was gonna say based Ava but…..
it is insane how many talented and hard working people it takes to make one dumb challenge video on a channel marketed towards children who can't put down an ipad
This is truly amazing how everyone overcame the trials & tribulations to deliver successfully. 💪👍
Wow, just wow!! The challenges faced by the teams in a span of 2 weeks are what most teams encounter on a 3yr build... and you left out the human-social issues.
As somebody who does special effects and engineering for theatre shows I can relate with the feeling he had about the crew and the stress you have for finishing shit like this before, during and after a show/shoot.
From my background as a project manager in IT - I'm wondering how did the two week time-scale get decided ?
Bravo to you and the team for all you achieved in such a short space of time, but it seems an unreasonable amount of pressure to put you all under.
I'm also concerned about the safety aspect - all those rooms have acrylic (flammable) walls and beds (also flammable) with hastily made electrical connections and not enough time to consider all the risks.
I also hope you were sufficiently remunerated for your efforts on this - after all, I expect this was a money-making venture by Mr. Beast.
re: the timeline - I'm guessing (wasn't there until crunch time...) that they were investigating off-the-shelf button/quiz show solutions, but none were able to scale to the plans they had without months-long customization timelines. So they dug through UA-camrs until they found a couple of us insane enough to try pulling it off in that timeline!
@@JeffGeerling Aha! Yes, I can see how that could happen. Well done for what you achieved in that time-scale.
@@gotmilkbuttJimmy is also learning on the go, no one has ever made videos like this at that scale and he has no technical knowledge, jeff could’ve said no like many others but he took on the challenge
You're such an underrated content creator and brilliant mind! This was great coverage.
BIRMINGHAM, AL needs a Micro Center
With all the NASA and NASA-adjacent techies, Huntsville has a feasible claim to being a better location. Also gives Nashville geeks a better shot at a day trip.
I am in the Charlotte area and with the amount of electronics Micro Center has I will be happy to have one in Charlotte because it’s like every single project. I need at least one electronic thing that I don’t have.
Why did you go to such lengths with the button bounce thing? Wasn't simpler to just program it so that it only allows the user to press the button once per second or something like that? Meaning, after you receive a "button_pressed", you just ignore that signal for the following second.
That's pretty much how I ended up doing it (but with like 100ms of delay per press). Check out the blog post linked in the description for a little more on the actual code!
Ah, I was wondering why no one else said this, but I just had to scroll down the comments a bit further...
Great minds think alike and all that.
Also funny how a million people tend to reinvent the same wheel.
...
And then realize later that 999999 people before them did the same 😅
@@ElMoonLite It is tradition!
@@JeffGeerling also I wonder, how did you find out this is called "debounce" ?
I find quite often during programming, or troubleshooting anything really, that you feel like surely you aren't the first person to encounter a problem, and you want to search for it, but even when rephrasing your problem a dozen different times you find nothing, so you come up with your own solution, and only way later realize the general consensus nomenclature was different but not straightforward and now that you know what it is called there turns out to be tons of information about that topic.
How do you (or other people) handle this?
You’re here making this technology when i can barely find the right cord.
waiting for the Miami location :)
Wish I could make it for the opening! I have a baby on the way so I can't go but if I'm ever in the area I'll stop by, the new locations get all the cool stuff :D