Embedded - 119: Do Your Neighbors Have Any Idea?

Episode Date: September 24, 2015

Ben Krasnow of the Applied Science YouTube channel talks with us about scanning electron microscopes, generating liquid nitrogen, and cookies.  Hackaday Conference is Nov 14-15, 2015 in SF, CA! Call for proposals. (Ben and Elecia are Hackaday Prize Judges.) Contact Ben through twitter: @BenKrasnow Applied Science YouTube channel (and don't forget the associated Patreon). Some specific videos we talked about: Cookie machine Electron microscope scanning vinyl record Faraday effect (control light with magnets!) LED contact lens (not for the squeamish) Other people's videos and projects: Brady Haran's Periodic Videos Veritassium channel Build your own waterjet Amscope microscope and low cost hot air rework soldering station

Welcome to Embedded FM, the show for people who love building gadgets. I'm Eliseo White, here with Christopher White. Guess who's here this week? Ben Krasnow. Should we talk about electronics? USB processors? Fire? Electron microscopes?

Wet chemistry? Hackaday judging? All of those things at once? Oh, that actually reminds me. Before we get started, Hackaday is having a weekend conference in San Francisco, November 14th and 15th. The conference will be quite inexpensive to attend, so it will fill up pretty quickly. Keep an eye out if you want to make it. Although if you really want to go, possibly even get some travel assistance, sign up to give a talk or workshop. I'll put the link for that in the show notes. There's no deadline. So once they get enough awesome talks, they're going to close their

proposals. Not an opportunity to procrastinate. Hi, Ben. It's great to have you here today. Thanks for having me. Could you tell us a bit about yourself? Sure. I'm a hardware engineer at Google Life Sciences and also run the Applied Science channel on YouTube. What's your background? Mostly electronics and mechanics. I guess I like the title hardware engineer a lot because it encompasses both, you know, mechatronics, PCB layout, mechanical engineering, all that sort of stuff. So Google Life Sciences. Yeah, we're the first alphabet company it's all public now i i i have no idea what that what google would do with life sciences they're

rewriting our dna or at least making it searchable yeah our mission is to transform health care from sort of a reactive kind of system now where you only go to the doctor if you're sick to something that's proactive so you know advancing advancing healthcare to be better detection. Okay. So probably leveraging the search and machine learning kinds of things that are already there to apply to health science. Yeah, it'd probably tie into it, yeah. Well, I want to ask about Google,

but I also really want to talk about your YouTube channel. Sure. And that's Applied Science. Tell us about, for people who don't know, what is it? I like, so it was originally called Ben Krasnow. That was the title until I changed it to Applied Science because I thought that was more descriptive. And even then people complained about the title change, but I'm glad I did it because the channel is really about applying science.

And so I like to take a topic and then show how you can actually do a sort of a real world demonstration of a scientific principle. And you do it from first principles. Yeah, more or less. It's very physics-y. Yeah, I mean, I definitely stay away from the math. You know, it's pretty rare that I bring out like a hardcore math equation, but I do like to sort of bring things back to, you know, more or less first principles. I did a recent series on like the Faradaday effect showing how you can control light with magnets and electrics it's pretty cool and you had uh polarization filters in order to get the light to go one way and i didn't even know

that was a thing you know i thought the effect was way too small to even visualize with sort of on camera without any sort of special equipment and so i spent quite a while trying to track it down but was happy to actually capture it and show it to people i loved watching that video because a long time ago i took a class in electro-optics and it was all very theoretical and you know there were some examples in powerpoint presentations of this is you know the kind of thing you can do with electro-optics they're all very expensive of course because exotic materials that was the first time i'd seen a demonstration yeah of anything electro-optical uh you know with household materials basically yeah exactly and so that's

that is sort of the whole point of the channel is just to show people and so if you search the internet for faraday effect or magneto-optical effects you would only find equations and sort of the driest of explanations of what's going on and so just having like a two or three or five minute video even just showing it in action is a pretty big win how long have you been doing it um probably since 2011 so my breakthrough project was the the home-built scanning electron microscope that was kind of the thing that just launched the whole channel off you know oh so that's i thought that was new well so that i have two now so the first one i built myself in 2011 and that was like

you know the key project and then recently someone donated a commercially made seven to me about a year ago okay so you are officially the first person i've ever met who has two scanning electron microscopes and really hilar hilariously, a day or two ago, someone offered me another one. And I said, you know, I think I've kind of hit the limit. Is this some sort of thing? Once you have one, people start giving it to you? How many at-at shirts do you have, Christopher? I don't think that rates. Enough to start getting comments at Chipotle.

Yeah. No, it's true. And so, yeah, I mean, eventually people will kind of associate you with an object or many of them, and then they figure, well, you must really like them, I guess. Exactly. You know, you're a collector now. Right, right. So you do it every week? That was kind of the where we talk to YouTube creators and kind of help you out and give you some tips, not knowing that I had recently started working at Google. And so I said, Yeah, I'm actually right across the street from you. You can just talk to me through Google's chat system or whatever. So

anyway, so they gave me some helper tips. And at the time, I was thinking maybe I would convert to a once a week format. And that just didn't last. I mean i mean i the time required with a full-time job was just no way yeah video is a lot of work yeah i mean just just having the discipline of doing like kind of a new project every week or at least just having some kind of an update every week was not going to happen so how do you choose it seems like i mean i was the example of the faraday effect did you come across that and say, well, okay, how could I show this? Or it seems like there's, you cover a lot of topics on the show from optics to electronics to mechanicals. And is it just stuff that strikes your fancy or? Yeah, I guess I'm trying to remember

how I exactly thought of the electro-optical stuff. A lot of times it'll be like a passing conversation that I had with someone where they're like, oh yeah, there's this cool effect, but I've never seen it. And then that's like, that's like my cue to basically start researching it right there. It's like, oh good. Something that everyone knows about, but no one's ever actually seen it. So I like doing stuff like that. How long do the videos take to put together? Some can actually be fairly quick. I mean, if I already have the thing that I want to show built, that could be a one evening kind of thing, basically.

But building the stuff and getting it to work. So I chased the electro-optical stuff for a cumulative probably a day or two. Like with magnets and vials of olive oil, basically looking for it over and over again and not seeing anything at first. And so do you script ahead of time or do you kind of wing it? It's mostly winged. I'll kind of pace around in the garage and think about what I'm going to say. And of course, it's edited. So I'll say things a few different ways and then take the one I like the best.

But I generally don't write anything down. So it's more or less off the top of my head. Well, sometimes you have the math areas written down. There was one about lenses and Fourier. You already had all the fourier ready to go right so if i do a drawing i'll make that for beforehand yeah and that's definitely a huge time saver there too like i've been tempted to do computer generated animations and sort of better graphics and i just don't want to let the genie out of the bottle because once you do that that's

like that'll probably double or triple my time budget at least and I you know it's I think that the hand-drawn sort of figures go along with sort of the the low mid-production value of the whole channel and so it kind of fits and I think people sort of like that so do you ever think about trying to make it glossy yeah I have thought of it um for a while I even considered doing it sort of full time. I still occasionally think of that a little bit, but I interview somebody from Google, I try to make the point, yeah, if you don't have a public life, don't come on my podcast. They might fire you. Yeah. I mean, what's really, really funny is that's how they hired me was through the channel,

actually. So someone at Google saw it and was like, hey, let's bring this guy in. And at the time, I was doing fine at Valve and everything was good. But it kind of planted the seed. And then later on, I realized that we had to start the conversation again. Okay. For those of you who just twigged on Valve and were like, oh, let's talk about that. We will get back to it, but I want to talk more about the videos first.

Why do you do it? Why do you do the videos? I mean, as a hobby, okay, but this is years of... Yeah. There's a lot of cool... I i mean there's sort of a lot of equally important reasons for me to do it like it's it's partially a good personal motivator so if i feel like you know people will send me messages like hey you haven't uploaded a video in a while what's going on so it's like okay so at least i can sort of force myself to stay focused

on projects because i feel like someone is actually paying attention. That's a really good personal motivator, or I guess it's an external motivator at that point. It's also a good personal project journal for myself, just a way to look back and see all the stuff I've built. It's pretty neat. I do it also because I do, in fact, get paid. So the YouTube sends me ads revenue, and I have the Patreon stuff going too. Yeah. How long have you had the patreon stuff patreon oh wait a minute christopher should explain patreon because he explained it to me i did initially yeah people give you money thanks maybe you should explain it ben yeah it's it's it's a service uh where people can directly sponsor a youtube creator

or any artist or any any sort of a creator and And they can either pay per video or just give you a monthly sort of donation. And it's really nice because it's direct from the consumer to the creator. And Patreon just takes a very small cut. So it has nothing to do with ads. It's just totally, you know, one hand to another. Yeah. And like you said, it's a continuum between tip jar to special stuff. Yeah, exactly. So people, it's kind continuum between tip jar to special stuff yeah exactly

so people it's kind of like i mean they have what sort of like kickstarter style rewards where if someone pays more than they can get you know special access or privileges or special topics or that sort of thing and so you what what why was youtube not enough why did you do the patreon too um i kind of like the model a little bit better. I mean, there's nothing wrong with being ad supported. But I mean, it's even nicer to have people just paying directly because they feel more connected to the content that's being created. If you pledge, you know, $3 a video, you have a real stake in sort of enjoying that video because you're $3 committed. Whereas if you're just ad supported, people just kind of drift in and out and they

aren't really as connected. And so I like my channel a lot because the viewers tend to be very interested in the topics that I'm talking about. So it just kind of fits better. Well, then you don't have that many ads. I mean, I only seem to see an ad once every four videos or something. That's something that YouTube controls. And so all I do is tell them that I'll allow ads on my videos and allow them to be skippable basically. So I can force people to watch the full 30 second ad, but that's the only controls that I have. Otherwise the number of ads and the type of ads are YouTube's thing.

All right. So why videos? I mean, I prefer, you know, voice, but also writing and you have so many diagrams and schematics they go by so fast sometimes yeah it's true there's a little bit of room i actually do have a blog too but it's just a collection of videos so so i kind of try it a little bit here and there if i'm going to post code or other still images they'll go to the blog which is again, again, not very often. Videos, I like consuming video the most myself. Like, I think YouTube's kind of 10-minute chunk is almost like the perfect size for myself. So 10 minutes is enough to get a topic across, but not so long that people kind of tune out

or kind of tune in and out as the video is going by. So I like YouTube's format quite a lot. I basically just wanted to create something that I like consuming myself. That makes sense. I tend not to be a video person. And so it's a little harder for me to get into the video. You know, what he's doing harkens back to like Mr. Wizard and Newton's Apple and Bill Nye and those sorts of things. Although with, you know, less. Yeah, no, definitely. In fact, I mean, Mr. Wizard, of course, a huge inspiration of mine since his show is with you know less yeah no definitely in fact i i mean mr wizard of course a huge inspiration of mine since his show is you know sort of similar to mine and that he would have segments right segment was about 10 minutes or less very very easy to stay focused

for the 10 minutes right and they were always some demo of something and it was this is how this works and if you want to learn more here's how to learn more yeah i definitely liked mr wizard so did you always do projects like this before the videos and you just started videoing them you had been doing this your whole life as just playing with stuff and pretty much my output definitely went up when i started doing the videos just because of like the the um you know the motivation factor there but yeah i've always done projects um not as ambitious and not as many and i and also i just haven't had as much time so i graduated in 2005 and then prior to that of course school just owns

you completely and so do you have favorite videos you've posted favorites um yeah yeah like the um the animation of the phonograph needle in the scanning electron microscope. That was a real popular video and also one of my favorites because I had never seen a scanning electron microscope animation before. So it's kind of funny. I mean, of all the CEMs in the world, as far as I know, no one's ever actually used them to do stop motion animation, which seems kind of like an obvious fit. And I've got a lot of other tricks I'm planning to do with mine, too, in addition to more animations. You did a dust mite too on the skinny electron microscope yeah that's yeah flea was it alive uh no so that

if you're gonna do because that would be cool too well so i'm trying so the problem is that everything you put in there has to be in high vacuum like really oh that's a little hard on the critters so there's that and then even worse than that if it's not conductive it has to be coated in metal right you coated the lp in some sort of silver thing oh well what are those weird creatures that that they they can live in on space little insects they look like uh it's a tardigrade that yeah that yeah yeah so i have already i haven't tried it yet but they're gonna be uh they're coming up. Yeah. Wait a minute.

What is it? I didn't. It's like a little water creature, kind of like a brine shrimp, basically. But they're very small. Yeah, exactly. It's of that scale and everything. And they're supposedly like the hardiest animals ever in the world. They can survive in deep space for a few months or something like that.

They can survive radiation that would kill everything. They're insane. Yeah. How do we know they survive in space is it because they come from space i think we took them there sure we took them there uh it's cool that it's a macroscopic like if you were talking about a a bacteria you'd be like well okay it can survive in extreme conditions but this is actually like a macroscopic little gritter, which is pretty impressive that it can handle all this. It goes dormant, too. So while it's frozen to death, it doesn't actually die. It just goes to sleep. And when it thaws out, it warms back up.

How are you going to make it conductive? Yeah, so you have to coat them in a very, very thin layer of metal. They'll be fine with that, I'm sure. Well, if the radiation doesn't kill it maybe the metal won't either you know also known as water bears or moss piglets i know what i want for my birthday yeah i can give you a billion of them yeah do you have any videos that are less favorite but are super popular the ones that kind of got out of control um i don't know if they got out of control but some of them were definitely misinterpreted and back when you have you know

like five or ten thousand subscribers it's not as big of a deal as when you have you know 100 or over 100 or even when you have 100 subscribers or something and so i did a video where i put an led inside a contact lens and so at at the time, I was working on augmented reality at Valve. And there was this news cycle where some researchers at University of Washington were showing how they put an LED in a contact lens and then put it in a rabbit's eye. And at the time, we were researching the topic quite heavily. And we kept trying to tell people, you know, putting the electronics in is not the hard part i mean it's really not the hard part it's really the optics that are the challenging part with augmented reality so i went home and took two of my contact

lenses and sandwiched an led in there a surface mount 0402 led with battery no with a coil and then i powered it inductively through a coil that i just held up to my to my eye and it flashes no you know it's not it's not it's not much power, and it's only over an inch. This is your worst nightmare. Yes, contacts alone. You shouldn't watch that video then, because it shows my eye red and irritated as I'm poking the contact lens in.

And it's blinking. And it's blinking. Oh, yeah, and it's blinking, right. And you can see the blinking, right? Well, so for the benefit of the viewers, I turned the LEDs that it was facing outward instead of in, as it would be for an augmented reality contact. And the point was supposed to be a joke.

I mean, I was basically showing people like, well, clearly, you know, I mean, not to poke the researchers too much, but the LED itself is not the hard part. It's really the optics. And a lot of people completely misinterpreted what the video was kind of about there. And it got a lot of people completely misinterpreted what the video was kind of about there. And it got a lot of views. Did you ever wear it anywhere without telling people? Because you could build the inductor as sunglasses and you could be a Borg.

It would be awesome. It would need a little refinement. If you do end up watching the video, you'll see why I couldn't do that. Just because it was horribly painful. My eyes are watering just thinking about it. I won't describe it in too much more detail. But yeah, it was people, the comments were a little bit out of control in terms of how they thought I was experiencing the pain of this contact being in my eye. Or did they think you were in more pain or not nearly enough?

Probably about the right amount, unfortunately. What other favorites do you have? Let's see. So the SEM stuff has been good. I guess the supercritical CO2 stuff has been like a perennial favorite. I was inspired by a video made by Brady Heron, who does all the periodic videos and 60 symbols and all kinds of other videos. And he interviewed someone at University of Nottingham who had a chamber with CO2 in it. And if you heat it up, the CO2, which

is in a liquid and gas phase, turns into a continuous supercritical phase. It's pretty weird. The division between liquid and gas just disappears. And then when you cool it down, it becomes turbulent and you see the thing separate back out into gas and liquid again. And so I thought that was pretty neat unfortunately his chamber was built with sapphire windows to contain this high pressure environment because it's a couple thousand psi in there so i did some engineering and built mine out of acrylic and it worked and so i made a video about that that seems scary yeah hey i engineered it you know formulas i even AI engineered it, you know, formulas. I even posted them to the blog, you know. That was one of the things I noticed with your channel

was that it was sort of bodging it together, building it out of scraps, maybe not buying that thing that you should have bought, instead just trying to figure out if you can make it. Yes, definitely. I like that aspect. And that's even kind of what I do in my day job. Like we're sort of a rapid prototyping group.

And so going fast and basically just making the core idea testable very, very quickly is my day job. That is the most fun part of engineering. And I'm super jealous. I can't believe you got that as your day job. I know. So that is why I decided to go full time. Everything's working great. And so, yeah, it's a perfect fit for my my approach there so moving on to valve and maybe getting back to

electron microscopes uh you were there to do part of the vr headset that they're coming out with this year the vibe yeah so the work i did did lead into that, yes. What should I ask you about it? I mean, Valve is known for being awesome games and... Well, I have a question. All right. Half-Life 3. Yeah, yeah. I'm sorry.

Yeah, I mean, it's... You know, the hardware team probably knew as much about that as anyone on the outside even. Do you... What can you tell us about their vr system or what you worked on there yeah so uh the the deal with htc uh hadn't happened uh before you know i left before that whole thing got off the ground so i don't know the details about the vive so that's the new one yeah so the vive is is the thing that htc was building in conjunction with valve and so but before i left we didn't know anything about the HTC stuff.

And so did you work with Jerry Ellsworth on the, well, not CastAR, on the Valve? Whatever it was. What became CastAR. Yeah, yeah. And so Jerry was their first hardware engineer hire and I was their second in 2011. And Jerry and I worked pretty closely

and basically built their hardware lab up. So they had a table and two soldering irons when we started. Oh, my God. And they said, we want to build a hardware lab. You're like, we're going to need a microscope. You're done. Yeah. You're already there.

So it was quite a journey. Yeah, we set up their machine shop and built basically all of their infrastructure up until later. Do you have opinions on different approaches for vr versus augmented reality yeah vr is kind of a subset of ar and so if you have like a working pair of ar glasses or headset you could just put dark stuff in front of it and it would be a vr headset basically so it vr actually makes it a lot easier to accomplish all the solve some of these technical problems because the electronics don't have to be see-through so vr is in my opinion sort of a stepping stone on the way to full ar an augmented reality is when you can see through the electronics and it pops up like i can

walk through a room and here's a troll all right that's sort of cool but vr if i walked through the room i wouldn't be able to see my furniture. Yeah, exactly. And so, right, right. And then there's, to make it even more confusing, there's annotated reality, which is like having email notifications that don't interact with the environment very much.

That's like Google Glass. Yeah, exactly. I shouldn't ask about Google because you either can or can't say. I'll just say I don't know. It's okay. Why is, what happened to google glass and why are we you're in the wet can i don't know i don't know okay how about why are there so many vr and ar um systems coming out in the next six

months what happened somebody figured something years ago that made it so that everything became possible? Yeah, I think there's two big things. And so one is high-resolution OLED displays. And so Samsung basically controls the flow of high-resolution OLED cell phone-sized panels to the entire world. And the resolution eventually got high enough where if you put one of those in front of your face

with the lenses, it started to look pretty good. So prior to phones being, you know, in the neighborhood of 720p, it wasn't at all pretty. But now with high resolution cell phones, it's pretty good. So that was the major tech change that made VR kind of a thing. And then also Facebook decided to invest $2 billion. And so that basically changed the entire business landscape. Well, Valve and CastAR and even Microsoft were getting started before the big Facebook deal. I know. But no one was writing articles about it. And none of the articles had the words $2 billion

in it. Well, that does help. Oh, it helps a lot. And so, I mean, it's the way the tech sector works. It's very much sort of, you know, watching what other people are doing. And so if Facebook feels that it's worth $2 billion, that has a pretty big effect on the entire market was there any advance in optics at all or is it just pretty standard stuff that they just happened to figure they could de-warp with gpus and yeah so actually that that is that the one innovation that helped and that wasn't even specific to optics per se in fact it was almost almost the enabler that allowed crappier optics to still look better

than they should, basically. Right. Yeah. So I would say in terms of tech innovation, it was mostly the panels, and that's like OLED deposition technology and Samsung basically just being masters at display. And what about head tracking?

Did inertial sensors help with that, or did it all go camera-based? Why is there less hurling than there used to be? Yeah, so this is actually where Valve's major contribution was, or is, or at least was when I was there. Head tracking is another pretty big ingredient in virtual reality. And so a few people at Valve, including myself, developed this lighthouse tracking system, which is very precise, six-stuff, fixed one-to-one to the world tracking. And I brought a little prototype here, which doesn't work as much on a talk style show, but it is pretty cool, let me assure you.

Looks a bit like a flux capacitor. It really does look like something that you put on your desk and everybody says, what is that? And you say a time machine. Yeah. So I designed and built that. So it has some influence. So describe it. It's this is the lighthouse tracking system that, that is public.

So we talked about it at maker. I should say, we valve talked about it at maker fair this year and all the technical details are out. It's basically a, an angle to time-time converter, and it measures the angle from this box that sits on your desk to your headset. And then based on that, it can figure out where your headset is in space,

and all that 6DOF tracking information makes the VR experience a whole lot better. And so it can not—6DOF is six degrees of freedom. Right. So it's looking—when I say say six dof it's usually accelerometers and gyros oh right right sorry so yes in this case six dof would be six rotational axes and six translation or sorry three translational and three rotational so when you're in the headset and you turn your head from left to right it can track that as well as movements left and right cool and the resolution is very good in addition

to the samsung panels being pretty important to vr also having this really high resolution tracking was also a key thing yes because that's really i mean the tracking uh the video and the head being off sync was what made people sick yeah exactly so you mentioned not hurling anymore a lot of the work that was done at Valve was to combat simulator sickness. And so making sure that people are in a one-to-one environment. So when you're in virtual reality

and you turn your head, the world must move the way that your brain is expecting it to, or otherwise there's this discongruity and you feel sick. It's really bad. It is really bad. And some people can't even take it

for like five minutes. So I heard at Valve that when you had headsets on that were in VR mode, so they were dark, people would come up and startle you. Were you part of that? Oh, I don't know. I mean, I think we did some interesting experiments with forward-looking cameras, and we found out that doing like picture-in-p and picture is actually very effective. It first seems like it wouldn't work. So the idea is that in VR, most of your field of view is taken up by the game. But you might have a little tiny window in the corner, which is actually reality. And it's very mind bending, because you start to feel like, you know, usually you're sort of aware of the real world and then having a small screen in

your field of view, that's a screen like an LCD monitor or whatever. But now you're sort of aware of the real world and then having a small screen in your field of view that's a screen like an LCD monitor or whatever. But now you're in a virtual world and the screen is actually the real world. Yeah, I was thinking I've seen some other examples where people put cameras on the outside of an Oculus. So you're wearing the thing and then the camera sees what you would normally see and then they pipe it back, which seems like a lot of work for, you know, just take them off. But it seems like you could do not augmented reality but altered reality things and you know warp things oh yeah yeah it seems like a hallucination simulator is really easy now no no definitely and so yeah people have come up with other phrases to it's like mediated reality or basically there's like other r's basically of kinds of reality but

yeah the biggest problem is latency. And so getting the frame from the camera processed into your thing, through the game engine, back out, I mean, it's a lot of work. So you're telling me I can't live in a cartoon yet, but soon. Pretty close. I don't know where the cutting edge, I don't work on it anymore, so I'm not sure where the bleeding edge is. But definitely moving that data is is challenging for any computer system i uh i have a friend who went to valve who arrived shortly actually after you left

electrical engineer phil king he was a past guest too uh he mentioned that there are various stories of your mad scientist uh antics at valve and then he sent out an email there asking what I should ask you, but I didn't warn you about this ahead of time. So feel free to pass on any of these. But one of the things the Valve folks suggested I ask you about, tongue butt troller? Oh, yeah, yeah. That was a good one.

Is that one word? That is all one word that is all one word i think it was well they were separate products but then they were put together yeah yeah so i um i uh so at separate times totally unrelated did videos using um a tongue and and your rear end as game controllers it's natural and yeah and made videos of this so the reason is that um when we're thinking of the ar stuff you know your hands are busy with the real world you're opening doors you're driving your car you're doing stuff so you need a way of controlling a ui without using your driving your car oh for ar this was before vr oh sorry i was thinking of playing a game while driving my car that's probably not the application you

were thinking directions definitely massive. Definitely maps and directions. Or, you know, you're just walking down the street or sitting on your couch or whatever. But yeah, so, you know, what else could you have, right? So for VR and for gaming type things, basically just rocking in your chair side to side or fore and aft is actually pretty natural. I liked that.

So I took like a $20 Target scale, bathroom scale, and just wired the sensors up to basically the buttons on a game controller. I liked that. So I took a $20 Target scale, bathroom scale, and just wired the sensors up to basically the buttons on a game controller. Because those usually have four weight things. Yeah, and they gang them together in the original application, but I split them up and just took the signals out. And then if you leaned in your chair, you could control the game controller, basically like the D-pad.

It was okay. It was pretty good for motorcycle games, I'd say. Yeah, I could see motorcycle games i'd say it's a pretty good motorcycle game especially exactly so i think so that was the first project that i gave to my intern there i was like all right you're gonna do the butt troller as we called it and uh they all thought it was pretty dumb and you know it is kind of a dumb idea but i mean you know i don't know it's the whole how did you hurt yourself i fell off off a barstool slash motorcycle. Yeah, yeah. For active games, you know, it's very, like, it's not... Anyway. But then the tongue controller was also thinking about sort of interacting with a UI on the go. So if your hands are full and you're walking or you're sitting somewhere or something,

you have a lot of muscles in your tongue, you can actually do very accurate movements. But you can touch every tooth in your mouth individually with your tongue right and so you're gonna put like a retainer sort of thing with electronics in there that's exactly what it was it was a retainer with like an optical mouse camera having a flashback to buckaroo bonsai right now yeah and i had both wired and wireless versions so we're laughing about this we're laughing about this but for for disabled people that could be a huge you know having an accurate it is and so they already exist uh unfortunately so i i didn't i looked into it and the the state of the art for disabled folks is they actually a pierced magnet in the tongue oh really and then they put hall effect sensors like outside the

mouth and it you know i didn't try one but it doesn't probably not something you're gonna sell to gamers no it's a little too invasive. And so I wanted to have one that you could just put in and out like a retainer very quickly and have tongue control that way. And again, it worked. It was okay. You know, probably not going to take the market by storm, but you know, it was there.

Well, then with the tongue controller too, now with less piercing. Yeah, exactly. And so then they put it, you know, so Valve I don't think was too crazy about me doing videos you know about tongue butt controllers i i learned right away that the media is not going to be very friendly to you and so if they are going to get clicks out of running a story they're going to run it in a way that gets them the most clicks regardless of how you come out

valve engineer makes ridiculous that's precisely the headlines they went with it's like hey thanks guys and they picked like the most ugly still frame out of me showing this tongue controller it's like all right all right i'm cutting you guys off well the butt controller is essentially the we fit board the standing board yeah with two that has two load cells so it only has two loads no because you can do front and back it's got four all right yeah well i had i had rotation too because you could spin around in your bar still at the same time tipping but you know yeah okay so the next thing to ask you about turkey slash furby x-ray machine oh yeah yeah we did x-ray a furby well who hasn't really yeah as someone who took bb8 apart wait why do you have an x-ray machine actually that was donated by sophie kravitz actually and so i i

i think i came down here and packaged it up and fedexed this uh dental x-ray machine back up to valve there why did sophie have an x-ray machine uh it's x-ray machines all the way down yeah i think it was you know a family member or friend of hers had it or something you know how these things get passed around i guess the question is why don't i have an x-ray because you're about to move and i'm making you throw everything away oh my god that's terrible do you want anything yeah so i mean yeah x-rays are great you know and it's yeah what What did you find out with the Furby? There's a lot of stuff in there, actually. It's a pretty, you know, heavily produced object.

I mean, there's actually more stuff going on than you might think. If you have your own fluorescent screen, you can actually do your own live motion x-rays, which is pretty cool. So we had the Furby dancing around while we were shooting it with x-rays and looking at its, you know, guts moving on the screen. You do have a video of that,'t you i'm afraid not it's too dim for video but um yeah it is very dim to show up on video but i think we should probably try something like that again i've got another x-ray setup at my place of course yeah as you do as one does

is your place where you take the videos is that part of your home garage or do you have a separate site? Yeah, it is. It is my detached garage in Redwood City here. And so I spent about a year and a half looking for a place in the Bay Area based solely on how big its garage was and finally found this place. Do your neighbors have any idea? No. No. In fact, at some point, not one of my neighbors turned me in to the state radiological health board because of one of my x-ray videos, actually. And this guy kept leaving his business card in my door, like, call me, call me. And I was, you know, kind of working in Seattle at the time, so I had an excuse. I wasn't home at all.

And eventually I caught up with him, and he was like, well, you know, it looks like you were shooting x-rays off without any kind of an enclosure there. So, you know, and he offered if I built an enclosure for it and added a key switch, he actually offered to come by and certify my custom built x-ray machine. And then I would have had a legal custom built x-ray machine. Would have had. Well, I told him I, you know, threw it all out. You know, if I ever, if I ever got another well i told him i you know threw it all out you know if i ever if

i ever got another one i would you know i feel like we should roll back in the show just a few minutes but okay we're not going to but you know if anybody has some lead that they need to give ben i'm sure he needs it we're gonna be x-ray safe in the future let's just say at least lead line the whole garage yeah yeah that's that's good start that go wrong with that plan and so you must have x-rayed a turkey as well yeah i got a i got a i built an x-ray backscatter detector so at the time when i was commuting from here to seattle all the time i would go through these backscatter detectors every week and so i was very familiar at the airport yeah yeah san jose funny enough they

don't have them anymore because they're stupid turns out they don't work yeah and and or they couldn't prove that the x-ray dose wasn't insignificant it was very unclear how that whole thing happened but they're gone now but at the time i was kind of interested in how they work so i built my own backscatter detector and x-rayed it or backscatter detected a turkey with hiding an allen key in a sweatshirt that I put on the turkey. As one does with one's turkeys. That video actually didn't get as

many views as I thought. I thought that was good, especially because it was like at the peak of the TSA and like people were like, I'm gonna have to. You may need to repost that for Thanksgiving. I'm looking that one up as soon as we finish. Okay, so the turkey slash Furby really were two separate incidents.'m sad i thought that maybe there was the furby inside the turkey like a turf for be very careful with that it's not gonna go anywhere you want

yeah no it was definitely a separate i don't recall ever doing a turkey and a furby in the same video okay senator i have no recollection of that for me caffeine of extraction oh yeah so that actually is related to the supercritical stuff so one of the ways that they decaffeinate coffee now is to force supercritical co2 through it and all the caffeine dissolves away in the supercritical co2 and then when you you know suck all the excess2 out, you have decaffeinated coffee. I thought they just ran water over it and I was essentially getting second brewed. Yeah, so that's called like the Swiss method or something like that. And so people were concerned that there's too much industrial processing.

So one way is to run water over it. Then you extract the caffeine from the water, probably with supercritical CO2. Then you run the water back into the coffee to put all the flavor components back in that you took out the first time wow this is why decaf costs more and i know that and i'm okay with it but so the thing now i'm less excited about i know and so it's worse is that up until recently the solvent that they used instead of super critical co2 was methylene chloride paint stripper and so that was actually the decaffeinating solvent that they used that's good forical CO2 was methylene chloride paint stripper. And so that was actually the decaffeinating solvent that they used up until... That's good for you.

Yeah, well, I mean, in theory, there's none left. So at the end of this process, you presumably end up with a whole bunch of pure caffeine. I had a little powder pile of caffeine. I still got it. Okay. Yeah. I just was

gonna compare caffeine to other drugs It looked pretty So, you know, thumbnails are a way to go It's not something you'd want to ingest Yeah Oh Well, yeah, it was probably a lot of cups of coffee

So, you know It wouldn't kill you probably but it wouldn't be great either Let's just say I remember in school we used to have 101 nights at one of the dorms that i didn't live in so i didn't really participate a north no um and they would lay out a 10 by 10

and put shot glasses on each one and there was 101 tequila night which of course a lot of people got really sick on and you tried to get drink as many in the hundred as you could. Oh, this. But people actually got sicker with 101 Dr. Pepper night. Because there was so much caffeine and they were pounding them. And so it went really badly. There was a lot more hospital runs that night. Wow. Dr. Pepper. There might be other reasons beyond the caffeine. Well, I'm not sure that it was

only Dr. Pepper sometimes. Okay. You did have people with access to a modern chemistry lab. Aha. Yes. And more time than sense sometimes. Although, of course, we love you, Mud. And West Storm. So, Cookie Perfection, actually, I already knew about. That one I didn't need a hint from the Valve folks because one time I tweeted snarkily on Twitter about being an on-demand cookie source for someone,

Christopher, and somebody pointed me straight to an actual on-demand cookie source. Yeah. Do you keep your cookie-making robot on standby? And somebody pointed me straight to an actual on-demand cookie source. Yeah. Do you keep your cookie-making robot on standby with butter and sugar ready to go? No, sadly, I mean, it is, people like the idea so much, but of course, you can imagine keeping like the butter and the eggs refrigerated kind of in the robot at the same time it's ready to go all the time is actually pretty challenging.

In fact, more challenging than building the thing. that seems like it's within your skill set i think you should well you know again there's always these ideas these side projects start taking off and people are like oh you've got to you've got to commercialize that thing it's a cookie vending machine yeah people just push the button it bakes your cookies i know so we've but i have talked about it semi-seriously with people and you know it's one of those ideas that kicks around it never really got quite enough traction to get off the ground but it's you know it's there the future where mrs field is a robot actually i think i saw that in futurama yeah so yeah the customization angle is one where it's like you you can tell it exactly what you want like you know chocolate plus chocolate chips plus

peanut butter whatever is your combination and the fact that it's freshly baked dr pepper pure caffeine yeah well that was one of the things that when i bake cookies i i start with the ratio method where cookies are one part butter two parts sugar three part flour or two parts butter three parts sugar four parts flour and then you know just go from there right and so people come over and they get cookies and they want the recipe and i'm like i don't know i remember i know you just saw me make them i it was either one two three or two three four and now we're done and i don't know there was some amount of vanilla i don't know do you do that generalization of recipe or do you have one that works really well in your machine? My starting point was the recipe on the back of the toll house.

Yeah, it's a good starting point. I know. And then, you know, not surprisingly, they've honed this in over, you know, 100 years or whatever. So it's pretty good. The whole idea with the customization angle is that some people might want to tweak it a little bit like, no, I like more chocolate chips or no no i want a softer cookie or a crispier cookie or that sort of thing and then you could have you wouldn't have to bake a whole batch of like one person's desired thing you could split the batch and do like you know oh yeah half and half or even just one of each you know one of each style or something like that and i think the the really interesting idea is if you scaled it up to like production thing, what you do is put like a QR code.

This is actually a real use for a QR code, probably the first ever. And people could vote. Is this a QR code in frosting or something? No, no. So here's what you do. So you bake different cookies. And then when the consumer buys it, they can scan the QR code and say, I rate that cookie a 7.8 or whatever.

And then in aggregate, you can track which cookie recipe is getting rated higher and have like a real-time evolving recipe. You can even scale it towards certain demographics or certain geographical areas or something like that. I thought that was actually pretty cool. That is an inappropriate use of data science. Genetic cookie recipe. Yeah, constantly evolving. Yeah. Although I think I saw on Serious Eats where he essentially did that. Really? Where he incremented the amount of baking soda versus baking powder

and warm butter versus cold butter because it's a different mix method. It was in cake flour versus regular flour. I'll send you the infographic on that. It was very amusing to see all of the different paths. But yeah. So how long does it take for your robot to bake cookies? I never built the mixer. And so what it does is it dispenses all the ingredients into a plastic cup and then you have to mix them together. And so right away, people fall to me like,

the butter mixing. I mean, come on. I thought it automatically baked them. No, I'm afraid not. It's just an ingredient dispenser. Oh, I thought you need to add it to one of those industrial toaster machines that it rolls around. That was talked about quite a bit when we were thinking about commercializing the thing for real. But that's actually not nearly as bad as the mixer. The mixer is by far the worst, most difficult mechanical part of the entire thing. Would you like to see my KitchenAid? Well, so the problem

is mixing one cookie's worth of dough it's like ridiculous like more sticks to the spoon than you in the cuff right oh no well because sometimes they make small amounts of cookies because i don't like to have a lot no there's you can get one of the little uh small mixer ones that you use for soup uh it's do you remember what you call what that was called it's a hand blender the hand blender but there's not there's the one with the two uh things that go around it comes one with one thing that goes around and that works pretty well okay and there are ways to get the dough off automatically oh okay well okay so we're gonna have to cut this part of the show because we might need to patent all of it.

Not so much. And see, I was kidding about fire in the intro, but you do actually have a liquid nitrogen generator video. Yeah. What does that have to do with fire? I just, it was the opposite of fire. Yeah. So you built a way to generate, can you go tell us how you do that yeah that was another one of my early maker fair projects that got a fair bit of attention and um i don't even know how i found the original uh there's there's this cryo cooler the heart of the liquid nitrogen

generator is this device that gets very cold all by itself it's self-contained and i don't know how i knew that one of these devices existed in an industrial piece of equipment for cell phone stations, but it does. And so in a lot of cell phone towers, they need a really, really tight notch filter to filter out the frequencies coming in from the antennas. And so there's this superconducting notch filter and it actually cools its electronics down. Really? Yeah. So it's a soft filter that's superconductconducting and so they need to have a little cryo cooler in there that gets the whole thing down to liquid nitrogen temperatures but the funny thing is no one actually knew what was in the box and so on ebay they were going for like 200 bucks it's like you know junk surplus and i bought one

and pulled the cryo cooler out and had this thing that if you pulled it out it would actually air would condense and drip off the the cold part of this thing so i ended up shifting the whole market after i discussed after i you know told people that's what's in the box the price you know doubled and tripled on ebay and it remains that way till today and you shut yourself in the foot with that one how do you extract nitrogen specifically yeah there is a membrane that's selective. And so you can actually get nitrogen enriched air out of it. But even if you don't... It's still cold.

It's still very cold. And then if you, I mean, the oxygen is even more fun to play with, right? So yeah, it's not a bad thing to... I did the classic dump liquid oxygen onto your barbecue. And that's always a lot of fun. So there was a guy that like, you know, this technique on on the internet before long before youtube and he's got like a two gallon bucket of liquid oxygen and goes up to this smoldering barbecue and just dumps it in one go you know there's somebody he we have to introduce him to

yes yes there is there's nothing left of the barbecue like it's just the metal is melted like it's completely gone that's yes wow okay now i have all these this raises so many questions so many questions uh have you done any uh videos about fire um yeah i had a recent one where i had some flash paper like magician's flash paper which is nitrocellulose and i wanted to build it which is also what ping pong balls are made out of yes so quit using ping pong balls to make your leds less annoying oh is that anyone ever caught one on fire they won't burn super fast they burn very fast and it's really fun someone with an led actually had enough power to light one up.

Oh, it has happened mostly because there's a short elsewhere in the circuit. But yes, if you go buy a six pack of ping pong balls, take them outside and light them on fire. It's really cool. Exactly. And there's some videos of people putting hundreds of ping pong balls. We didn't suggest any of this. I said go outside.

Isn't there a disclaimer that's like, well, if you blow yourself up, you know, it's too bad, basically. What Ben just said. Don't try any of this at home. Yeah, basically. Or do try it at home, but with the fire extinguisher

and only one ping pong ball at a time. Yeah, that's reasonable. On a skewer. I didn't really appreciate the skewer, but I think that was because I had a beer first. I'm sorry, I'm on the phone with a lawyer. But it's true. I mean, burning stuff is fun. And so flash paper is basically just a very thin ping pong ball.

And so it has a high enough surface area where it burns really rapidly. And I wanted to see if I could do computation with the fire. So if you cut the strip into a specially made pattern and then light it, potentially you could have it burn in such a way that it does computation. So this sounds kind of silly, but let's say you had like an encoding system.

No, that is entropy winning. Fire is entropy winning. It can't. Well, so the energy is in there and then you're using it. So it seems like the energy is flowing downhill. So as it's on its way down, you could use it to do work, you know, computation maybe. And the nearest thing that I got to is one of these domino computers where you set up dominoes. And someone did do this on YouTube.

It was another Brady Harn video, actually. And by tipping the dominoes over, you can have it knock like a gate over. And then when the chain of dominoes gets around to that gate again, it's open and the signal doesn't go through. So you actually can build logic gates with dominoes. And you can do the same thing with fire. Basically, it burns the fuse out, and then it can't go through there again. So you could do that. But I wanted to have something more like superposition, where you put like a strip of paper on top of another one. And then the combination of the two burning would give you some kind of an additive effect. I'm not actually a data scientist, so it just sounded like a good idea.

And was it? I didn't get it to work. Sounds like building Minecraft computers in the real world. Exactly. It's like, there's got to be something there. And yeah, it's the energy source. Go back in time and suggest it to Babbage. Yeah. Yeah, gears, I don't know, man, but fire. Fire, I know. That's where it's the energy source and go back in time and suggest it to babbage yeah yeah gears

i don't know man but fire higher that's where it's at it makes for better video content there was a group that actually had cool readout and so what they did is they put metal ion salts on their strip of flash paper and then as it burns linearly it changes color because it's going through all the metal ion salts you know like purple for potassium and that sort of thing. Hey, and you know, coconut water has a lot of magnesium. Really? So you just wait for that to evaporate and it'll all be so much more fun. I don't think it has enough magnesium. I don't think it has that much magnesium.

I'd rather light my vitamins on fire. Huh. Well, actually, so the nitrate, when you do burn a ping pong ball, you notice that it has a yellow flame. Yes. That's mostly from sodium impurities left over from when they washed the nitrocellulose with sodium bicarb they neutralized the acid from the creation process and so i think the actually true nitrocellulose burns almost colorless that's mostly the sodium ion that you're seeing in those things kind of disappointing yeah i hate

things that burn colorless not only are they dangerous, they're just not that much fun. Hydrogen fires. Methanol fires. Yeah. Going back in time, as somebody just mentioned, how did you get the second scope? If you built the first electron microscope, how did you acquire the second? Yeah, a very generous viewer said, well, you know, it looks like, you know, you could use one of these. And he had been storing it in his apartment in Sweden since his university wanted to throw it out. So it's actually surprising how often a university or even an industrial research group wants to throw away, you know, a once million dollar piece of equipment.

It happens actually fairly regularly. And so he rescued it from the trash in Sweden and kept it in his apartment, but he was moving and had to get rid of it. So he said, do you want it? If you pay for shipping from Sweden, it's all yours. So how much would that sort of microscope cost now? It's a good question. The market is kind of weird for old, out-of- date scanning electron microscopes since no university would want it. But hobbyists would. So it's kind of a, but they're hard to move around. So they aren't easily sold. You might figure a few thousand dollars on eBay or Craigslist, something in that range. How big is it?

About the size of this desk that we're at, just about. So three by six, three by nine? Big. Yeah, big enough. And it weighs four or 500 pounds. It's on casters and, yeah. Requires cooling water. What makes, I mean, scanning electron microscopes,

I thought, you know, that was kind of, there are some newer technologies that can look at more atomic things. Right. But, I mean, once you get to scanning electron microscope, there's this wall there. How do they get better? Why is yours old instead of... Yeah, it's a good question.

A lot of it comes down to the source of the electrons. So the earliest source is just a hot tungsten wire. It's the same thing that vacuum tubes use in lots ncrts and the tungsten wire gets hot and it boils off electrons and you focus them down to a point but the problem is you can't make the tungsten infinitely sharp like it still has size to it so later on they came up with better and better methods that have like sharper and sharper points to launch the electrons off so the next revolution was lanthanum hexaboride cathodes, which is this beautiful purple crystal that has, you know, like a two micron point on it.

And that shoots the electrons off the tip of that. And then even that wasn't good enough. And so now the modern tech is field emission, which has like a, essentially a one atom thick tip that it launches electrons out of. Could you replace your tip? Maybe. It would require probably redesigning some of the

other parts in the column, but yeah, potentially. I even have some lanthanum hexabori that I was thinking of throwing in there, but the vacuum requirements are higher too. So you need an even bigger vacuum pump, you know. That's probably what most of the equipment size comes from. It's true. It's just generating the environment is the worst thing basically for a SEM. And you've looked at LP records. Yes. And little tiny, tiny fleas. Yes. What else have you looked at? What is the most surprising thing?

The most surprising thing. Probably DLP mirrors are pretty cool. So a DLP chip has these tiny little 10 micron mirrors on it. And each one has a hinge. So you can look at that under the sun. That's pretty cool to see like an entire array of these 10 micron mirrors all hinging. What I really wanted this, of course, was animated to see the thing. Well, unfortunately, at that scale, the flip takes about a microsecond or 10 microseconds or something on that range.

So you can't possibly view it in real time. You'd have to do it in stop motion and then freezing the mirror like making the mirror go to a partial position is not easy either so anything you've done i haven't looked at the microscope stuff but i'm assuming you've married this to modern digital capture stuff so yeah the yeah so the scope i have is like early 1980s vintage or even designed in the late 70s. And so their entire board is replaced by just code on a microcontroller. It's actually shocking how much stuff is in there. I mean, it's all discrete logic so that the scope in its original form has no firmware whatsoever.

It's entirely solid. And you can replace hundreds of chips, of course, with one library that takes, you know, a couple of lines in a microcontroller. When you used a Teensy? Yeah, I did. I used the Teensy LC. And so that's what's doing your capture? Yeah, so right now the scope is still taking care of its own scan generation, and the Teensy is just picking off the data stream. But eventually I want to do full scope control with a microcontroller. Will that let you go faster?

Yeah, then you could arbitrarily pick different scan rates. You could even have different scan patterns. There's no reason to raster it left to right. You could do spiral raster scans and all kinds of cool stuff. Why would you care? It's sort of a trick that MRI machines use. You can spiral scan and then you get more image data from the center of your image quicker. Oh, all right. Which is kind of neat. Yeah. And then you pump that out to a USB controller. Yeah, so I use the Teensy's own USB interface to do basically a serial port link back up to the computer. How's that working for you? Pretty good.

So the theoretical limit for USB full speed is like a megabyte per second. It's like 10 megabit or something, 12 megabit. And as it is, I'm not even close to that. So as fast as the ADC can sample the stream, it's not filling up the USB pipe. Would a more precise ADC give you better data? Where is your bottleneck here?

What's your noise limit, I guess? The SEM itself is actually limited in how um good of an image you can get out of it in a given amount of time right so if you scan faster you get less signal and it's an interesting trade-off between how good you want your image to look and how fast you want to go and the trade-off is actually much harder to deal with than people might think like you actually don't get very much signal out so to get one really sharp frame is about 10 seconds of scanning so that you know the bandwidth is of course no problem with that scan rate and so that isn't an adc problem that's just a how fast it can shoot off it is in fact yeah

it's a physics problem so if you just bombard the sample with more and more electrons you eventually destroy your sample so you can't actually extract more signal out of it without literally destroying it. Because at that size, the electrons are little tiny cannonballs. Yes, yeah. Little tiny cannonballs. It's the ultimate limit of quantum mechanics. Yeah, you've observed your target way too hard. Yeah, and you're counting basically every photon

or every electron that converts to a photon out of that thing. That's weird. I mean, when I think about electronics, I don't really think about electrons. Yeah, shot noise is a big problem. So when you're scanning fast, you're getting down to the absolute bottom limit. Well, I think we should, I want to keep you here pretty much all afternoon. But in case you don't have the time, we should talk about Hackaday. Sure.

Welcome to the judges' corner. Yeah, yeah. It's going to be great. Have you peeked? Are you looking at the things ahead of time? Are you waiting until we get our notice? No, I'm waiting until we get our notice so I can see them all at once. Are you looking forward to it?

Yeah, of course. No, it's going to be great. And have you looked at the judging criteria? I have. That's good, because I hadn't before. Well, it just recently came out, right? Actually, it was part of the rules. Oh, right. Of course. I knew that.

We should read the rules just to find out. So, the non-judge, is it different from last year? Yeah, there are some differences. Primarily, they wanted to focus on saving the world this year. And so they have some new judging criterias on that. And fewer or no judging criterias on how connected it is, which was a little tough to decide last year. Because, I mean, it connects to USB, that's connected, right? Or it connects to satellites, that's connected, right?

And so it was a little hard to say right on that one. But it's still open and wow factor are the same. This year, is the project reproducible and could be extended for other uses? And does it address a wide-ranging problem? Mm-hmm. So how long do you think each entry is going to take you? I'm the sort of person who would probably spend a long time on it. You know, I don't want to shortchange anyone, give everyone equal attention and everything.

So yeah, I'll probably look pretty carefully at them yeah i think we have 50 entries to look through each it's 100 entries are in the judging pool and each judge will get assigned 33 okay so on that order yeah 33 so you know i don't know how many minutes each but i think it's suspected to take a day or two to kind of work through them yeah yeah i think 50 I think 50 last year, it took me two, two and a half days. This year, they want us to give feedback, constructive criticism. That's a new thing. That is a new thing. Yes.

And I think it will be hard because you're going through all these things and it's hard not to say, but you weren't like the last person i looked at or that sort of thing and flipping through them so fast it's hard to give really good detailed criticism without coming off like a jerk yeah that's the uh age-old question of like interviews do you actually tell a candidate why they weren't hired most companies have decided not to of course but, but some do. But it is, I mean, I think that as judges, we really have to remember they are learning something that is always part of Hackaday. Right.

And they're doing this in public, which is terrifying. So we have to at least applaud their courage, even if they're making something that we don't deem socially good or technically good or even uh you know it's some other criteria yeah i mean that's i think it's going to be fine i think most people you know don't take it too seriously and so it's not going to be a disaster if their product doesn't make it but even getting to the finals is a pretty big honor oh it is in financially too yeah there you go yeah i mean they they have some nice prizes yeah yeah yeah we should talk about judging after we do the judging okay we could give them some feedback you know i don't think i don't think people want feedback from here i think they'd rather

get their feedback in private oh Oh, it's private feedback. Oh, that changes everything. Well, I think Hackaday gets to see it. I see. And I think there was some mention that in the previous round, not all of the feedback was nice, and so they didn't give it all back. Oh, wow. We have to be nice.

Wow. Or we may get censored. Christopher has taught me the ham sandwich method of constructive criticism. I'm pretty sure it's a turkey sandwich. All right. Where you have a compliment and then the criticism and then a compliment. Right. So you just have to make a sandwich.

I like bread, though. So, I mean, you know, what's wrong with, you know? Oh, and last year there were some that I don't think I could have said anything negative about. I was just like, this is amazing. Good job. Your hair looks nice today. You're fired. You forgot the other piece of bread.

This open face. Don't let the door hit you on the way out. So Andre from the Great White North north asks what tools should everybody have screwdrivers hammers pliers scanning electron microscopes what else do you recommend wow well for uh uh for what kind of work i mean basically any old packing any old building yeah for for electronics firmware okay like hacking you do yeah so i'm definitely more comfortable with hardware so anything i can do to make my software life easier i immediately latch on to so of course the teensy paul stoffergan's awesome libraries that save you

know hours upon days of time i absolutely love um the the usb stuff that he's done makes it so easy to send fast stuff and the ADC libraries and everything else is super helpful. So I would say if you're a hardware person, find a really, really easy to develop with stack. And so for my computer side of that sanding electron microscope project, I used processing.org,

which some people laugh at, but man, for prototyping graphics, I mean, it's like 10 lines of code and you've got a graphics window up. And so if you do this through Visual Studio, you'll have, I don't know, probably a few hundred lines of code to get a graphics window up. And so it's, by the time I downloaded Visual Studio, I already had a working prototype in processing.org. So I was like, yeah, I think I'll probably just take the easy way. So as far as tools goes, I would say get yourself a Teensy. It's always a good, super rapid prototyping base.

Cool. That's it? No, I have a long list of tools. I mean... Soldering iron, microscope. Yeah, I like Hakko soldering irons. That's kind of been my standby for a while um american amscope from amazon they yeah yeah they're really cheap but

they're really good too i mean the quality is just surprisingly high on those for the longest time when you went to the amscope um site on amazon it said it was a two-pack a two-pack it had a little two-pack symbol and we finally got one and i kept hoping there would be two in the pack and there was there was probably two eye pieces that was that's a binocular scope you think no no no i don't think so but i've ordered many of them and so i bought them you know for my home shop for for my lab at work, for lab at Valve and everything else. But yeah, they're really great and pretty reliable. Get yourself one of those. And also the hot air rework stations.

This is another funny thing where it's like the cheap ones are actually pretty good and the expensive ones are only marginally better. So you can get a hot air station for like 50 bucks on Amazon. And then the next one up is like a Hakko for like 800 or something. It's like, no, it just blows hot air. Really, that is not one we have. And it is something that I wanted recently. Very handy. I was even doing BGA and LGA rework at work and you don't need an expensive, you know, you can do it with a $50 hot air gun. It's a conspiracy of big solder i know what don't you have that you want oh that's a good yeah so you know a water jet cutter is really nice oh yeah so you know don't shoot small when you're faced with the question of what do you want i've you know seriously considered it hard

to build too yeah it is very hard to build and i have looked quite a bit into that of course you have oh yeah so there there is a guy so if you search for diy water jet there is of course someone out there who did do it it looks like a like a senior project or maybe like a thesis or something like that and it works he's even showing it cutting quarter inch aluminum so it's not a joke um it's just you know combining it with the cnc's controls and even compensating for taper of cut and all that sort of stuff is what you're really paying for with the big machines. And they're very, very energy hungry.

It's like a 30 horsepower motor that runs those things. I know. And most of the energy, of course, it's just gone into heating the water, but you do need that much power to actually cut even weak materials. So DIYing one would require starting with like a 30 horsepower induction motor and getting getting worse from there well and his only cost five thousand dollars i know it's an impressive build no i really liked it i think you're oh we're conflicting over the definition of only there well have you seen the price of actual commercial machines? Oh, yeah. It's six figures or above, right?

Well, I think some of them are a little less than six figures, but it's bad. Yeah, and they're very heavy, of course. I saw one run at the tech shop. It was really, really cool. The size of a small room, too. Oh, and it filled the whole room. It has a couple of super nice benefits. One, it doesn't heat up the cut zone,

so you can cut any materials that would otherwise be affected by a laser cutter. It can cut any hardness just about up to stone even. We were cutting glass. There you go, yeah. It can cut brittle materials like glass. You don't have to clamp the work, so you can cut brittle materials that would otherwise be destroyed by clamping. It can cut almost five axes. It's like wherever you aim the jet, it will cut.

So you can even do undercuts and things like that. And it's a pretty cool machine. Do you have a laser cutter? No, not at home. I'm tempted by those sometimes, but mostly because I would cut paper or foodstuffs. Yeah, it is tempting for that. I mean, in acrylic too, if you do a lot of like...

Yeah, the prototyping boxes. Yeah, exactly. Flat construction is super good, yeah. Yeah, I do have a lot of my yeah the prototyping boxes yeah exactly flat construction is super good yeah yeah i do have a lot of my prototypes live in cardboard boxes yeah our tennis ball containers hey it was sitting there you talk a lot about science and electronics together why the combination why is that what applied science is about um i guess mostly just personal interest i mean electronics are um electronics are sort of a way that people make a living with engineering that's much more common than sort of applying basic science

principles in demonstrations and so it's kind of nice to have a, like a really practical side of the channel and also a little bit more of a kind of flash demonstration, whiz bang kind of side of it. In fact, actually that in my most recent video, I talked a lot about firmware and electronics and some people, some people like it and some people don't. I think the channel has audiences that are both centered on much more of like electronics engineering, and then also just sort of like a topical science channel what did you say about firmware um i went into the details of how the the structure of the thing that i wrote for the teen c works and so it's not you know a

great feat of firmware or whatever but it is actually going fair a fair bit beyond like the typical arduino hobbyist crowd um the timing with the interrupt loops was pretty critical, and also just ordering. It uses the nested vector interrupt controller on the thing there. It ordered which priority was important and all that kind of stuff. Well, if somebody's trying to reproduce what you've done, that's certainly helpful to go into that little detail. Yeah, exactly.

And so I did post the code and everything, even as bad as it probably is and all that sort of stuff it was code it's a skinny electron microscope yeah um so i was wondering could you scan some cake i've always wanted to see what cake looks like cake yeah yeah you can scan cake get a metal coated but it'll it'll work um it'll be slightly dehydrated but that might affect the structure a little coated but it'll it'll work um it'll be slightly dehydrated but that might affect the structure a little bit but it'll more or less be okay there are some metallic coated cake sprinkles that's right actually those would probably work just fine as is do you get color out of your skin electron microscope nope it's just surface texture which is um you know the electrons don't

know what the color of the thing that they're hitting, but they do know, you know, they are affected by surface topology. And that's a benefit and a curse. Like, it's nice that you actually aren't distracted by surface, like, coloration or absorptivity. But if you're interested in that, of course, you don't get it. So, yeah. What energy is the photons that are emitted? Are they invisible or? The electrons coming down the column, like the incoming.

So once they hit, presumably it knocks off a photon and that's what you detect? It's actually more electrons. And so it makes like an electron cascade. Okay. And there's a couple, two distinct signals that you get. One is actually very slow moving electrons. So the ones coming in are maybe 10 kilovolts. And then they create this plume of like low energy like 2 to 50

volt electrons so they're basically if the ones coming in are going three tenths the speed of light right the ones that come off are just swimming i mean just drifting around in comparison even though they're still going thousands of miles an hour in comparison they're just drifting and so then they're attracted over to detector and you get the signal out that way okay so you interrogate each yeah each spot okay it sounds very similar to laser scanning yeah exactly it's it's and so you build up the image pixel by pixel basically it's a lot of fun yes yes okay so you've been to our office when can we come over to yours yeah yeah before you guys move out of town or even after if you want to come

back for a visit or something yeah yeah it good. I really should put together more kind of shop open houses, but I haven't yet to date. Ooh, party at Ben's house. Yeah. Applied Science Live. Yeah. Yeah, that'd be a good one. So have you found that having the YouTube channel is an enabler for other things? Oh, yeah, totally.

In fact, it's by far the single best thing that I've ever done in my entire existence. No joke. I mean, so the YouTube channel is how I basically got the job at Valve and the job at Google. So yeah, it's been pretty cool. It's a case of if you build it, people will come. I think that it, yeah, it's a case of just basically sharing with what you have. People always say, don't worry if your code is ugly, just post it. And it's kind of the same idea.

It's like, well, just show what you're working on, and then people can see it and comment on it and don't worry about the criticism, basically. And keep doing. Yeah, yeah. What do you think has made you successful? I mean, there's a lot of YouTube channels out there. Oh, probably sticking with a niche audience.

Like I don't, I've had a lot of temptation, well, not a lot, some temptation to make the channel much more topical, like a more generalist sort of science channel. So I purposefully leave lots of details in and I purposefully don't have animations and I purposefully don't do topics that I know that a lot of people

would like but you know aren't interesting to a small group of people so even though 170,000 subscribers you know sounds like a lot it actually is niche compared to like Smarter Every Day or Veritasium or one of those guys I like their channels too it's just sort of a different audience yeah and it's fun to do what you do yeah and not necessarily to try to do what other people do yeah i think exactly so that is in fact that's a that applies to like all kinds of businesses if you try to um worry too much about like you know what you think people want then you know you've kind of lost your mission whereas if you just concentrate on what you think is good then you have a coherent vision of what's going on i agree with that do you have any

other advice for people interested in starting youtube channels um no i think they should do it i think it's a great medium it's a good way to you know get exposure it's a great way to get comments i'm actually really happy with the comment section on my videos and i do almost no policing like i've deleted maybe three or four comments ever in the history of my channel. Otherwise, everything that everyone writes is up there. So that's been super cool. I think if I were giving advice to someone, I would say, keep it short and edit.

And that's about it. Cool. Well, I have certainly not kept this very short. Ah, sorry. And I don't think there'll be much editing but so that we can have a little bit more afternoon do you have any last thoughts you'd like to leave us with hmm well i knew this was coming and so i've been thinking sort of off and on throughout the uh this uh show here what my final thought was going to be i actually thought

that the questions from valve were going to be much more invasive than they were. That was barely scratching the surface. I thought that was just the warm-up to sort of, you know. So that's my final thought. Now I want to know, what should your Valve coworkers have asked about? I don't know. I didn't know what to expect,

so I was expecting that preparing for the worst hoping for the best i warned ben that he didn't have to answer those questions if tongue butt troller was embarrassing but oh no he answered all of them and so yeah apparently there were some that were embarrassing that i didn't get yeah it could have gone much further south and phil you have disappointed me. Next time, do better. But for now, my guest has been Ben Krasnow, hardware engineer at Google Life Sciences, whose Applied Science YouTube channel is pretty amazing.

If you don't learn something neat watching one of his videos, double your money back. Don't forget about the San Franciscoisco hackaday conference in november and propose a talk today thank you ben for being on of course thank you to christopher for producing and co-hosting and of course thank you for listening if you'd like to say hello or send us robots wow that must have been copied from last week. But, you know, if you want to send us robots, sure. Do hit the contact link on Embedded.fm or email show at Embedded.fm. We will talk to you next week.

But I do have a final thought before we go. This one is actually a quote from Ben Krasnow, which he talked about building the liquid nitrogen generator. Building things to practice building things. when she talked about building the liquid nitrogen generator. Building things to practice building things. I'm sure I could have bought a commercial air dryer for cheaper and quicker, but then you don't learn anything.