Transcript
John S: Welcome to the business of machining episode number 408. My name is John Saunders.
johngrimsmo: And my name is John Grimsmo.
John S: And we have done this this podcast for so many years just to get to today's episode, which is not sponsored by Willem & McAdel. But in so many respects, it is because if there's one machine that I just love, it's the 408 MT from Willem & McAdel, which is a mill that has some lathe turning ability.
John S: And ours has been an absolute beast. And I think yours is cranking too.
johngrimsmo: Mm-hmm.
John S: what's ah What's on yours right now?
johngrimsmo: So honestly, mine is currently off because we made enough clips because for a while.
John S: There you go.
johngrimsmo: um And then we're just about ready to need more and turn it back on and get back into production.
John S: Yeah.
johngrimsmo: But i mean my guys have been busy doing other stuff. I've been busy doing other stuff. So I kind of don't feel bad. But I do look at it longingly sometimes and go, it's kind of lonely over there.
John S: Yeah. That... I've heard the the machines like the Willowman, there's competitors as well. they are They do great when they're run. In other words, it scares me to think about, oh, we'll leave it idle for three days or three weeks, which I hear you totally like, makes sense, don't reproduce, et cetera.
johngrimsmo: e
John S: But you haven't had any issues just picking up in the middle of...
johngrimsmo: I mean, mine's been off for years because I got it in 21.
John S: yeah,
johngrimsmo: um i I don't know. I've certainly had issues. I've had to replace various things and um it's an older 20 year old machine.
John S: Yeah.
johngrimsmo: So I don't know. It's a good question.
John S: But no, I mean, if you, you wouldn't, if we, if we hung up and you went over turned on, you wouldn't expect like, Oh, I might have to clear an alarm or figure out if there's a,
johngrimsmo: Yeah.
johngrimsmo: There's like this B-axis ref alarm that just means you have to turn it on twice.
John S: Oh, I think we used to have that too.
johngrimsmo: Yeah.
John S: That's, that's funny.
johngrimsmo: I remember Marcus said there is a fix for it, but I don't care.
John S: I vaguely remember that.
John S: Yeah. Yeah.
johngrimsmo: Yeah.
John S: Right. Interesting.
johngrimsmo: But your new machine, like totally solid. Rockstar, no no problems.
John S: No. ah
johngrimsmo: Yeah.
John S: Well, I shouldn't say that. Grant, ironically, was at PMTS, that trade show, yesterday, so the so it didn't run yesterday.
johngrimsmo: Oh, good.
John S: um But, and he, no, he that's not true. Like, we've definitely had some tool changer things. He's worked out with Florian. the We actually just fatigued through, not even fatigued through, that's wrong, abraded through an airline with a chip, like ah an exposed airline.
John S: Yeah.
johngrimsmo: ah
John S: And so Grant was one easy fix. he did He found it and then was able to, ah think, trim it and just shorten it up. And I think he already had an idea for a cover, but I was thinking like, well, you could even do ah maybe a better option, like even some heat sh shrink tubing to give it an extra layer.
John S: Although, frankly, chips don't run through any of that stuff over time or or whatever.
johngrimsmo: Yeah.
John S: So...
johngrimsmo: Well, there's about 600 airlines in this machine.
John S: up
John S: Well, these are, I think, feeding the laser. So it's an actual airline that's line of sight to the cutting tool, if you will.
johngrimsmo: Interesting. Yeah.
John S: ah But no, it's been it's been great. And my only thing I've said before is I wish more... I wish I had more like it. Like, I wish people... like There's been, like, these sort of unspecific rumors that Brother will come out with something that's Willemann-ish.
John S: But, yeah...
johngrimsmo: yeah
johngrimsmo: Well, like you see these medical shops, usually medical, that have a sea of Willamins's.
John S: Yep.
johngrimsmo: Like that's where mine came from. Yours too, your old one too probably. like um Because it's one of those things, like you get to like it and you build your whole workflow around it you're like, this thing can make very specific but very cool parts fairly quickly.
johngrimsmo: um And yeah, people just like them. I mean, CJ's up to three now, if I believe.
John S: Yeah, right.
johngrimsmo: am Am I right about that?
John S: I think so. But just the flexibility of in today's day and age where we talk so much about automation and what you know basically what can you get out of the time you have in the day in the shop, that our ability to flip, um switch out quick vise, we don't switch tools ever.
johngrimsmo: Yep.
johngrimsmo: Yep.
John S: And so it's just a question of really the subsequent device, maybe the main call Chuck, and she just runs.
johngrimsmo: Yep.
John S: It's great.
johngrimsmo: yep it It is a bar-fed 5-axis. You know?
John S: Yes, exactly.
johngrimsmo: And it's it's one of the coolest machines out there. um It's kind of in its own league. Yep. Yeah, love it.
John S: Which is a segue to where I'm struggling, which is and some of its self-imposed because most most of March I spent, took some time off, but then also just doing stuff, which is totally fine. Except in February, I had made some really good progress on learning robotics.
johngrimsmo: Mmm.
John S: um We have the UR on the Williman and then I have the Tormach Z6 in a sandbox here. And, um, By no means am I frustrated or giving up, but it's kind of that, okay, you had came into this project with a lot of optimism and confidence. And um i want to pick a simple task. We get 500 or a thousand pieces a time of round bar delivered to us, cut into 20 inch sticks or something.
johngrimsmo: Oh,
John S: So it's a very, it's it's slightly bigger than than half inch, 13, 14 millimeter. So each piece is very lightweight, no big deal. But they're each one we de-mag by hand.
John S: We ground four at a time, but I thought, okay,
johngrimsmo: ah wow.
John S: That's not something I need to automate, and I wouldn't spend five figures to build a cell that just did that. But that's a good project. Like, perfect example of a manual labor. um The material is usually delivered to us stacked, although sometimes it's in ah sometimes it's spaghetti, but usually it's stacked, and we could be more deliberate with our vendor on that.
John S: So, $40.40.
johngrimsmo: What's the steel?
John S: $40.40. sort of ah four for you
johngrimsmo: Okay.
John S: And so there's no single point of non-starting point here. Like if it's delivered in a bin, you've got a g gripper, the gripper can grab the part, lift it up out, and then run it across a demagger and then put it in a different bin. Great. You could even...
John S: cheat by taking the time to put them into a feed hopper, like uh, think of a funnel, except instead of a funnel coming to a point, it just comes to a slot because that way the, they're, they're all, at the but there's different ways you could do it.
johngrimsmo: Yeah. good
John S: But, um, And some of this, frankly, might be easier with UR. haven't touch spent as much time with their GUI, but they have some better pattern sorting put away tasks.
John S: But um I'm kind of at this point where oh man, like how do you find the material? You can't, the camera systems, you could do this if you spent your life, six months of your life focus on it.
johngrimsmo: Yeah.
John S: But it's kind of one those examples where it's just like, doesn't, this isn't great, so.
johngrimsmo: juices not juice not worth the squeeze kind of thing.
John S: Right, which was part of me trying to do it both myself and it's kind of a proxy for like, why don't you see robots all over the place?
johngrimsmo: Yeah.
John S: um They're not particularly expensive. Totally get it if you have a rack of 15 fixtures or 15 grip points like the Auroa and all you're doing is moving from one of those 15 points to one other point.
John S: um Totally different animal, but um yeah, it just seems... frustrating that that's not the barriers are still fairly high in my opinion on that
johngrimsmo: is Well, my Auroa doesn't have a robot arm inside it. It's a up, down, rotate, front, back.
John S: yes sorry it's the shuttle and out thing totally totally right
johngrimsmo: Yeah, yeah. um But to your point, because it's simpler and that's all it has to do. it It just has to go up, pick the rack, move forward, pull out, rotate to the next position.
johngrimsmo: And it's brilliant. It's like simple, but yeah.
John S: Yeah. Yeah.
johngrimsmo: Sometimes I stare at the Auroa and I was like, you know, I think about building a robot arm cell system for some speedios or something like that. i'm like, they put a lot of thought into making this a ROA. Like they had to solve a lot of problems that I, you know, it would take a lot of time to solve.
johngrimsmo: Um, if you tackle this yourself and not that i'm against doing it one day, but it's, it's a lot of work.
John S: Yeah.
johngrimsmo: You know, what do I want to spend my time doing?
John S: Right. Right. and know Arguably, it's a software company that makes hardware to complement what they want to do because that's where the mojo is. The hardware is not particularly innovative or complicated.
johngrimsmo: Yep.
johngrimsmo: Yeah.
John S: um remember seeing a master of MAM studying it and then realizing it's got that track motion. if forget how to describe it, but one chain pulley shuttles out a slide and the other slide is linked to it so that they you get kind of double with extended motion.
John S: So when the bottom slide moves out three inches, the top one moves out a total of six and they aren't stacked.
johngrimsmo: Yep.
John S: So you can... You can have this telescoping effect to go in and grab something put it away.
johngrimsmo: yeah That's how the Aurora works.
John S: Yeah, super cool, right?
johngrimsmo: Yeah. Yeah. It's weird to see. It's kind of trippy.
John S: Mm-hmm.
johngrimsmo: Yeah.
John S: Yeah.
johngrimsmo: So you're not going to robot all of the things because you have robots now.
John S: Correct.
johngrimsmo: um Do you have, like, what's your next theoretical project for either the Tormac robot or like, like what do you want to do next?
John S: The next step is to be a surgeon, which I have not chosen to not do, like in the kind of spirit of don't don't pretend to yourself. You can do something you just don't have time for it right now. and In a good way. We're doing some other great stuff right now.
John S: um but I want to start by just picking up a piece of material. It's about the size of a deck of playing cards, move it over to pneumatic vice, close the vice, put it back and do that 300 times.
John S: And if it doesn't fail, we'll we'll try the next thing.
johngrimsmo: Yeah.
John S: If it does fail, it's like, hey, what happened? Where did I go wrong here? um So um just building off that confidence of knowing what you've been through. the goal would be to make a better decision on buying and integrating a automated five access that can do In a perfect world, it can do both material loading and work-holding loading. We've talked about this before, so I don't want to belabor the point.
John S: um And their shortest the shortest list I could say on three options, one would be complete total DIY, probably a mistake, but it's an option.
John S: Two would be somebody who's proven in this specific space, which I think the only company I've heard of is Lights Out, and NFG, that are really doing this.
johngrimsmo: Yep. Yeah. Mm-hmm.
John S: um or something that's truly turnkey and very, very robust, which is the Hermla RSO5.
johngrimsmo: yeah
John S: um But even that, you need to answer a bunch of questions to them about the parts, the drawer size, the configuration, the end effectors. don't have the answers to those right now. Cause I don't know if we can do certain things.
John S: Do I want to buy, i can ramble here. I won't. You could do, but you buy different shaped material and window machine it? Or how do you opt one, opt two? Do you want to lay it flat? um These things have much, I mean, it's like you and me developing a product these days.
John S: It's just much more cascading long-term effects than the garage John's had.
johngrimsmo: Yep. Yep. ye
johngrimsmo: and And you try to put all your engineering experience brain power into it upfront, but then inevitably you come up with all these roadblocks throughout the way that you could never have anticipated before.
johngrimsmo: And sometimes that derails the entire purchase or something.
John S: Yeah, right.
johngrimsmo: So yeah, I'm definitely more hesitant on investing in, you know say anything over $10,000 to make sure it's gonna do exactly what I want and meet my expectations and actually you know succeed.
John S: Yeah.
John S: Yeah.
John S: The last thing I'll say that is kind of cool, there is some, and this is, I guess this is why I'm wanting to be a little bit of a nerd on a robot is you can build sort of assets that become reusable. So you can switch between user frame and tool frame. So it's kind of like, um,
John S: You don't need to reinvent the wheel with every program. and And what's cool is you can tell the robot, hey, I want you to get to approximately this point. And you can control that with the the angular and go angular motion of all six. And then you can switch to a much more precise mode where it's going to switch into a relative from that distance mode. And it gives you more functionality. And how you jog it, you can switch because if you're thinking it, if you pretend my arm is the robot. If I'm pointing straight in front of me, then let's say that Y is you know where I'm looking.
John S: But then if I move my arm to the side, like I'm turn signaling on a bike, well, what is that now? I don't care if I'm just programming a pass-through point and it can kind of handle it on its own, but I might want to jog from this point, in which case it's way more natural to think about, do you want that to be Y positive or X positive? And so you can actually go to there and then switch your offset to being a local offset to that new location.
John S: So you're jogging in a natural
johngrimsmo: Okay.
John S: instance and it actually will jog and just like tool center point control jogging you'll actually jog and move all the axes it needs to synthetically create that motion which is freaking awesome
johngrimsmo: Yeah.
johngrimsmo: This is on the Tormach robot.
John S: Yeah, it does not know you are as well, yes, exactly.
johngrimsmo: Yeah, similar, but yeah, yeah.
John S: z i sorry Z is probably the easier one. Normally Z is just going to go down toward the ground, but like if I'm flipped the robot arm over um or to the side, maybe to the side would be a better example.
John S: Z might be actually moving toward like as God forbid the L word lathe, Z might be moving toward your part that's horizontal to the ground, right?
johngrimsmo: hmm weird
John S: Yeah, right.
johngrimsmo: so were robots like prior to you are come and hit the scene and they were all plc programmed right and you'd have to know all these touch points and like figure it out i don't know
John S: I'm not the right person to answer that. Um, I do believe FANUC robots with the arm pendants had a waypoint modes where you could just go to certain areas and then say, go from here to here. So it'd be almost like the a bubble sphere gets at this point then go to this point.
johngrimsmo: Yeah.
John S: Um, yeah, I bet don't know.
John S: Yeah.
johngrimsmo: Let's see. I've been finalizing... um So the hard milling on our fjell more geometrically to get it to do exactly what I want so that it flips out perfectly.
John S: Yes.
johngrimsmo: um Because we've been having some problems where some knives don't want to open. Like that when you try to flip it out, it's stuck.
John S: OK.
johngrimsmo: And we're like, nothing changed. It's the same measurement, same everything else. as you know What's different about this one than the other one? And since we have the button for the button lock, we can swap buttons, we can test, we can tune, we can do all this stuff.
John S: Mm-hmm.
johngrimsmo: And I finally, i basically, this little tiny feature that's about maybe, you know, 0.150 inch diameter, like very small um feature under the microscope, you can start to see what's going on. That's where the button engages into the blade and causes your detent.
johngrimsmo: um
johngrimsmo: I was able to machine a one thou radius, 3D machine a one thou radius on the edge of this corner.
John S: o Okay
johngrimsmo: See if I can do it. um And I basically programmed the one or like modeled the one thought radius and then used 3D contour, I guess, to do two tenth steps.
John S: Yep.
johngrimsmo: And then work my way up with a little corner radius end mill. And i was like, I mean, i think it'll work. Even under the microscope at full zoom, um I can almost not tell that there's a one thou radius, but there is.
John S: Yeah, right.
johngrimsmo: So I made a two thou radius and that's very visible. And somewhere in that range is the sweet spot of exactly what we need because that having it be a sharp edge was causing us problems. And then the tumbler would actually smush a tiny burr into the inside of the pocket because it was a sharp edge and now it's not, which would grab onto the button even more.
John S: Oh.
johngrimsmo: And then
John S: Yeah, right, right,
johngrimsmo: it This led to such a rabbit hole of like the specific heat treat on our buttons. You know, how do we go harder? Because if the button is denting um from this sharp edge, et cetera, et cetera. And then it also led to some geometrical anomalies that were basically a decision I made like a year ago and never looked back on.
John S: right. Okay.
johngrimsmo: um Like the the circular shape of how the button engages with that pocket on the blade. And they just didn't match properly. So anyway, all these things started to add up and the past week has been spent further understanding and eliminating every one of those one by one by one.
John S: The button is done this is a turn part.
johngrimsmo: Turn part, yeah.
John S: and But you're surfacing this.
johngrimsmo: Surfacing the part of the blade where the button engages with.
John S: Excuse me. Got it.
johngrimsmo: Yeah, this is this is a visual conversation that's much easier to show than to talk, but yeah.
John S: Oh, yeah, sure.
johngrimsmo: but yeah
John S: Huh.
johngrimsmo: It's all coming together.
John S: And if you want to fill it, not a chamfer, like just you want chamfer the chamfer tool.
johngrimsmo: Correct.
John S: Interesting.
johngrimsmo: Correct. And I do not want to use two separate tools because there's always a mismatch, no matter how you touch them off, no matter whatever. Tool to tool, touch off to touch off.
John S: yeah
johngrimsmo: So I can't use two tools. If I can use one tool, and it's consistent, high quality Japanese tool, Moldino, that's like the radius is what they say it is, and everything's accurate.
johngrimsmo: um then maybe I can actually surface a 1,000 radius and I can. Like that's insanely small and it worked and it worked great.
John S: It's tiny. Yes, that's crazy.
johngrimsmo: But that also leads to what if you're, blade is a few tenths thicker what if your tool touches off a few tenths different than before or like that one thou disappears real fast and becomes either an undercut or a nothing um so there's a bit more probing checks that i put into place right now and uh as of this morning i think it's working fantastically well so yeah the first hundred we had some variability um
John S: Oh yeah.
John S: Yeah.
John S: Good. That's awesome.
johngrimsmo: And some of them Eric had to like massage with a machine stone, like to, and that's just not the way to go.
John S: Yeah.
johngrimsmo: So.
John S: And these are doing this in a hardened state and the idea with ah the 1,000 rad is it helps the transition or helps get rid of the burr.
johngrimsmo: both. It helps get rid of the burr. The tooth I read is I think a little bit too big because now the transition isn't as snappy when you flip the blade.
John S: oh
johngrimsmo: um going to pull it back a little bit, but the fact that I can pull it from two to one and a half thou is like kind of wild.
John S: Yeah.
John S: Yeah. That's awesome.
johngrimsmo: Yeah. And then the, this little pocket that we've mentioned on the blade, the depth of that pocket is critical to how much snap the blade has, how much it feels when you, when you flip it.
johngrimsmo: um And literally a change of four tenths in depth is very feelable in the detent, the flipping action of this knife.
John S: Right.
johngrimsmo: So that it has to be, insanely consistent, and we learned that the first 100, we had a range of a thou, sometimes more than a thou, and they would be all over the place.
John S: Right. Most makes me think you want to make these two mating parts on the same fixture on the same machine at the same time, but they're not one to turn part.
johngrimsmo: Yeah, exactly.
John S: So you can't really do that. Although
johngrimsmo: Yep, so you make it, it's made on the tortoise, very accurate, very consistent, holding great diameter. And then, so you just have to deal with the blade side.
John S: Okay. Got it.
johngrimsmo: So that's where I'm at now. but
John S: I just finished an episode on the Acquired podcast, which they tend to do these like two to four hour long form type technology.
johngrimsmo: the Rolex one?
John S: Yes.
johngrimsmo: Somebody mentioned it to me.
John S: You listen to it.
johngrimsmo: They're like, I know Saunders likes the Acquired podcast. You've got to listen to the Rolex episode. I haven't listened to it yet.
John S: So I fully expected to try it out and turn it off because just to be blunt, Rolex has nothing for me. Like i just nothing.
johngrimsmo: Yep.
John S: um But And in some respects, I am much more appreciative of what the brand is because Punchline, they make... Well, the numbers are it private and they don't they don't share, but the there's some pretty good sleuthing and they make probably a million watches a month.
John S: So 30, yes, 30,000 a day, which is ties exactly into why I personally don't care for it because it's kind of that like...
johngrimsmo: A month.
johngrimsmo: Yeah.
John S: why would you spend five figures on a watch that they've done a great job at making you think, and I've now gonna be rude and like offend people that love Rolexes, but like, it's not luxury. They want you to think it is and it's not a wait list, but like they make, they sell,
John S: they sell a thousand of these watches every day um or something like that. yeah Even if I'm off 20%, like whatever, is it's there's nothing like, I think they started to talk about the truly like Richard Millet and Patek Adam Mars, Piquet, like those are like 5% of that volume, if you will.
johngrimsmo: Yeah. who
johngrimsmo: Definitely.
John S: But forget about all that, the manufacturing processes of, and and and there's also, I think some irony that Rolexes aren't even that great at keeping time, but nevertheless, like still, it is a robust, durable, but you know, incredible pride in manufacturing piece of equipment that is made in higher volume than, you know, almost anything else I know of ah and that in terms of that level of intricacy and so forth.
johngrimsmo: the
John S: It's pretty cool.
johngrimsmo: Good episode.
John S: listen to it.
johngrimsmo: Yeah, yeah, i think I will.
John S: I mean, it's not your jam, it's not your jam. I do it in time and a half.
johngrimsmo: Yeah.
John S: But ah yes, I mean, I would recommend it.
johngrimsmo: Sweet going on the list.
johngrimsmo: Yeah, they have, I'm not mistaken, Willemann makes custom machines for them, like custom designs, right?
John S: They wouldn't say, but but there may or may not be companies that use machines that are also totally debadged.
johngrimsmo: Maybe. i thought I've seen pictures.
John S: um
johngrimsmo: Yeah.
John S: So yeah.
johngrimsmo: Yeah.
John S: yeah
johngrimsmo: Yeah, it's, ah they've got figured out. That's for sure.
John S: Yeah. Cool, right?
johngrimsmo: Yeah, it's that's really cool. I do respect that.
John S: Yeah. It makes me, ah it's ah it's a big piece of humble humble pie. Like I'm proud of what we built, but that's, um it's a whole nother level, John.
John S: Like I don't, I, yeah.
johngrimsmo: Yep, yep. That's many other levels higher than we're at. um But can we learn from that? Can we not just be inspired, but can we actually like use that to drive us to make better decisions kind of thing?
John S: Yeah.
John S: Which I'll throw this out as a wear random segue. There was also an episode dated on Nike. Not exactly a technology company, but I wasn't complaining because it was also a great episode. And there are these Nike 10 rules that are, I guess, famous. I've heard of them to anything, but they're.
John S: There are 10 rules that seem to have transcended just the internal corporate culture. um And one of the things that I have been chewing on, because I'm not sure what to make of it, was, well, okay, a couple of them to set the tone. Number one, our businesses change.
John S: Number two, we're on offense all the time. Like Nike's aggressive. They are doing stuff that may be... They're just, they're aggressive. ah The number three is the one that jumped out at me. Perfect results count, not a perfect process. Break the rules, fight the law.
John S: bunch of ways you can take that, but you and I have talked so much about goals and products that we make being the byproduct of the internal recipe, the process, the team, the culture.
johngrimsmo: Yeah.
johngrimsmo: Process, yeah. Yeah.
John S: They're sort of saying, F it all, figure out, you know, get the result perfect, quit worrying about all the process stuff. And I'm not sure where I fall on this.
johngrimsmo: Sure, sure. I think ah something I've been thinking about is do whatever it takes to get the results that you want and then optimize.
John S: For sure. Absolutely. That's a ah great takeaway of, of and you and talked so much about that, like, figure out how to get it to work hard mill the thing or, or, or do it more steps, but know that you've got one.
johngrimsmo: Yep, yep.
John S: That's going to be great. But that's also sort of talking about the process.
johngrimsmo: It is, yeah. Yep.
John S: you You know what I mean? And again, not to be pedantic about it, but um I also like the breakthroughs fight the law. I kind of like, i don't really care. I'm not going to, you know, not talking about endorsing illegal activity, but like, just do what you need to do.
John S: Like make it work, figure it out.
johngrimsmo: Well, there's all these all these conventions, all these you know rules to for machining that you're like, why?
John S: Yeah.
johngrimsmo: Sometimes it's real. Sometimes you have to find out that you're not limited 150 surface feet for certain materials.
John S: Yeah.
johngrimsmo: Because you're like, no, like,
John S: Yeah. Yeah. and And look, joking aside about my disdain for lathe, there are a ton of things that we do um on ma machine centers with the right machine centers, the right tooling, recipes, and work holding that mean i don't need a lathe. It works great.
John S: Yeah.
johngrimsmo: So why why do you have ah hundreds of 20 inch bars of 15 millimeter material if you don't need lads?
John S: I didn't think we're going the pen business. Okay.
johngrimsmo: Or is this strip loaded in a horizontal?
John S: We make ah almost all of our Modvice small accessories that are square are made out round bar.
johngrimsmo: You're one of those guys. No, I like that.
John S: Yeah, works great.
johngrimsmo: um There's benefit to it, right? it's cheaper sometimes. It's a
John S: Cheaper, it's easier to actually, it's way easier to hold. It's just more reliable.
johngrimsmo: interesting.
John S: more consistent. Yeah, the grain direction, a whole bunch of reasons.
johngrimsmo: Yeah, yeah.
John S: Yeah.
johngrimsmo: Cool.
John S: Price is a big one though, for sure.
johngrimsmo: Yeah, it's like significantly cheaper. I remember Ken talking about that years ago. Mm-hmm.
John S: Yeah. And there's a bunch of alloy issues of availability and steel issues when you get into certain sizes of, there is no rolled bar product. It's effectively all cut out of sheet. And then question of how is it cut?
John S: But when you end up cutting, um on two or four sides, then you have inconsistencies, lack of parallelism. Lead lead time costs just all, you know, it's not like aluminum where you can go by extrusion of aluminum in almost every eighth inch or quarter inch variety up to two or three inches, you know.
johngrimsmo: Yeah.
John S: But you de-mag, magnetism comes from ah usually the upstream sawing or even the trucking, just bouncing around or getting slammed in.
johngrimsmo: No way.
John S: We've kind of proven it um through that way. We've actually had them take Gaussian measurements throughout the process on the vendor side.
johngrimsmo: ah way.
John S: um And the short answer, kind of like, you know, perfect results count, not the process. I don't have a way to solve this process. you know we buy a lot of it, but not enough to where the, we can, you know change,
johngrimsmo: Yeah, yeah.
John S: Whatever. That's not the bottom of my face. So we do that. It's not a huge deal. It's just annoying to do it.
johngrimsmo: Mm-hmm. Mm-hmm.
John S: Yeah.
johngrimsmo: I was almost wondering, as we were talking about it earlier, is it easier to mount, you probably have one of those little brick-sized DMagger units, is it easier to mount that on a slider and then have a rack of, like a bar feeder rack, of material that rolls down to the next one and then you slide on this rail, linear rail or whatever, the DMagger across it and then pull the bar out, get the robot to pull the bar out, next one rolls down, something like that.
John S: Yes. Yep.
John S: Actually, that is a great idea. Now that you say that, it would almost be ironically using a magnet, which is an issue because using a magnet will...
johngrimsmo: Yeah.
John S: add to or or creates magnetism in the part, but you could use a magnet on a robot to pick up the material. That's a lot easier than trying to grab it. And then one at a time, you can have it drop it into a fixture that will auto or like a, uh, they call them flip, flip fixtures, but it's just like a little, think about like an iPhone stand that you'd take your iPhone down on your desk. It would hold it up at an angle.
John S: You know if you put set the material on something like that, regardless of how it's oriented, it'll end getting oriented correctly.
johngrimsmo: yeah
John S: Then you could have a
johngrimsmo: yeah
John S: the robot could have two end effectors. This is a but million of us could demag it. And then you could, um, yeah, you can handle it again and drop it off.
John S: can have it rolled down into a tub or something. Yeah. It's actually a good point, John.
johngrimsmo: Do you actually roll the the the bar like and deemmage different areas of it, or do you just swipe it across a couple times and you're good?
John S: Sorry.
John S: We will swipe it and then you kind of hold them. We'll do them four at a time. You swipe you swipe them over. you can go quick. And then you flip endo and flip them 180 and do that side. And that's good enough.
johngrimsmo: Endo them instead of rolling them?
John S: But it would...
johngrimsmo: Wouldn't that be the same thing? Yeah.
John S: Because you do four at time, so you can't really roll them.
johngrimsmo: yeah
John S: um Yeah, you're right. The better way to do it would be to get two... The demags aren't expensive. To get two of them and then you could almost dip them into the demagger...
johngrimsmo: like a v or uh two little paddles there's ways to do it although two i don't know anything about magnetic fields would two of them face to face cause magnetic field weirdness i don't know yeah yeah i remember watching this japanese video of d-mag
John S: this room Yeah, right.
John S: It's great question. chin I don't think so, but there is some truth to all that. Yeah.
johngrimsmo: was looking for suppliers for D mags just to see what's out there. And I watched this, you know, product video and the guy like took his part over the thing and he moved it like, like four feet away to get out of the zone to really maggot.
johngrimsmo: um Cause we tend to just kind of go above the middle point and near me, me and it, you know, makes the noise and calls it good when I'm doing, trying to D mag the button and the tiny little spring that goes under our button.
John S: Yes.
johngrimsmo: And I did that just two hours ago because they were magnetizing together. so I go to the D-Mag and i swipe it across, meow, meow, meow. And it, they still stuck to each other. And was like, hold on.
John S: Mm-hmm.
johngrimsmo: And then I did the trick and I, you know, it's all in a big paper towel. So I moved the paper towel like 12, 18, 24 inches away. And now they don't stick together anymore.
John S: you can't You can't pull it away. Spencer Webb explained this, and I mentioned it on the podcast, and I can't remember why, but you can't jerk it away. You have to slide it off.
johngrimsmo: Hmm.
John S: But ah once you're an inch or two off, I don't know that there's, on our $100 DMaggers we're buying, there's a difference.
johngrimsmo: Yeah, that's what we have too. Yeah.
John S: Yeah.
johngrimsmo: Yeah.
John S: Yeah.
johngrimsmo: Fascinating. It's one of those things I don't need to know more about, but I need to know just enough to do my job. And.
John S: Yep.
johngrimsmo: Hmm.
John S: Yep. Well, last thing I'll say about robots is Johnny Five, which is um his left and right track drives are almost done.
johngrimsmo: Hmm.
John S: Definitely the electronics was more learning than I expected. on Silly stuff like how to do, you know, we're crimping four gauge wires into these Anderson connectors.
John S: It's like no joke. Like that it's a gate, like this is like what you'd feed your main house with. Like it's ah
johngrimsmo: Yeah, okay.
John S: heavy stuff and then smaller 24 gauge with crimp terminals and trying to do this all nice and neat and there's it's a tight fit in some of these areas but um left and right track drives are are i think they're literally done and then i need to test the motors out now make sure i like them before i put them back onto what's called the cradle base which is kind of like his butt and that hooks everything together and then that will let me put in the rc system with which means you'll be able to like track drive um track drive him around, but he's, he looks like a headless Johnny or a torso list Johnny five.
John S: It's just the two track drives in the center section.
johngrimsmo: Yeah, yeah.
John S: But one of the other guys has got the project who did this. I mean, once that's done, like me and one other the person could stand on that center part and drive around on it. Like it's got some torque and power and all that.
johngrimsmo: Whoa.
John S: And just seeing him properly operate with an RC, um, I just bought the RC controller, which looks like it is one of those like helicopter style ones. And it's got a ton of buttons and it has the ability to then play through scenes.
johngrimsmo: Yeah, yeah.
John S: So you can have these pre-automated scenes that control the different servos. It's all starting to come together. um So I should probably do a little update. We put some stuff out on Insta over the time, but it's been fun.
johngrimsmo: Can you toe tip with this setup?
John S: Yep. Toe tip is part of the, what's called the cradle base. That's part I have to reassemble. And that was already, toe tip was already working. So that's just a question of hooking them up to the RC side of it.
johngrimsmo: Sweet.
John S: Yes. Yep.
johngrimsmo: That's exciting. It's nice ah to have actual progress. like Like, yes, all the work goes towards the end goal, but when you finish a ah module, a you know, part of it, you're like, I can drive it now.
John S: Yes.
johngrimsmo: Yeah.
John S: Yes. And I'm cautiously optimistic that that was actually, it was a ton more work than I thought, but there isn't a ton to his center section. And then the head and the neck are, I spent six months working in them last year. So they're, they'll need some TLC, but they're not, it's not like I'm staring at a blank slate there.
johngrimsmo: Yeah.
John S: Um, and the arms will be a ton of work, but I don't right now, I don't care if the arms just hang there, right? Like that's another project for the second half of this year, which I can't believe i'm saying that already. It's April.
johngrimsmo: Yeah.
John S: Second, my holy cow. Yeah.
johngrimsmo: First quarter done.
John S: Right.
John S: Yeah.
John S: Yeah. Bought the new bamboo, which I was, we needed a bigger, we needed another printer for the shop and I wanted a bigger printer.
johngrimsmo: Yeah.
John S: um Those were the two criteria. And then, and then there's a million opinions out already on it. Rob already has his, so good shout out to him on that. But, yeah. It still blew me away that it's like, I mean, effectively the same prices in each, the Carbon X one.
johngrimsmo: Yeah.
John S: Yeah. Right.
johngrimsmo: That's wild.
John S: Yeah.
johngrimsmo: Yeah. And i I saw a bunch of videos. I didn't fully watch them, but you know, you see pictures and you see posts and you're like, okay, so it's bigger. They say it's bigger until I saw a side-by-side picture.
John S: Oh no, she big.
johngrimsmo: Like, yeah, it's like a CNC kitchen had a side-by-side as his thumbnail. like, holy moly, that's way bigger.
John S: Yeah. the my favorite of the build plate comparisons. And now it's a little bit because the dual head means you can't, there's a hack to get it to print the full volume.
johngrimsmo: Oh yeah.
John S: You have to have film it in each left and right and blah, blah, blah.
johngrimsmo: okay
John S: But I think I'm okay with that. um
johngrimsmo: It's still way bigger. Nice.
John S: It's way bigger. It's not as big as some people people were hoping for.
johngrimsmo: Two.
John S: And part of me doesn't know why they didn't go a little bigger because man, but for a bunch of the Johnny five parts um that I want to get done, it'll be great.
johngrimsmo: nice
John S: Yeah.
John S: Yeah.
johngrimsmo: to
johngrimsmo: o
John S: What's been on your mind or radar or to do?
johngrimsmo: um
johngrimsmo: Fixing old problems that i that I didn't know were problems. um I think I talked about this bit last week on the Mori, how I was going through the old code and I was like, this is, I would redo this now.
John S: Oh, yeah. Yeah.
johngrimsmo: So i've I've run into a couple other scenarios like that. um So like on the current, we've been making the RASC now for five years. um And some of that code is four years old, like runs fine, doesn't need to be changed.
johngrimsmo: Some of the logic, the probing, the, um, one of five axes when you tilt it sideways and you probe on a sideways tilted face like when my tombstones are at b90 or negative 90 um hide nine has to enter this mode called plain spatial that is weird because when you tip it over x is not x anymore but plain spatial yeah exactly that kind of helps figure it out but um
John S: Yes.
John S: yeah
John S: like the It's like the robot. Yeah, yeah.
johngrimsmo: The way I probe certain things, when I probe certain things, the way I use that data, either update a tool offset or update the work offset or the datum table, which is a way to hack the work offset and put a temporary modifier into it, which is super cool.
John S: Okay. do
johngrimsmo: I think I've been doing it wrong.
John S: you mean?
johngrimsmo: like i tried it. um Hard to explain. Hard to wrap my head around, honestly. But let's just say for things like I'm trying to machine a pocket that is exactly 50,000 deep from surface down, not necessarily from fixture surface up, because material thickness might change, right?
John S: DAVID MALAN- Understood, yeah. Right.
johngrimsmo: um So I'm probing the fixture to establish you know perfect, accurate. We've got six different tombstones, and they're all slightly different in size by you know many tenths. um So do you go bottom up or do you go top down? Do you touch your material and you establish your work plane from there down?
johngrimsmo: um So I spent quite a bit of time in the past few days testing, tuning, seeing what the difference is between those two methods, seeing how doing it within plane spatial gives me the result I want or not.
johngrimsmo: And basically wondering why I'm often constantly making a part about ah one thou under my nominal.
John S: Oh, yeah, we're right.
johngrimsmo: And it works. We've been making it work. But we have a mating feature in there that's also 50 thou. And if one's 49 and the other's 51, then now you have a 2 thou stick out.
johngrimsmo: And that causes problems.
John S: Right.
johngrimsmo: um
johngrimsmo: So I don't have a solution yet. But I have been banging my head against the the control trying to figure that out. you know
John S: So if you if you machine a big plane, like a big ah dice, we'll make an example. yeah and then and you're on a five axis, but you do that without moving A, B, C, et cetera.
John S: Then you want to drill precision hole or bore on the side of the dice. So you tip it to B90. And if in a perfect world, and especially current, everything's perfect. But let's say don't want to rely on that. You want to come with your probe, and let's assume the probe is perfect.
John S: You probe at B90, you probe the top of the dice, and then you say, from there, I want to go down exactly 50 thou. that what you're talking about doing?
johngrimsmo: That's basically what I'm talking about doing. Yeah.
John S: Okay.
johngrimsmo: Yeah. And it's a version of that is what I have been doing, but, um, over over the past few years, I would sometimes have to put like, a manual modifier into that logic.
johngrimsmo: So like, I know it says 1.95, but make it 1.96 and it does what you want kind of thing.
John S: Got it.
johngrimsmo: And it's, that's really bugging me now. I'm like, I'm cheating and ah that's I shouldn't have to cheat.
John S: You're cheating. Yeah.
johngrimsmo: or Like I need to understand why.
John S: Right. Right now.
johngrimsmo: And at the end of the day, does it matter if I get the result I want? But, um,
John S: ah Perfect results on perfect process.
johngrimsmo: Exactly. Right.
John S: Right.
johngrimsmo: So anyway, I'm there. Um, and doing more of this on the speedio, more probing, more ultra accurate, you know, trying to hit a 10th, regardless of thermal deviation tool changes, whatever.
John S: Yeah.
johngrimsmo: Um, and I'm doing it. It's just got me to rethink a lot of our processes going, okay, John, four years ago was clever and did it this way. Would John today do it differently? Is it worth the effort to do it differently now?
johngrimsmo: Um,
John S: Yeah.
johngrimsmo: is the Is the result what you want? Can you ignore this and move on? Or do you need to dig in and get this done?
John S: Yeah.
johngrimsmo: Yeah, because we're noticing in the assembly of these knives, you know, so sometimes making up numbers here, so say it takes 10 minutes to do final assembly on a good knife.
John S: Yep.
johngrimsmo: Some of them might take an hour. And i don't like that.
John S: Yeah, the variability is sure.
johngrimsmo: That is a total waste of opportunity.
John S: Right.
johngrimsmo: um Can I solve that upfront is my the real root of my question here. you know Does it fall back to, is it a machining inaccuracy? Is it stock variation thickness?
johngrimsmo: Is it the way I'm probing? Is it toolware? Is it how accurate the laser touches off the tools? is you know Yada, ya yada, yada.
John S: Yeah.
johngrimsmo: So i'm trying to eliminate all these variables so that we have a process that's perfect so we can get the result same every time.
John S: Are you... do you run Kindomax in the current very often?
johngrimsmo: Not very often. And we probably do it, honestly, a couple times a year.
John S: Got it.
johngrimsmo: um And then it does log the results for all of history, so you can really see. And it doesn't move much.
John S: Got
johngrimsmo: um Always do it after a bump or a crash, which we haven't had in a long time, knock on wood.
John S: it. It reminds
johngrimsmo: It would be nice would be nice to do it monthly, but.
John S: Yeah. Reminds me of something that i frankly have just glossed over because ah you pick your battles, but Laurens and Dennis will often dive into these pretty respectable conversations about five axis calibration cubes.
John S: And it's kind of like the it's like the two inch cube with you machine it one way, then you flip it over or rotate B and do a feature there.
johngrimsmo: Yep.
John S: um And when I read them, I'm like, I'm not following this. And then that's where I'm like, John, this is not a problem that's on your plate right now. So just kind of move on. But I'm a little embarrassed to say, ah which i I wish I understood what they were doing there.
johngrimsmo: the
John S: And it's probably not that complicated, but nevertheless, I was, um could you could you teach what they're doing there?
johngrimsmo: yep
johngrimsmo: Put
John S: Do you know, you understand?
johngrimsmo: No, no, exactly.
John S: Okay, yeah.
johngrimsmo: And it's ah it's a mentality, a mindset, a methodology to, there's a right way and a wrong way to do it.
John S: Yeah.
John S: Yeah.
johngrimsmo: There's probably 10 ways to do it. But, you know, guys like that will talk about their experience and be like, no, this is the way to do it because it gets the result. You know, if you machine the cube, you machine a circle in one side, you flip it over, you machine a circle in the other side, does a dowel pin fit perfectly through it on both sides all the way, right?
John S: Yeah.
John S: Yes. Yes.
johngrimsmo: um Things like that.
John S: Something this may not help, but it's been ah top of mind for me privately lately, which is thinking about what my role is here and what we do and um the products that we make and
johngrimsmo: Mm-hmm.
John S: The thing that helps me get outside my comfort zone is if we need it, let's say we sell a thousand of these a year, if we sell 20,000 and money in the, in the, and then it's ah for whatever reason that like i pick a number, 3 million, 10 million, 30 million, doesn't matter.
John S: Like your job is to allocate capital, to buy the equipment you need to do this that you can do 20,000 of them. And it's not, you gotta be a hero on this, part or you got to inspect it here. And, um, we actually, ah to our credit, it was a huge win last week when, um we pulled a whole batch of puck chucks into assembly. So like they had been through machining, heat treat, hard milling, grinding, and I didn't do anything.
John S: It that was the first time they had been handled by the team.
johngrimsmo: Oh.
John S: um I was involved. I helped some. but like it was now We're now at a point where like the next time we do it, I'll probably sort of say, hey, let's get a second set of eyes on it or like happy to remind people.
johngrimsmo: Yeah, yeah.
johngrimsmo: Yep.
John S: But like that was huge. And we didn't have a single QC failure.
johngrimsmo: Nice.
John S: um So that's... in it in in um I'm pleasantly surprised because you're always surprised on the flip side.
johngrimsmo: course.
John S: It's what is how we built that workflow to to be. um
johngrimsmo: Yep.
John S: And that was great. That really felt great. Now we can't make 20,000 a year, but we're setting it up to build so that we have the potential to lot of them.
johngrimsmo: Yeah. Yeah, that's awesome.
John S: Yeah.
johngrimsmo: Yeah, that took a lot. Like, I remember two years you've been talking about the the workflow, the planning, the process.
John S: Yeah.
johngrimsmo: I know. Trust me.
John S: Shame.
johngrimsmo: Yeah. 12 years on my end to get this field done. It's just how it goes sometimes. Yep.
John S: Yeah. Yeah.
johngrimsmo: No, that's good.
John S: and there's And that's the question. we What do we, ah for the puck chuck, what's the equipment? What do we need? we what period Period. And then, um and there's a lot to be said for
johngrimsmo: Yep.
John S: you know, lean means just eliminate it. Like don't don't figure out ways to do it better, just get rid of it when you went and where you can. um That's one reason why we iterated so quickly to the Gen 2 Puck Chuck was we quickly realized how it's worth taking that step and de-featuring a bunch of things. It's so much cleaner and simpler.
John S: still need to figure out, sorry I'm rambling, where how we sort of like market the pitch around it. Cause the, what I want to say is like, Hey, this is a very accurate, but also not over-engineered. It's not a frilly.
John S: um This is a very utilitarian, excellently well-made zero point. That's at a great value.
johngrimsmo: Absolutely.
John S: um That's what I'm pitching for versus um the kind of shiny, the Rolex. They're like, Hey, shiny bells and whistles.
johngrimsmo: Yeah, this is the best on the market. Yeah, totally.
John S: Yeah, exactly.
johngrimsmo: Question. Do you see people using Pupchucks not on a Saunders picture plate?
John S: yeah we sell them. We're making them actually about half and half right now. We call it the standard, which is the fixture plate versus the rotary. The rotary has a bolt hole pattern that accommodates effectively every A-axis, fourth axis.
johngrimsmo: e The fourth axis?
John S: Yeah.
johngrimsmo: Yeah, okay.
John S: um I forget the the degree increments that works that way. but um And then you don't need a fixture plate to even use either one of them. um Works well that way. But yeah, no, we're not... um We want to use them as a great complement to the fixture plate, but we're also not needing to force you ecosystem.
johngrimsmo: Of course. Yeah.
johngrimsmo: Well, it's the classic thing where... especially on machines like Memori and MySpeedio, where the tool can't touch the table, you always have to build something up.
John S: Right.
johngrimsmo: um And you have T-slots, and then you've got to work up from there kind of thing.
John S: Yeah.
johngrimsmo: just To just buy a puck chuck and put it on a T-slot table, you've got to interface it somehow.
John S: Yeah, exactly.
johngrimsmo: you know
John S: CJ was a great early validation of when he was like, hey, I put all the puck trucks on the fixture plate. They work great. And then I needed to change something. he's like, it took me, it's not only 10 minutes, but it just doesn't stress you out. Like, it's just like, you just move them from here to there.
johngrimsmo: yeah Yeah, yeah,
John S: um It's like Lego. It's not like, oh boy, I need to align it and think about it. Or you go to a Shunker Lang and it's literally $10,000 for the Vero type systems that are on the bigger grid plate.
johngrimsmo: yeah. Yep, yep.
John S: So yeah.
johngrimsmo: Cool.
John S: What's on tap today?
johngrimsmo: um Finalize the Fjallblade, literally have it on my list. Move it into Steven's hands. like
John S: Okay.
johngrimsmo: This job is going to be not mine by the end of today.
John S: Okay.
johngrimsmo: and I'm close. I've got to make a little bit deeper, tweak a couple things, and then and then he can run. We've got a stack of 20 blades for this hard milling step that is waiting on me to finish this.
John S: Yeah.
johngrimsmo: that's That's my only real thing. Yep.
John S: um We are, ah we are, actually I'm making the first part on our 1500 MX and it's great. um
johngrimsmo: Yeah.
John S: Yeah, ah prototype, um actually a product a lot of folks have asked for. We're now finally in a position to make it, which is a six inch insert for our mod vice.
John S: So it would extend one inch over the left and right sides.
johngrimsmo: Oh.
John S: um At some point we might make a bigger jaw that's gets more complicated with the um base screws on the mod device. But um yeah, it's perfect example why we have the 1500. It's like, hey, drop a fixture on the fixture plate. Tools already in there, cams there, and then prototype this thing up and,
John S: um we actually made extra up to fixtures that go in the horizontal. I realized can grab one of those extra replacement fixtures and drop it on the fixture plate and actually make the whole thing. i thought I was going to opt to it on the horizontal, which is great. Cause I don't want to be prototyped on the horizontal.
John S: Um, takes time away from it, crashes it like it's Garrett's machine.
johngrimsmo: Yeah.
John S: I don't, then have mess around with it. And so it was literally the first part I'd really cut with a 1500. It's great. It's like, ho it's just great.
johngrimsmo: Yeah.
John S: Yeah. Yeah.
johngrimsmo: Wicked.
John S: Yeah.
John S: Yeah.
johngrimsmo: Sweet. Yeah.
John S: See you next week.
johngrimsmo: All right, man. Have a good week.
John S: Take care.
johngrimsmo: Okay, bye.
