Transcript
John Saunders: good Good morning, welcome to the Business of Machining, episode 442. My name is John Saunders.
johngrimsmo: And my name is John Grimsmo.
John Saunders: John and i just ah share our passion for manufacturing, talking about running small shops, and what's on what's top of mind, good and bad, right?
johngrimsmo: Exactly.
John Saunders: Yeah.
johngrimsmo: what you got for today?
John Saunders: ah Okay. The but thing that's affecting me the most in a good way is, um, is just like calming down, slowing down. I think a lot of that does tie into the the phone and social media. I'll come back to that. And i also don't want to ramp ramble on a soapbox, but it's been great. Um, otherwise,
John Saunders: Brother Progress with Automation, which is super awesome. Organizing life thoughts over the course of a year, i want to talk about. And Johnny Five a little bit. And then 2026 Capital List. And then Home Assistant for the automation stuff.
johngrimsmo: Incredible looking forward to it. So on my list, um, maybe a touch on hard milling, uh, trip to California where I will, I will see you, um, an email from our buddy, Cyrus, uh, including white light into varometer and fringes, um, some grinding wheel stuff and, more grinding or stoning.
John Saunders: Yes.
John Saunders: Grinding and stoning.
johngrimsmo: Yep.
John Saunders: Okay. I like it. Uh, kick us off.
johngrimsmo: Um,
johngrimsmo: Pretty much, man, I just realized all we do on the Speedio is hard milling. We actually don't do any soft milling on the Speedio at all.
John Saunders: Was it ever meant to?
johngrimsmo: Not necessarily. Well, it was meant to supplement, um but I just never put those words out of my mouth. Like, wow, all it does is hard mill.
John Saunders: but
johngrimsmo: It's great. It's fantastic.
John Saunders: Yeah. Uh,
johngrimsmo: Actually, it does one titanium tab cutoff on our fuel handle, but I might move that to the current anyway.
John Saunders: OK.
johngrimsmo: But ah it's it's been great. um trying to hyper obsess about hard milling perfection is, is it's all that little back and forth, like tweak CMM feedback, tweak again, probe, deprint, verify, see if it's doing the thing like consistently.
John Saunders: Yep.
johngrimsmo: Is it the machine moving around? Is it the tools wearing? Is it, it's something, it's always something. Right. But we're getting super good results and it's good.
johngrimsmo: Yep.
John Saunders: Awesome. This is the, what you talked about before with just understanding the temperature of the machine, thermal movement.
johngrimsmo: Awesome. It's like,
johngrimsmo: Mm-hmm.
John Saunders: Okay.
johngrimsmo: Yep. So now my thermal probing, thermal comp probing and writing to G 53 has been fantastic for the past week.
John Saunders: Okay. Yep.
johngrimsmo: Um, I'm measuring trends and graphing it all in my Python program. Um, it's all coming together. Great. And then I asked Steve today, I'm like, how is, how are the bevels going? Are they consistent? Cause they're the ones that will walk around a lot. And he's like, oh, they've been fantastic.
johngrimsmo: I'm like, okay.
John Saunders: This is the geodesic chop grinding style.
johngrimsmo: Yeah. Yeah.
John Saunders: Okay.
johngrimsmo: Jig grinding. Yeah.
John Saunders: Yes.
johngrimsmo: And it's going great. But if the machine moves, you know, by five tenths this way or five tenths that way, like it really makes a difference and, uh, it's pretty good. So the, the actual goal, like I keep asking myself, what are you trying to accomplish here?
johngrimsmo: Like bigger than just the day to day, um, is palette changing, you know, 10 palette, 10 blade pallets and a hard grinding, all of them and having them be within range within tolerance.
John Saunders: Yes.
johngrimsmo: So, That is my next step. And I had a big chat with Eric and Angela yesterday about what that project looks like, like what does done look like.
John Saunders: hu
johngrimsmo: And now that I think I have thermal under control, I think the next step is just to like load five pallets and run them and see what happens. you know
John Saunders: You don't run the brother lights out then.
johngrimsmo: We don't because currently we're running like one pallet at a time, which could be 20 60 minutes.
John Saunders: Yes. But if this all works, then you would run it at least unattended during the day, but hopefully you're probably also,
johngrimsmo: Yep. Yeah. Even into the night, um, freely as many parts as we want to make, like, I just wanted it to work.
John Saunders: yes
John Saunders: Yeah, no, I hear it. It's funny how similar that journey is on our side of like, we're literally at the point of um testing out Modvice production on the Brother and the, it's like, it's so easy to break it down.
johngrimsmo: I think so.
johngrimsmo: Nice.
John Saunders: It's like, we can run, We can run each operation manually, like handload hand load, hand load, then you can add the robot in, then you can add the robot in unattended during the day. There's actually more micro steps between these two, which is even nicer. um You can choose to do probe verification if you want to do just extra cautious stuff. And like, I,
John Saunders: I'm not a super confident type of person in the sense that like, I don't get cocky and think like, we're going to, yeah.
johngrimsmo: Send it.
John Saunders: Like I'm always like ready to eat, eat some like, Oh, struggle bus home, humble pie. Like this is to kick our butt, but I just feel like we will hit those hiccups. That's, but that's okay.
John Saunders: Like I know, you know what I mean?
johngrimsmo: Yeah, yeah.
John Saunders: it feels really good.
johngrimsmo: Yeah, the the key is finding the failure modes and making sure they're not going to cause cost money, like cause damage.
John Saunders: Yeah. Yeah.
johngrimsmo: um But otherwise, you need a big enough sample size to give yourself good enough data. like
John Saunders: yeah. He.
johngrimsmo: Like running two parts and being like, sweet, automation works. That's not a big enough sample size. But you run, you know, you run five and then you run 10 and then you run 20. And it could take weeks to get to that level. But you're you're getting enough data to to prove it out and to find more failure modes.
John Saunders: Yeah, yeah. We were grinding some OPTU fixtures. So this is probably the single biggest concern I still have that we haven't truly proven out, which is can we flip a steel part over in OPTU?
John Saunders: um
johngrimsmo: using the URL robot.
John Saunders: whether Yeah, yeah that well we can flip it with a robot, that I'm not worried about. But whether it's the robot or by hand, um can we push it down and clamp across the part with a screw vice and maintain the nominal parallelism and flatness spec that we want?
John Saunders: Which is not a big tolerance.
johngrimsmo: Yeah.
John Saunders: And we're doing that right now in the horizontal for years. We've done it with inserts that we make on the Willimon that we screw into the part and then we clamp through the part through those screws, which is it's kind of like an ID expanding clamp.
johngrimsmo: Thank you.
John Saunders: And it works great, like just great, but we don't really have that luxury. And so um part of the process when we were sort of interviewing the lights out guys and helping them understand, is this going to be a good workflow? Here's the goal.
John Saunders: They taught, they said something that, on paper now seems obvious, but I'm going to repeat it because it it was very awesome. It's like you need to buy a, I don't think it has necessari has to be Kurt Brand, but needs to be a high quality angle lock style vice that has a monolithic body.
John Saunders: So the fixed jaw is not a bolt on with a key, but it's monolithic.
johngrimsmo: Mm-hmm.
John Saunders: And then we took the jaws out of it. Now you could, you can violate any of those if you want to, but I'm like stacking the deck in our favor. And so we did that. We bought the, it's a shorter version of a Curt Weiss, think a 3620
John Saunders: We don't need all this stroke. And sure enough, we got it here. We put it on a table and we took the jaws out. so you have a hard body without the jaw and then you have the movable and we put a granite parallel in there and we clamped a part while we had indicators on the part and there's zero clamping deflection.
johngrimsmo: Interesting.
John Saunders: I don't know if, I think we talked about this and I know you you, I know you like believe me, but I've been doing this for now 15 years and like, I would not have believed so somebody when they told they told me that.
johngrimsmo: Probably. Yeah.
johngrimsmo: Yeah.
John Saunders: You just, you always can eat like on a, these and these are millimus like either one micron or, sub 10, don't know what it is. These indicators, you know, somebody sneezes next door and they move like, and it was dead nuts.
johngrimsmo: Nice.
John Saunders: So now what we did is we machined these um little inserts that are like our own version of parallels that have pressed ball bearings in them. And then we ground small flats on the ball bearings.
John Saunders: That's what we were doing this morning. And the those are going to serve as our Z datum when the robot pushes the part down. So I was showing Ed Diokamoto, he, other than a quick like warmup demo two, three months ago, he has not run that machine.
John Saunders: so sorry, there's a long version of selling you.
johngrimsmo: no
John Saunders: The reason that he asked me a question that don't, I wish I could have answered better, but it's also, whatever the, we were jogging down and I thought he had hit thou, but it was actually the default Okamoto one is actually tenths.
John Saunders: And I think there's even one that's less than that. So you're, you're, I think you're jogging down in 10 million increments.
johngrimsmo: Yeah.
John Saunders: Um, I have never bothered to put an indicator on there. I should actually to see what sort of, you stiction or backlash hysteresis, is that the word? um But as you're talking about the brother and just eking the most out of it it just kind of looked at me and was like, if you dump bump it down 10 millionths, do you get 10 millionths? I'm like, I mean, I like can't tell answer that other than like, this is a well-built high-end Japanese machine where what they're giving you isn't even necessarily like,
John Saunders: Like, I don't think they're not putting there as like a reach stretch value. Like it's, it's real.
johngrimsmo: Yeah, yeah.
John Saunders: That make sense?
johngrimsmo: you trust You trust it, the machine listens to you, you know?
John Saunders: Yeah.
johngrimsmo: Yes.
John Saunders: Yeah. So fun.
johngrimsmo: That's great. So, man. Because you're with the you're talking about jaw lift, pretty much, on on most devices anyway. Like, does the jaw lift when you push it up? And and or does the clamping mechanism move your part while you're clamping it?
John Saunders: Bingo.
johngrimsmo: Yeah.
John Saunders: Yeah. Like, Right, you have to have something that's kind of perfect to begin with. And nothing's perfect, that's the problem.
johngrimsmo: Hmm.
John Saunders: But can the rope...
johngrimsmo: But you can stack the deck in your favor, you know?
John Saunders: You can't Well, that's what we're going to find out. You know, I think we can grind something quote unquote perfectly parallel. but Within, i think Ed measured the the pads that we just did and said they were, ah think 40 millionths was the greatest deviation he saw.
John Saunders: So like great, for a quick quick grind job, I'm fine, like no complaints there. But it's it's still an amount, right? So if we just lay apart on top of that, you won't see a significant increase in deviation over those pads, right?
John Saunders: They should inherit that instability.
johngrimsmo: right
John Saunders: And then it's really a question of them when you, the robot can, it's actually really cool. The robot can push down like a finger would, but we'll see. We'll see.
johngrimsmo: nice do you have so that's your z do you have an x y or and like a single direction move to like bump it up against the stop on the left or something
John Saunders: Good question. um I don't think we do right now ah The robot might be able to place it close enough because in this case, all we're doing is chamfers, but the chamfers need to look nice too. But it's not a crazy critical tolerance there.
John Saunders: We could also probe, although I'd much rather not probe something like that if we don't have to. So I don't know. I um i really and feel grateful for the team. I've continued to step back and let them run with this. And even just like one of the other products on that, I was going to assume we were doing it this way. And then I saw How they actually decided to fixture it and it's just like I'm so I'm so glad I got out of the way like that's such a great And it's not to say I don't bring value or add like help out but like um i'm so I'm saying out of it.
johngrimsmo: Good for you.
johngrimsmo: But let them do it too, right?
John Saunders: Yeah
johngrimsmo: Yeah. That's awesome, man. And again, you need a big enough sample size to maybe the robot flip will chamfer accurately enough. You know?
John Saunders: Yeah
johngrimsmo: Maybe you don't need overthink it.
John Saunders: What's funny, cause I don't, i I think everybody listening knows I'm not full Grimms mode, but I'm also like, ah, Chamfer's man, you can see fourth three four three or four different Chamfer, you can see it.
johngrimsmo: Mm-hmm.
John Saunders: um
johngrimsmo: Yep.
John Saunders: Yeah, we'll see.
johngrimsmo: And does that bother you or not is the next question, but yeah.
John Saunders: mike My answer there is something I continue to embrace, which is I don't mind it if it's a one-off, like we would probably still sell that product, but it doesn doesn't mean it's not accept it's not acceptable for everyone.
johngrimsmo: Well, the point is if it's happening on everyone, you can fix it for everyone.
John Saunders: Yeah, that's for sure.
johngrimsmo: That's what I look at.
John Saunders: True.
johngrimsmo: If we are consistently wrong, we can also be consistently right.
John Saunders: because it's really right Right. Yeah. No, no, that's really, that's
johngrimsmo: And we see that all the time. They're like, yeah, it's always like this. And i'm like, not for long.
John Saunders: yeah, that's true. Well, good. why You're, you're, you're, you're getting that data set of to give you those confidence.
johngrimsmo: Yep.
John Saunders: Yeah. It's awesome.
johngrimsmo: Which is great. And the thermal movement is not as much as I thought it was now that I'm actually looking at the data. It's probably i got to look at it closer, but under a thou movement in each axis, depending on temperature and things like that.
johngrimsmo: And each axis doesn't move at the same rate.
John Saunders: Yeah. Fair
johngrimsmo: But I need to look more to say any further.
John Saunders: enough.
johngrimsmo: But it's it's cool to have the data and to be able to slice and dice it however I want, as I want. And maybe I'll turn it off at some point, but for now, it's like it's just nice to collect it.
John Saunders: Yeah.
johngrimsmo: But i was talking with Angelo yesterday about failure modes. And as I'm making this stuff more complicated with the probe coming in all the time and the deep print sending into the Raspberry Pi, Angelo's like, if you're gone for a month and something breaks, how do we revert to like something stable?
John Saunders: Mm-hmm.
johngrimsmo: And you know so put put some basic programs on there. well it It won't work. It won't make a good result if I take off all the things that are making the good result. But I do have to figure out ways.
johngrimsmo: Like, imagine the Raspberry Pi just dies, and I'm not here to like flash the new image and to whatever.
John Saunders: Oh.
johngrimsmo: Because if it tries to deprint to a file or to a computer that's not there, I think it will alarm. And just the whole machine is bricked at that point because every program deprints something now.
johngrimsmo: So do I need to put like a little flag in there? So if macro 500 is a one, de-print everything. If it's a two, de-print nothing. So that so that the the guys could change that to a two.
John Saunders: Interesting. Oh yeah.
johngrimsmo: And all of a sudden, if the pie breaks, now everything is, don't know.
John Saunders: Yeah. Well, there's like, you're on vacation is one scenario, you die is another scenario. You on like a long, you know, you going to the Arctic Circle out of pocket for three weeks. That's like probably the worst case scenario.
johngrimsmo: Yeah. Yeah.
John Saunders: Not that that's worse than you dying, but like if you, if you die, then like, okay, we just have to like move on without John. Like we know.
johngrimsmo: yeah Do it our own way kind of thing.
John Saunders: Yeah.
johngrimsmo: Right.
John Saunders: But if you're, if you're on vacation and something bad like that happens reality, I would think you're just like, okay, here's what we need to do. Order this. Like you can get involved.
johngrimsmo: If I'm, if I can be involved, say I'm not, say I'm on vacation and then really offline or something like that. Not that I do that, but
John Saunders: Well, that's what I'm saying. Yeah.
johngrimsmo: But it is it's an interesting thought process to give the team the power and the ability to not have a bricked machine until I get back because costs a lot of money and makes them feel bad.
John Saunders: Sure.
johngrimsmo: And like they just feel useless when my complication makes them unable to do their job.
John Saunders: Yeah.
johngrimsmo: um
John Saunders: Yeah.
johngrimsmo: Yeah. <unk> I don't have an answer yet, but it's interesting.
John Saunders: For now.
John Saunders: Yeah.
John Saunders: California sire. Oh, here, I'll go. I'll take a turn. So um I just want to re there when you start seeing research um about social media and you know how like when we were growing up, there started to be the ah certainly in America, the kind of final pushback against tobacco, nicotine, cigarettes and like they, you know, smoking in restaurants ended and blah, blah, blah.
johngrimsmo: Yeah.
John Saunders: And you know, it's kind of like for like 30, 40 years, we knew this is, you know, not great, but whatever. And it's like, to be clear, you know, smoking, social media, alcohol, it's all choice. I'm not judging people, but it's choice. But like I, the research you see now about like social media kind of rewiring your brain for that addiction, if you will. Um, I just want to share on a positive note. It's been like wonderful. really thought it hard and I've struggled with other quote unquote addictions or like wanting to check sales or, I don't know, whatever. But, uh,
John Saunders: It has been just great. I can't say anything other than that. So ah I don't want to ramble. I'll leave it at that. But if anybody wants a word of encouragement, give it a shot for a week or two. i still am logged into the social media stuff on a computer that I can remote into, which keeps me connected, but it avoids the...
John Saunders: what do you call it, like the idle time scrolling and the like doom scrolling. It's not nearly as fun to scroll it on a desktop. And it's a bounce because I like being part of community and I don't want to like turn my nose up at it, but um definitely definitely glad I did it.
johngrimsmo: Yeah, years ago, I kind of fell into the trap of feeling the need to like keep up with everybody, to check everything, to see all the things.
John Saunders: Mm-hmm.
johngrimsmo: And then, like, with your friends is one thing, but with the general, like, your feed kind of thing, once it got bigger than I could manage on a, you know, it sort of occurred to me as, like,
John Saunders: Yeah.
johngrimsmo: it's not my job to keep up with this stuff. Like there will always be too much cool stuff to look at. know I mean? Like there, there will never be an end to like checking it.
johngrimsmo: So something in there kind of made me realize that whether it's, you know, five minutes or five hours, get what you can and go.
John Saunders: Yeah. They're good at, yeah. I give you, yeah you, you, you figured that out way before I did like good for you.
johngrimsmo: Yeah.
John Saunders: Yeah.
johngrimsmo: Yeah. Yeah, and I don't and don't like the Doom Scroll anymore. um
John Saunders: Yeah, great.
johngrimsmo: i haven't haven been for a while, which is good. So it's like you cutting it completely is, it's different for me because I don't i don't need to.
John Saunders: Yeah, no, dude, you've got rock and roll.
johngrimsmo: But but it's when maybe I'm being oblivious and like using it more than I think I am whatever. But no, it's it's interesting it's it's good to be introspective and like try things, you know?
John Saunders: Yeah. Love it. um We also, on our shop update video, talked about the, as just a PSA, about the Haas machines having the ball bearing on a glass plate sensor.
johngrimsmo: Good.
johngrimsmo: Mm-hmm. Mm-hmm.
John Saunders: And I um want to continue to keep the discussion going by also You know strongly strong convictions loosely held I will happily eat crow and admit that I'm wrong if somebody can show me that I'm wrong But I don't think I'm wrong and that is I'm tired of people saying five axis machines or accurate machines are controlled by ITAR They're not there's an export regulation thing that could apply but this idea that like any machine that meets certain build specs or capabilities immediately by default becomes ITAR controlled is as best I can tell just not true.
johngrimsmo: Mm.
John Saunders: So Mitsutoyu or Mazak or Haas that have relocation sensors or I've heard of some DMGs but I've also did some research to confirm a variety of machine tool brands, high-end brands from Doosan or but but vari auma duusan hermla many others don't necessarily have it it at all, or it's just not the case.
John Saunders: And so I think I, there's part of me that enjoys trying to break down this, this old wives tale reputation comment of like, it's all right, it's like, no, it's not like it's such, it's a, it's a private choice, whether it's complying with the sanction or going above and beyond as an internal insurance policy to avoid further sanctions.
johngrimsmo: and
John Saunders: But, um,
johngrimsmo: Right.
John Saunders: Yeah, I just, um if anybody can argue with me, love to hear it, but ITAR doesn't, what?
johngrimsmo: It's not a blank blanket statement.
John Saunders: Yeah, exactly.
johngrimsmo: Hmm.
johngrimsmo: But you still believe that all Haas machines, all American-made Haas machines, have this sensor now?
John Saunders: They do, yep.
johngrimsmo: Yeah. Wild. Wild.
John Saunders: Three access machines do, it requires service visits. This is Haas. And we posted, there's some comments in the YouTube video, the shop update video we posted from me and and from others that show DOJ PDFs on the sanctions.
John Saunders: And i mean, Haas told us, like they are now building a registry database of all Haas machines. And so I believe I believe, but don't know, but I believe that even an older Haas machine that doesn't have this sensor, if you need service on it, they will then service it or sell you parts, but they need to come on site. They'll take pictures of it. They'll log who you are, what you're kind of doing, job shop, OEM, et cetera.
John Saunders: And they, they are building a database and I, you know, I'm a, I don't like, don't like lists.
johngrimsmo: I thought you were going to say they ah they install the sensor on older machines.
John Saunders: No, they don't they don't force the retrofit to my knowledge, for sure.
johngrimsmo: Yeah, yeah.
John Saunders: Yeah, yeah.
johngrimsmo: Still no.
John Saunders: And obviously, completely in support of of national security of both the US and all the Western world. Like I have no desire to sell a machine that ends up in an axis of evil state, but there's a big difference between a sensor that stops me from moving at three feet and something that, you know, GPS or other type of sensor that stops it from going on a boat across the world.
johngrimsmo: right to a location or whatever yeah yeah you feel like you're getting penalized for nothing yep and financially too well that sucks
johngrimsmo: okay well i will be seeing you in what two months three months three months
John Saunders: Yeah.
johngrimsmo: For the toolpath, Mammoth Mountain ski trip. um Super excited.
John Saunders: Super. Yeah. Someone on the summit.
johngrimsmo: Yeah, super.
John Saunders: um be It's going to great group of folks there.
johngrimsmo: Yeah, I was texting with Tucker yesterday and he said he's going to be there. He's like, I want Iron Man to ski the slopes. like He's going to bring the costume and everything.
John Saunders: oh
John Saunders: Yeah. Awesome. I'm looking forward to it there. If folks want to join in, there definitely still spots open, although I think they're well over a hundred registrations.
johngrimsmo: Nice.
John Saunders: Um, but if you just Google toolpath summit on the summit at Mammoth, California, it's, it's skiing, but you don't have to ski to like, to enjoy the there's panels, speeches, you know, that kind of stuff.
johngrimsmo: here
johngrimsmo: Yeah. Yeah.
John Saunders: Um,
johngrimsmo: you're You're giving a talk. Lockwood's giving talk. Pearson and a bunch of other people are giving talks.
John Saunders: yeah.
johngrimsmo: Super looking forward to all of it.
John Saunders: Yes.
johngrimsmo: Actually, I was texting with Rob about like project planning and leadership and stuff like that. and he's like, you're going to have to wait for my speech.
John Saunders: Yeah.
johngrimsmo: was like, oh, come on.
John Saunders: No, one of the things I talked to Al about when when Al was putting this all together, i was like, please, please, please make this a about the community and learning and do not, like, I don't want any sales pitches. I do not want a ah tool, a software demo. Like, I want it to be...
John Saunders: hate the word networking, but I want it to just be about learning and education and paying forward and all that. So I think everything, there's definitely some corporate sponsors if you will, but like, I think you and I have been to events before where you're like, why am I sitting through an hour demo on ah whatever productivity software that is just, oh, why, you know, yeah, right.
johngrimsmo: Exactly.
johngrimsmo: Exactly.
John Saunders: I'd rather run a lathe.
johngrimsmo: Yeah.
johngrimsmo: Can't wait. um And then, yeah, so I was booking my tickets last night and I texted you guys about it.
John Saunders: Yeah.
johngrimsmo: And I'm like, do I just get a one-way ticket and then go west, go to the coast and like figure out something to do in like LA or San Francisco or something like that. I thinking about it. and there's There's a shop I want to see in San Francisco.
johngrimsmo: So I'm like, okay, but I don't think I need to fully confirm, but I don't think there are flights going from that Bishop airport to San Francisco at that time of year. They start a month later.
johngrimsmo: So like like if you go to Expedia or whatever and you go, okay, Bishop Airport to San Francisco, they take you to Denver like two hours this way and then three hours over to Denver.
John Saunders: Okay.
John Saunders: Yes. du ba back Oh,
johngrimsmo: And like you can, so that's a seven hour total trip. You can drive in six hours. And so if anybody out there is going that happens to be driving from Mammoth to San Francisco area and and I can bum a ride from, you know, just just shoot me an email, something, let me know because I'm kind of looking for a seat.
John Saunders: yeah. Oh.
John Saunders: Yeah.
John Saunders: I actually bet you figure that out.
johngrimsmo: Yeah, yeah, I'm sure I will.
John Saunders: No joke. Yeah. Okay.
johngrimsmo: Yeah, it's like a six hour drive or something like that. I could rent a car, but like if people are going anyway.
John Saunders: Yeah, exactly.
johngrimsmo: Yeah.
John Saunders: There's definitely people coming from the Bay area to your point.
johngrimsmo: Right. So.
John Saunders: Okay. Good call.
johngrimsmo: Cool.
John Saunders: What's the, what's your Cyrus email?
johngrimsmo: So picked, I told you last week, I picked up the white light interferometer again. I'm like trying to get, trying to get fringes. I'm trying to get interference and like light canceling itself out to measure flatness and my, my rinky dink, you know, um, desk set up at home that like vibrates when somebody walks by, um
John Saunders: Mm-hmm.
John Saunders: yep
johngrimsmo: I have not been able to get these fringes. So then I emailed him and I sent him a video and i was like, Hey man, this isn't working. And he basically said, my buddies and I are thinking of making a blues song called ain't got no fringes because it is so common.
John Saunders: you go Love it. Love it.
johngrimsmo: And, and he basically told me the window that you're looking for of, of distance between the two components and parallelism is 20 microns. So like eight tenths.
johngrimsmo: And I'm just like sliding parts back and forth and like angling them and trying to get within this 8th window to even see like bad fringes.
John Saunders: Okay.
John Saunders: Oh, window. Okay. Okay. Okay.
johngrimsmo: And once he told me that, it kind of clicked. And I was like, oh, this is this is a bit more precise than I assumed it needed to be. like I thought I could just see something on the screen. so I don't know if you ever saw at CMTS one year, I got ah a Williman Vice, like a mini tiny...
John Saunders: Yeah, I have one too.
johngrimsmo: yeah Yeah. So it's been useless at our shop forever. And I saw it and I was like, that's my new stage, like my linear stage. So I double-sided tape that down to my, my, my whatever the thing. And then I'm using that to advance the one mirror and,
John Saunders: Okay.
johngrimsmo: with a piece of putty on top, Radico, and the reference mirror, which is a gauge block on top, and I can like move that around. And I got it. I was able to see really bad fringes if I zoom in close enough, but like huge win.
John Saunders: Great step. Yes. Yeah. Yeah.
johngrimsmo: Epic, epic, epic. And then I was like, okay, so I did that for a couple times in a row. you know I get the kids, and I'm like, look, look. light That's light interfering with itself and canceling itself out. Isn't that cool?
johngrimsmo: They're like, cool, Dad. I'm like, yeah.
John Saunders: All right.
johngrimsmo: I have no shame with sharing with this with my family.
John Saunders: right
johngrimsmo: like
John Saunders: I care, John.
johngrimsmo: Thank you.
John Saunders: ah
johngrimsmo: a
John Saunders: But you're basically saying like your focal range, it may not be the right word, but like it's a very, very narrow.
johngrimsmo: It's called coherence length of white light.
John Saunders: Okay. Okay.
johngrimsmo: Of laser, it's like huge, like really long. But for white light, the coherence length is, call it 20 microns, depending on whatever.
John Saunders: Okay.
johngrimsmo: Super, super short. So the parallelism of everything and distance... between the beam splitter and your two parts, like has to be really, really, really tight or else you just won't see anything.
johngrimsmo: And so I got that. And then next couple days I'd go back and I'd be like, you know, turn it on again and see if I can get it again kind of thing. And then what I did the other night is I i had some really good fringes.
John Saunders: Yeah.
johngrimsmo: I was like, okay, let's see if the rest of the system works. Can I record a video of my screen with fringes? What you want is you want the fringes to scan across the entire surface at a very consistent speed, like call it, you know, one inch per minute or whatever, like steady.
John Saunders: Mm-hmm.
John Saunders: Okay.
johngrimsmo: And then you record the video and then you run it through an algorithm, a bunch of software, which i already have. I figured that out months ago with Cyrus's videos that he sent me. And then you're basically watching every pixel of the of every frame of every video, right?
johngrimsmo: for light to dark transition as the fringes scan across the part, light, dark, light, dark, light, dark.
John Saunders: Yeah.
johngrimsmo: And you're generating a topographic height map of every pixel. So it it goes into every scratch. It goes into every deviation, every speck of dust, every everything to nanometer Z-level precision.
johngrimsmo: So I was able to kind of push on the putty and like stress my part one way and let the putty kind of like just ooze itself back.
John Saunders: Yeah.
johngrimsmo: And I had good fringes and I was like, record. And so i recorded this video and the fringes, they go kind of fast and then they slow down and slow down and slow down. So I ran that through the algorithm and I brought it into the software, Gwydian.
johngrimsmo: And I have an actual 3D super accurate height map, but where the fringes slowed down, like a flat surface ramps way up.
johngrimsmo: Does that make sense?
John Saunders: Okay.
johngrimsmo: And anyway, like huge results. And what i'm measuring is a gauge block, a Michetoyo gauge block with a little engraving on it, like super shallow laser engraving.
John Saunders: Okay.
johngrimsmo: And I can see that in the scan and it's really not good and kind of crappy right now, but it's progress.
John Saunders: No, it's huge.
johngrimsmo: So it's like, I can make this system good. Now it all works. That's where I'm at right now. It all actually works.
John Saunders: can i Can we rewind and go back to the explain like I'm five?
johngrimsmo: Yeah.
John Saunders: So if I buy those, what are they called? The glass hockey pucks that you put over with ah with the ping pong ball and the in the height of the sodium lamp. the The optical, optical flats.
johngrimsmo: Oh, the optical flats, yeah.
John Saunders: Okay. So if you if you have a monochromatic light source and an optical flat and you have a pretty close to perfectly flat part, you can put the op...
johngrimsmo: Yeah.
johngrimsmo: And shiny.
John Saunders: And shiny. Okay. Interesting. And you put the optical flat down on there. The fringes will help you read the situation, but you know, good.
johngrimsmo: Yeah, so go. What you're doing there is you're creating a thin layer of air pretty much between the optical flat and your part. You're touching, you're contacting, but there's always a wedge.
johngrimsmo: It's always touching on the left, high on the right. And what you're measuring is is how many thicknesses of light are in that wedge from where it's touching, right?
John Saunders: or between that's a great, it's insane way to put it.
johngrimsmo: Right.
John Saunders: Yeah.
johngrimsmo: it You know, light's 500 nanometers thick or whatever, 532 for green.
John Saunders: Yeah.
johngrimsmo: um So it's touching on one corner because everything has a contact point at some point. And you can like mess with it and like push it on one side and you lift and you you feel for it.
John Saunders: Yeah. Yes.
johngrimsmo: So every fringe is half a wavelength of light. And you can like measure the flatness, but you're not always looking for how many fringes there are. You're looking for how straight they are.
John Saunders: yes
johngrimsmo: which is what shows the deviations and the contours and things like that. With an interferometer setup, it's non-contact.
John Saunders: Yeah.
johngrimsmo: So instead of having the flat being the shiny object and your part being the shiny object and they're touching, you now have two shiny objects that are distances apart from the beam splitter and they have to be exactly the same distance apart.
johngrimsmo: you know One has to be one inch to the left and the other has to be one inch down to 20 microns or eight tenths.
John Saunders: Yeah.
johngrimsmo: like perfect and square to each other or else you're not going to see anything.
John Saunders: But that's, so that's what I'm trying to make sure i understand this. You could probably a dumb idea or even dumb comment, but you could take a you know, still shot picture of a fringe pattern on a, through an optical flat and put it into the software, but it's kind of, okay, I gonna say it's probably done, but like you could, you're trying to build a system that can actually scan.
johngrimsmo: Yep. And people do.
John Saunders: So like in your gauge block example, let's say that they like really cranked up the power on the fiber laser and they like almost like laser etched out the Mitsutoyo logo. So it's actually, you know, half a thou,
John Saunders: deep, in other words, quite deep, and and it's a they fully etched out the whole logo. The M is like a block M, so you've got like a pretty big area. Your white light can not only tell you, not only can it give you data that the M is embossed or etched lower, but it can kind of quote, unquote, go into that M and tell you the flatness of the M floor.
johngrimsmo: Yeah, exactly.
John Saunders: That's insane.
johngrimsmo: To basically a millionth of an inch resolution.
John Saunders: but it doesn't it have to re adjust the focal length. know it's the wrong term to go into that M because it's now a different, it's more than two tenths away 20 micro.
johngrimsmo: depending on depending on how the system is designed and which part is moving, like you need one axis to move. You need a stepper motor or a piezo movement sensor or whatever.
John Saunders: Okay.
johngrimsmo: I'm using the putty to like squish and like move itself by gravity, but you need a way to drive this forward at, you know, a tenth at a time or something like that at a very steady rate across your scanning range.
John Saunders: Okay.
John Saunders: Yeah.
johngrimsmo: Um, But on that example, not only could I tell you how deep the M is, i would be able to see and measure how much burrs that are popped up on the edges.
John Saunders: Yeah. Yes.
johngrimsmo: going build
John Saunders: Does it have an auto focus feature? Like, is it? No, I'm going to keep moving the stage until I that's honey.
johngrimsmo: one. i'm go to build one
John Saunders: Okay. That's like, I'm telling you, John, this is
johngrimsmo: I literally asked chat yesterday, i was like it was like, how does autofocus actually work in the context of my white light interferometer?
John Saunders: Yeah.
johngrimsmo: Because I asked it in the folder of my interferometer setup. So it knows everything I'm trying to do. And it goes, well, because you're going to have encoders on the axis, like this is going to be easy.
John Saunders: yeah
johngrimsmo: And it's all about like FFT and pixel mapping and all this ah autofocus stuff. And they're like, yeah, yeah, we can do that. I'm like, heck yes.
John Saunders: So cool. so So, so, so, so, so cool.
johngrimsmo: So yeah, it has to be in focus and then basically imagine how I'm figuring it out. Your parts can be on the table, the whole lens and the camera and the beam splitter and the light and everything are going to be moving in a Z stage up and down.
John Saunders: Okay.
johngrimsmo: So I will focus that down onto the part until I'm in focus and then I'll pick my upper and lower scan range.
John Saunders: Mm-hmm.
John Saunders: Yep.
johngrimsmo: And then I'll sweep down and I'll see fringes and I'll record the whole video. So with my setup, I'll be able to scan like a sphere, like like a ball bearing, and I'll get the whole top half of the ball bearing.
John Saunders: yeah
johngrimsmo: um So it's not just for flat, flat objects. I want to use it for contours, for features, for features for machined things.
John Saunders: Doesn't the light have to be normal to the surface though?
johngrimsmo: No, no, there is a drop-off zone.
John Saunders: Wow.
johngrimsmo: Like, i don't know how many degrees, but probably less than 45 degrees before it like just doesn't work anymore.
John Saunders: Okay.
John Saunders: Yeah.
johngrimsmo: But from, call it from zero to 30 degrees, I'll be able to see everything.
John Saunders: Okay.
John Saunders: Dude.
johngrimsmo: So it's going to be awesome. And then if you angle your part to like see it better, um you can scan sideways features.
John Saunders: Yeah.
johngrimsmo: Or i even thought about putting, you know if yeah say, you have to scan the sidewall of a mod vise or something like that, maybe put a 45 degree mirror under the camera so that the mirror is then seeing sideways to your part.
John Saunders: Yeah.
John Saunders: Okay.
John Saunders: Yeah.
johngrimsmo: That might work. I'm not sure yet. But
John Saunders: Dude, this is so cool.
johngrimsmo: yeah, going to awesome. so um I'm excited. it's It's becoming a real. And yeah, it's it's getting there.
John Saunders: Yeah.
johngrimsmo: It's getting there. And what's really fun about like having it at home on my desk in the middle of the dining room, family's walking by all the time. My dad came by. I showed him. um It's fun explaining it to people who don't know what they're talking about, like my kids, because it's alert it's teaching me, okay, you got 18 seconds to explain this concept to 12 year old.
John Saunders: yeah
johngrimsmo: So it's teaching me how to explain it quickly, better in a way that makes sense, which is fun. Because i always like to learn how to communicate better and I want to make YouTube videos about this. And like, you know, I could explain it in two hours, but how do you get the gist in a quick, quick intro?
John Saunders: Please do.
John Saunders: What is the light source?
johngrimsmo: It's an LED.
John Saunders: Okay.
johngrimsmo: yeahp At first I used a flashlight, but it has its own reflector, which I think caused problems, but didn't.
John Saunders: That's right. We talked about that.
johngrimsmo: I think that wasn't my problem. My problem was I was moving things too fast.
John Saunders: Okay.
johngrimsmo: Yeah,
John Saunders: I'm just, I'm getting your projects confused because the, what?
johngrimsmo: yeah I've got a few on the go and they're all the same. I've got a few on the go and sometimes they're all the same.
John Saunders: No, no. yeah and I know they're different, but like, it's just hard to keep track because this is, this has of sort of nothing to do with the telescope lens. It doesn't have depth of field.
johngrimsmo: the vision vision system, um still using the same lens, the telecentric lens.
John Saunders: Correct. That's like sort of.
John Saunders: you you You use a tele-acentric lens on the white light of primer.
johngrimsmo: Yeah, yeah, yeah.
John Saunders: Okay, yeah, I am confused.
johngrimsmo: Which you don't you don't have to do, but it'll be better for bigger Z stage height movements, things like that, edge drop off and all that stuff.
John Saunders: Okay, got
johngrimsmo: And then once I realized, once I like got these things kind of working, my vision system and my white light interferometer are basically the same machine. And I can build one to do both.
johngrimsmo: And I'm like, oh my gosh, this is this is insane.
John Saunders: it. Okay.
johngrimsmo: So now I'm thinking about an XY stage to like scan bigger parts and stitch all the images together.
John Saunders: Yeah.
johngrimsmo: And the program that I'm making is going to have two tabs.
John Saunders: Yep.
johngrimsmo: One's for vision system, one's for white light interferometer, because there's so many overlaps with like camera settings and all this stuff.
John Saunders: Yeah.
johngrimsmo: So yeah, it's it's so cool.
John Saunders: Love it.
John Saunders: Love it. ah I'll build off your sort of like progress comment, which is again, Johnny five freaking, what is satisfying is it's going exactly how I wanted it to.
johngrimsmo: I think.
John Saunders: I got burned out in October.
johngrimsmo: Nice.
John Saunders: I said, it's okay. I had a bunch of travel between October and kind of like, the new year So I i took, i I put the project on hold, i built it r two d two which is super fun.
johngrimsmo: Mm-hmm.
John Saunders: And now I got back into J5, kind of like last week. And I showed a few clips on the shop update video, if folks want to see it.
johngrimsmo: I saw it, yeah.
John Saunders: like, i mean he's he's working, like there's, pretty awesome. But I actually did take a lot of him apart, um, just to start really tidying up. And, um, what do you call it? Like, there's a, not, there's a, uh, term I can't think of that. I want to think of like robustness testing, like, uh, hardness testing or something, but like, let's really make sure he works and I'm not going to have, um, current spikes or ah RF issues or who knows what, um,
John Saunders: You know, a build like this, every time you have a quick connect fitting, you've got to think like, could something happen where an axis moves and actually takes apart a quick connect fitting? And then you're just like, you're in hours of taking them apart to figure out, is it a fuse blow or servo failed? Unfortunately, there are a ton of components in him that will break. like Gears can strip and, you know, there's just a million different Chinese servo drivers and potentiometers, et cetera. so trying to hardness test him and then build him back up.
John Saunders: But the breakthrough I had this morning was um I had been planning on controlling a fair number of his motion and axes with the remote control. And what I'm learning is it's going to be way better to really do almost nothing with the remote control.
John Saunders: ah The remote control is one of those big fancy RC ones that has, you know all these different things on it.
johngrimsmo: Okay.
John Saunders: as it
johngrimsmo: You're not hardwired to him. You're remote. He's just self-s...
John Saunders: Correct. He's already RC'd.
johngrimsmo: Okay.
John Saunders: Yeah. And it's a 20 channel RC, blah, blah, blah. But I'll use the remote control to basically only drive his tracks. Everything else is going to be run through the Polulu Maestros, which are these control software where you can have scripts and sequences.
John Saunders: So for example, his he has two different ah linear motors that cause him to kind of stand erect versus to crouch down.
johngrimsmo: Mm-hmm.
John Saunders: Those were both previously on the remote control. So I could like make him stand up and... sit down with a remote control. That's gone. And this morning I got those moved over to the maestros. So now I'll have him always like go to sleep and kind of a crouched down position. And then when I like, when you and Leif come visit next time and you walk into the room, I'll hit a button and he'll kind of wake up and he'll stand up. And as he's standing up, he'll move his, moot he'll move his arms and he'll,
John Saunders: move his head and he'll look at you and say hello. And that's all thing. I think I could get that done tonight. Like just to be clear, like where I'm at.
johngrimsmo: No way.
John Saunders: So super happy.
johngrimsmo: Yeah. Yeah. So is that like a subroutine that does ah a movement, like a series of movements?
John Saunders: Yes.
johngrimsmo: Okay.
John Saunders: Yep. It's really simple. Most of these are sort of like PWME type things that have like a range from zero to 2000 or 500 to 2000.
johngrimsmo: Okay.
John Saunders: And you just, it's a scripting language that you can also WYSIWYG it by, you can actually move it to where you want and hit capture, capture this frame.
johngrimsmo: Hmm.
John Saunders: And then you can just play back the frames. um But it's, it's, you would, I could teach you this in three minutes, like or at least the basics of it.
johngrimsmo: yeah I mean everything's easy once you break it down and understand it a little bit better
John Saunders: So yeah. just
John Saunders: But that's that's to the thing I keep coming back to that like, just and like all my goal this morning was to get one axis working on, the moved it from the remote to the pull Maestro.
johngrimsmo: what totally one thing at a time paul all i want to see is fringes and then
John Saunders: And I told him, I said, well, yeah.
John Saunders: Yvonne's making breakfast with the kids and I tell everybody and everyone's just like, yeah, like the slow clap.
johngrimsmo: yu yeah you Yeah, literally.
John Saunders: And they're like, oh
johngrimsmo: like like
John Saunders: I'm kidding. Yvonne who listened, I love you. She's listening.
johngrimsmo: Exactly. You're amazing. um Yeah, I sat down with Meg and Claire yesterday for an hour, and they told me everything about their life for an hour. And then like right at the end, they're like, okay, I'm going to bed.
John Saunders: Yeah.
johngrimsmo: I'm like, can I show you guys one thing? It'll take 30 seconds. I just want to show you my new thing.
John Saunders: Yeah. Yeah.
johngrimsmo: It's fun.
John Saunders: No, it's awesome. What was your grinding and stoning process?
johngrimsmo: Um, um, um, um. Quick thing, the grinding wheel we use for chop grinding on the Speedio works great. And since we have the laser to laser the diameter of the wheel, and I did have the guys like mic the wheel diameter with little metal shims so you don't scratch up the mic because it's a one-inch diameter wheel.
John Saunders: Okay.
johngrimsmo: So they would mic it every day and update the offset, the diameter based on the mic because we weren't so much trusting the laser. And I was like, that's dumb.
John Saunders: Okay.
johngrimsmo: Let's trust the laser. um So... One of the problems in failure modes is next day after the coolant around the wheel has kind of dried and has this like, like bigger, bulby diameter, the laser doesn't like that.
John Saunders: Oh.
John Saunders: Mm-hmm.
johngrimsmo: And so it measures it measures big when it measures or it fails because it's like reflecting light and this weird bubble of oil around it. um So I re-strategize the programs to be like rough grind first, don't worry about touching off and then touch it off when it's wet and then finish grind.
johngrimsmo: And like, I think that's great now. So that's good.
John Saunders: Can you have it rough grind to like get knocked the cops up cobblers off thing? Could you have it move over to an air blast and blow it off so that you get most of the coolant off of the.
johngrimsmo: When the coolant's wet, it doesn't matter. But when it's like dried and coagulated after a day, that's when it's a problem.
John Saunders: Oh, okay.
John Saunders: Got it.
johngrimsmo: and And basically, i'm um now that I'm tracking the wheel diameter over time and the tool life, like the grinding wheel is never going to be bigger ever.
John Saunders: Got it.
johngrimsmo: It's only going to get smaller, like unless it loads up or things like that. But it doesn't in our case. um So.
John Saunders: or run out.
johngrimsmo: But not in use, like it'll just it self dresses kind of thing.
John Saunders: Could you not think about if it got somehow accidentally overloaded and could that induce shank run out?
johngrimsmo: Maybe? don't know. But my point is my data shows like very consistent, you know, a few tenths where every cycle with the odd five thou bigger.
John Saunders: Mm-hmm.
johngrimsmo: And like, no, that's not right.
John Saunders: Yeah.
johngrimsmo: So do you have an outlier elimination to like, if it's bigger than one thou, cut it off?
John Saunders: Yeah.
johngrimsmo: but then you don't want to get into this range where it's stuck outside your zone and can't correct itself because you're cut off in both directions is bad. So I'm just letting it run right now because it always self-corrects on the next next scan anyway. But but yeah, the the whole dried coolant thing saved a lot of headaches. and And now the laser is trustable and working and saving everybody time.
John Saunders: Yeah.
johngrimsmo: And it's good. enough
John Saunders: You what that it reminds me of? Okay, good.
johngrimsmo: Yeah. No, finish.
John Saunders: You didn't you do something when you were cutting the foam on the router?
johngrimsmo: Yeah.
John Saunders: You, you would bring the, the foam cutting tool over to like a piece of mahogany.
johngrimsmo: Yeah, a piece of block of wood. i don't know what kind of wood.
John Saunders: Yes.
johngrimsmo: But yeah, and I don't think we're doing that anymore.
John Saunders: Okay. Does that.
johngrimsmo: I think the guys are just picking the foam off um ever periodically and just keep grinding it.
John Saunders: Okay.
johngrimsmo: And I think Grayson told me he's on the same tool for like the past year plus.
John Saunders: Yeah.
johngrimsmo: I'm like, whoa, really?
John Saunders: Oh, great.
johngrimsmo: is that Is that good? Or is that... I know. If it works, it works. Go for it.
John Saunders: Yeah.
johngrimsmo: um But the second thing about grinding is the sharpening stones that Marv made for me.
John Saunders: Oh yeah.
johngrimsmo: And... So using his super scientific sharpening stone method, where he's basically buying the best diamonds and CBN stones he possibly can and creating his own resin and creating his own molds and like purity and perfect.
johngrimsmo: So I had him make me these little tiny um nubs for polishing our bevels that we put in a dye profile machine that like oscillates and circulates the stone.
John Saunders: Okay. Okay.
johngrimsmo: And we've been doing that for years using aluminum oxide stones. So I wanted to try Marv's stones and he made me a custom size in all these different grits. And so I sorted them and I organized them and I gave them to Eric and I said, you can 3D print a tray, but here's the sizes and here's the whatever.
John Saunders: like
johngrimsmo: And try them out let me know. I don't have results yet, but I actually got Eric excited about this once he saw the stones and once he saw some of the results and things like that.
John Saunders: okay
johngrimsmo: And what Marv has noticed is that the amount of junk in in grinding wheels worldwide is like it's like so many impurities, so many other things, big grains, small grains.
John Saunders: Oh, really?
johngrimsmo: He's like, no wonder why people have trouble grinding. So him taking the scientific approach, the German engineering, like... like phd in surface finish um that he has he's like well what if we just have pure diamonds like and like and like a clear a clean resin that breaks away at the right it's that whole grinding theory is like like the grit is held in just enough that when it starts getting a little bit dull it pops off and self-sharpens and not that not that up and yeah so
John Saunders: if It's the answer to everything.
John Saunders: Yeah. Yeah.
John Saunders: we need We need a blues song about fringes. We need a poem about the fracturing nature of of bonded grinding wheels. And then I've, years ago, like maybe even the New York days had started to write an ode to Scotch-Brite.
johngrimsmo: moon
John Saunders: And i don't know it was gonna be like a spoken spoken song or a poem. i I could maybe dig it up, but like there was a point in time where I didn't know what Scotch-Brite was and i discovered it and you're just like this stuff.
johngrimsmo: Right.
John Saunders: is to quote my kids, what do they say? No cap.
johngrimsmo: Yeah.
John Saunders: um
johngrimsmo: Yeah. Telling like I've told people in person who've never heard of Scotchbrite about Scotchbrite and they're like, what?
John Saunders: Yeah.
johngrimsmo: Yeah.
John Saunders: Yeah. I got a tip of the hat. got to give a tip of the hat to Grant. So we have a brother s seven hundred It has a well, I'm calling it a lights out system. A lot of the components are from lights out, but we're doing our own integration, but we're got a UR 10 lights out stuff on it.
John Saunders: And that includes an auto door. The auto door can be controlled from the UR tablet. But right now, the way we did it is um there wasn't, you couldn't manually open and close the door because it was now actuated by a pneumatic cylinder.
johngrimsmo: Yeah, yeah.
John Saunders: Side note, I thought there was a way to buy like a five port that has an exhaust and that you could always maintain manual control. but for whatever reason, that's not what we have, or or maybe it's not possible.
John Saunders: Regardless, um we were talking about like, let's just build a physical button because that's what we want. It's like, I don't care about opening the door by hand. If I can walk up at any point in time and hit a button, frankly, even better.
johngrimsmo: Yeah.
John Saunders: And to his credit, Grant bought a click, another click from from Automation Direct, PLC, and then started doing some research. I think it was a combination of you know research, internet chat, but then also Marcus at Willimon, who's done some work on them. And then Troy and the team at UR, who, um those guys are awesome. They were like UR's integrator the year two years ago. Like they know their stuff on on automation, PLCs, all that.
John Saunders: They were like, we don't know. Marcus had an idea and then Grant figured it out. And unfortunately I can't i can't speak to the quirky technicalities of like the Modbus and how you interface, but um Grant figured it out and then like said, was what I love and makes me so proud is Grant then was sharing that back with Troy was like, No way, that's so useful.
johngrimsmo: Nice.
John Saunders: Like, that makes me so happy that like we're helping each other out because I kind of want that street cred because we're going to get our butt kicked on some automation stuff where I'm going to hey, will you help us out? um And I'm just proud that like now you have a 3D printed thing with these buttons. You walk up, you push it and it opens the door and it's just glorious.
johngrimsmo: That's epic.
John Saunders: Yeah, it's really proud.
johngrimsmo: Solving problems every day, man.
John Saunders: Yeah.
johngrimsmo: Love it.
John Saunders: That's what I got.
johngrimsmo: um Last quick thing on R. Willamon. It started having a tool change error where it couldn't see the tool in the in the carousel. And there's a little sensor that c has like laser light that bounces off the tool and hits the sensor.
johngrimsmo: And we're thinking that was flaky.
John Saunders: Yeah.
johngrimsmo: So... but my guy credit to jeff jeff who like totally figured it out um he's like playing with it and he'd he'd update me every now and then and i was like looking forward to seeing yeah how you figure that out and he determined it wasn't the cable it was the sensor itself so we bought a new sensor 300 bucks from williman came a couple days later installed it works great no problem so like will williman williman yeah our
John Saunders: Oh, yeah
John Saunders: yeah. Yeah.
John Saunders: Wait, Willimon or Kern? Oh, you said Willimon. Sorry, I'm thinking blank. Yeah, yeah, yeah.
johngrimsmo: you know Our 20-year-old Williman, which we run every day now.
John Saunders: Yep.
johngrimsmo: He does. And it's fantastic.
John Saunders: Good. Are you making pen caps?
johngrimsmo: um um um
John Saunders: Clips?
johngrimsmo: Fjell, knife, pocket clips.
John Saunders: Yeah, pocket clips. Okay, great.
johngrimsmo: Yep.
John Saunders: Do the saga clips still get made on the Nock?
johngrimsmo: They do still get made on the Nock, yeah.
John Saunders: Okay.
johngrimsmo: They could be made on the Williman, which we haven't yet.
John Saunders: Yeah, sure, sure.
johngrimsmo: So for the whole week, last week, we pretty much had all three lathes down.
John Saunders: Got it.
johngrimsmo: um
John Saunders: That warms my heart.
johngrimsmo: yeah Which is so weird because Jeff runs all of them plus the surface grinder. And he's like, I mean, he's busy trying to fix them, but he feels so bad because he's like, he's like, all my machines are down at the same time.
John Saunders: Sure.
johngrimsmo: What the heck? This never happens.
John Saunders: With the knock and the Taurus redoubt?
johngrimsmo: Tornos has this weird um live motor servo alarm that we can't figure out. We emailed Tornos. They haven't gotten back. And then the NAC started of making a weird bearing noise in the live tools.
johngrimsmo: And so we dug that in, pull all the tools out, talk to service.
John Saunders: Okay.
johngrimsmo: You know, it's a thousand dollar part. and And then we're like, okay, send it to us. And then I'm like, well, how do how do I install this thing? And they're like, oh, we're going to have like, not quite take the turret off, but almost take the turret off. So we're going to take two days to come in and do that. And so we haven't committed yet, but we're like, oh no.
John Saunders: So it's not a BMT bolt on thing. It's actually inside the turret.
johngrimsmo: It's... it It is a bolt-on live tool, but what drives it inside the turret, I think, is the problem.
John Saunders: Yeah, exactly. That's what mean.
johngrimsmo: Yeah.
John Saunders: Yeah, yeah, yeah. Okay. Oh, I'm sorry. That sucks.
johngrimsmo: So that's all stuff.
John Saunders: Yeah. Yeah, I hear Yeah.
johngrimsmo: And all three machines, I'm out of the loop. Like, I could CC on the email, and i'm like, I advise when I can, but I got my other things going on.
John Saunders: Yeah.
johngrimsmo: I've got interferometer to build.
John Saunders: Dude, love that. Well, speaking of feeling like we're doing a good job on leadership, last year we did our or kind of 2025 improvement list. There was an aspect of that that drug on a little bit longer and took up more of my time, which is ah okay, but like it was real work. This year, we're crushing it. one of our Our quick laundry list, put an automatic garage door opener on our bay door,
John Saunders: Quote is in and they've already, as Serena's handled it, they're coming to install it. We needed small metrology stuff, new mics, new pins. The team just ordered what they needed for NRC.
John Saunders: We had a plumbing repair that art needed to happen with a double venting, something that wasn't causing a pipe to condensate, like annoying because nothing was wrong, but it was in certain climates causing a leak.
John Saunders: And that's already been in a fix. We're having a flooring guy come to floor our utility closet because it's the old tile from pre-
johngrimsmo: Yeah. i saw that in your video. Yeah.
John Saunders: Yeah, that's fixed. Um, and then Garrett did a really good job. I should put this up on Insta. He did a really good job at making a custom 3d printed gridfinity bin that like open and close, it's like a little toolbox that has gauge pins for certain products. It looks so dialed.
John Saunders: Um, and so it's kind of like, here are all these things we wanted to do. And the goal when we sat down was like, let's pick what we can do and pick an arbitrary date, February 15th. But like, don't let's, we're not messing around here. Like, let's do it.
John Saunders: We're done. Like, Yeah, like the garage door hasn't come, but I don't care.
johngrimsmo: insane it's happening yeah that's awesome yeah exactly that's awesome and then it's like what's not what's the word not crowdfunded but it's like team ideas like you all came together you all wrote them down you listed them out you prioritized and you executed and it's it's done here you are two weeks later yeah man
John Saunders: Like it's all good to go. Yeah. So it feels, now we get back to work.
John Saunders: Yeah. Yeah.
John Saunders: Yep. Feels good. So now we gotta get the horizontal sold and then get more robots and verticals in here.
johngrimsmo: and any bites good
John Saunders: Uh, short answer. No. Um, I'm going to keep listening online before I up leveraging a ah broker. Yeah.
johngrimsmo: Sweet.
John Saunders: Thanks for asking.
johngrimsmo: I mean, is that kind of, that's a fairly common machine that's like should be fairly well. But, you know, sometimes we look at like weird machines that are like perfect for us and imperfect for everybody else in the world that would be impossible to sell.
John Saunders: Yes.
johngrimsmo: Food for thought. i don't know, but not that.
John Saunders: It's a big, no, no, you're right. It's a big machine for sure, but it's also a very standard 400 millimeter horizontal with a appropriately sized ABC and pallet tool crib. Like it'll sell.
johngrimsmo: Yeah, yeah
John Saunders: But no, it's not, I mean, if it was an I would have sold it that afternoon. Like, yeah, yeah.
johngrimsmo: Exactly right. Yeah. Yeah. Yeah. right, man.
John Saunders: See next week.
johngrimsmo: See you next week.
John Saunders: Take care.
johngrimsmo: Okay. See
