Become a Creator today!Start creating today - Share your story with the world!
Start for free
00:00:00
00:00:01
#398 Python script to create the perfect toolpath!
0 Playsin 7 hours
TOPICS:
- NYCCNC has been purchased by DSI
- Hardmilling
- Kolsterizing process by Bodycote
- 5 axis automation and robotics
- DOITS finally work
- - Shrink the project
- - Cannot think about it
- Grimsmo grinding improvements
- - Python script to create the perfect toolpath!
Recommended
Transcript
Introduction and Entrepreneurial Life
00:00:00
John S
Good morning, welcome to the business of machining, episode 398. My name is Sean Saunders.
00:00:07
johngrimsmo
And my name is John Grimsmo.
00:00:09
John S
And John and I talk each week about life. What we're doing well, what we're not doing well, and we happen to record as a podcast.
00:00:18
johngrimsmo
Yeah.
00:00:18
John S
Yeah, right?
00:00:19
johngrimsmo
I try to center it around business, but, uh, business is life for us being owner, owner, entrepreneurs, you know?
00:00:21
John S
Yes. Yeah, I don't think there's, not a big separation between those two, right?
00:00:27
johngrimsmo
Yup. Yup. Yup.
00:00:30
John S
Um, can I just go test her out for that announcement?
00:00:31
johngrimsmo
And I go for it, please.
NYCCNC Website Transition
00:00:35
John S
Um, I think I've sort of like sort of teased about this or talked about it a little bit, but now it's kind of like formally fully out there. um The DSI folks have taken over the and NYC CNC website and are basically going to port it into their DMC.
00:00:49
johngrimsmo
interesting yeah yeah
00:00:51
John S
Yeah, so this is such a natural conclusion and um I like to be super candid and honest about these things rather than like paint the rosier picture. I wasn't putting my heart and soul into NYC CNC in terms of like actual active new content and really,
00:01:06
John S
um It's funny, because now that it's all over, you look back, you're like, oh, could I have built out more of a team? But I didn't. And um what I found, frankly, again, being super candid, was I was reaching out to Phil, to Devin, to CJ, or Angelo, to ask them, hey, what is this latest and greatest thing in Fusion? um Meanwhile, they're putting together their digital manufacturing collective, which is kind of a much newer, fresher website around the Fusion community. And so.
00:01:35
John S
um They are moving all that content over to there. And I think nyccnc.com is still up, um but it'll soon redirect into the DMC.
00:01:44
johngrimsmo
Can you update me on what nyccnc.com is and has been for the past six years?
00:01:50
John S
Yeah, totally fair. um We obviously have our YouTube channel, NYCCC, we still have that channel. ah Six years ago, I realized there's so much stuff that I wanted to add to that goes above and beyond just what's on the YouTube.
00:02:03
John S
So like, little examples would be most and I think like, of the six or 700 WordPress articles, only 40 or something were behind the paywall.
00:02:12
johngrimsmo
Hmm.
00:02:12
John S
But um Most of the stuff was free, some of the stuff was behind a paywall, and the idea was like, hey, here's the Renishaw code, or here's the F3D file, or here's the example, here's the PDF, here's the Sandvik book, all these kind of content articles as a resource to look up how to do stuff.
00:02:28
John S
You could go there and you could look, hey, here's every video we've done on post processors. um Again, most of that was most that was hosted on YouTube, but obviously this gave us the ability to control how it was laid out, bingo.
00:02:32
johngrimsmo
Yes.
00:02:37
johngrimsmo
Mm-hmm. Categorize it. and Yeah, I've been there for thread milling calculations or something like that.
00:02:41
John S
um
00:02:45
John S
probably Alex said that, probably the best home run example.
00:02:47
johngrimsmo
Yep.
00:02:48
John S
So, I mean, literally thousands and maybe tens of thousands of people have downloaded that treadmill sheet. um
00:02:53
johngrimsmo
Yeah, the Excel spreadsheet. That's right.
00:02:56
John S
And we had business stuff that was, I mean, it was good, but it wasn't something we were refreshing a lot about, hey, accounting for manufacturing, actually, it really is good stuff. Why you should or shouldn't patent ah books to read at different points of the journey.
Fusion Fridays and Community Building
00:03:08
John S
I love, I'm still passionate about it. I just wasn't, you you kind of realize you've got to pick your, yeah, bingo.
00:03:13
johngrimsmo
Can't do it all. Yep.
00:03:16
John S
So shout out to the DSA guys. What I like about what the team are they're doing is trying to build a community around fusion. So they're hosting free fusion Fridays where you can get on and ask questions about where you're stumped. They do. It's a great also to be a natural um What do you call it like cross sell for them because they want to build this community they want users they want to add content but also for folks that want more you can engage them.
00:03:43
John S
Like you have john right like hey oh this is all free across but i want to actually pay you to help me on this part or this camp file or this fixture or this post processor right.
00:03:52
johngrimsmo
Yep.
00:03:53
John S
Um, that's all right. Frankly, i I haven't really used them a ton for the paid stuff. Um, and so I wanted to see, cause you said they've done your post on the post.
00:04:03
johngrimsmo
On the current, yeah, just just recently, Phil finished it up last week, um moving myself away from Camplete and towards ah using fusion simulation, machine simulation directly.
00:04:06
John S
Okay.
00:04:13
John S
Directly. Yeah.
00:04:15
johngrimsmo
um I haven't had a lot of time to debug and test and properly finish, like like test out the posts, um but I've played with it.
00:04:24
John S
Okay.
00:04:25
johngrimsmo
The code looks good. I just still need to run it on the machine. um But it's great.
00:04:30
John S
Sweet.
00:04:30
johngrimsmo
I mean, Phil is, what did Phil say?
00:04:31
John S
Good.
00:04:33
johngrimsmo
He's made 278 posts this year or something like that. like
00:04:36
John S
That's crazy. and
00:04:36
johngrimsmo
It's insane. um He knows what he's doing, and so does the
Gratitude to NYCCNC Community
00:04:40
johngrimsmo
rest of the team. So if it's just so good to have them kind of in your corner, um either Phil as a friend of ours, but a DSI as a company.
00:04:46
John S
Yeah. Yeah.
00:04:50
johngrimsmo
I've had multiple phone call video chats with them, and they really know their stuff.
00:04:55
John S
It's awesome.
00:04:55
johngrimsmo
And being a customer of theirs, now that I'm ah buying my fusion licenses through them, It's, you know, at first I was like, why would I do that when I could just go straight through Autodesk? But it's really starting to make sense and pay off long-term to have their connections, you know?
00:05:11
John S
And that's, ah I frankly need to ask them because we do the same. We buy our fusion licenses from them, but I think that might get you some kind of complimentary support, but I think there's a different level of support where it's like, Hey, can I have an hour long call to help me with this pencil tool path where I'm stumped on it?
00:05:17
johngrimsmo
It does.
00:05:26
johngrimsmo
I don't know.
00:05:27
John S
I don't, yeah.
00:05:27
johngrimsmo
Yeah.
00:05:28
John S
And look, in ah and a big thank you to everybody along the way who was a, you know, either leveraged, utilized, supported, was a member at NYC CNC. I appreciate it. I love that community. We had a good forum. um Lots of, you know, a big part of the story that I'm, um I want to continue over there on the on the d DMC.
00:05:49
johngrimsmo
Was
Future of NYCCNC and YouTube Content
00:05:50
johngrimsmo
the forum still active at all?
00:05:53
John S
Yes, it was not as active as it was two years ago, probably.
00:05:55
johngrimsmo
Sure.
00:05:56
John S
um And the forum was, do I?
00:05:57
johngrimsmo
Do you ever look at it? Yeah.
00:06:00
John S
I don't, I, someone could probably find, ah call me out on this being slightly incorrect, but I don't think in the six years it was up there was ever a post I didn't respond to.
00:06:11
John S
That was kind of the idea, too.
00:06:11
johngrimsmo
Oh, okay.
00:06:12
John S
like when know and the hey Oh, absolutely.
00:06:12
johngrimsmo
Yeah, so you did.
00:06:14
John S
Every time.
00:06:15
johngrimsmo
Yeah, I got, got it.
00:06:16
John S
In the heyday of our YouTube a days, that the idea was if you want to actually you know get our attention, just become a member over there.
00:06:23
johngrimsmo
Yep, yep.
00:06:24
John S
And it worked great. like Lots of people did, and some people didn't, which is fine.
00:06:27
johngrimsmo
Sure.
00:06:28
John S
um So yeah, it was good.
00:06:30
johngrimsmo
That's awesome. So the future for NYC CNC, are they just taking the articles? You keep the name on it, or are they living on the brand of NYC CNC? I don't know.
00:06:41
John S
they They get the domain and they're gonna leverage, I mean that, I think that name has some cred um or recognition.
00:06:46
johngrimsmo
Yep. But it's you as well.
00:06:49
John S
um Again, I kept the YouTube channel and we'll still do some stuff with that, not as much as we used to.
00:06:52
johngrimsmo
Yep.
00:06:54
John S
um Yeah, that was the punch line.
00:06:58
johngrimsmo
Cool.
00:06:58
John S
Yeah, yeah.
Johnny Five Robot Project
00:07:00
John S
So it was good. um The other really fun thing I was gonna mention last week and then we got and a great, conversation but I spent like a lot of time over the holiday break on Johnny five and it is he's coming along super well um I'm not documenting a ton of it I will at some point at least catch folks up on it publicly I put up some stuff up on Instagram for sure but um
00:07:13
johngrimsmo
Oh yeah.
00:07:21
johngrimsmo
Yeah. Yeah. What kind of things?
00:07:27
John S
Oh, like the head, elect electronics in the head are mostly done. So the eyes, the irises open and close, the lights turn on, the heads, eyeballs move in and out. They articulate. I last night got the hydraulic system, I haven't charged it with fluid, but the motors for the hydraulic system are articulating, which causes his neck to move in and around.
00:07:47
johngrimsmo
yeah Yeah.
00:07:48
John S
The eye flaps like, this is now like, it's It's real. like Lots of electronics, lots of servos, master controllers, software. It's super fun.
00:07:58
johngrimsmo
Man, watching you work on this, I mean, I love the movies as a kid, but I never thought too deeply about it.
00:08:02
John S
Yeah.
00:08:03
johngrimsmo
Watching you work on it and talk about it makes me realize how utterly insane these were in the eighties to be made for a movie in in like how long did they spend making them?
00:08:07
John S
yep Yeah, you know it really was. Yeah. Yeah.
00:08:14
John S
months I mean, and they made, I have a bunch of like, the input ink team has compiled a bunch of footage of stuff, but I mean, there were 18 robots.
00:08:16
johngrimsmo
Oh my gosh.
00:08:25
John S
And now those were, many of those were partials, but still like it was, it was incredible.
00:08:27
johngrimsmo
Sure. That's insane.
00:08:30
John S
Absolutely John, like nuts that they did all this.
00:08:33
johngrimsmo
Wow.
00:08:33
John S
Huge respect. I guess not my world. I have no experience with that. My guess is it's also like, you can do a lot when you have budgets that add multiple more zeros on it than you know we do.
00:08:42
johngrimsmo
Yeah.
00:08:44
John S
And a lot of the parts were vacuum formed that were actually machining. But we also, I mean, don't get it wrong, 2025, you got all the advantages of the world between CNC, 3D printing, you know PCBs that are, I mean, so cheap, they're almost free.
00:08:52
johngrimsmo
Yup. Electronics and yep.
00:08:57
John S
like yeah It's really cool.
00:08:59
johngrimsmo
That's cool.
00:09:01
John S
so
00:09:01
johngrimsmo
So do you still have the same kind of goal for Johnny Five as you did on day one, which as I remember was, you know, obviously share the whole project, but take it to schools, get kids into robotics and like have a cool thing to show.
00:09:14
John S
hundred percent 100%.
00:09:15
johngrimsmo
Yeah.
00:09:16
John S
Yeah. i My goal had always been a telemetry suit to let a kid maneuver him around. that's There's enough momentum behind how a couple of the other build guys have built it with a...
00:09:22
johngrimsmo
Hmm.
00:09:27
John S
ah You have these remote controls and then you have preset scenes and things where he does actions. um So that's the route I'm going to go because I know how to get that done.
00:09:39
John S
I don't think that actually precludes you from later adding something like a telemetry suit, basically like a jacket a kid could put on.
00:09:45
johngrimsmo
yeah
00:09:45
John S
But um the motion of him alone will be plenty good. And so he's he is the head and the neck with the main focus.
00:09:49
johngrimsmo
okay
00:09:56
John S
There's the track drive system. I've got to do some electronic work on it in the center torso, but
00:10:00
johngrimsmo
Yeah.
00:10:01
John S
What I'm doing now, literally William and I, my wife is like super happy with me right now. There's been a huge mess in our living room at home, which is where I'm now working on him because it doesn't work for me to work on him at Saunders.
00:10:12
John S
I just, um when I'm here, I'm here.
00:10:12
johngrimsmo
Yeah, yeah.
00:10:14
John S
So I've been taking him home in small parts and then it's just a mess. And so last night William and I assembled a Harbor Freight US General Toolbox. They're actually really nice for the and price.
00:10:23
johngrimsmo
Yeah.
00:10:25
John S
So we now have a brand new toolbox in our living room. um But that's going to help me organize a lot of his parts and the tools I use. And then um the headless robot has been sitting right there.
00:10:38
johngrimsmo
Okay.
00:10:38
John S
um And so ah he's actually going to come home, probably. I don't think I'm going to get to it today, but like.
00:10:46
johngrimsmo
Wow.
00:10:47
John S
So that was a little fun update on Johnny five.
00:10:51
johngrimsmo
Love it.
00:10:52
John S
Yeah.
00:10:52
johngrimsmo
big toolbox in your living room. Tell me you own a shop without telling me you own a shop. like
00:10:59
John S
Yeah.
00:11:02
johngrimsmo
That's exciting.
00:11:05
John S
All right, what are you up to?
Arc'teryx Collaboration
00:11:06
johngrimsmo
um Cool thing happened today, sort of not work related, but sort of work related too.
00:11:06
John S
What's in your mind? What's in your mind?
00:11:12
johngrimsmo
um We got an email last month from Vincent, who is the senior product designer at Arctyrix.
00:11:19
John S
No way.
00:11:19
johngrimsmo
you're familiar with the clothing brand Arcterix, right?
00:11:20
John S
Yeah. Yeah, sure.
00:11:22
johngrimsmo
And he's like, dude, love your stuff. Need some knives. You want some gear? And we're like, yes.
00:11:29
John S
Yeah.
00:11:29
johngrimsmo
So we we just coordinated an exchange. And some of our guys are huge Arcterix fans. They've bought all kinds of stuff from them. I've never purchased them, but ahve I've known about them since the 90s.
00:11:42
John S
It's Canada too, right?
00:11:43
johngrimsmo
Yeah, they're in Vancouver. Yep. And so we worked out a deal and sent a bunch of knives and a bunch of pens. And they just, like just today we received the box with, you know, total Gore-Tex outer shell, the beta AR, down the atom, ah the Squamish you know she windbreaker shell.
00:12:01
johngrimsmo
I got some, some under layers, some underwear, long underwear, and
00:12:02
John S
Perfect.
00:12:06
johngrimsmo
Eric got this down-filled beta AR, huge puffy super warm jacket, bunch of gloves, like it's pretty cool.
00:12:13
John S
Sick.
00:12:14
johngrimsmo
Pretty happy with it.
00:12:15
John S
That's awesome.
00:12:16
johngrimsmo
And they received their knives yesterday and they're like, holy cow. Like I knew they'd be good. I'm totally blown away. And they love that it's Canadian and I love that they're Canadian and you know, it's it's really cool.
00:12:25
John S
Yeah, it's awesome go that's awesome.
00:12:27
johngrimsmo
So that was just a feel good, awesome, you know, friendship that's made.
00:12:32
John S
awesome. Yeah, sure. No, that's always always funny to me too. Like our characters is a real brand. Like I own two of their jackets, like the hard fleece stuff.
00:12:38
johngrimsmo
Absolutely. Yeah.
00:12:40
John S
They're super nice. But like, sometimes you forget that there's like real people, even decision makers that are like, Hey, like you love your stuff.
00:12:46
johngrimsmo
Yep.
00:12:49
johngrimsmo
Yeah. lot you You know about us? Like what?
00:12:51
John S
Yeah. Right. That's awesome.
00:12:53
johngrimsmo
So that has worked out really good. And they, you know, I see them as almost, if not the top of the food chain, as far as outdoor gear, like not only hype, but quality and, uh,
00:13:04
John S
Here.
00:13:04
johngrimsmo
an actual use case, and they now see our knives as the same thing.
00:13:08
John S
Sweet.
00:13:08
johngrimsmo
And
Fjell Knife Production Journey
00:13:09
johngrimsmo
they're like, I've owned other names, other brands, you know, this, that, and the other, and they're like, I'm never, yeah, now now they're hooked.
00:13:15
John S
Good for you.
00:13:16
johngrimsmo
So that's fun.
00:13:17
John S
Awesome.
00:13:18
johngrimsmo
So that was fun. Let's see, in other news, dude, Fjell is making insane progress in headway. um I don't even know how many parts we've made.
00:13:28
johngrimsmo
Probably almost 50 handles are machined.
00:13:31
John S
Awesome.
00:13:32
johngrimsmo
And they're going great. I was worried about like the high-feed end mill that's doing the big slot. um We just replaced like, I think the first one and it still looks really good.
00:13:43
John S
Nice, okay.
00:13:44
johngrimsmo
And a lot of tools, like some of the tools we're replacing on a life, we're learning the life, it's going great. So the handle's awesome and it's holding tolerance um kind of better than I was expecting. So we're having to massage it every,
00:13:54
John S
hundred
00:13:57
johngrimsmo
10, 15 parts by a 10th, as opposed to every other part, like I was worried about. So that's awesome. I've been working on hard milling the field blades the past few days, and it is going so stunningly well.
00:14:09
johngrimsmo
Like with the right mold, you know, hard mills, it's just, it's a point and shoot.
00:14:12
John S
yeah yeah
00:14:14
johngrimsmo
And just this morning, I pulled the tools out of the machine, inspected them out of the microscope at full zoom. And I'm like, have I used these yet?
00:14:22
John S
Good.
00:14:22
johngrimsmo
They look absolutely brand new. They only have about 40 minutes of life on them so far, but that's machining 61 Rockwell stainless steel.
00:14:28
John S
It is crazy.
00:14:30
johngrimsmo
sealess deal like they look I have to grab a brand new tool and hold them end to end on a microscope just to see if the little imperfections that I'm seeing, yeah, they're on a new tool too.
00:14:42
johngrimsmo
hey
00:14:43
John S
That's why our Rego Fixation, we 3D printed a end mill holder with new and old.
00:14:45
johngrimsmo
Because you can't tell.
00:14:49
John S
That way when you swap tooling, you don't like you just don't have to think about it. You're like, okay, great.
00:14:52
johngrimsmo
Yup, yup.
00:14:54
John S
It's funny.
00:14:55
johngrimsmo
So that's going phenomenally well.
00:14:55
John S
yeah I thought you, I knew you were gonna just love the hard milling stuff.
00:15:00
johngrimsmo
I used, I'm not hated, but I used to find it so unreliable and so difficult to just count on tool life, all that stuff. But like, wow, with the proper tools, the, I mean, these are $80 end mills for a two millimeter little thing.
00:15:16
John S
Yeah.
00:15:16
johngrimsmo
It's, it's just, holy cow, just phenomenal.
00:15:20
John S
Yeah, good.
00:15:21
johngrimsmo
I'm so blown away. I want to hard mill all the things now, like, yep.
00:15:26
John S
Yeah, no, I know, right? What really helped us too because it helped us relax some of our manufacturing tolerances downstream prior to heat treat because it's like, hey, it totally does weird things and you just don't know.
00:15:34
johngrimsmo
Yeah. Yeah. Totally. Cause hatred does weird things. Yep.
00:15:41
John S
um So it's kind of like, oh yeah, this makes sense.
00:15:45
johngrimsmo
yeah and And not only consistent weird things, it does variable weird things. So it might warp features. It might expand surfaces a different part to part.
00:15:54
John S
Yeah.
00:15:54
johngrimsmo
and And you start to look down literally to the molecular grain structure and the distribution of chromium, vanadium, all the metals in the steel that will affect all this.
00:16:05
John S
Mm hmm.
00:16:05
johngrimsmo
And nothing's perfect. um Yeah.
00:16:09
John S
Did you ever randomly made me think back? You were talking to a company that has some kind of a kind of kind of call unusual process for stainless is ah whatever it was aimed at.
00:16:17
johngrimsmo
Mm-hmm.
00:16:19
John S
And did that go anywhere?
00:16:20
johngrimsmo
So it's for 17, 4PH stainless steel. It's called Kholsterizing by the company Body Coat.
00:16:25
John S
Yes.
00:16:27
johngrimsmo
And I've had a couple of people reach out over the past year or whatever it's been remembering. I mentioned it and then asking like, I've got an application. Can you give me their contact? And, um, it didn't work for us because the, this is for the button on our field knife.
00:16:42
johngrimsmo
The substrate was still 45 Rockwell, but the surface is like 70 Rockwell or something.
00:16:42
John S
Okay.
00:16:48
johngrimsmo
So I would still be able to dent through the hard surface and dent the substrate because it was still kind of soft. So for our application, I didn't like it. However, I have some other applications like the button on our pen um and some other things that would super benefit from this. And I sat through like an hour.
00:17:07
johngrimsmo
um PowerPoint presentation with one of the guys, and he explained it all to me. And my super wrong layman's example is molecularly, they're basically shoving carbon carbon atoms in the surface of the part between all the open atoms that are there.
00:17:23
John S
Yeah.
00:17:24
johngrimsmo
And it just makes it harder.
00:17:26
John S
Yeah.
00:17:26
johngrimsmo
um I don't know if that makes sense to a metallurgist, but that's what I got. That's that's what I walked away from the conversation with.
00:17:31
John S
Yeah.
00:17:33
johngrimsmo
And does it ever work? um
00:17:38
johngrimsmo
total massive change in surface hardness.
00:17:39
John S
yeah
00:17:41
johngrimsmo
It's like case hardening, I guess, but better. um And they've got all these tests and experiments on the the hardened layer shearing off, like if they break the part in half, and it's getting electron microscope of that surface versus a case hardened part versus whatever.
00:17:53
John S
Oh.
00:18:01
johngrimsmo
And the adhesion is insane.
00:18:01
John S
Yeah. Sounds like it's a thin layer of 70 HRC though.
00:18:07
johngrimsmo
Yeah, yeah, fairly thin. I forget they have a number. I don't remember what it is, but fairly thin.
00:18:11
John S
I got a couple foul though, really, like it's more like I'm making again, I don't know, I'm tired here, but like it's more like it's scratch resistance with light with a sharp tool that's lightly pressed upon it.
00:18:13
johngrimsmo
Yeah.
00:18:21
John S
Like when you try to take a file on a hardened block, it doesn't cut.
00:18:24
johngrimsmo
Yep, yep. Yeah, e yeah it might it might fail a Rockwell test, you know?
00:18:25
John S
But to your point, like the whole idea of a Rockwell test is how far can it dent a ball into a part. But if you're, if you've got a tooth out right tooth out thick hard skin, but you're pushing a ball in it's going to just inherit the softness of the to succumb to.
00:18:35
johngrimsmo
and
00:18:39
johngrimsmo
Mm hmm. So it's application specific, because if you have point contact load, you're still going to possibly deform the thing.
00:18:43
John S
Yeah.
00:18:47
johngrimsmo
But if you have a bearing load where it's flat, two flat surfaces or two rubbing surfaces, like ah a shaft against a rotating thing, they basically have rotating shafts with no lubrication.
00:18:47
John S
Yeah. Right.
00:18:58
johngrimsmo
And they just see how long it lasts versus the galling effects of stainless steel to untreated parts and then to treated parts.
00:19:02
John S
Right.
00:19:05
johngrimsmo
And they have videos on their YouTube channel about this. It's like, what
00:19:10
John S
Cool.
00:19:10
johngrimsmo
Anyway, why do you ask?
00:19:13
John S
Oh, it literally just made me think as you were talking about heat treat. And then um what I remember what stood out was that process claimed to have like almost zero deformation to the part size or like changes
Colsterizing and Material Hardening
00:19:24
John S
to the part size, which I kind of wanted to call BS on, but it sounds like I'm wrong.
00:19:24
johngrimsmo
Yeah, we measured it. Totally. No, I couldn't measure a single tenth of difference and visually they look the same too. Like no color change, no, no scale, no nothing.
00:19:34
John S
Yeah, that's crazy.
00:19:36
johngrimsmo
I was like, I almost can't tell the difference. Like I literally can't tell the difference now other than the one little baggy says the decay on it, it says colsterizing, which is impressive.
00:19:44
John S
ah Yeah.
00:19:48
johngrimsmo
Like no change in appearance, no change in size and massive change in surface hardness. So it was very cool.
00:20:00
John S
That's what changed topics, but the, had a guy reach out and actually had a good conversation this morning. He had kind of had a DIY system or became DIY because frankly, the integrator just, you know.
00:20:13
John S
the integrator got in over their head and basically gave up, which is a story I feel like I hear kind of, not too often, but you tend to hear about frustrations from these.
Automation Exploration with Robots
00:20:23
John S
Regardless, what he has, it kind of made the light bulb would go off for an automated five-axis system. His system has one robot, and that robot can load, ostensibly, more than one machine. I don't recall how he has his running.
00:20:36
John S
but um Take XYZ brand five axis, I don't care if it's the Haas or Herma, like you put a zero point system on it, it and it can now receive pallets. Those pallets could have, let's say that we have five different pallets that have different size style or size of work holding on it, like you know small part vice, big part vice, round part vice, whatever.
00:20:55
johngrimsmo
Yep.
00:20:59
John S
So the robot loads in pallet number one, two, three, or or five, then that same robot assumedly can change its end effector, can then start loading in material for that part and unloading it when it's done and flipping it between op 10, op 20.
00:21:13
johngrimsmo
Okay.
00:21:16
John S
This is kind of what the Hermla high-end system does, like the RS1, but it's you know unobtainable price-wise. And the beauty and simplicity is you can you know yeah I understand it's simple.
00:21:31
John S
Like the key, what I'm learning with automation is you need to keep it simple. So um having one pallet that has one work holding on it that has the jaws and the stroke set up for one part is simple.
00:21:44
johngrimsmo
Yeah.
00:21:44
John S
That's easy to swap in and out. Then loading parts material in and out of that, that's not necessarily simple, but I understand the simplicity of that path.
00:21:46
johngrimsmo
yeah
00:21:53
John S
um So that's kind of what I'm, versus, okay, so versus the harmless system that I'm still learning about called the RSO5. It's great, but...
00:22:04
John S
It would be a a home run if you were only doing clock clock slides or some one part size. It can change the grope robot gripper. That's really cool. No problem to do that. It can even change the vice jaws.
00:22:17
John S
It's on the same vice, but it can actually use that machine.
00:22:17
johngrimsmo
you
00:22:20
John S
The robot can actually go in and remove the vice shots and put new ones in. So part of me is like, Oh, that's a slam dunk. Except it ends up that I, and I haven't gotten super details, but number one, I think the price goes up a bunch.
00:22:31
John S
Number two, the complexity and inevitable, like. If there's a frustration or need to reprogram it, it's just more complicated.
00:22:36
johngrimsmo
Yeah.
00:22:39
John S
So the idea that, why can't we just switch out the work holding when you switch between these different five-part families and then the same robot can load and unload the material?
00:22:39
johngrimsmo
Mm-hmm.
00:22:48
johngrimsmo
Mm hmm. And would are you leaning or considering a DIY ish route?
00:22:57
John S
Still learning. um I actually want to call a couple of people I want to call these people except I want to wait till I learn more to ask better questions.
00:23:05
johngrimsmo
Yeah, yeah, yeah.
00:23:06
John S
um The reason, of course, no. Of course, you'd much rather have a proven solution, for sure, and for conversation.
00:23:12
johngrimsmo
Yep.
00:23:14
John S
But you know the the takeaway from this guy's story this morning was like kind of like the integrators know more than you. You're going to get posed. like i didn't struck I didn't clarify in the contract, which is you know well into the six figures, that there needs to be a runoff that includes it being able to run X number of hours unattended, which probably, he didn't say this, but probably led to some of the, like, frustration about whether they considered the project done.
00:23:37
John S
in And then it ends up, well, I now, you know, I think the point is if you're gonna have a robot system to be a key part of your shop, you need to know enough to tweak it, adjust it, just like what you see, like Lawrence does.
00:23:37
johngrimsmo
yeah
00:23:48
John S
CJ even with his yeah UR, but I think Lawrence is probably a better example of using a lot of different type of robot stuff. You need to know how to ah adjust it and so forth.
00:23:54
johngrimsmo
Yeah, yeah, for sure.
00:23:57
John S
um So, Again, I don't want to DIY it, but part of me also, Oh, are you our robot came for the Wilhelmin last week? Um, it's sitting in a box, but they're kind of, they come in.
00:24:05
johngrimsmo
Oh, no way.
00:24:08
John S
I think two weeks to do the install. We weren't, we weren't in a rush show.
00:24:10
johngrimsmo
Ooh.
00:24:12
John S
It's going to be a while, but, uh, we, we had the yellow FANUC robot when we bought that use Robo drill, but we never did anything with it.
00:24:12
johngrimsmo
Is that your first official robot? Robotic arm?
00:24:18
johngrimsmo
Oh, that's right. Yeah, yeah, yeah. Yeah.
00:24:22
John S
Um, part of me is wondering. because I don't think we're going to use that robot a ton on the Wilhelmin, just when you need it, you need it. So part of me is like, well, wait a minute. I wonder if I could use that robot to build my own proof of concept automation system.
00:24:36
John S
Because I don't need the five axis. I just need to load, pick up and move one tombstone.
00:24:40
johngrimsmo
It's a process, yeah.
00:24:41
John S
Yeah, exactly.
00:24:42
johngrimsmo
Put on a VF2 or something and just prove it.
00:24:45
John S
No, I wouldn't even put on a machine.
00:24:45
johngrimsmo
Mm, sure, on bench.
00:24:46
John S
I would put it, yeah, exactly.
00:24:49
johngrimsmo
yeah Yeah, I've seen a lot of guys prove the process out, like a Zach, Zach Smith, the the Watchmaker guy, um got his UR on his current Evo or whatever, and was bench-loading palettes 100 times in a row just to see if it works.
00:24:55
John S
Yes.
00:25:02
John S
Yeah. Yeah. Cause if that works, then maybe I buy a used, you are just to set up a real system offline. Again, just put in an empty office room and just say, can it run for three days straight loading and loading material pallets, all that.
00:25:13
johngrimsmo
Yeah.
00:25:16
johngrimsmo
Hmm.
00:25:18
John S
Um, yeah.
00:25:18
johngrimsmo
Now would that process prove out a, you are path for you or could you apply, could you put a FANUC in its place, but then the programming is all different, you know?
00:25:27
John S
Yeah. I'm saying you are because it just seems like it's become the Kleenex of robots or co-bots.
00:25:30
johngrimsmo
Yeah.
00:25:33
johngrimsmo
Yep.
00:25:35
John S
I think if you're gonna move an actual pallet with a vice on it, you probably need something heavier duty than your the UR's that I'm aware of, although maybe they have a 25 kilogram one.
00:25:46
johngrimsmo
They have a 30. You are 30.
00:25:47
John S
Is that right?
00:25:48
johngrimsmo
Yeah.
00:25:48
John S
Okay. I don't know, John. I i think the UR interface I've heard better things about compared to some of the other robots
00:25:55
johngrimsmo
is Yeah.
00:26:00
John S
Yeah.
00:26:00
johngrimsmo
I've gone down this rabbit hole. Probably a lot of times over the past six years or so, like whether it's you are looking at other brands and automation and and how to program it.
00:26:11
johngrimsmo
And what do I want to do? And do I want to make a little pallet rack cell for one of our other machines?
00:26:12
John S
Yeah.
00:26:16
johngrimsmo
um totally doable, even even our speedio, I was like, okay, a speedio with a a DIY small palette robotic changing cell would be a pretty effective, fairly low cost, as opposed to another current aroa cell.
00:26:26
John S
yeah
00:26:32
John S
Yeah.
00:26:33
johngrimsmo
Because that's what I'm looking at. like I'd love to buy another current an aroa, but that's like big bucks. And I can get a speedio and build a URL palette changer and literally make parts all day.
00:26:44
johngrimsmo
Um, and so what I, what I need.
00:26:45
John S
like what Dennis talks about. It's like, Hey, you are with a bunch of ro ah brothers is a compelling offering for some parts when you don't need the grove or you don't need the Ozda.
00:26:47
johngrimsmo
Yep.
00:26:56
johngrimsmo
Yep. Yep.
00:26:58
John S
Yeah, so I'm actually, I'm excited for it. And look, some of that, like, you know, hit the recording, turn off the recording right now, so nobody hears this, I'm joking.
00:27:01
johngrimsmo
Good.
00:27:07
John S
But like, some of it's just like, I want it, like, I enjoy this part of it.
00:27:10
johngrimsmo
Yeah. Yeah, totally. The research phase, the figuring it out, the logic, the.
00:27:14
John S
Yeah, yeah. um And I'll leverage the the network.
00:27:15
johngrimsmo
Me too.
00:27:17
John S
you know i could Some people that are, two people come to mind that are very seasoned expert, but like, let me, it's kind of that campsite rule of like, let me put more effort into showing something first before I go.
00:27:29
John S
Just ask. um
00:27:30
johngrimsmo
Yeah, yeah, totally. give Give me the answer. I don't know what I'm asking kind of thing. No, you got to wrap your head around it.
00:27:36
John S
Yeah, but it does surprise me and in what I've seen from automation systems and the aros of the world, so forth.
00:27:37
johngrimsmo
Yeah. hundred
00:27:43
John S
I don't recall seeing systems that blended workload loading and material loading.
00:27:52
johngrimsmo
he but But, I mean, you need ah an end-effector changer. And there are some systems I've seen, or integrated systems I've seen that do that.
00:27:58
John S
Yeah. Yeah.
00:28:02
johngrimsmo
It's all the technologies out there.
00:28:04
John S
yeah Oh, zero doubt that the tech's out there, like for sure.
00:28:09
johngrimsmo
Yeah, exactly. But it's the repeatability in the system and, yeah.
00:28:15
John S
Yeah, let's remember.
00:28:16
johngrimsmo
Obviously, Puckchucks would be a part of this.
00:28:20
John S
Yeah, for sure.
00:28:22
johngrimsmo
to to use not to make obviously obviously you want to use it to make pukchaks but um but but to use them in the system maybe your end effectors and pukchak like quick change maybe seriously I've seen people use little shunk palette shunk Vero systems but the little ones on the end effector and that's your end effector changer
00:28:26
John S
Right, right, right.
00:28:34
John S
That's funny.
00:28:37
John S
Bureaus on the end.
00:28:41
John S
i I mean, look, that's I love that idea. I think I would go as robust as I could upstream. Like, I would buy the correct robot. I'd buy shunk tool hardware for the end effectors or robotic or whatever, whoever makes it.
00:28:53
johngrimsmo
yeah
00:28:56
John S
I don't need to create more stress around that.
00:28:57
johngrimsmo
a
00:28:59
John S
um But robots can load parts. I know they can load them incorrectly. So it's a question of like, how does does it fail gracefully? Or what happens when it fails?
00:29:06
johngrimsmo
yeah
00:29:07
John S
And are you using probing on every part? There's all these questions, but I'm kind of fired up about it.
00:29:10
johngrimsmo
see
00:29:13
johngrimsmo
When you think of all these questions, what is your process? Do you just kind of sit there and mull about it? Or do you open up a notepad or piece of paper and like, just write down every idea so that you can look at it all?
00:29:24
John S
Absolutely has to get rid of that.
00:29:25
johngrimsmo
Yeah, yeah, me too.
00:29:26
John S
Yeah.
00:29:27
johngrimsmo
I mean, I think about most of it, but every now and then I'll just be like, I gotta write this, I gotta get this out of my head so I can see it all.
00:29:27
John S
to go
00:29:33
johngrimsmo
And then I just brain dump,
Shop Revamp and International Inspiration
00:29:34
John S
Yep.
00:29:35
johngrimsmo
and then I can delete, organize, sort, have questions to answer.
00:29:40
johngrimsmo
Some things are thoughts, but not real questions, they're just, you know?
00:29:42
John S
Yep. I use a lot of voice memos on my phone too because then you don't need to type.
00:29:47
johngrimsmo
I never do that, ever.
00:29:49
John S
Yeah, it's great.
00:29:50
johngrimsmo
Yeah, I don't know if I'm just resisting, I've never tried, it never started, but I know you've talked about it for years.
00:29:55
John S
Yep, works well.
00:29:56
johngrimsmo
Does it convert to text or do you have to listen to it again?
00:29:59
John S
um Now it does because AI, everything, but like the point was the point for me was actually to have the audio because sometimes I need that ah energy or inflection to myself to listen back to my own message.
00:30:02
johngrimsmo
Okay.
00:30:11
johngrimsmo
Okay, fair. Yeah.
00:30:15
John S
yeah
00:30:16
johngrimsmo
And then how are they labeled categorized? Can you title them?
00:30:19
John S
Ah, then I delete them. um I only use voice messages when I want to talk through something to myself or can't type like if I'm, you know, this will happen to me when I'm like, you know, walking out to the car with the kids or something.
00:30:29
johngrimsmo
sure Interesting.
00:30:33
John S
And it's like, oh, that's a great idea. And then you write it down or when I travel, I find a lot of inspiration from getting outside of my daily routine for whether you're a trade show or a shop.
00:30:44
John S
um Most notably for me, fortunately, the tours I've seen overseas, like in Europe, where you just have such a different cultural role approach to investing in systems. America is much more bargain bin, in my opinion, or what I've seen.
00:30:58
johngrimsmo
yeah
00:30:58
John S
So you try to pull some of that inspiration in, and in fact, kind of a convenient segue. um We already gone through the first phase of what I talked about last week of all this investment.
00:31:09
John S
Already had an electrician come here and kind of do an overhaul, look at the shop.
00:31:13
johngrimsmo
Oh.
00:31:13
John S
painters on their way, gonna have a plumber come, buy new racking, buy new tables, um buy new granite surface plates, just like, actually, that while we were on the podcast, a guy was out front to look at getting us another sign, like, all these little things of, we don't have to do them all, but at least look at what these things cost.
00:31:28
johngrimsmo
Yeah. You're talking about these are the employee led kind of wishlist ideas.
00:31:35
John S
Yeah, and also just like, yes, that.
00:31:38
johngrimsmo
Yep.
00:31:39
John S
Yeah.
00:31:42
John S
Yeah.
00:31:43
johngrimsmo
That's good. That's awesome to hear.
00:31:45
John S
Yeah.
00:31:48
John S
Okay, last thing I'll say and then I'll shut up. Do it is going great.
00:31:51
johngrimsmo
Oh, yeah.
00:31:52
John S
And here's the key, John. Here's the key. Because this is, a this is we are both guilty of this. You have to, you have to do
Productivity Insights and Task Management
00:32:00
John S
two things. Number one, you have to shrink down the size of the project.
00:32:05
johngrimsmo
Okay.
00:32:06
John S
And number two, you can't think of, you cannot think about it.
00:32:07
johngrimsmo
he
00:32:10
John S
If you're sitting at your desk and you're like, when should I do the do it? Or what is it on today? at 1045 or 215, you just get up and you look at what the do it's that's been assigned to you.
00:32:22
John S
I mean, I signed under myself is and you just go do it. You thinking is the most dangerous thing.
00:32:25
johngrimsmo
Yep.
00:32:27
John S
It's like the gym. If you think about going to the gym, it's a nightmare. Once you're at the gym, it is good to go.
00:32:31
johngrimsmo
Yeah. You build that habit, the um resiliency to just get up and do it. that No hesitation.
00:32:37
John S
Just do it. You just don't think about it.
00:32:39
johngrimsmo
Yep.
00:32:39
John S
Just go to, in terms of simplicity, the example would be, instead of saying clean the kitchen or even saying clean the sink, it's like clean the faucet stem that makes the, where the water comes out of.
00:32:39
johngrimsmo
Yep.
00:32:51
John S
Cause it's like, I understand how to start and stop that task, how to complete that task and how to make it look great. Super, super specific.
00:32:56
johngrimsmo
yep Yup.
00:32:57
John S
So it's okay. um And even if that's small things, okay, John, there's 250 working days a year. Like if we each spent five or 10 minutes doing one of those things each day, you'd look back and be pretty darn happy.
00:33:09
johngrimsmo
Yup, yup. I love that.
00:33:12
John S
Yeah. Yeah. So for you, because you shrink it down, don't make a feel handle.
00:33:13
johngrimsmo
Those are good rules. Shrink the project and you can't think about it.
00:33:17
John S
It's just like just tackle where you are. You're crushing on this right now. Great.
00:33:21
johngrimsmo
Yup.
00:33:21
John S
But like whatever the next part is.
00:33:22
johngrimsmo
No, but I... Yeah, I do think about that. Break it down into the next, um, the next accomplishable task. Like, yes, there's a lot, lot to do, a lot going on.
00:33:29
John S
Yeah.
00:33:31
johngrimsmo
It's overwhelming. Um, but what is the very next thing? And the thing I'm really guilty of is, um, not doing the most important thing. I kind of do either the thing that's in front of me or the thing I want to do or the thing that's exciting.
00:33:45
johngrimsmo
And sure, some of that's fine, but, uh, When I do properly, kind of list everything out and I'm like, okay, step back. If I had to explain this to my wife or to somebody at work, like the most important thing that I'm going to do is this one.
00:33:55
John S
Yeah.
00:33:59
johngrimsmo
And if that's the one I keep avoiding, I'm not helping anybody.
00:34:05
John S
Yeah, we all been there. um You know, I remember talking to Al Watmo when he made the career change kind of from Autodesk to toolpath and was like. I thought I was being the hero and a good leader by doing all the stuff that I hated doing only to find out later.
00:34:20
John S
Other people love doing that stuff. And it was, frankly, my own you know fault for not realizing I should just have some other people do it.
00:34:21
johngrimsmo
Yup. in
00:34:27
johngrimsmo
he
00:34:30
John S
so So what do you hate doing? What's on your list?
00:34:32
johngrimsmo
It's not that I hate, um I just think, I think I keep myself busy doing important things, whereas really, there are certain things I should be doing first, because then I can clean them off my plate.
00:34:32
John S
Let's get into this. What are you procrastinating on?
00:34:47
johngrimsmo
um And then the one I've been pushing to the end is producing field clips. Um, I've made prototype ones.
00:34:51
John S
OK.
00:34:53
johngrimsmo
I know they work. I know, I know they work. Uh, I just haven't done it yet. And that's, that is literally the only last thing on my field list is producing clips, um, which I'll be starting now.
00:35:10
John S
I think you're probably scared that they may they may something may frustrate you when you make them.
00:35:16
johngrimsmo
Small part of it. Yeah. But I think it's dialed.
00:35:18
John S
But just OK.
00:35:20
johngrimsmo
Yeah.
00:35:21
John S
So the do-its would not be make field clips. The do-its would be get through the first four cam operations on like, so that means you got the work set up, the machines on, the coordinates are set up, the cam is posted, and you just rough it out.
00:35:30
johngrimsmo
Yeah, yeah.
00:35:33
johngrimsmo
It's like, or I mean, the program's there.
00:35:33
John S
There's no anxiety about that, right?
00:35:37
johngrimsmo
It's, so the do it's would be make one clip, but that's the cycle.
00:35:41
John S
That's too much.
00:35:42
johngrimsmo
Like I would just hit start and that's the cycle, right?
00:35:43
John S
Okay.
00:35:45
johngrimsmo
You know what I mean?
00:35:46
John S
Oh, I thought you were still, I thought you're still potentially like proving out testing's operation tool.
00:35:50
johngrimsmo
Not really. I, I did one tweak.
00:35:52
John S
Oh, dude.
00:35:53
johngrimsmo
I lengthen the ears a little bit. So I need to make one and install it in the handle and see if it works. Um. So I do have one more test to do. And that's it. I make one clip, see if it fits. If it does, we hit go. And then we just maintain tool life.
00:36:08
John S
Okay. And I'm asking, this is not a trick or a loaded question. What, when we hang up, what's stopping you from, is the Willman powered on? Like what's stopping you from literally just hit cycle start.
00:36:18
johngrimsmo
I got to put the jaws on. I've been making a different part on it. um Put the jaws on.
00:36:21
John S
Okay.
00:36:23
johngrimsmo
I think that's it. Make a part.
00:36:25
John S
Material college change out dude.
00:36:26
johngrimsmo
Materials. Nope. It's already set up. Yeah, like ah kind of obvious, but also I've been doing other stuff.
00:36:35
John S
No, work but I totally hear you. i'm I'm in this with you.
00:36:37
johngrimsmo
yeah yep Yeah.
00:36:38
John S
Like, I'm not throwing stones here.
00:36:40
johngrimsmo
But it's cool to like literally look at my list and go, you know what? That is the last thing that I am responsible for.
00:36:46
John S
It's great. Oh, yeah.
00:36:48
johngrimsmo
I've been, I've been handing everything else to the rest of the team. It's dialed. We're we're maintaining, um, I'm still fine tuning the hard milling even though we are hard milling but we're kind of in process tweaking it um to get the blade sit the handle exactly how I want Angela is going to make a CMM file for it.
00:36:59
John S
Yeah.
00:37:05
johngrimsmo
Just so we can validate as we go, and like the goal is that these are being made, you know, while I'm not here.
00:37:13
John S
Well, that's what I was going to say too. Why are you, why isn't somebody else hitting cycle star in the little, there we just runs it.
00:37:18
johngrimsmo
Nobody else runs it yet, but they will. Like this is the perfect opportunity for both Jeff and Angelo to really dig their hands into it.
00:37:20
John S
Yeah. Yeah.
00:37:25
johngrimsmo
Um, they've both worked on it ah mechanically quite a lot, but, uh, even like how to unclamp the jaws and do stuff.
00:37:28
John S
Yeah. Yeah.
00:37:34
johngrimsmo
Um, they just need practice and experience. And this clip is the perfect opportunity for them to take the time and do that.
00:37:41
John S
Do you see Grant's quick change Willem and Jaws he made?
00:37:44
johngrimsmo
Uh, yeah, but I can't picture it right now. Explain it.
00:37:47
John S
He made um some hardened jaws that have a cylindrical ID that he can just quickly swap in and out different sized.
00:37:57
johngrimsmo
Yeah.
00:38:00
John S
Like I think what we're using, we made our live center right now for a part. um And I think that, mm-hmm, it's kind of cool.
00:38:05
johngrimsmo
Really? Can you make me one?
00:38:09
John S
Yeah. i Ask him. Yeah.
00:38:10
johngrimsmo
Yeah.
00:38:10
John S
Um, and then like, let's say you have a really, really quick and dirty prototype, you know, not, you don't want to make or consume a set of jaws. Now you can just maybe 3d print or make an aluminum insert that does what you need.
00:38:21
johngrimsmo
Yeah, yeah.
00:38:23
John S
And then it just sets screws into there and you're done.
00:38:26
johngrimsmo
Cool.
00:38:26
John S
It's like soft jaws for soft jaws for soft.
00:38:28
johngrimsmo
Yep, yep.
00:38:29
John S
Yeah.
00:38:31
johngrimsmo
Super soft jaws is what they're called.
00:38:33
John S
Yeah. Flexible, flexible. Sorry.
00:38:36
johngrimsmo
Yep, yep.
00:38:36
John S
Flexible. Yeah. yeah
00:38:40
johngrimsmo
Mm-hmm. um So I tend to fall down these rabbit holes of interesting topics that are totally related and totally needed, maybe not needed more important than making the clips kind of thing.
00:38:47
John S
Okay.
00:38:54
johngrimsmo
But I'm going to open myself up and share a really cool one that I figured out.
Grinding Wheel Optimization with Python
00:38:58
johngrimsmo
So when we're grinding the blades, both Fjell and Rask, Like we we the on the Brother, um we put the bevels into the blades using our, it's a little one inch diameter, about the size of a quarter, like thin little CBN abrasive wheel.
00:39:04
John S
On the brother grind. not Yeah.
00:39:13
John S
Mm hmm.
00:39:19
johngrimsmo
And we've been doing that for years. We used to do it on the current, now we do it on the speedio. And we do it on the, on the Rask every day. And now that I'm making VL parts and bringing them into the the world, I'm like looking closer at that, you know, again.
00:39:32
johngrimsmo
And the finish has always been acceptable, but our finishing guys are spending, I think they said about 22 minutes on average, um, stoning the, the bubbles to get all the machining lines out.
00:39:41
John S
It's a long time.
00:39:43
johngrimsmo
It was a long time on every single knife. um And it's always kind of bugged me, but it's it's a need because the grinding isn't that good that you can just tumble it and call it good. And our our expectations are super high, right? And the toolpath I'm using from Fusion is the Geodesic toolpath, where I just click on a surface. I actually extruded a clean surface from a clean sketch with radiuses. So I'm just extruding like a sidewall.
00:40:10
John S
Mm-hmm.
00:40:11
johngrimsmo
that has radius, flat radius.
00:40:12
John S
Okay.
00:40:14
johngrimsmo
And then I pick that as the geodesic toolpath. And it kind of, it traces, it goes down, over, up, over, down, over, up.
00:40:20
John S
Yeah.
00:40:21
johngrimsmo
So I'm grinding down and up. I'm chop grinding, but in both directions, which is really cool.
00:40:23
John S
Yep.
00:40:26
johngrimsmo
Wheel wear is insanely consistent um because it wears the bottom and the top of the wheel evenly.
00:40:28
John S
Hmm.
00:40:32
johngrimsmo
So that it gets smaller, but it's even and it's self-dressing.
00:40:37
johngrimsmo
It just slightly gets smaller every day. So I have the laser and I laser touch off. um Sometimes you get erratic readings where it'll measure tooth out bigger.
00:40:37
John S
Yep.
00:40:47
johngrimsmo
And I'm like, the grinding wheel is not getting bigger, period.
00:40:49
John S
ah Yeah sir Yeah
00:40:51
johngrimsmo
So I had to ah incorporate outlier elimination to be like, if it's bigger, don't worry about it. If it measures smaller than four-tenths, also ignore it. So only make a change if it's within this four-tenth range, but you can still get locked out of that range if it jumps down, say a thou for whatever reason, and now you're more than a thou from your starting point and it's stuck and it's wrong.
00:41:16
johngrimsmo
So anyway, I did a bunch of stuff on that over the past two weeks. Got it pretty dialed, pretty awesome. So now it's self compensating, which is really cool.
00:41:23
John S
Yeah, that's cool.
00:41:25
johngrimsmo
Before Steven would make one blade at a time and he would be like, oh, that one's a little under so I'm going to comp the macro and then dial it in and every blade he'd have to dial it in. So now it's pretty automated, which is great.
00:41:37
johngrimsmo
but The surface finish itself was never making me you know all giddy. And it turns out the geodesic toolpath is interpolating that radius line radius with some variability to it.
00:41:52
johngrimsmo
So when it commands a Z down move, it's also putting an X 50 millionths every now and then.
00:41:57
John S
Really?
00:41:59
johngrimsmo
And it all depends on your tolerance setting, your smoothing setting.
00:42:01
John S
Yeah.
00:42:02
johngrimsmo
And I'm literally one tenth change, ruins everything.
00:42:06
John S
Yeah.
00:42:07
johngrimsmo
So like in fu into fusion, I turn on show points, which reveals all of the secrets, holy cow.
00:42:12
John S
Yeah.
00:42:15
johngrimsmo
And the grinding wheel is so um grindy that it will see a point. If there's a point halfway down the bevel, even if it's a straight line to a straight line, you'll see a dot.
00:42:21
John S
Yeah.
00:42:24
John S
yeah
00:42:25
johngrimsmo
And it's wild.
00:42:27
John S
Grant and I were literally just talking about this because we're making a shtimp, I wish I had it, and here's a cool part.
00:42:27
johngrimsmo
um
00:42:32
John S
It's a shtimp works breather valve for a Porsche, vintage Porsche.
00:42:34
johngrimsmo
Yeah.
00:42:35
John S
And actually, fun fact, DSI was helping Grant on the geodesic toolpath on this stem. And we were talking about like, if you have a hole in a surface, even if you patch that surface, so it's,
00:42:47
John S
perfectly patched. Where the patch meets the solid body and the hole, it will create a cam tool tool point. And every machine will create ah create some sort of a dwell mark there.
00:42:54
johngrimsmo
Yep.
00:42:58
John S
even yeah you know Even a Mitsui Siki, like you're just going to see something.
00:43:03
johngrimsmo
Because the toolpath is putting it there. it's it's And um we see this on the mori.
00:43:05
John S
Bingo.
00:43:08
johngrimsmo
We see this. Yeah, if if you guys listening have never turned on turn on points, they will match up to imperfections in your surface.
00:43:14
John S
No. yeah
00:43:17
johngrimsmo
You're like, why is it always fascinating? Well, it's because that's where all your points are. um But then I'm like, okay, the geodesic does what I want, but it does it a little bit too many weird points every now and then, or it's it's doing an XY move by 50 millionths or a 10th or something like that.
00:43:36
johngrimsmo
So you're you're kind of telling the machine to kind of move into its backlash range, which is almost nothing, but it's something.
00:43:43
John S
Yeah.
00:43:44
johngrimsmo
And so it's causing some variation. So chatting with ah Greg Koenig from, who's like runs Aluma Labs, um
00:43:52
John S
Yeah, super good brother guy, right?
00:43:54
johngrimsmo
Instagram and he works at Yamazin and he knows his brothers. And I was like, dude, I've got the high accuracy mode. I've got five digits inch on the machine.
00:44:01
John S
Yeah.
00:44:01
johngrimsmo
I can read to 10 million seven inch. And I'm like, what about moving to metric? That'll give me 2.5 times more accuracy. And he said on my machine, cause the C zero is zero control, the resolution of the encoders is almost at the limit of where I'm already at.
00:44:15
johngrimsmo
So the, the benefit is not, but if I had the newer brothers, the encoders are so much better.
00:44:15
John S
Okay. Interesting.
00:44:22
johngrimsmo
And on a brother, you can't just be like, this is a metric program, run, gee, whatever it is, 19, eight, whatever.
00:44:29
John S
Yeah.
00:44:29
johngrimsmo
um You can't do that. You have to turn the whole machine to metric and run a program that is not defined as metric.
00:44:34
John S
Oh.
00:44:36
johngrimsmo
So he's like, um maybe don't do that. No, actually he said, everybody should be running metric all of the time because have you ever seen an inch ball screw? Like the machines are metric.
00:44:46
John S
Well, I mean, but I don't really see metric ball screws either. I just see my machine make parts.
00:44:50
johngrimsmo
Exactly. Anyway, so metrics not going to work. um And then I forgot how we said it, but I eventually got Fusion to output clean code of just a 2D arc, straight line, arc, and then a lead out arc.
00:45:06
John S
Yeah.
00:45:07
johngrimsmo
Four lines of code, arc, line, arc. And then wrote a Python program to make that into a geodistic toolpath, and into a chop grinding toolpath.
00:45:14
John S
Stop pattern. pattern. Dude stop.
00:45:19
johngrimsmo
So using chat GPT, obviously, it interpolates the arc using IJK values and it breaks the arc mathematically into 4,000 segments and they are perfect.
00:45:34
John S
Dude.
00:45:34
johngrimsmo
Whereas the geodesic toolpath would like kind of round and guess and and and grinding shows everything. So once I did this, it took me the whole weekend, this past weekend, um and it's freaking phenomenal.
00:45:48
John S
Dude.
00:45:49
johngrimsmo
massive. So it's not that the speedio wasn't accurate enough with my high accuracy chip, it's that literally the toolpath is feeding it, you know, slightly.
00:45:52
John S
Yeah, yeah. Dude, we're feeding it.
00:45:59
John S
Yeah.
00:46:00
johngrimsmo
Anyway.
00:46:00
John S
Wizzy Wig, right?
00:46:02
johngrimsmo
Yeah. And then you take your, your PFG stones and you stone the surface.
00:46:03
John S
Dude, that's just...
00:46:06
johngrimsmo
And I'm like, Oh, baby, this is great.
00:46:11
John S
I got to say, because i part of me knows there's millions of people that don't know chatgpt or use it, and i'm but I'm also not using it as much as others, but I had to write this quote and I couldn't, like it was four sentences, I didn't love it, I knew what I wanted to say, I wrote it, tried rewriting it, then I asked chatgpt to write it for me. um and it did a great job, except it was exactly not the tone I was looking for, but I grabbed a couple of its kind of phrasings and then I rewrote it myself. And then and then I literally, and I'd never done this before. This was kind of like my aha moment. I literally pasted it back. I wrote to chat GPT. I said, what you wrote sucked, but ah what I'm writing has sucked too. Here's what I've come up with. Will you please just completely rewrite what I already wrote? John, it rewrote it for me. And I'm like, this is perfect.
00:47:00
John S
I mean, it in like it's i know i'm just crazy.
00:47:05
johngrimsmo
Yep, yep, just crazy.
00:47:05
John S
It's just crazy. Don't, nuts.
00:47:07
johngrimsmo
And I don't know almost anything about Python programming. I've read some code, but um I knew what I wanted and I knew what was right and I knew what was wrong.
00:47:16
John S
Yeah.
00:47:17
johngrimsmo
And I could look at the code and I could tell if it was right or wrong. And I could use and NC Viewer to visualize this code.
00:47:22
John S
Yeah.
00:47:24
johngrimsmo
And it took a ton of iterations back and forth, but me and chat GPT created a phenomenal code. So I output the 2D contour from fusion and run this Python script on that file.
00:47:36
johngrimsmo
It creates a new file that goes down, over, up, over, down, and it's perfectly articulated over this arc.
00:47:39
John S
yeah
00:47:43
johngrimsmo
like 100% perfect. And last thing I'll say, um Greg had this amazing suggestion.
00:47:47
John S
Yeah.
00:47:49
johngrimsmo
He goes, or you could go the old school route, if backlash is a concern or whatever, um and back off on every toolpath.
00:47:57
John S
And come back in.
00:47:58
johngrimsmo
um And then go back in, in the same direction. So you're always feeding into the backlash.
00:48:00
John S
Yeah. Preloaded. Sure.
00:48:02
johngrimsmo
And and mean we're talking a tenth of backlash or something. And I was like, well, let me add this as ah as a variable flag to my Python script.
00:48:08
John S
Yeah.
00:48:10
johngrimsmo
So chat GPD, I want to add the option. Like once I got a clean toolpath, I want to add the option backlash elimination. and Greg said on the speedio there's a G09 as an exact stop mode so it's like move in X to here don't do anything else until you've exact stopped at X and then do Y and he said break out every movement into linear motions so it goes down and it goes out in Y exact stop over in X exact stop out in in Y to your spot and then back up in Z
00:48:28
John S
interesting
00:48:41
johngrimsmo
And you can literally see the machine like micro pause on every geo nine. Like it's, it's slower.
00:48:46
John S
Interesting What's you can work on that later like it's still you're eating great parts now I
00:48:47
johngrimsmo
It adds a few minutes to the toolpath. I don't care. Cause it's.
00:48:52
johngrimsmo
Yeah. Yep. And I did an AB b test with geo nine on and geo nine off. And I didn't, I couldn't really see a difference. So maybe it's not that, but the backlash elimination makes a big difference. The Python toolpath makes a big difference. Like, and I haven't heard feedback from the finishing guys, if they take less time to finish, but I know they're going to take less time to finish.
00:49:13
johngrimsmo
So that was one of my my rabbit holes, but I'm really happy with it.
00:49:13
John S
Dude, that's so cool, right?
00:49:16
johngrimsmo
I'm like super stoked because imagine saving 10 minutes on every knife of a manual hand stoning, like massive.
00:49:22
John S
Oh, yeah, not 10 minutes of cycle time, 10 minutes of hand labor.
00:49:27
johngrimsmo
Of human, yeah.
00:49:27
John S
Holy nuts.
00:49:28
johngrimsmo
Yeah, exactly, so.
00:49:30
John S
i am We're still working on this Raspberry Pi project. I got my box printed with my disconnectors and then I was looking up the
00:49:34
johngrimsmo
Yeah.
00:49:39
John S
Same thing, rabbit hole-ish. I can go down as far as you did, but looking at what the latest and greatest is. And then you find these like cursor, I think you gotta pay for it, but it's an AI code site and you showed it code and you're like, hey, this is like the example that this guy gave on YouTube was he wrote this script that like parses through the first 10 comments on YouTube. And then he just in like effectively in VS code with this plugin for cursor, whatever it is. He's like, do the same thing, but for YouTube or Instagram or TikTok.
00:50:09
John S
Just spit out all this code to move the whole thing over with the API calls like it's it's I Don't think I still appreciate how powerful this stuff is Yeah, just look up cursor AI you'll find it you can't miss you can't miss it Yeah
00:50:19
johngrimsmo
Yeah. Cursor, yeah, I'm gonna look into that. Cursor what code?
00:50:26
johngrimsmo
Oh yeah, I see it, yeah.
00:50:29
johngrimsmo
Cool, yeah, I was just looking last night on how to integrate different things, like put data from this into a spreadsheet or ah ah QuickBooks information to basically what Zapier does, which I've never used, but I've heard of it.
00:50:42
John S
Yeah.
00:50:43
johngrimsmo
Because I think we'll have some serious applications where we want to tie this to that without anybody touching it.
00:50:48
John S
It's cool.
00:50:50
johngrimsmo
Yeah, create a spreadsheet with the top comments from all of our videos. Like, that's weird stuff.
00:50:55
John S
yeah are you just Yeah.
00:50:55
johngrimsmo
You know? Why not?
00:50:59
John S
It's crazy.
00:51:00
johngrimsmo
ye Cool, man.
00:51:02
John S
Crazy.
00:51:04
johngrimsmo
So, what's next?
00:51:04
John S
Good.
00:51:05
johngrimsmo
What's next for you?
00:51:06
John S
What's next for me? ah What is next for me? um
00:51:15
John S
I did my duets today. Actually, my, my to do list is clean. So I'm doing some, I get her head started on some of the year end stuff.
00:51:18
johngrimsmo
oh Good for you.
00:51:21
John S
I've got a couple of documents I got to read. um We've been more deliberate about. getting me out of the picture of some of the machines up here.
00:51:32
John S
So like we had a new test fixture, Garrett handled it and Grant's testing it.
00:51:38
John S
I've been putting off, I actually have to reach back out to the Fiverr developer. He has a script that'll auto boot the PI, but he needs to remote into it. And I put it up on a guest network.
00:51:45
johngrimsmo
Yeah.
00:51:47
John S
I'm a little bit, I don't think a Fiverr guy is like setting this up for a sting off on like hacking into our network, but I'm a little bit sus about like giving a stranger
00:51:55
johngrimsmo
Yep. A hundred percent.
00:51:56
John S
um So I put on a guess network, which I think is takes it makes it notably harder.
00:52:00
johngrimsmo
Okay.
00:52:02
John S
I mean, if somebody really wants to mess with us, I guess they could.
00:52:05
johngrimsmo
Yeah.
00:52:06
John S
Yeah.
00:52:07
johngrimsmo
Cool. Actually, it's pretty, uh, auto boot the PI or auto load your program.
00:52:08
John S
How about you?
00:52:14
John S
Yeah, just on a load of Python script, which I would think is super easy. um
00:52:19
johngrimsmo
Yeah, it is. I took notes on that. Pseudo CH mod or something, I can't remember. But I've done that.
00:52:27
John S
Yeah, yeah i'm going to um I'm either going to watch what he does, or I'm going to put a GoPro up and just screen record what he does, because I want to i want to build more of these.
00:52:32
johngrimsmo
Sure, yeah. Yeah, you want to know.
00:52:35
John S
And um I don't know if he's, whatever, I'll figure it out.
00:52:39
johngrimsmo
Samba? Yeah, there's a Samba thing. I think you got to install the Samba library and then it can whatever you put in that file auto loads those scripts.
00:52:45
John S
Notice, yeah.
00:52:47
johngrimsmo
Yeah, that's how we do it for tool management and it's great.
00:52:47
John S
there's
00:52:49
John S
I was watching another Raspberry Pi YouTube video and there's like a app that can auto load that hides the cursor. So if you're using it as a touchscreen, you don't see the cursor on boot. Like there's so many crazy little cool.
00:52:57
johngrimsmo
It's so cool.
00:52:59
John S
Yeah. and from um Okay.
00:53:00
johngrimsmo
Yep.
00:53:04
John S
Yeah. So what do you do to us today?
00:53:06
johngrimsmo
Okay, I'm gonna take a page out of your advice. I'm going to install the soft jaws to make clips.
00:53:12
John S
Good. I'll tell you, it's all you gotta do.
00:53:13
johngrimsmo
That is step one because honestly, i'm I'm making this clamp which we need hundreds of I'm making a clamp But every time I look at the machine, I'm like I have enough time to put a new bar in and make three more clamps But you know that the I gotta stop this project and say no.
00:53:15
John S
Right.
00:53:23
John S
Yeah, I know.
00:53:28
John S
Yep.
00:53:28
johngrimsmo
No now is the time to make clips I think I'll do that today. I've made enough clamps That's that's what I'm doing I mean See you next week
00:53:32
John S
Good. Good. Hey, I'll see you next week. Take care.