Become a Creator today!Start creating today - Share your story with the world!
Start for free
00:00:00
00:00:01
#294 Making Custom Porsche Parts, Chip Management on the Speedio and MORE
185 Plays2 years ago
Topics
- Saunders running his horizontal mill 18 hours a day
- Making custom Porsche parts
- kids in the shop
- Grimsmo crunched the probe tip on his Speedio with custom programming
- Chip management on the Speedio
- probing, manual programming
Recommended
Transcript
Introduction and Hosts
00:00:00
Speaker
Good morning. Welcome to the business of machining episode number 294. My name is John Saunders. And my name is John Grimsmo.
Experiencing vs. Hearing Business Stories
00:00:07
Speaker
John and I keep each other honest, but also talk about how it's different when somebody tells you something versus when you live it as a business owner. Yes. I mean, that relates to hiring and firing. It relates to implementing machine tools, to making parts. Yes.
00:00:25
Speaker
Yeah, you hear all the stories, you can read all the forum posts you want, but until you do it, it's not comfortable. It doesn't make sense. You're like, yeah, people have been saying this, but I haven't done it yet.
Automation's Impact on Efficiency
00:00:36
Speaker
which this makes me a hypocrite because I'm about to tell you how much I love automation and what it's done for us. But I don't know if I why I would acknowledge that any listeners would sink in that when I'm already admitting that it wasn't until I embraced it that I realized what it means. So
00:00:57
Speaker
hypocrisy abounds yet I can't tell you enough how heartwarming it is to log into our webcam regularly now with no effort. I was joking in a video that we just filmed on the lessons that we've learned buying and implementing a horizontal over the first six months, how we're like amateurish or like one step above amateurish horizontal operators.
Implementing New Machinery Challenges
00:01:20
Speaker
And maybe that's not true because I think you and I use things like probing and more complex cam. Just because it's what we grew up with compared to so many of the installation of machines out there that are more relics or just 10 years, 20 years older. But nevertheless, I've got a lot left to learn. But still, it's no stress at this point to have that machine running 18 hours a day. Like just not. That's incredible.
00:01:47
Speaker
And Yvonne needed a 12 by 8 Saunders Machine Works palette yesterday for an order. We didn't have any. And it was the material, op 1 and op 2 were already loaded up. We just hadn't finished them because we switched other products. And so we literally just hit a button, moved it back into the cycle. And with zero additional operator effort, it was done in like nine. I mean, it came up. It was done in like an hour and a half. Yes. But you could schedule it in. It was ready. The workflow was there. Oh, yeah.
00:02:17
Speaker
Yeah. Imagine on another machine, when you'd have to change the whole setup, you'd have to move a bunch of fixturing around, pick up your origin, et cetera, et cetera. Oh, yeah. Touch off tools or reset up tools. Yes. Having a magazine big enough with just all of your production tools are in there. You know they're good. The tool life is fine. You're just loading material and hitting go. And the logic in the machine knows, OK, this is your offset. This is your tombstone.
00:02:44
Speaker
The checks, this is everything. Yes. Yes. Yes. Yes. That's awesome to see you getting there. Yeah. Yeah. The last week was really rejuvenating. I'm working on these valve covers.
Company Introduction: Stimptworks
00:02:55
Speaker
I think I've alluded to it. We're starting up a company called Stimptworks, which is manufacturing vintage Porsche parts. Oh, yes. We have an informal partnership with a person in this industry. They're sort of our first customer, if you will.
00:03:09
Speaker
Um, and it wouldn't, it wouldn't be an endeavor that I would feel as comfortable taking without his guidance. He's doing test fit, helping us with some product design stuff. Cause it's like, kind of like, Oh, look, we're, we're engine, not engineers, Alex isn't in here, but we're machinists. We know to make stuff, but we aren't, um, as much as I'd like to enjoy becoming one. We're not car experts. He is. So, um, he actually already placed his first order. So we've been running, um, sort of these, uh, nine 64 valve covers. Uh, we're running the nine, nine threes next. And, um,
00:03:38
Speaker
For me, it was just wonderful to sit down and program parts. I just love it. That's awesome.
Understanding Porsche Flat Six Engines
00:03:52
Speaker
I don't know Porsche engines too much. Is it two valve covers? Like flat four kind of thing?
00:03:57
Speaker
No, so notwithstanding all the like SUVs and other nonsense, 9-11s are flat six engines, period. So there's on a traditional vehicle, you have, I believe, identical lower and no, are they mirrored? I don't think they're mirrored. I should know that. But there would be a set of lowers and a set of upper valve covers.
00:04:22
Speaker
Cool. Yeah. Yeah. Cause on the Volvo engines, it's just one valve cover, just top four cylinder. Yeah. It's an inline four, right? Yeah, exactly. Yeah. So easy PC.
00:04:33
Speaker
I'm hardly the one to start professing like car expert engine knowledge, but a flat six is not like a V6 or V8. A V engine is literally in a shape of like a V, whatever, 30 degrees or something. I think of what most people probably image pops into their head when they think of a car engine.
00:04:52
Speaker
You know, your straight four is kind of like just, you know, four years of corn sticking straight up in a row. Porsche engines are flat six. So the Pistons go straight out parallel to earth ground. And they are that helps keep the center of gravity quite low on the overall engine. Didn't think about that, which is one reason why the cars handle so well and
00:05:21
Speaker
Yeah, it's also called a boxer engine. Subaru's have them. Volkswagen's have them, I believe. And most of the old ones, in fact, really all of the old ones until 1990, eight or nine were air cooled. So the engines have what effectively looks like heat sinks or radiator fins on them because there's no
00:05:39
Speaker
radiator fins, you know what I'm talking about, little like blade fins to help increase the surface area for engine cooling because there is no radiator or water pump or coolant like there is in any modern car.
00:05:53
Speaker
And, but like mine, for example, it takes, it's a 3.6 liter engine, but it takes 11 quarts of oil. So dry sump oil system where effectively is also using the oil itself as a means of cooling. Interesting. Is there an oil cooler? Yeah. Yeah. Okay. This was fun.
Design and Prototyping Collaboration
00:06:13
Speaker
I like it. That's cool. Yeah. I love it. And it's, is it your design for the product or you worked with the guy to help design it?
00:06:22
Speaker
So we designed it and then sent him some and then it's the stuff that I love. He's like, you know, every time you install these, this little flange gets in the way of this tool or I want a way to mount a clip to hold the wire harness here. And it's like, yes, I can do that. Yeah. And I think that's what from a business philosophy, what helped
00:06:41
Speaker
us attract him a little bit is the fact that we would just make a change and FedEx him a replacement in like two days. So here's a guy who's been doing this for 50 years or 40 years. The parts that we're sending him are just cherry. And it's awesome to see like, holy cow, there's just coming my way. And then even little things like we were really playing around with the gasket fit tolerance, which has to do with the width and the depth of this slot.
00:07:08
Speaker
And so we just machined five different test pieces with the depth varying by three tenths on each one.
00:07:16
Speaker
labeled them, engraved them, sent them to him and said, hey, would you mind? And he has a PhD engineering buddy that he's done work with over the years. So he's got some horsepower on that engineering side as well, pun intended. And it was great because he like just was all over this like, okay, that's great. I took your five different gasket tests. We measured expansion, torque values, what we thought was what I looked to see when I'm seating a gap. Like it was just like, no, it was just awesome. Yes. That's fun.
00:07:45
Speaker
Good for you. Thank
Staffing Changes and Business Impact
00:07:46
Speaker
you. So cool. In the spirit of being honest, though, the last 10 days have been pretty tough. Yeah. Yeah.
00:07:54
Speaker
So we're still kind of adjusting to the sort of like new Saunders, if you will, with having both Julie and the temporary person that we had in shipping no longer here. And then my wife has stepped on board, been absolutely solid. And then we have a new part time shipping person who works at another local establishment.
00:08:16
Speaker
not in the industry. And he comes by at about 1.30 and works for two and a half hours every day. So that's great. He's been getting very much up to speed on assembly, shipping and so forth. But it's kind of easy to say, much harder to do to remember what it's like to onboard somebody. And frankly, most of our onboarding in the past would have happened with
00:08:39
Speaker
somebody like Julia or Ed taking the reins. But Ed's busy, Julie's not here. So I've been having to step into that role, which is something I very much don't enjoy. Just being honest, it's not that I can't do it. It's just that I
00:08:52
Speaker
I like having somebody else who knows how to do it who can take as much time. If I'm being just very introspective here and very impatient because I want to go do other things. So, kind of one of those lessons of like know what you're good at, know what you're not and that's not easy for me.
00:09:11
Speaker
But then we have this huge labor swing of like last week we had a full-time intern, actually it's an intern who's going to be at our shop for a full week and then he's back at his high school machining program for a week. So he's alternating. So he was here all last week. We had another new intern who's also been great.
00:09:30
Speaker
We had the new part-time worker start Yvonne's here. So we had all this help. And then this week, one person's out sick. One person is next door because we're running training classes right now. So like yesterday and today, it's like two people in the shop and we've gone from like seven to two. And it's very difficult. I can see that. It's hard to keep a good flow of steady work process. Yeah. And just put a lot more on your plate variably.
00:09:58
Speaker
Because while there's many things a business owner needs to spend his time doing, the parts have to be run as well. Yes, yes.
Adapting Roles and Responsibilities
00:10:10
Speaker
It's kind of like, you know, it's like alcohol. It's like the first step is admitting your alcohol. It's like, I have to admit, like, John, these aren't going to be days where I get to go play in Camland for three hours. I've got to recognize, OK, your role has changed today. Don't don't fight it. It's OK. Yeah, exactly. And like today, Angelo's off. He's just taking a planned day off. Go have some fun. So I'm running the current. I checked the current on the webcam last night right before I went to bed. Still running. Great.
00:10:39
Speaker
come in this morning, broken drill. And I'm like, okay, no big deal. But okay, that's my problem right after the podcast to take care of, no problem. I also have my son with me today.
00:10:50
Speaker
So he's going to help me, uh, indicate a probe tip that I crashed. We'll talk about that later on this video. Um, anyway, yeah. So like, similarly, like my role today is not playing cam land and design the things that I want to do. Uh, today is kind of hit the critical points and, uh, take care of life and teach them some stuff. And I'm like, oh, if you're, if you're going to be here, let's learn some things together. Um,
00:11:19
Speaker
because he's got a very scientific mind. I feel like he's going to want to do big crazy things in life, but it's the follow-through of a nine-year-old. How do I focus that? Actually, let's do some things together. Let's learn something together. The two-second attention span is
00:11:42
Speaker
It's like, if you miss the spring pass on the 2D adaptive, come on buddy, adjust your talent setting. Exactly. Exactly. I don't have time for that. Exactly. Exactly. So it's the balance. That's wonderful. Yeah, it's good. It'll be fun. Like I want to actually make a point of instilling some skills and some
00:12:00
Speaker
whether it's useful information or just like context to how mechanics work and how things interact. He certainly loves going up to the current when it's free and just hitting the jog pendant and like jogging it around. And like we were here on a Sunday and he did that.
00:12:15
Speaker
And he was getting close to the object. I'm like, just careful, careful with the probe in it or one of the tools I forget. And he's like, I know, I know which way it is. I'm like, I'm going up. It's fine. And I'm like, scary, but okay. He's smarter than, you know, you give him credit for sometimes. Um, but even on a five axis, like he figured out X, Y, Z, B and A and B and C like real fast.
00:12:38
Speaker
That's like, oh, that moves that way, that moves that way. Okay, yeah, I know how to move it away. Yeah, so be cool. That's awesome. William loves hitting the Okuma commands to like rotate palace out or start a program, which is that's innocent. That's not even like jogging, because you can't really crash doing that. Right, right. It's just there's something that's in our blood about controlling machinery that's just wonderful.
00:13:02
Speaker
And even just through osmosis, the kids being around us, you and I with our nerdy personalities, they're absorbing stuff. They're picking stuff up. Just how the world works and how things are made and injection molding seam lines and plastic toys and all this stuff. They're starting to know this stuff. Do you think at all about the future of Crimson Eyes as it relates to your offspring?
Children's Workshop Interests
00:13:32
Speaker
Not really, no. I mean, Lave talks about it all the time. He's like, I'm going to have a workshop way bigger than your workshop, and I'm going to make machines and just give them to you. I'm like, you go, buddy. Yeah. But he's like, no, maybe I'll work for you when I'm a teenager or something to learn my skills, and then I'll go build my shop and do crazy things, build spaceships that travel at the speed of light and stuff.
00:13:56
Speaker
I encourage that as much as possible. We're going to need more people like you, buddy. Yeah, right. Can we can we go on a really unrelated tangent for a second about the speed of light?
Mind-Boggling Space Distances
00:14:06
Speaker
Yeah. We were outside looking at the stars, talking about Andromeda. How Williams asking how far away it is. And I don't know this stuff. And I'm like, William, I think it's like two million light years away. So we pulled up on Wikipedia and I'm regurgitating this from memory. So I'm wrong. Right. Doesn't really matter the numbers. But it's like the speed of, you know, so we can send
00:14:24
Speaker
Voyager 1 is moving at 40,000 miles an hour or something. That's pretty fast. The speed of light is something like 186,000 miles per second. To get to Andromeda, which apparently in the Galactical solar system is pretty darn close to us, you would have to travel at 186,000 miles per second for 2 million years to get there.
00:14:53
Speaker
So it's like Gail is just doesn't it? Yeah. Like Williams, she is old enough to like gets that. But he realizes, you know, OK, it's not just like we would have to try hard. It's just it's like way different than that. Yeah, it's different. And yeah, life was telling me, I think he heard, I don't know if you misheard, but like the closest solar system, not ours, is four light years away. I'm like, no way. It's got to be way more than that. Yeah. So I don't know all these facts, facts, but
00:15:22
Speaker
Yeah, it's fun to see kids try to wrap their brain around the scale and even grownups around the scale and magnitude of space in the universe and like what's beyond Pluto kind of thing. Oh, yeah. It's it's bonkers. Like it is unfathomable. Yeah. It's like that movie from the 90s Contact where they talk about like folding space time to get to the other side with the Jodie Foster one. Yeah. Yeah. That's a great movie out on my list.
00:15:50
Speaker
Okay, we should get back to real machining talk. No, this is our podcast. We'll talk about whatever we want. No, it is really cool. It's just crazy. Actually, I bought a poster for the shop that I haven't put up yet. That's like, I don't know if it's Hubble telescope exactly, but it's one of those like space out into other galaxies.
00:16:11
Speaker
pictures blown up on a pretty decent sized poster and I want to hang it up somewhere. It's just fun to stare at and be like, oh, we are insignificant. We are absolutely. It's like you have the pale blue dot thing somewhere, don't you? I love it. You don't ever see it, but it's actually what I look at. It's like right behind me. Right there. Look at that. Yeah. And then I have a copy of the Voyager record right there. Oh, cool. Yeah.
00:16:39
Speaker
Now you just got to get current to give you one of those James Webb telescope, many things that they made. Yeah, although I wonder what the. I did, they just did that for fun, because I think I think so. Yeah. But Sui Siki was flexing on everybody by having like almost no marketing, but being like, we made the James Webb telescope. Yeah. Yeah. What do you like? Mic drop. Yeah. Right. It's like Japan versus Germany. Yeah.
00:17:07
Speaker
cool.
Repairing Damaged Equipment
00:17:08
Speaker
I crunched the bloom probe on the speedio yesterday. Which is just a stylus. I think of it as a rain sharp stylus probe. Yes. Same thing. I'm pretty sure the body is still fine, but the tip was gone.
00:17:25
Speaker
Oh, you really crushed it. Yeah, I'm not sure if the body functions yet, because I haven't turned it on and tried it yet. But I had extra styli, stylus, ceceses, and I put it in yesterday, tightened it, and today I'm going to dial it in for run out and then calibrate it all again. I think it still functions fine. They're pretty burly, but I crunched it in Z and the cutting that paper phenolic rich light material.
00:17:53
Speaker
and it is buried in the material like I don't know how deep. The whole body? The tip. Oh, okay. Like the stylus and there's an adapter that kind of flares up a little bit bigger. Yeah, it's in there and we couldn't get it out.
00:18:10
Speaker
So it was part of a... You weren't jogging, I guess. No, I was running first run. I wanted to probe each of the four corners, log the height of each of the four corners, have some logic to find the lowest point, because this material is a little crooked. So I'm going to find the lowest point, set that as my origin, and cut 5,000 under that.
00:18:31
Speaker
That's really smart, John. Right. So that was my goal, because the first time I touched one corner and I was like, yeah, any materials flat enough. If I touch any corner, go down five thou. And then it didn't clean up because the materials tapered by 10 thou or whatever. Yeah. So I want to find the lowest point and I want to face it from there.
00:18:48
Speaker
And took quite a bit of logic to be able, like it's easy to probe four points. It's fairly easy to log the Z height at each of those four points into like, I think I did one 51, 52 53 and 54 after each probing log. But then it took some, it took me a couple hours to actually find the logic to like find the lowest of the four values. How are you doing it? It's basically if one 51 is less than 52,
00:19:17
Speaker
then 51 wins. If one 51 is less than 53, then one wins and go through that whole logic tree. Um, quite a few lines of, of if statements to be able to filter out in any scenario, um, which is lowest and, uh, it works. So that works great. And then running it like the first visual run-ups I'm watching it probe, probe, probe, probe. I see it's picking up the four points. Then the next line at the next offset or, um, macro variable is, uh,
00:19:46
Speaker
which one is the lowest, which was number two in this point. And so that works. And then the code says go to point number two and actually probe and set my G55 work offset. Oh, okay. Okay. It actually finds the lowest and then it re probes it. It re probes it because the first four are not adjusting the work offset. They're just fine logging the number. I see. And I forgot the K value in the probing line of code.
00:20:15
Speaker
which is an offset value. So it was trying to probe completely through the part and it for some reason didn't listen to protective positioning move something something and I'm watching it and fingers right on the feed hold and a lot of swear words and yelling immediately afterwards. But I didn't know you swore. Never. Only when necessary.
00:20:41
Speaker
And yeah, I watched it crunch, but it stopped probably half inch before the body hit the material with the probe body. Why wouldn't the body be fine then? I don't know. I mean, internals like shoving up and bottoming out inside was my biggest concern really.
00:20:59
Speaker
Was the stylus still actually connected to the stylus threads into the body? It was, but the adapter was bent and was loose. So it had mildly unthreaded itself. So I think it's fine, but until I watch it probe and verify that it still functions, then.
00:21:19
Speaker
It's actually, it's an interesting style I designed that we discovered when we were just making our own Heimer tips, which we actually still do.
00:21:30
Speaker
We have a little 3D printed jig and it works fine for, especially for the tolerances that we need in those use cases. But the way the Heimer tips are designed is the ball is attached to a pointed cone. Is that the correct turn? Like an upside down ice cream cone. So it's a pointed, comes to a tip.
00:21:51
Speaker
with the tip pointed up toward the spindle. If you crash in Z with the tip, the pyramid causes the ceramic to shear open and destroys the ceramic shank altogether, which is what you want to avoid shoving the ball into solid ceramic that shoves it up into the body.
00:22:16
Speaker
Cool. Can we talk about the probing logic? Yeah. Just because I don't know how to do this. I'm more just asking, talking out loud. So if you probed four datums, you'll get four values. You can query those values with code. I live in the FANUC Akuma Haas world, but I'm not brother. Same as FANUC.
00:22:38
Speaker
So it should be easy. So let's say you always pick the top left as your coordinate system because that's what's harder is to like decide which one you're going to pick. So let's say the top one has a value of two inches and then it proves the others. You should be able to take the other three and do a simple min and find out
00:23:02
Speaker
Like if one of them was six thou under, then it would be 194. And that's the lowest point, which means you could do a delta switch means six thou. So you only need to take your top left corner. It was always your datum and do a G10 type adjustment or is it G11, G10 by six thou, which will keep the top left always be your coordinate system, but adjust it down by the lowest point it found. That makes sense? I think so.
00:23:31
Speaker
The whole, when you touch the surface once, you can't touch it four times and amass data naturally. You have to manually apply the Z value, the probed results to its own. Two's a new variable? New variable. You have to probe once and write it to 151. Probe again, write it to 152, probe it. And then, so you have four sets of data now, or you have four numbers.
00:23:58
Speaker
I couldn't find a way to actually query those. In Heidenhein, you can. You can create a whole list and you can say literally a query function or whatever it's called and do some logic. You can have thousands of things and it'll find the minimum in like a microsecond. Yeah, exactly. In Heidenhein, I didn't see anything like that. You have to do it the brute force way where it's like
00:24:22
Speaker
And with four values, it was not too hard, but if there was a hundred, maybe there's a better way. I just, I busted open the FANUC macro programming book and I was Googling a little bit and I didn't find anything interesting that I liked, but it was to the point where it's like, well, I see a path. If I do F statements, I know I can logic my way to the end of this. I just did that. Um, and it eventually worked. Um, the first time I tried it, cause you can do,
00:24:48
Speaker
In FANUC, you can do and statements or even or. So you can go if 151 is less than 152 and 151 is less than 153 and 151 is less than 154. That should work. But for some reason, the brother didn't like it. So that didn't work. Yeah. Nested conditionals like multi-criteria conditionals. Which is really cool. That saves a lot of repeated lines of code to bust them all into one thing.
00:25:18
Speaker
But then I finally had one that worked, except it was giving me wrong results because it was creating a false positive, basically. I forgot why, but so you were a lot of playing. You wrote this code that found the lowest of the four and then it re-probe that one. You were then you're not using the fusion probing your handwriting, your probing because the fusion probing works for the brother, but it doesn't apply
00:25:48
Speaker
So the way the fusion outputs the probe is if you probe the surface and offset your origin to there, it's Z0.
00:25:59
Speaker
You can't make it Z two inches, like if the material is two inches thick, let's say. I thought you could. Well, in bloom anyway, it's a K value that needs to be applied to that line of code, which I knew, but I forgot. And I used the fusion outputted codes and it basically tried to go two inches lower than it was. Yeah. But that's also going back to the classic like.
00:26:24
Speaker
you only crash because you're messing with datums that aren't the same as the surface you're probing. We bake mistakes as well. Even to this day, if I'm probing the top of the surface and it's one inch thick and the datum is actually the bottom and I just can't reach the bottom, I will literally
Consistent Coordinate System Setup
00:26:49
Speaker
shove a hammer down my pocket to remind myself after I probe it to update it because that's just ripe for how we goof. Right. What I'm trying to do is establish a single coordinate system for this fixture system, which is at the top of the pallet system, bottom of the pallets themselves. Yeah, right.
00:27:12
Speaker
and which I think is 2.45 inches in this case with the vacuum pallet and the rich light on top and everything. So in a K value, you should have said 2.45, negative 2.45. I forget. But it didn't. So it tried to crunch through all that. And I don't know why it didn't stop itself, but it didn't. Yeah, that's weird because normally the fusion defaults are the protected position move starts at a retract plane that's higher than the height of the tool.
00:27:40
Speaker
Hm. Why did the machine finally stop? Did it reach the position it wanted to be at? I don't think it got that far. I think I stopped it. Oh, that's impressive. Were you on slower rapids? I think so. Although it's frustrating, because you watch it probe four times, you gain a little bit of confidence. And when it goes to probe that fifth time, you're like, it's guided. I can go full speed now, even though it's a completely different probing op. And I shouldn't have. So I forget what rapids I was on.
00:28:09
Speaker
Yeah. That instant panic. We're like, no, I'm sorry. Yeah. Does Bloom have the body program like Renishaw? I don't know. I was wondering about that, too, as it happened. And then I took a break and I got some food and I was asking the guys about that, too, because, yeah, Renishaw has that like give or take thousand dollar replacement program, which we used once last year. Yeah. I would hope that Bloom has something similar, but I don't know yet.
00:28:38
Speaker
PSA on that too. We called Renishaw and could have definitely done it direct but you had to set up billing stuff and paperwork and I realized MSC with whom we do business works with them as well and we're set up with it. So I would not that I ever plan on crashing the probe again or anybody out there would but it's worth
00:29:00
Speaker
knowing that because like when we crashed a probe, it was a pretty quick process of being like, okay, like let's get into an order within an hour. There's no point in, it's not going to heal itself. And spending an extra day or two for paperwork is no fun. I'm sorry.
Chip Management and Filtration Challenges
00:29:20
Speaker
Yeah, that happens. So we have been cutting quite a bit of rich light and the chip management is an absolute nightmare.
00:29:28
Speaker
Yeah, you said that last week. It's worse this week. Oh, really? We flooded the shop last night. John? I mean, not the shop, but certainly out the speedio and out the garage door a little bit. Yeah. After I left, funny enough, I was like, OK, guys, just started a cycle. Should run for about an hour, hour and a half later. Yeah, yeah, yeah. And I got a text immediately as I'm driving to pick up the kids from school. It's like, ah, flood. We're managing.
00:29:58
Speaker
But it's, I mean, I'm using a corn cob buffer. So it's making the huge volume of chips is now a small volume of ships, but it's creating quite a bit of dust. And some of the finishing operations that I'm doing is just creating this dust that is clogging our 20 micron filter paper. And maybe we need like a hundred micron filter paper or something, but it clogs it so quickly within like half an hour or an hour that it just creates a big bathtub in the, the,
00:30:26
Speaker
chip tray of this video and won't drain faster than the speedio shoves coolant down its throat. And the speedio moves so much coolant that, uh, yeah, I'm trying to figure out what the best solution is here. So, um, my current best thinking is like, you don't have any paper filters in your shop, do you?
00:30:49
Speaker
With a big roll of paper that feeds? Correct. We do not. Okay. My current has it and our Okamoto surface grinder has it. And we even put one on the tumbler to filter the coolant and get all the sludge out from tumbling. So we're quite familiar with them. And I think I want to buy one and put it on the speedio, even as our kind of chip conveyor.
00:31:09
Speaker
for all of time. Oh, yeah. John, that could work. That's totally good. Because I remember touring Miltara's shop a couple of years ago, and they were cutting turbine blades out of aluminum and just making mountains of chips. And they're like, two things we figured out. Chip management sucks.
00:31:27
Speaker
And if you make small chips with the right tools and the right toolpaths, then you can just pump everything out within a slurry and then through a paper band filter or some sort of external filtration system. And that's what they did. They just pumped all of the chips out. No chip conveyor at all. And I think that might be what I want to do on this video.
00:31:50
Speaker
Haas had a, I didn't see it, Amish saw it and sent me a picture right after I left IMTS, which I appreciate, but Haas had a prototype paper band filter, which I- It was like a mini conveyor, like a mesh filter mini conveyor. Yeah, it's exactly what I want though. Amish and I have talked about this a lot since then. Well, then Amish showed that he had one designed. This is the same thing you and I put out the challenge to. I had two people say they're going to do it, nothing happened.
00:32:20
Speaker
But honestly, if Haas comes out with one, it's not that I don't want it to be excellent, but I'm not expecting it to be the Rolls-Royce functionality. I just need a, it's like a treadmill or conveyor belt that's perpendicular to the flow of the coolant underneath the waterfall, which Haas makes it easy because they have that exposed waterfall. Yeah, it's video too.
00:32:39
Speaker
Okay, great. Easy-ish, but I mean, I've been staring at the speedio for hours trying to find the most optimal kind of what's possible, what fits, what could be made, what could be bought. And yeah, that kind of mini mesh conveyor could work, but it wouldn't solve our dust clogging issue on the speedio.
00:33:05
Speaker
but it would get the bigger chips out. Chip management is a serious game. You look at some of the bigger chip conveyors and they have their caveats, but they're also designed for a reason in a way. See the size of the motor on them and they're really powerful and they have to be able to push through a clog or stuff getting in the gears and the belts and the wheels and all this stuff. It gets deeper than it looks on first go.
00:33:31
Speaker
Well, that's how our Concept 2000s work is they have a drum internally that filters it and it's self-cleaning with an inside wash bar that's pushing out on it.
00:33:41
Speaker
Yeah, I still think that's the coolest design. I like that. And I would I would want that over any of the LMSs that I have just because they don't self filter any of that little stuff. It just ends up in the coolant tank. But behind this video, we have like a wall in the way and most ship conveyors want to go sideways, not straight back once you get it custom made. But John,
00:34:08
Speaker
You're cutting rich light. Go buy an off the shelf router. Yeah, it would be better. Totally capable. Use a vacuum for dust extraction, which is exactly what it's like. I don't know much about wood or rich light, but but like you're not. I think you're fighting the wrong battle with micron level filters. Yeah, I kind of agree with you. And I mean, our router is shipping this week, I think the one we've been building for so long. So that'll be cool.
00:34:37
Speaker
And I'm definitely open to putting this job on the router when it gets here and when it's all up. But for now, trying to manage what I have. And then also trying to think, in this video, we want to cut carbon fiber. We want to cut G10. We want to grind rasp blades. We want to create dust.
00:34:59
Speaker
in a way that's like filtering the coolant is easier than dealing with dust extraction. Because I've seen like robot drills and speedios that cut wood dry, and they're an absolute mess. Like that dust, even with dust suction, it just gets everywhere. But if you've flooded all through the coolant and just deal with filtering the coolant, then you're at least isolating the problem. Yeah, I don't know though, John.
00:35:27
Speaker
level of filtration you might need and how often you're changing filters might be futile. Maybe. I mean, the paper band filters are self feeding. Yeah. And they were, they work quite well. And there's actually a company in Ontario, like an hour 15 away that manufactures them. So I called them twice. I emailed them. I haven't heard back. It's been like two, you know, a day and a half. I'm like, come on guys. I want to hear.
00:35:55
Speaker
But even though you'll have to get rid of inline filters because they'll just get clogged. Yeah, so you kind of stage the filtration. You hope the filter paper automatically removes most of it and then you can put bag filters after to make sure even less goes through the spindle, like through the nozzles.
00:36:15
Speaker
It would be called a closed loop system that's basically separate from the machine. It sucks out, cool it from one end to the other end. It puts it back in the other end. It's just like its own bolt-on cord. That's exactly what it would be. I would put a big rubbermaid tub under the waterfalls of main coolant and that would slurry pump out to the paper filter and it would all flow through that and then collect in the tank underneath and then pump back into the speedio and the various float levels to activate it all.
00:36:45
Speaker
Theoretically, sounds amazing. Yeah. The graphite EDM machines we've seen, though, that mill electrodes just use, they put boxes around them to contain the volumetric area, and then they have dust extraction systems. Yeah. I think you're not one to let people tell you you're crazy, but I think you're crazy to try to filter lots of carbon fiber and G10. Yeah. Yeah. Maybe. I have been told I'm crazy before. Yeah.
00:37:14
Speaker
Well, I feel like you're fighting the wrong battle. Like it doesn't need to be that way. Yeah, it's part of it's like, you know, this is me. Certainly getting a handle on this video and using it and figuring out, you know, having a good like thing to sink my teeth into, which has been fun, but also frustrating because it's just fighting an uphill battle. And I do have all this rich light more coming today that I've committed to.
00:37:39
Speaker
making and using. And currently this video is kind of the only slash best machine we have to do that. Do it dry. I haven't. I tried it briefly just to see. I don't know how nasty it would get. Maybe. Maybe we should try it.
00:38:02
Speaker
Can I see a fusion probing? Well, it's not a question. It's a thing I realized that I'm embarrassed. I didn't know, but I want to make sure I'm not crazy here. It's really cool that it works how I think it works. So let's say you finish off one on a part and you need to flip it for up to up to every piece of machine geometry is occluded or hidden because of the hat top.
00:38:25
Speaker
you with me? But you have a through hole that you could drill all the way through and we actually interpolate it or remit at the end just so it's better than just a drill hole in terms of its quality and circularity. I then, when I'm setting up my OP2, I probe that hole because it is my coordinate system and it's located in some random area of this part.
00:38:52
Speaker
And that is my coordinate system. Now, in a perfect world, that should be fine. But the reality is, in this case, it's like a three eighth inch hole. So it's a pretty small hole that you're probing. And for whatever reason, it doesn't seem to be perfect. I'm talking about being off of a few thou over like a 14 inch part for chamfers and so forth. What we realized that I didn't know, and I guess I kind of want to confirm that you knew this or it works and I'm not crazy here, is we set that hole as our datum. That's easy.
00:39:22
Speaker
We then face off and do a rough adaptive to clear away the hat top. So with that adaptive, we're leaving 20 or 30 thou, plenty more to accommodate for any misalignment at this point. Then in Fusion, we use the free in process probing WCS to probe the left edge, the X edge and the Y edge, which are not the coordinate system, but Fusion knows where those are in the CAD world.
00:39:49
Speaker
And they were made in op one. We just couldn't get to them until the hat top is gone. It then updates your coordinate system, making the adjustment by a fower to make the part perfect. Did you know that worked?
Probing Workflow in Fusion 360
00:40:04
Speaker
I have not done in-process probing at all in Fusion yet. Because most of my playing has been on the current, which goes through complete and it's different.
00:40:13
Speaker
Oh, that's right. Right. Right. So I haven't, I mean, the speedio has been the first like three axis I've sunk my teeth into in a good while program wise. Um, but no, I didn't really know that that's, that sounds great. And so the whole isn't good enough or it's well, so we have a six millimeter probe tip and it's only going into a 0.4 inch hole. So there's not the probe moves like 0.05 inches. So I don't necessarily think it's
00:40:37
Speaker
it's not like a flawed workflow. It's just once I've got access to a nice clean machined big surface, I can come in and touch an X and that is perfect. And they are moving by a thou or two, which
00:40:51
Speaker
Well, for that matter, you don't care about accuracy of the first facing op to get that off. You don't even need to probe it. It's just like, oh yeah, it's all five thou. Who cares? Zip, then probe. Well, we have to do an adaptive because the facing would face off true hat top areas where the face mode would tear the material out. OK. So but yeah, I hear you. I guess my point is I didn't realize fusion probing was that smart where you're probing features that aren't the actual coordinate system.
00:41:17
Speaker
There are other just random features on the part, but those features, because there's obviously a CAD distance between that, say, outside edge in the hole that was our coordinate system, it knows, oh, no, that's supposed to be 4.24 inches over. I'm updating your coordinate system by one pal.
00:41:32
Speaker
I do that on the Kern quite a bit, but manually, like I wrote all the probing code to do that with the offsets. And like when we're grinding a Rask blade, we rough machine it and then probe to see how much is left to rough machine. And then we semi finish machine it so that there's always like whatever it is, 3000 left or 2000 or something. And then we grind it and then we probe again and then we finish grind it. Although now I think we just grind once.
00:42:00
Speaker
But there's all these little shifts that are, I want the bevel to be exactly the right position in space to a few tenths. And we're there now. But the tool wears down, the end mill, as it gets dull, it has more deflection. So there's always this fine shift, a couple thou back and forth. And tracking that over time is kind of cool to see, because as the end mill wears, like a three eighths, five flute end mill,
00:42:26
Speaker
the offset gets bigger and bigger because more deflection needs it to go in further the next time kind of thing. Right. Right. So I can watch the diameter where grow from like one thou to 10 thou. Didn't you write like an alarm that once it hits a certain. Yeah, that's so cool, John. Yeah. So it.
00:42:47
Speaker
If it goes beyond that, then something's wrong. And normally, within a normal span of tool life, 100 minutes or whatever this tool lasts for, it doesn't get beyond, say, 12 thou wear. Which is crazy to think about 12 thou wear, but hard milling with a tool that wears out, it works.
Tool Wear and Monitoring Importance
00:43:06
Speaker
I don't know how it works, but it functions. And then the grinding wheel cleans up the rest of it. That's cool. That's cool.
00:43:15
Speaker
So yeah, it's dangerous though, like all this probing mid op handwritten codes, the fusion way, if it works is cleaner because you can just post one file, one code. But kind of like what I'm doing on this video, like the second you get in there and start to fat finger things, test carefully.
00:43:35
Speaker
Yeah. And you even have to come up with like contingencies. So like what if the materials just not there, not installed and it's still probing? What's it going to do? Is it going to probe air? Is it going to hit the surface and know it's wrong? It's all this little things you got to think about. But it's super duper powerful to be able to do all this crazy stuff. So it's fun. Yeah. No, I love it.
Delegating for Growth
00:43:59
Speaker
And I do think it would probably be
00:44:03
Speaker
It's kind of the thing I don't think you want to hear, but for you to grow your company, to take a step back away a little bit, to train up other teams, you've got to stop writing these concepts. I kind of agree with you. Like craziness. You can flex on all of us on how good you are at it, which is awesome. But it's not good for the company. I'm starting to feel it, yeah. It's good because it makes the parts that I want to the tolerance and quality that I want. And it's like an available resource that gives me the result that I want.
00:44:33
Speaker
At this time, I don't know how to achieve that result without this complication, but that's my kind of John Grimsmo. What's obvious to me is like way over complicated to somebody else. Oh, yeah. It's just too much tribal knowledge in. Yeah. Yeah. I hear you. I hear you.
00:44:50
Speaker
And like, hey, a bloom probe and a speedio, probably no big deal. But John, I mean, this could be a $50,000 or $100,000 current day. Yeah. And that's like, I don't know, it's kind of one of those like, those machines are never the same if they're crash hard. Yeah.
00:45:07
Speaker
On the flip side, our lathe runs a bunch of custom-ish code. Now, we're just doing pass-throughs nowhere near as complex as you are, but they're still nevertheless hand-coded stuff, and one of the times I did crash the lathe on it years ago, just to be clear. The problem now is that Fusion has made so many advances in multi-access turning in the post-processor that I don't want to use those because my thing works, don't want to change it.
00:45:31
Speaker
I locked our post down from a post like 2019 and like heck, I'm going to go spend a couple of days risking and testing the newest, latest and greatest iteration of all that. I kind of hear you, but I would say try to start using fusion probing where you can.
00:45:50
Speaker
I think that's what led to part of this problem was I was using the canned fusion post to update the speedio for the probing op, but it didn't have that extra value that shifts my offset up, my work coordinate up 2.45 inches. Anyway. Yeah. I would do it as a second line because that should be, I want to say it's in a G10L
00:46:14
Speaker
something that unless you just, like that could just be a secondary NC pasture that makes it really obvious to you and future people like, okay, it's going to do the canned probing from fusion, but then we are going to make an adjustment afterward. Yeah, I might do that. Or I just make it super easy and say G55 is top of the part, not bottom of the part. It's just like this op is.
00:46:37
Speaker
Yeah. That too, John. Ed was saying something that I'm, I'm just throwing it out there. I don't, I have not looked into this, but something like the height planes and fusion probing can actually make your offset change in a way that's pretty horrifying. Although it might be the feature you want. I don't know if that's true. I don't know if that's only true on like, say like the Haas or Kuma posts, but, um, kind of a PSA slash that's out there over, over time. Phil's taking the boat around the island.
00:47:08
Speaker
Nice. Up to anything cool today? Running parts. Finishing up the last set of valve covers, horizontal. And now it's fighting for time. Now it's like, OK, I want to run some stuff during the day when we can listen to it, watch it. Yes. Do the easy stuff at night. And then honestly, some other good business stuff that I had on my list to talk with you. But the last two podcasts, we've just been busy talking about good machine stuff. So next week, I guess. Next week, yeah. Let's dig in.
00:47:38
Speaker
Yeah. Just, uh, get everything fixed today. Okay. Yeah. Run parts. Okay. I got a pretty cool picture of the tip buried in the rich. I'll post that up on Instagram. Yeah. Cool. All right. I'll see you. Take care. Bye.