Become a Creator today!Start creating today - Share your story with the world!
Start for free
00:00:00
00:00:01
#432 Raspberry Pi Home Assistant and webcams
0 Playsin 1 day
Topics:
- Stress relieving titanium
- Robot flip fixtures
- Foam update
- Raspberry Pi Home Assistant and webcams
- Probing too much
- Telescopes and lenses
Recommended
Transcript
Introduction and Topics Overview
00:00:01
John
Good morning. Welcome to the Business of Machining episode 432. name is John
00:00:06
johngrimsmo
And my name is John Grimsmo.
00:00:08
John
John and I have lists of things that we need to talk about and you guys get to listen in. Um, my list, actually my list starts off with you, which is stress relieving titanium and steel. And then, um, reverse osmosis for shops, raspberry pies, ah brother controls and what causes them to crash.
00:00:25
John
Um, 3d printing with materials added like steel and wire. That's going to be an interesting and quick one. And then finally, um
00:00:32
johngrimsmo
OK.
00:00:35
John
Our paths continue to have similarities yet maybe a slight divergent. You've got your watches in the last sort of two
Stress Relieving Metals
00:00:44
John
weeks. William and I, my son and i discovered telescopes.
00:00:47
johngrimsmo
Oh, I saw a little picture you posted and i was like, which telescope is that? Okay.
00:00:51
John
So that's my list.
00:00:51
johngrimsmo
Yeah. Okay, cool.
00:00:52
John
what
00:00:53
johngrimsmo
Yeah. My list real quick. Give you shipping update. Also stress relieving titanium, a little foam update. ah Talking about time studies in the shop and then ah but probing too much.
00:01:01
John
Okay. Yep.
00:01:06
John
Yeah, yeah, that's a good one.
00:01:06
johngrimsmo
Like, like a blue Miranda shot kind of probing too much.
00:01:07
John
Ooh, that's actually good. added a note too on something. Okay. ah Well,
00:01:17
johngrimsmo
Send it.
00:01:17
John
what good all the go shoot shoot your shot on Strasil Leaf.
00:01:23
johngrimsmo
So um because our we we've seen a lot of titanium, a lot of steel, there's a lot of heat treating involved in the steel. um Everything warps, everything bends.
00:01:34
johngrimsmo
Trying to learn, we've talked about this for many years, trying to learn how the molecular you know makeup of a steel makes it want to warp or bend or stretch or and what stress relief kind of does and how to do it.
00:01:48
johngrimsmo
um So there's both stress relieving are blades, which is a ah before heat treating step.
00:01:56
John
Interesting.
00:01:56
johngrimsmo
so
00:01:57
John
that a thing?
00:01:57
johngrimsmo
Some people, yeah, absolutely.
00:01:58
John
Pre-heat-treat stress relief?
00:02:00
johngrimsmo
Some people do it before machining so that it doesn't move as much during machining.
00:02:04
John
Yeah, that makes sense.
00:02:05
johngrimsmo
I'm, I think, well, I forget if I do it, if I want to do it before anything happens, like raw blank kind of thing. or if I want to do it after machining, but before heat treat to, I don't know i forget.
00:02:21
johngrimsmo
It's been a while, but for titanium specifically.
00:02:22
John
Yes, sir.
00:02:24
johngrimsmo
So we have our integral handle, which is a solid chunk of titanium with a slot milled out of it.
00:02:28
John
Yes.
00:02:29
johngrimsmo
And the two leftover walls, I think they're moving a little bit.
00:02:31
John
Yep.
00:02:34
John
I guarantee they're moving.
00:02:35
johngrimsmo
ah For sure.
00:02:35
John
How much? Mm-hmm. One.
00:02:36
johngrimsmo
But I think they're moving. Well, they're not moving consistently. That's the thing. Because we have a way, because we undercut the the place where the bearing goes. bearing Bearings go on both sides, the little bearing cages.
00:02:48
johngrimsmo
um We have a way to measure that gap and say, I want it to be one, two, five. And sometimes it's one, two, four, three, two. Sometimes one, two, six.
00:02:58
John
well
00:02:58
johngrimsmo
So we have maybe a two two or three thou, three to four thou variance handle to handle, which I don't like.
00:03:05
John
Okay.
00:03:05
johngrimsmo
um it's not machining i don't think anyway i mean it's like literally handle to handle and then it changes and goes back and there's no like it's not really thermal the current's thermally stable um it's not too aware it's i'm leaning towards it being the material itself and i'm wondering if a stress relief step before any machining
00:03:06
John
Yeah. Rory.
00:03:27
John
Mm-hmm.
00:03:28
johngrimsmo
loosens all the grains inside and just lets them all, you know, be happy together. um So that when they machine the machine more consistent. So I found a a few heat treat documents on stress relieving titanium, one of which was from allegedly the SR 71 Blackbird project.
00:03:46
John
i'm I'm literally reading the skunk workbooks. Like I'm, I almost finished it this morning in this exact, well, not stress relief.
00:03:51
johngrimsmo
Yeah.
00:03:54
John
Well, sort of the the only story.
00:03:55
johngrimsmo
Well,
00:03:55
John
Good.
00:03:56
johngrimsmo
This document didn't mention the SR-71, but the guy got it from thought it might have.
00:03:58
John
Yeah. Yeah.
00:04:00
johngrimsmo
um But it's like, I forget the year, like 50s or whatever that came from. And yeah, so it's it's not necessarily up to date, but it has some very interesting information about it.
00:04:11
John
But in some respects it is because it's like so much of what we figured out then has been the basis of, i mean, that was the first widespread commercial use of titanium in an airplane.
00:04:13
johngrimsmo
I don't know.
00:04:23
John
And really like they sourced it through a, see this is a commonly known, the fact the thing I heard was that the CIA quote unquote smuggled it with CIA agents and operatives in Russia.
00:04:34
John
It actually wasn't quite as sexy as that. They just created a whole bunch of shell companies.
00:04:38
johngrimsmo
Hmm.
00:04:38
John
The last of which was like, you know, J and J's mining in Asheville, North Carolina, that just bought the stuff from Russia. Like they were rare.
00:04:46
johngrimsmo
yeah
00:04:46
John
Maybe it wasn't, maybe it was an intermediate country, but like they just bought it commercially from Russia, but hide it who the end user was.
00:04:52
johngrimsmo
Hmm.
00:04:52
John
But the crazy punchline was, that other that other Kind of one of those, like, I was laughing about how you and I would just be, we would be talking about this on a business and machine episode is they did all this machining with titanium. They had to figure out how to new do new grinds and drill bits. This is pre-carbide um machining it. And they figured all that out.
00:05:12
John
And they made the they made the the darn plane or prototypes. And then like weeks or months later, they started having issues, which were because the, now I don't remember exactly what it was. It was like the chromium or the fluoride and the water and the coolant or something caused...
00:05:29
John
Initial stressors that took weeks. Oh, no, it was the heat and annealing cycles or something combined with the coolant use or something that caused the stuff. And they basically had to like, can you imagine, John, you made either one of us made these products. And then later we realized something in innocuous that we did or didn't know about this new material is now causing us not only to fix our product, but our product involves manned flight.
00:05:51
John
Like, oh, anyway, sorry.
00:05:52
johngrimsmo
Yeah, yeah. And there could be like cracks in it or whatever. um Yeah, so I'm trying to figure out, you know, what it looks like. ah I forget the exact number, but I think it's 1100 degrees Fahrenheit for two hours, something like that.
00:06:07
johngrimsmo
um So relatively easy to do, you know, it's not a super high height heat. um And I actually wanted to know, was like, I know we can heat anodized titanium. If you take a blowtorch to it, it turns blue or purple or or green, depending on the temperature you get it to.
00:06:18
John
Mm-hmm.
00:06:21
johngrimsmo
So the first step was I had my guy, my heat treat guy, Larry, um just throw a piece of scrap titanium in the oven for ah a while, like an hour at this temperature and see what color it turns. You know, does it anodize? Does it turn black? Does it turn, you know, gray or something like that?
00:06:36
johngrimsmo
And it just anodized the lightest blue. which tells
Heat Treating and Machining Consistency
00:06:40
johngrimsmo
me as far as the anodizing spectrum go, it's like super low on the oxide layer. and And I was worried it would grow such a black oxide layer that it'd be hard to machine through or bad for the tools or something like that.
00:06:50
johngrimsmo
No, not at all. Just like a super thin blue layer, like light blue. um So I had him heat treat or stress relief 10 of our integral handle blanks.
00:07:02
johngrimsmo
And we've machined one of them, but I don't know... I think we have a measurement from it, but what I really want is to machine all 10 and see the see if the trend stops, like they're all the same size now, or if they're still all over the place and the difference is somewhere else.
00:07:17
johngrimsmo
um I think, like my gut says, this is a good thing and it's relatively easy for us to do, ah but I need to get through all 10 of them to like see data and I don't have that yet.
00:07:24
John
who
00:07:28
johngrimsmo
So stay tuned.
00:07:29
John
Do you put it in a seamless baggie?
00:07:32
johngrimsmo
ah No, we chose not to because in the atmosphere at that temperature, they they don't really oxidize enough to care about.
00:07:38
John
Okay, doesn't matter.
00:07:39
johngrimsmo
Yeah. And we're cutting away all the skin, all the outside, like cutting through everything.
00:07:40
John
Can ask you?
00:07:43
johngrimsmo
So it doesn't matter.
00:07:45
John
Oh, it all gets net machined away anyways.
00:07:46
johngrimsmo
Yeah. tough Yeah.
00:07:47
John
Yeah, we're right.
00:07:47
johngrimsmo
Yeah.
00:07:48
John
If I can ask a favor or suggestion, at least one or two, if you have the tooling to rough machine the slot, but leave 10, 20 thou,
00:07:59
John
Then also do that. I don't even think it matters if you, and it wouldn't hurt pre start to relieve that before that, but really machine that away, then stress relieve it, then finish it.
00:08:10
johngrimsmo
Yeah, Angela suggested that too. That's what he really wants to believe. um Because in in aerospace, what he used to do a lot is that rough machining, leave 20 thou, come back and finish.
00:08:21
johngrimsmo
um And i'm sure, it's kind of annoying, though, to like, mid process stop, you know?
00:08:27
John
No, I do it.
00:08:29
John
I agree, but that's something you and I have to get over on our stubbornness level of like, if you just have to seek the truth.
00:08:33
johngrimsmo
Yeah.
00:08:35
John
And if it's annoying, we'll figure it out.
00:08:36
johngrimsmo
True.
00:08:36
John
Like just, yeah.
00:08:37
johngrimsmo
yeah Well,
00:08:39
John
But I worry that it's such a it's such a deep channel shape in the end.
00:08:42
johngrimsmo
Yeah.
00:08:45
johngrimsmo
I think we currently slot it and then leave five or 10 thou per wall. And then I do finish machine those walls, but all in the same operation. um
00:08:54
John
And you're, but you're just holding the blade up from the bottom, like on a dovetail.
00:08:58
johngrimsmo
Yeah, exactly.
00:09:00
John
Yeah.
00:09:01
johngrimsmo
So with the dovetail, there's no locating feature, so I can't reliably take it out put it back in. Yeah.
00:09:07
John
You don't lie. Oh, you could line chunk it type thing. You know, that actually would be repeatable enough here.
00:09:10
johngrimsmo
Yeah. And that would repeat. Yeah, you're right. Yeah, for sure. Yeah, we waterjet the blanks to be the exact width of our vice. So we just kind of finger align it side to side with the vice and make it centered.
00:09:21
John
Yeah. Yeah.
00:09:24
johngrimsmo
And that works great for initial setup, but I wouldn't repeat that.
00:09:27
John
Yeah, for Yeah.
00:09:29
johngrimsmo
Yeah.
00:09:30
John
ah
00:09:31
johngrimsmo
But it's like, you know you can have five different ideas, one of which has to be tested.
00:09:37
John
yeah
00:09:37
johngrimsmo
um so I'm going test this one, get all 10 done. If we see a trend, awesome. If we see similar all over the place results, yeah. And this is fixable down the road. like We're not scrapping handles because of it, but it causes more work for the finishing guys to like deal with a bigger gap or a smaller gap.
00:09:56
johngrimsmo
You have to roll in the bearings, like roll in races pretty much to make them bigger. If it's too tight, yeah.
00:10:02
John
Oh yeah, right, right, right, right
00:10:03
johngrimsmo
ye And I want consistency across everything. So if if it's as simple as heating up the blanks for two hours, great. If it's more complicated, yeah
00:10:12
John
right. There's ah also an aspect of the stress relief that feels very full Grimmsmo of like, hey, you I'm going to hypothesize without any but basis to this that you if you ship a titanium PL that has stresses built up in the handle, I'm going to hypothesize that over the course of EDC for four years, it'll naturally stress relief.
00:10:36
John
I may be wrong on that, but I'm um thinking about when we did the scraping class and we had our cast iron process.
00:10:37
johngrimsmo
Yeah.
00:10:42
John
um hugely stressed cast iron castings and you would hang them up and hit them with a hammer and let them sit there overnight. And that changed the stresses.
00:10:52
John
um
00:10:53
johngrimsmo
Crazy.
00:10:54
John
Yeah. And I mean, um this is Richard King, who is like 90% expert, 10% just bat-esque crazy. But I'm going to go ahead and say it's probably not a bad idea.
00:11:03
johngrimsmo
Yeah, yeah.
00:11:03
John
yeah. so
00:11:05
johngrimsmo
Yeah, and Angelo mentioned that too, because I wonder if it moves over time. Like if you machine the slot and you measure it and then the next day or a week later or something, if it's different, I don't know.
00:11:17
John
Yeah. you could um mean get Part of me gets so excited about just building crit traps. You could just build a little like alarm clock. you book Go buy an alarm clock for $10 off Amazon.
00:11:27
John
It has a little ah ringer.
00:11:28
johngrimsmo
Like a little pecker. Yeah.
00:11:29
John
pe Yeah. And peck the bejesus out of a fjell. Put it in a sound box.
00:11:34
johngrimsmo
So...
00:11:35
John
Yeah.
00:11:36
johngrimsmo
interesting what you said about ringing the cast iron, because I've heard of that too. I think I did some scraping research a few weeks ago, and or months ago, and they said that. And I was like, no way.
00:11:47
John
Yeah, that's how they shot. they There's a modern Marvels episode on when they rehabilitate M1 Abrams tanks. um And they hang up the tank, like pick it up vertically, like in a building that kind of looks like the Space Shuttle Assembly building, like huge.
00:12:02
johngrimsmo
Yeah.
00:12:03
John
And then the doors close and it shot peens the tank. I mean, you would die immediately. It's like a 12 gauge going off a million times a second at these things. And it stress releases it.
00:12:15
johngrimsmo
Yeah. Yeah. Shot peening is a different type of stress relief. Cause you're, you're peening the surface and you're, you're creating micro dense all along the entire surface to spread loads or something like that.
00:12:22
John
Yeah.
00:12:30
johngrimsmo
I don't know. I feel like my gut says that's different than a heat treat stress relief where it's all internal and internal stresses so get relieved.
00:12:36
John
Oh.
00:12:37
johngrimsmo
Right.
00:12:39
John
Yes, way above my pay grade. My point was shot peeing probably has some analogy to tapping with a hammer.
00:12:45
johngrimsmo
ah yeah Okay.
00:12:45
John
to um That was my thinking, but... ah
00:12:48
johngrimsmo
Yeah, for sure.
00:12:49
John
um That's kind of like, yeah, if you cut that channel in that slot, the way you're doing it, I guarantee you it's it's either collapsing in or opening up, which is either way going to affect your finish.
00:13:00
johngrimsmo
But based on what? What is the variable there? and I'm thinking it's material grain structure part to part, hence stress.
00:13:06
John
Yeah. Yeah. So I was just talking to Courtney, our intern about this, because we're building fixtures. we're sorry now We're designing fixtures for our new One Piece Flow robot brother system.
00:13:20
John
And we had, she has it in her office. um We have a fixture that is just a simple riser. Think of a one inch one inch candy bar. It's at a square end. It has tapered ends so the chips can wash off of it.
00:13:34
John
And instead of that being like a parallel, the bottom has four holes for for either S7 or maybe carbide ball bearings. And the top has three ball bearings.
00:13:47
John
Three gives us the plane on the top for part. And what we need to do is we need to flip the part with the robot for an opt two and we need to maintain that parallelism and nominal dimension pretty darn perfectly.
00:13:54
johngrimsmo
Oh.
00:14:00
John
And we do that on our horizontal now by clamping through the part, but that's a manual process. And in a weird way, you actually could potentially have a UR robot actually thread in fasteners and like fixture fasteners that then get clamped through the part, but we're going to try it a different way. I think it's going to work. And we've been talking to lights out our integrator and showing them samples and prototypes and videos. And they're like, they were,
00:14:23
John
and you know, we're still early on in working with them. So there's, I think it's important to say there's the honeymoon phase to things, but like the reality is if they say they're not, they don't seem like the kind of guys who are going to feed feed you a lot.
00:14:33
John
Like, Oh, that'll work. And then all of a sudden we start doing it. We're like, Oh, we can't
Titanium Processing Methods
00:14:36
johngrimsmo
So
00:14:36
John
and do this.
00:14:37
John
So sorry, I'm rambling here. That candy bar can't be a candy bar though, because it has to have a U channel cut away in it for the um end effector to reach in and grab the part, which means things
00:14:48
johngrimsmo
you're creating like a flip fixture.
00:14:50
John
just It's just a fixture to lay the part down in the vise.
00:14:53
johngrimsmo
Okay.
00:14:54
John
It's just a big parallel. But instead of instead of being parallels, we're setting the part down on three ball bearings.
00:15:00
johngrimsmo
Got it.
00:15:00
John
Three ball bearings forms a plane. But sorry, sorry, to your point, yes.
00:15:03
johngrimsmo
Okay.
00:15:05
John
This is flipping it from Op 1. The robot comes in grabs the part. Sorry, I know but people who listening can't hear me, but like I'm putting my fingers out front of me like a pair of scissors.
00:15:09
johngrimsmo
Yeah.
00:15:13
John
That's the end effector. It grabs the part out of Op 1. It lifts it up, then the robot arm twists 180 degrees, and then it sets it in the second vice. That's all it does.
00:15:21
johngrimsmo
Yeah.
00:15:22
John
But when it sets it down the second vice, it has to have clearance for the end effector to not crash with that solid candy bar parallel.
00:15:29
johngrimsmo
so So you're you're basically creating a planar floor
00:15:36
John
Exactly.
00:15:37
johngrimsmo
um And then your XY location is a separate issue.
00:15:37
John
That's a simpler way of saying it. Thank you.
00:15:43
John
Correct. Actually, it doesn't.
00:15:44
johngrimsmo
Yeah.
00:15:44
John
Yeah, exactly. Yes. um And so we ended creating effectively a fjell blade, except so we have a deep slot, but we have a much wider thickness on each side.
00:15:55
John
But regardless, that will, and so we were having this conversation about hot rolled versus cold rolled.
00:15:58
johngrimsmo
Yeah,
00:16:01
John
And, um, I, again, I'm not expert. I'm actually, I'm thirsty for someone to correct me or teach me more, but cold rolled steel gets, it gets hot poured into a block and that block gets rolled back and forth in a rolling mill, like a building that might be a mile long.
00:16:16
John
And it literally goes under super strong rollers to thin it out. to So if you want to buy it to two inch thick material, they roll it down to two inches in cold.
00:16:19
johngrimsmo
Yep.
00:16:24
John
So it kind of seems logical to at least the all of us that are born with this mindset for metal that, yeah, if you do that cold, it's going to be really stresses built up in
00:16:34
johngrimsmo
Yep. it's It's flowing and it's moving where it wants to. And you're putting weird strain between different grain boundaries and all this stuff. And I'm pretty sure our blades and definitely our titanium is cold rolled to thickness.
00:16:47
johngrimsmo
um
00:16:48
John
No way.
00:16:49
johngrimsmo
I think so.
00:16:49
John
i Well, so I thought certain materials were only ever hot rolled. Let me look at that.
00:16:55
johngrimsmo
a
00:16:55
John
Let me risk it. Because if you were if it was cold rolled, John, this thing would be opening up on you 30,000. Fair. Okay.
00:17:01
johngrimsmo
Maybe just hot rolled and they just blast the outside like roughly because it doesn't have that hot rolled like steel finish, but it does have a blast.
00:17:11
John
fair could be happy
00:17:12
johngrimsmo
so I don't know.
00:17:13
John
Okay, most titanium pleats plate and sheet starts with hot rolling. Caveat, enter, this is chat GPT. ah Most titanium plate sheet starts with hot rolling. Some is cold rolled after hot rolling to achieve tighter thickness tolerances, smoother finish and higher strength.
00:17:29
John
um But it requires immediate annealing steps to prevent cracking. It's common for thin sheet or foil applications. So structural and plate is hot, thin sheets and precision is hot, then cold.
00:17:44
John
I'm going to speculate that yours is Hotman, but again, that's guessing.
00:17:50
johngrimsmo
That's a good question. I have to ask my supplier. I know they're a rolling mill, like Niagara Specialty Rolls. That's their job.
00:17:58
johngrimsmo
And they yeah and they they thin it out.
00:17:58
John
Isn't that crazy?
00:17:59
John
Yeah,
00:18:00
johngrimsmo
um I just don't know. and they probably They would probably have both hot and cold options in in-house.
00:18:03
John
that's
00:18:08
johngrimsmo
um
00:18:11
johngrimsmo
Writing that down.
00:18:11
John
but But obviously even hot is going to have for sure variations in stresses.
00:18:18
johngrimsmo
Mm-hmm.
00:18:19
John
But I'm going to... I'm going to speculate, like if you and I were making an offline side wager, that even even if you have hot rolled material and even if you stress relieve it prior to, you're going to put stresses back into it as you remove what is effectively 80 to 90% of that center area, of that slot.
00:18:36
johngrimsmo
Yeah. yeah
00:18:37
John
So I'm sort of saying I don't care what your material is or what state it's
Probing in Machining
00:18:41
John
in. You need to stress relieve it after that rough machining before you finish machining.
00:18:50
johngrimsmo
So we have options. First step, I'll do the test of stress relieving first. And if that provides inconclusive results or bad results or whatever, then I will figure out how to stress relief mid process.
00:18:58
John
yeah
00:19:07
John
Yeah.
00:19:07
johngrimsmo
as long as you have a way of repeatedly putting it back on, you know, theoretically you're leaving 10 thou everywhere. So you should be able to locate within 5 thou and then finish everything.
00:19:19
John
Yeah.
00:19:19
johngrimsmo
Yeah, so.
00:19:20
John
what What sort of fixture is the current fixture holds these?
00:19:24
johngrimsmo
I mean, it's an Aroa Pulse.Base, but it's small vices, like three inch wide ah little vices that that dovetail, that hold the dovetail.
00:19:34
John
Do they have teeth that would orient, like would they would sort of locate on some X or y
00:19:38
johngrimsmo
They do, yeah. yeah yeah
00:19:41
John
Okay. I bet you it repeats pretty, once you crunch it in that you don't need the Lang stamper. Like as soon as you.
00:19:46
johngrimsmo
I machine a dovetail so it's not crunching anything. um It's literally taper gripping the dovetail so that the teeth don't really do anything.
00:19:55
John
I'm sorry. so Oh, okay.
00:19:57
johngrimsmo
But you could figure something out.
00:19:59
John
Yeah, it's easy.
00:19:59
johngrimsmo
Yeah.
00:20:00
John
Got it.
00:20:00
johngrimsmo
yeah So anyway, that's my stress relieving.
00:20:02
John
Oops.
00:20:07
John
I want though, I like, there has to be a way, you know, you're talking about, um continuous casting or not like automate, uh, automation plus foam molds. It's like, there's no reason why you can't have your own robot that pulls a fixture out of the Auroa, puts it in a heat treat oven, especially you don't have to put it in in a stainless bag.
00:20:29
John
Um, and then reloads it back in.
00:20:32
johngrimsmo
That's a crazy idea.
00:20:33
John
but it's not.
00:20:34
johngrimsmo
That's just funny to think about.
00:20:38
John
I'm guessing just like an open flame isn't enough. Like has to be 1100 degrees for an hour.
00:20:40
johngrimsmo
Yeah,
00:20:41
John
Is it something you're going to accomplish with induct? Well, maybe you could induction.
00:20:46
johngrimsmo
yeah I don't know. You want consistency, you want thorough evenness. I don't know. But it's ah it's also not difficult to unclamp them, put them in an oven, reclamp them.
00:21:00
John
Yeah.
00:21:01
johngrimsmo
Interesting.
00:21:06
John
Um, can I ask to learn, um I need to learn brother machines.
00:21:11
johngrimsmo
Yep.
00:21:12
John
Can you remind me the quirky thing that you were ticked about that caused you to crash your probe?
00:21:20
johngrimsmo
Oh, oh um
00:21:25
johngrimsmo
the, i think i was using old probing macros that are, that differed from the documentation.
00:21:37
johngrimsmo
Like when you type, when you type a probe statement that says, you know, a is your Z distance, W is your whole size or whatever.
00:21:37
John
Oh, okay.
00:21:46
John
Okay.
00:21:48
johngrimsmo
one of the number one of the letters is, are you using ah fixtures like WCS up or using surface down kind of thing? And it was that difference that I got wrong um because my macros differed from what the documentation said.
00:21:58
John
Yeah. Okay.
00:22:00
John
Okay.
00:22:05
johngrimsmo
And i had to I had to dig deep to find that.
00:22:05
John
Okay.
00:22:07
johngrimsmo
So I don't know if that's a problem for everybody, but it was a massive problem for me.
00:22:12
John
Yeah, that's actually great to hear because I remembered incorrectly thinking that it was because I remember feeling like somebody else like oh, that happened to me too. But it's like, no it was something more crazy. Like everybody could do this if you, when you move to a brother from a FANUC or a HOS, it's like, this is a common, this is a common thing people do and it stinks.
00:22:24
johngrimsmo
Yeah.
00:22:28
johngrimsmo
who
00:22:28
John
But it sounds like that's not, I'm misremembering.
00:22:30
johngrimsmo
The other thing that Yama's in his installing it for you probably knows this, but my guys in Canada, it didn't seem to know this is the speedy only has one, what's called skip signal.
00:22:42
johngrimsmo
So when you use two probes, you need to skip signals because one is the, for the spindle probe ones for the table probe. Um, on most FANUC machines, there are two skip inputs, so it's easy.
00:22:55
johngrimsmo
But on the Speedio, you have to create a flip-flop relay that only does one or the other, and the logic of the probing routines tell it to flip.
00:22:59
John
Interesting. OK.
00:23:05
johngrimsmo
um Now, my install guy didn't know that, so he spent days here. Like, why isn't this working? ah Until we installed the relay and got it figured out and got it working.
00:23:16
johngrimsmo
um And then i told you know some guys like Ken and he goes, oh yeah, that's people know that. like Well, not everybody knows that.
00:23:22
John
ah Okay, but the the relay kind of reroutes the train tracks, meaning if you go to use your spindle probe to touch off a part, and when that happens, you accidentally bump your tool probe, it's not going to think the Z height is where the Z happened to be bo jogging as you touch the tool probe by accident.
00:23:41
johngrimsmo
No, it's not gonna fail like that.
00:23:43
John
Okay.
00:23:44
johngrimsmo
um But if it, say the relay doesn't trigger, then one of your devices is not gonna work. um
00:23:50
John
Yeah, sure.
00:23:51
johngrimsmo
Or you might get one device working when you're installing and the other one just doesn't work and you can't figure out why something like that.
00:23:51
John
Okay.
00:23:56
John
Okay. That's good to know.
00:23:58
johngrimsmo
So yeah, yeah, they call it skip signals, which whatever exactly that means, I forget.
00:24:03
John
Yeah. I think it's just an interrupt.
00:24:06
johngrimsmo
Yeah, maybe.
00:24:06
John
Um, it's actually a pretty fascinating thing. I dug up the Renishal like white
Shipping Updates and Miscellaneous Topics
00:24:11
John
paper of years ago on it because it's why feed rate matters so much, both the nominal feed rate, but also being consistent because, um, there is,
00:24:22
John
noise or delay, more so delay. So like when your Renishal goes to touch off, the machine has continued to move before the signal can make its way back.
00:24:33
johngrimsmo
Yeah, it takes time.
00:24:34
John
And so you, yeah, I mean, it's small, but, but these probes are are accurate so they can comp for that.
00:24:35
johngrimsmo
Yeah.
00:24:41
John
So long as you're consistent and you have true, and I mean, this is again, way above my pay grade, but like you need to make sure that the skip signal can interrupt at the right point in time and not wait for some other microprocessor behind the scenes service to occur or happen.
00:24:53
John
Cause that variation of microseconds will change the values.
00:24:57
johngrimsmo
Yep. Yep. And then I think in calibration, it's sort of auto accounts for those values because you calibrate based on fixed gauges kind of thing.
00:25:04
John
Yeah. Right. Right. It knows, but it's weird. ah you If you extra extrapolate that out, it's pretty weird because it's kind of like anybody here has ever shot sporting clays.
00:25:14
johngrimsmo
Right.
00:25:17
John
It's like you're shooting in front of the the clay pigeon a lot and it feels weird, but you know, your shot is going to get there when the clay pigeon happens to be there as well.
00:25:20
johngrimsmo
Okay. Yeah.
00:25:25
John
Or a football guy thrown to a running back who's going left to right or a wide receiver.
00:25:29
johngrimsmo
Yeah. yeah
00:25:30
John
Yeah.
00:25:31
johngrimsmo
Huh.
00:25:32
John
Yeah.
00:25:35
johngrimsmo
Give you a quick update on the foam. um I was battling with the the mixture of the foam and it wasn't going together and it was curing rock hard. um I think I figured this out after our podcast last week.
00:25:47
johngrimsmo
I ordered the wrong foam. I ordered the rigid foam, not the flexible foam. I'm like, why is this rock hard?
00:25:52
John
Oh.
00:25:52
johngrimsmo
Why is it brittle? Why is it falling apart? I ordered the rigid foam because I got confused.
00:25:57
John
of
00:25:57
johngrimsmo
And so anyway, new stuff's coming in the next couple of days.
00:25:58
John
Yeah, right.
00:26:00
johngrimsmo
Um, there are some definite chemical concerns, um, in the shop. Like you read the SDS and everything's going to kill you. Um,
00:26:09
John
Yeah.
00:26:10
johngrimsmo
But so we're taking that seriously. um One of them is very common chemical in all foam products, especially spray foams, truck bed liners, things like that.
00:26:20
johngrimsmo
So it's actually a very well researched chemical problem in the foam industry.
00:26:20
John
OK.
00:26:25
johngrimsmo
um Just put all the right precautions into place and make sure and and obviously put it on the way scale of is it worth us doing this, you know, given all these concerns.
00:26:29
John
Yeah.
00:26:34
John
Right.
00:26:37
johngrimsmo
So I'm just collecting all data, but. Yeah.
00:26:41
John
Did you three you, you were testing it a 3d printed mold.
00:26:46
johngrimsmo
At the time, i was just testing it on a paper plate. um Just like watch it foam up and see.
00:26:49
John
Okay.
00:26:51
johngrimsmo
and then it dried rock hard.
00:26:51
John
Yeah.
00:26:52
johngrimsmo
And i'm like, what the heck? So there's a lot to it. like it's It's easy to do simply. But if you're trying to scale for production, if you're trying to think of this whole process as like, no, I want this to be fast, easy, cheaper, easier, um safe.
00:27:00
John
Yeah.
00:27:06
johngrimsmo
you know, is this viable to make, you know thousands of foams year kind of thing?
00:27:08
John
yeah
00:27:11
johngrimsmo
Is it faster than the way we're machining it? So I don't know yet. I, my gut says yes, but we'll see if, uh, if I end up scrapping it.
00:27:19
John
a while back we were touring a, like a film set with the kids. And I like saw the prop room around the corner. I like wanted to peek in and you're you're like on this rigid sort of formal rigid tour.
00:27:27
johngrimsmo
Yeah.
00:27:30
John
And I looked in and I saw two things. I saw boxes from smooth on and I saw a row of bamboos and there's just something crazy.
00:27:34
johngrimsmo
Yep.
00:27:37
johngrimsmo
Yes.
00:27:38
John
Cause I just feel feel like you and I are right.
00:27:39
johngrimsmo
That's all they need.
00:27:42
John
It's crazy.
00:27:43
johngrimsmo
Yep. You can make anything with those two objects.
00:27:46
John
Yeah. Yeah.
00:27:47
johngrimsmo
So the in production, I'll be making 3D printed container and then a silicone like sleeve mold or whatever you call it inside because the foam won't stick to the silicone.
00:27:55
John
mm-hmm yep
00:27:58
johngrimsmo
So you can peel it off easily. But for quick and dirty testing, like doing that whole multi-layer mold is kind of a lot of designing and programming. And I don't know where to put all the vents and all that stuff. I don't really know yet.
00:28:09
John
okay yeah yeah
00:28:09
johngrimsmo
So quick and dirty, I just figured I'd print a mold positive or negative or whatever, just like with a cavity inside, two halves that just kind of lay together.
00:28:19
johngrimsmo
And then I'm going to mold release spray on the inside and just fill it with expanding foam, let it come out the holes. And that will tell me the density of the foam that will let me see it, feel it, you know, without worrying about silicone at all.
00:28:28
John
Yeah.
00:28:31
John
It seems very smart.
00:28:32
johngrimsmo
Yeah, it's gonna be great. So next couple days, I'll have that.
00:28:33
John
Yeah. Good. Cool. You said you had a shipping update?
00:28:40
johngrimsmo
um We're shipping. That's it.
00:28:42
John
Oh, okay. All good.
00:28:43
johngrimsmo
Things are pretty good. Yep. We have a limited number of shipping carriers because of a lot of reasons. So we are on FedEx right now. um We haven't shipped a ton of stuff with FedEx. I know you have ah a close working relationship with FedEx, but yeah, it's it's going well. They ship fast. It's more expensive. It's like noticeably more expensive.
00:29:01
johngrimsmo
So
00:29:02
John
This is compared to the Canada Post though.
00:29:02
johngrimsmo
so Come in to Canada Post and even UPS.
00:29:06
John
Oh, is that right?
00:29:07
johngrimsmo
Yeah. But maybe it's because, yeah, maybe it's because we don't, <unk>t I don't think we have like a formal account with FedEx, like like with terms and all that stuff.
00:29:08
John
Ask for better rates.
00:29:15
John
Oh, John, you got a call.
00:29:17
johngrimsmo
Yeah, 100%.
00:29:18
John
like Sorry, call them today.
00:29:19
johngrimsmo
Yeah.
00:29:19
John
Good grief.
00:29:20
johngrimsmo
I think might think my guys tried to and i don't we're working on it.
00:29:23
John
Okay.
00:29:23
johngrimsmo
But that's because we've only shipped maybe 10 packages ever through FedEx.
00:29:23
John
Yeah.
00:29:27
johngrimsmo
So we haven't needed it. But as of now, like, yeah.
00:29:28
John
Yeah. Yeah. Good.
00:29:31
johngrimsmo
And we do have a formal UPS account. maybe that's why.
00:29:34
John
Yeah, sure. Those, I mean, there still are frustrations with, even for us with FedEx, but like if you tell a FedEx salesperson, like, hey, we're with UPS, but we're, we're considered switching um you, that is their full-time job.
00:29:50
John
Like to, like they you are like ah a cat fish on the line.
00:29:53
johngrimsmo
It's good to know. You're golden goose, yeah.
00:29:54
John
Yeah. Yeah. Yeah. yeah yeah
00:29:58
johngrimsmo
Cool.
00:30:00
John
I was um watching, excuse me, I was listening to the wonderful PFG Live. And then he was talking about Tuck. i'm I'm sure this was Tuck's Garage. Had a timing belt go out on an older machine. Does this sound familiar?
00:30:17
johngrimsmo
No, not yet.
00:30:18
John
Okay. Well, if it's not him, it I guess it doesn't matter. But this idea of, hey, I have this machine, new or old, and I don't know how to replace this profiled belt. Now, on the podcast, Robin Ranzidi was listening. Of course, Robin is like, that was made by Perkins Wilker and like, whatever, this line or something. And that's incredible. That's clearly the best answer, full stop. But um this is probably a There's probably a bad idea here. There might be a good idea as a nugget or gem that I'm not sure what it is, but um um my thought was you could use something like TPU, which you can purchase in different durometers and you could have bamboo pause multiple times and you could lay down
00:31:03
John
electrical wire or some other steel spring wire in even in potentially little grooves that you are could control the slicer to have that maybe leave you a groove. And basically you could make your own wire.
00:31:16
John
Uh, what do you call it? Like a wire laced belt, like a belt with wiring in it.
00:31:21
johngrimsmo
okay yeah oh like uh like timing belts what you're saying
00:31:22
John
Makes sense.
00:31:25
John
timing belt that has wire in it to strengthen it over just what TPU would be.
00:31:27
johngrimsmo
Yeah.
00:31:28
John
And part of what also inspired me is like, Lawrence was just showing something that I don't remember enough, which is it's not hard to have your printer pause at certain layer heights and you can insert a stiffening rod of steel, you could insert magnets and then it just resumes and you now have this integral, again, strengthener, magnet, whatever.
00:31:39
johngrimsmo
Yep.
00:31:41
johngrimsmo
Yeah.
00:31:47
johngrimsmo
Yep. i've I've definitely heard and even used the magnet trick. And I was just recently thinking about the stiff rod idea. And i was like, that could be really cool to add like just a quarter inch rod of steel along the length of this part.
00:31:53
John
Yeah.
00:31:59
johngrimsmo
And like, yeah, that would add some juice to it.
00:32:01
John
yeah Yeah, 100%. Or like dowel pins, cheapest could be off McMaster.
00:32:02
johngrimsmo
Yeah.
00:32:05
John
You can have a whole smorgasbord of diameters and links and then just drop them in as needed.
00:32:09
johngrimsmo
Yeah. Especially at all the weak points, the sheer planes or whatever, like, yeah.
00:32:13
John
Yeah. My only other two updates but aside from Telescope is Raspberry Pi, Home Assistant, super easy, like really was. You can download Home Assistant.
00:32:26
John
You can download the Raspberry Pi OS from the Raspberry Pi website as the Home Assistant version. So like anybody can do this. There's nothing to it.
00:32:35
johngrimsmo
Okay. I
00:32:36
John
Boost up to Home Assistant, um logged in, bought an i but an IoT switch, a power switch on Amazon. What's weird to me that... what I wish somebody, I wanted somebody to kind of explain to me is like, when you buy these IoT switches, you still have to use that IoT switch company's app to like, yeah, exactly, to QR code it, to work it, but and then you get it working and then Home Assistant can see it and take over.
00:32:54
johngrimsmo
was wondering.
00:33:01
johngrimsmo
Okay.
00:33:02
John
but that's part of what I hate is like, I don't want all these sketchy domestic or foreign apps and things.
00:33:05
johngrimsmo
Yeah.
00:33:07
John
Um, you know, home assistant is right on that cusp. So many people have said good things about it and I'm going to keep using it, but it's just like, I don't want to be flashing firmware on cameras. I don't want to be getting super hacky with all these random so websites that I have to like create accounts for to activate things.
00:33:24
John
I kind of just want like stuff that works.
00:33:24
johngrimsmo
Yeah.
00:33:26
John
Um, but it's,
00:33:26
johngrimsmo
Mm-hmm.
00:33:29
John
Pretty awesome. Like there's a video I saw that's an inspiration right now. I'm going to see if I follow through with it. But Home Assistant, you know, under $100 webcams that are no longer tied to freaking monthly subscriptions and other shenanigans.
00:33:45
johngrimsmo
ye
00:33:46
John
Then there's an AI platform. AI person detect. And then there's an, also an LLM summary so that you can say once a week, it sends you a summary. That's like, Hey, between the hours of midnight and 4am, I saw a raccoon.
00:33:59
John
I saw a person and here are the timestamps.
00:34:01
johngrimsmo
no way oh cool
00:34:02
John
Oh yeah. And here are the timestamps. You can program it to ignore your dog, but not others maybe.
00:34:08
johngrimsmo
what and
00:34:09
John
And none of this, like, I don't think anyone's gonna call me out on this. Cause well, first off, there's a guy doing on YouTube, but like, it all makes sense if you start to think about it. Like, yeah, we can do this now, but um few aren't.
00:34:16
johngrimsmo
Yep. Mm-hmm.
00:34:20
John
So it might actually be a good solution. We've had some dumpster divers at our shop that I don't and don't really care about, but I wouldn't.
00:34:26
johngrimsmo
I was with you once when you pulled in and and just confronted them.
00:34:28
John
Oh really? Yeah.
00:34:32
John
Yes.
00:34:32
johngrimsmo
But it was
Raspberry Pi and Home Automation
00:34:33
johngrimsmo
good.
00:34:33
John
Yeah, it was funny. um i i get I don't lose sleep over of that stuff anymore on the flip side. I wouldn't mind having video footage saved with the dumpster divers in case whatever.
00:34:42
johngrimsmo
Sure. Yeah. In case, whatever. Yep.
00:34:45
John
Yeah.
00:34:46
johngrimsmo
boom
00:34:49
johngrimsmo
Yeah, we've been using a um Raspberry Pi program called MyEye. It's like M-Y-E, E-Y-E or something like that.
00:34:55
John
Yeah.
00:34:58
johngrimsmo
um
00:34:58
John
Remember you've seen that?
00:34:59
johngrimsmo
Super free. It just used the Raspberry Pi webcam, like the easy thing. And you flash your Raspberry Pi with this MyEye firmware, whatever it's called. And it does track, take so many video clips when there is motion.
00:35:09
John
Yes.
00:35:16
John
Yes.
00:35:17
johngrimsmo
um And we don't actually record anything. We currently only use it for a ah parts transfer camera. So the... the between the two buildings, when the machine shop finishes parts, they put them on this table, there's a webcam right above it, and the finishing shop has a TV screen constantly showing the live feed.
00:35:28
John
ah
00:35:35
johngrimsmo
So when somebody puts it you can see their hand go down, you can see like parts get placed.
00:35:39
John
Love it.
00:35:39
johngrimsmo
So there's parts ready for us. Okay, let's go get them, kind of thing. Because years ago, that was a point of contention, was like, I didn't know they were ready, or I didn't, who's supposed to blah, blah, blah. So this made it super easy. And it does record, but we never use it.
00:35:52
johngrimsmo
um
00:35:53
John
You never use a recording or you never use any of it at all.
00:35:55
johngrimsmo
never use any of the recording. We just, we just use the live feed.
00:35:57
John
Okay.
00:35:59
johngrimsmo
But the fact that it does recording, there are a couple of times, especially early on when it was fun and exciting, when I would look it up and like, yeah, it happened at like 8am.
00:35:59
John
Well, that's i mean. Okay, that's fine. Yeah.
00:36:07
John
Yeah.
00:36:08
johngrimsmo
You said it was 7am or something like,
00:36:11
John
The other thing I learned, which I am, i don't know, I guess I'm a little bit frustrated. I didn't couldn't figure this out earlier, was when we got the horizontal three years ago, maybe, ah PS, still for sale.
00:36:24
John
oh I think, knock wood, I think we've sold two of the four machines, which is great.
00:36:27
johngrimsmo
Nice. Congrats.
00:36:28
John
Yeah, but um I wanted a way to have a video feed of the pallet changer area so that you could see it from the control.
00:36:38
johngrimsmo
Yeah. And you you did, you had like the super basic basic offline.
00:36:44
John
We did it with a security camera, an HDMI cable, and a small video monitor, which actually was a great solution, probably 200 bucks total for all that.
00:36:53
johngrimsmo
Like no, no internet, no nothing, just hardwired.
00:36:56
John
Bingo.
00:36:56
johngrimsmo
Yeah.
00:36:57
John
But I still, it kind of bothered me that I couldn't find a better solution.
00:36:57
johngrimsmo
i
00:37:00
John
And so sure enough, when I was looking up some of this Raspberry Pi webcam stuff, because i was trying to figure out like how does home assistant interface, if it's not through a service like Ring or like Wise, which are these sort of very paywalled, siloed gardens, um,
00:37:17
John
And ends up, it's what's called RTSP, real-time streaming protocol, which is a thing that some, but not all cameras support. And it's exactly what I wanted, which means I could set up a um webcam somewhere else.
00:37:29
John
And then I could have a Raspberry Pi with a, so I guess technically this would probably cost more, but regardless, it's actually better more of what I wanted, especially because it doesn't need a cable connection between the camera and the monitor.
00:37:42
John
um And then you just have ah the Raspberry Pi show the RTSP video feed um which is awesome. Like that's what I wanted.
00:37:49
johngrimsmo
So not recording anything, that's not the point.
00:37:52
John
but Well, the Raspberry Pi could still record, but I just wanted to know, it's a security camera, but I would be using it for to see the state of the machine tombstones from 10 feet away where it's more convenient.
00:38:00
johngrimsmo
Yeah.
00:38:03
John
But I just always wanted a real-time video feed without necessarily having to run wires between them.
00:38:10
johngrimsmo
And then I was wondering for your security camera theory with the the home assistant, um what is the storage device? Is this cloud based? Is it on the SD card?
00:38:20
John
um I need to learn more. It appears that the cameras, so like I just bought one called Reolink, R-E-O-L-I-N-K. It's like 60 bucks. It has one or two SD slots that can hold up to 256 or 512.
00:38:33
John
um There was incon unclear which one was the max, which again, by the way, blows me away that you buy a 256 gigabyte SD card, even like a class nine.
00:38:42
johngrimsmo
micro SD.
00:38:43
John
Micro SD, thank you. is Yeah. um For... $21, $24 for the extreme grade from Sandus.
00:38:48
johngrimsmo
Jeez.
00:38:51
John
It just nuts.
00:38:52
johngrimsmo
Yeah.
00:38:52
John
But, um, and I don't know how much time that would hold, but it certainly would over, i mean, it could self overwrite after days, I would think.
00:39:00
johngrimsmo
Yeah.
00:39:01
John
Um, I would then assume that there's a way in the RPI to automatically scrape or download or offload the footage that you wanted to keep for purposes of logging. Like, Hey, save anytime you have, uh, save five minutes at any time you have action.
00:39:14
johngrimsmo
Yeah.
00:39:14
John
Um, And our Raspberry Pis are limited with home assistant on how many cameras they can run. And then you need to step up to a more beefy build, or then it gets confusing.
00:39:20
johngrimsmo
Okay.
00:39:23
John
a guy named Kurt, shout out you, Kurt. You've been awesome. He's been emailing me about this. There is, um, there is what's called, um, USB hardware accelerators, um, which can like help like literally like it's almost like a graphics card for a camera or RPi working offload.
00:39:42
johngrimsmo
What?
00:39:42
John
Yeah, it's like all this stuff I don't know, i'm learning.
00:39:44
johngrimsmo
Yeah.
00:39:45
John
Soaking up here.
00:39:48
John
The big part of this will be not only for personal use at home and for practical use here at the shop around like the traditional cameras in the shop stuff, lights and all that, but really having a separate home assistant that's going to be probably for each automation cell Because like I said, we want to record.
00:40:07
John
um It's very easy now with Home Assistant. If you have a tower light go red, it flags it and you can then look at the footage that occurred 10 minutes leading up to the red flag.
00:40:12
johngrimsmo
Hmm.
00:40:17
johngrimsmo
That would be helpful.
00:40:17
John
i I absolutely want that. want to walk in and know why why at 2.48 a.m.
00:40:20
johngrimsmo
Yep.
00:40:22
John
m my system failed.
00:40:23
johngrimsmo
Yep.
00:40:24
John
Without having to log in and scrub through footage and blah, blah, blah. Like, nope, feed it to me.
00:40:29
johngrimsmo
Yeah, give me the last 10 minutes before and the two minutes after kind of the incident.
00:40:34
John
Yes, yes.
00:40:35
johngrimsmo
Yeah, yeah. Smart, I like it
00:40:43
johngrimsmo
it. Okay, let's talk about um probing too much. So I'm a huge fan of using the probes for all of the things and measuring surfaces and comping tools and, you know,
00:40:56
johngrimsmo
trying to hit the perfect diameter is using the probe. Sometimes, especially on our lads, I've found it it falls into a cycle, a loop of like chasing its tail kind of thing. So it's better to just leave it a alone and let the machine do what it's going to do.
00:41:10
johngrimsmo
And you kind of watch the thermal and then comp accordingly. Specifically on the Speedio right now, I programmed this many months ago when we're machining the Fjall blades hard.
00:41:21
johngrimsmo
um We're machining the critical detent surfaces. I've got it touching the top of the blade because that's my reference i want to measure top down for the this feature i want to go exactly 11th out deep like 11 0.
00:41:30
John
yeah
00:41:35
johngrimsmo
so i want to measure the top of the blade because everyone could vary by tenths that's fine i'm gonna measure the top of the blade i machine the pocket 9th out deep and then i probe the top and the pocket the ninth out again and the logic says that should be ninth out it was actually nine one so comp the tool by one tenth and
00:41:51
John
Interesting. So the same tool then comes in to finish.
00:41:55
johngrimsmo
Exactly. And this happens within the span of 30 seconds.
00:41:57
John
Okay.
00:41:59
johngrimsmo
So there's not really thermal issues between the two operations. um Like the machine's not going to grow in 20 seconds. um not Some people would argue yes.
00:42:10
John
Yeah, no I hear you.
00:42:10
johngrimsmo
but And it worked fantastic. We were hitting less than one-tenth ah repeatability measured online and offline um of of pocket depth.
00:42:22
johngrimsmo
ah
00:42:22
John
Mm-hmm.
00:42:22
johngrimsmo
using this method. And it worked great. So I just forgot about it. And the guy's been running it. And then I go and check it today. And I looked at one of the parts and I'm like, this is measuring 13 thou. It's supposed to be 11 zero.
00:42:33
John
Whoa, whoa, whoa.
00:42:34
johngrimsmo
Where did that two thou come from? ah Hold on, hold on. Stop everything. That was five minutes before the podcast. And they're like, can we run more? And i was like, no, I don't know what's going on here.
00:42:43
John
yeah
00:42:46
johngrimsmo
So yeah, that was a stable process. And I don't know when it went unstable.
00:42:51
John
You're allowing it to comp the tool any amount though?
00:42:54
johngrimsmo
I think I put limits on it, thou. um
00:42:57
John
Okay, well then, but it should, i'm not I don't mean to sound like I'm challenging you.
00:43:01
johngrimsmo
No.
00:43:02
John
If you if you tell it to machine 9,000 and then it comes in and probes it again and it's actually more than 9,000 plus one, shouldn't it it would stop, right?
00:43:08
johngrimsmo
Yeah. I forget what happens. i don't I don't have an answer for you.
00:43:11
John
Okay. Okay.
00:43:13
johngrimsmo
Either it would ideally alarm out and be like, ah, that's weird, stop. um It could ignore it, which would be not great. um I don't know.
00:43:24
John
Are you using dual contact holders?
00:43:26
johngrimsmo
Yep. Pretty, pretty much everything.
00:43:27
John
Okay. Yeah. That's a big, it's, I've been front center for me on the, which i my understanding is not only are they more rigid because of the dual contact, but also they will have more repeatable Z.
00:43:30
johngrimsmo
Yeah, for sure.
00:43:39
johngrimsmo
Yep. Cause your taper is not sucking up into the cone as much. Um, yeah, I think the price difference from Mary tool, I forget what it is, but it's like small enough. Then you're like, I'll just do a dual taper everything.
00:43:51
John
um Well, keep me posted. I don't know that that's,
00:43:57
John
I am also a proud member of the Approbe Too Much Club, although I have been starting to ask about how to cancel my membership.
00:44:01
johngrimsmo
We
00:44:06
johngrimsmo
and need t-shirts to the probe too much club.
00:44:09
johngrimsmo
Hello.
00:44:09
John
Right. Yeah. The former Renishal, I had a whatever i funny analogy when you were talking a minute ago, I can't remember. but ah I obviously it's still a wonderful tool.
00:44:21
John
There's a lot to be said for designing better processes and not, and for sure, not getting yourself in feedback loops.
00:44:22
johngrimsmo
the
00:44:27
John
Grant grant was just going through this with the Willamette about when the warmup cycle, how aggressively can we start getting onto stuff with it? And, and it was kind of like, Hey, let's just run the warmup longer and see if that solves the problem before we try to go make lots of changes for the first 30 minutes of operations that are just the wrong kind of we the wrong kind of work and in copying that we wanna do.
00:44:48
johngrimsmo
Yeah.
00:44:50
John
And on a lathe, those darn things, you know, you've you've got, you're doubling your error because of radial cuts.
00:44:56
johngrimsmo
Yeah. Yeah.
00:44:57
John
The reason number 782, those things Yeah. Yeah.
00:45:00
johngrimsmo
They're amazing.
00:45:00
John
um
00:45:01
johngrimsmo
And what a lot of machines, once they warm up and you hold the heat, they're super consistent.
00:45:05
John
yeah
00:45:06
johngrimsmo
They're like super reliable, but it's that that's the swings, both the initial warmup and also you break for lunch, you open the door, you, you you mess with a part for 10 minutes and now the machine's colder.
00:45:12
John
yeah
00:45:15
johngrimsmo
and These are real things, especially on lathes. um Maybe it's part of it's because the enclosure is so much tighter and everything's like generally smaller than a big mill that has a lot more airflow and a a lot more metal. I don't know.
00:45:30
johngrimsmo
Like I know with our, with our Nakamura, when we, but we were battling thermal growth issues, it was growing by three thou. And then we installed a mist collector, which sucks in cold air through the chip conveyor and exhaust out the tops. You're constantly sucking cold air into the machine that cut our thermal growth by more than double.
00:45:48
johngrimsmo
Like, yeah, it's huge.
00:45:49
John
It's awesome. there In a good way, you mean?
00:45:51
johngrimsmo
Yeah.
00:45:52
John
Yeah, yeah, right, right.
00:45:53
johngrimsmo
Yeah.
00:45:55
John
somewhat related, but also not, we, I've been starting to work on little projects that I've wanted to learn myself and have the team help with. And so, um,
Advanced Probing Strategies
00:46:03
John
I've been kind of coming up with these little scenarios and projects. And one of the ones that, um, Caleb just ran and it's awesome is, um, and it sounds simple.
00:46:12
John
Um, but it's ah a little, it was a little bit harder to implement than I thought. So Let's say you put a, well, actually your example here works quite well. Let's say you put a Fjell blade that's been rough machined. And then let's say it's been heat treated and you bring it, or press treated, you put it back in the machine.
00:46:38
John
You don't know exactly where that blade now exists in space and you don't care. Like plus or minus two, three, fourth out, doesn't matter. You're machining everything away. So it doesn't matter.
00:46:47
johngrimsmo
i
00:46:49
John
But, What you wanna do is you wanna pick one datum or one point. Let's say that there is a ah point on it that isn't um going to be machines.
00:46:56
johngrimsmo
yep
00:46:57
John
That part point will remain there at the beginning and end of the operation. I wanna probe that point and I wanna log the y dimension or maybe x and y of it.
00:47:05
johngrimsmo
Yep.
00:47:06
John
I'm gonna do all my work and then I wanna come back and I wanna touch that point off again to see if it moved.
00:47:12
johngrimsmo
Mm-hmm. Mm-hmm.
00:47:12
John
This has nothing to do with my offset. This has nothing to do with comping the part. I just want to know if the point the point was the same at the beginning and the end. and we And we did it. We got it working.
00:47:22
johngrimsmo
that actually solves one of my problems that we were just dealing with two days ago um exactly because we have we do exactly that we do that every day for various other blades we have the pivot hold of the blade that's uh sometimes machine only soft sometimes re-machined hard afterwards depends on the blade um but yeah we we fixture it as accurately as we can and then we probe that hole and that is our holy grail of coordinate we do um
00:47:34
John
Mm-hmm.
00:47:47
John
You update the corn system though, there for that. Yeah.
00:47:51
johngrimsmo
And then I want to know if the blade moved because we are getting bad parts. And I'm like, well, the process that I created it is perfect, but it's not obviously. um But if we're probing the hole, why is it off by so much?
00:48:02
johngrimsmo
Either the probe's off and there's incompatibility there, inconsistency, or the part is physically moving due to the cutting forces. And I don't have an answer yet, but... um
00:48:10
John
o
00:48:11
johngrimsmo
Yeah, because I probe it um before machining, I could very easily probe it again after and write those results to a separate file and like compare them.
00:48:12
John
Yeah.
00:48:22
johngrimsmo
Yeah. and like
00:48:23
John
Yeah, Fusion. So if you have a coordinate system, this is easier for sure. It was trickier for us because we didn't actually want to mess with our WCS.
00:48:31
johngrimsmo
yeah
00:48:31
John
um But Fusion has a great, it's part of the extension, but you can just sort of say, hey, after every operation, I'm going to have a inspection operation that's going to come in and check. And if it's out by more than a value I set, pause the program.
00:48:45
johngrimsmo
Oh, okay.
00:48:46
John
So when you're doing something where you're like, i want to know where it's moving, you're just going to duplicate that operation after every one and then sit back, let her go to work.
00:48:51
johngrimsmo
i Oh.
00:48:53
John
And then it should tell you when it sees the part move.
00:48:56
johngrimsmo
it in In real life, like when you're machining, it'll stop.
00:49:01
John
So it'll it'll it'll drill a hole and then it's going to probe.
00:49:01
johngrimsmo
Yeah.
00:49:03
John
And then it's going to mill a pocket and then it's going to probe.
00:49:04
johngrimsmo
Yeah.
00:49:05
John
and it's going to face it and then it's going probe. Then it's going to chamfer, it's going to probe. And if any of those probings see that the part moves, it will' so it'll stop.
00:49:11
johngrimsmo
Yep. And if it gets to the end all fine, then you're good.
00:49:14
John
And your part shouldn't have moved.
00:49:15
johngrimsmo
within the limits of what you set, plus or minus 1 thou whatever you told it.
00:49:19
John
Yes.
00:49:19
johngrimsmo
Yeah.
00:49:20
John
But what was pretty cool, because Caleb set the limit to, I think, $5,000, that may be the default. And he ran it first off, the first one, which of course is just establishing it. Then he put a piece of masking tape or painter's tape over it and checked it and it continued. And then he put a second piece of tape over it and it failed.
00:49:37
johngrimsmo
okay yeah
00:49:38
John
Pretty cool.
00:49:39
johngrimsmo
OK. Yeah. I like that quick and dirty proof.
00:49:41
John
Yeah.
00:49:41
johngrimsmo
Huh. hu
00:49:43
John
like I'm going to do a video on that. It's certainly thing we're going to keep like secret. um So if anyone's if anyone's interested, they can reach out to me directly. I'm happy to share it now. Otherwise we'll get a little YouTube video of that up.
00:49:53
johngrimsmo
Yeah, I think that'd be really good.
00:49:56
John
Okay. Can I share my Hubble
Astronomy Interest and Hubble Telescope
00:49:57
John
thing or telescope thing real quick?
00:49:57
johngrimsmo
Yeah, please, please.
00:49:58
John
Okay. Okay. Um, well, short answer is my son and I are sort of getting into astronomy. We'll see if this goes anywhere, but like being able to pick up a hand-me-down 20 year old telescope in our backyard. And we saw the rings of Saturn on our first night is pretty freaking cool.
00:50:12
John
Like I'm sure anyone who's listening, who's done astronomy will understand the sense of joy and awesomeness. Um, and if anyone's not like, I would encourage you, like I'm already, we went to one astronomy club event.
00:50:23
John
I'm sensing that these are going to be the kind of people that I like nerding out with.
00:50:23
johngrimsmo
Nice.
00:50:27
johngrimsmo
Yeah.
00:50:27
John
Um, And then that got me going down the rabbit hole of reading about the most infamous event in telescope his news in my childhood. i don't know if it was as big a deal in Canada as it was in the U.S., but the failure of Hubble.
00:50:40
John
They built this telescope for you know hundreds of millions, maybe billions of dollars, took 10 years. They launched up in space. They take the lens off, s snap a picture, and it's bad.
00:50:50
johngrimsmo
Yo.
00:50:50
John
And ever I remember this well as a childhood and it was ground to the wrong tolerance. It was ground wrong. And, um, now that we're machinists and and there's lots of things to hear when it it comes to like beam splitters and lights and metrology and, um, you know, but monochromatic light sources that are all in the same overlapping realm of telescopes and and metrology, which is super cool.
00:51:11
John
Um, but the two things I've already learned, number one, the, Manufacturing flaw with Hubble was because the um setup that they were using to drive their grinding had three or four lenses, and one of those lenses was installed backwards.
00:51:25
johngrimsmo
ah
00:51:26
John
And so that's why they were thinking that they were doing it correctly. And that's why I'm reason why they could fix tell hubble because they knew exactly how incorrectly they ground it.
00:51:33
johngrimsmo
Right. okay
00:51:34
John
um Kodak was actually hired to make the backup lens. The backup lens was done correctly. There's lots of politic political BS about how they wanted to check each other's work. And there were egos involved that stopped that, that would have discovered this. And you know, our,
00:51:48
John
this came up at our astronomy event. And even the local astronomy guy was like, I've ground lenses. Like I have equipment that could have told me that Hubble's was ground wrong. Um, which you kind of have to laugh about, but here's the, here's the punchline. I remember hearing this and thinking that like they ground it wrong, like in an embarrassing manner, it was off by 0.002 millimeters.
00:52:11
johngrimsmo
Two microns.
00:52:11
John
That is, uh, less 20 minutes.
00:52:15
johngrimsmo
Zero zero zero two is two micron for sure.
00:52:19
John
It was about eight tenths of a tenth.
00:52:21
johngrimsmo
Yeah.
00:52:23
John
Okay. I mean, that's not a big amount. It's not like, don't know.
00:52:25
johngrimsmo
No, no, super tiny.
00:52:27
John
Maybe in the optical world, that's a lot. But I don't know Growing up, I got the impression that they like were off by like an eighth of it something.
00:52:29
johngrimsmo
Yeah.
00:52:33
johngrimsmo
Right.
00:52:34
John
bad um
00:52:34
johngrimsmo
Yeah. Have you seen, um, yeah.
00:52:36
John
I don't know. And this thing's huge. It's like, don't know. ah Is it eight feet wide? Yeah.
00:52:42
johngrimsmo
Yeah, two microns over that distance is is wild, but important, clearly.
00:52:45
John
Yeah.
00:52:47
johngrimsmo
um Did you see Huguen's Optics latest video? on the
00:52:52
John
Don't even know it.
00:52:53
johngrimsmo
So he's he's like the optics guy who comes up in the interferometry searches.
00:52:58
John
Okay.
00:52:58
johngrimsmo
um Definitely watch this latest video because it's all about telescope mirrors and it's all about making a tunable telescope lens with Huygens, H-U-G-U-Y-N's, whatever, Huygens optics.
00:53:07
John
Interesting. What's the name this guy? Sorry.
00:53:11
John
Yeah, I'll fire it.
00:53:15
johngrimsmo
And so he makes a six inch telescope mirror or whatever with maybe 20 actuators that can flex it
00:53:22
John
Yeah.
00:53:23
johngrimsmo
And he he explains how this is used because when we're on ground and we look up in the sky, there's atmospheric turbulence. And this video explains beautifully.
00:53:29
John
sir
00:53:31
johngrimsmo
and I'm not even that interested in telescopes at this time, but I watched it and i was like, so that's really cool. And they can compensate for the atmospheric turbulence within our atmosphere to get a crisper image.
00:53:43
johngrimsmo
um It's a beautiful video. I suggest every every nerd out there who's even vaguely interested in this, definitely watch it.
00:53:47
John
yeah
00:53:49
John
Okay. I am hooked. In a weird way, like there's part of me that's like, I need a new hobby. Like i need a hole in the head.
00:53:56
johngrimsmo
Yeah.
00:53:56
John
On the flip side, I'm like, it's so cool.
00:53:57
johngrimsmo
right
00:53:58
John
And you don't have to spend money on it. Like we got a hand-me-down telescope.
00:54:01
johngrimsmo
Yeah.
00:54:02
John
It's a great thing to do with my son. Like I'm loving all that part of it.
00:54:05
johngrimsmo
yeah There's a a DIY mostly three d printed telescope project that Lainey Machine Tech reposted.
00:54:10
John
Oh, you're killing
00:54:13
johngrimsmo
ah forget what it's called. It's got a funny name, but it's pretty cool. But if you already have one, it's probably the same thing.
00:54:18
John
Well, that was the next thing I went on. I was like, and I wasn't going to reach out to Adam, certainly not yet, but like, I know that telescopes are part of his curriculum. And then I'm like, wait minute here.
00:54:25
johngrimsmo
Yeah.
00:54:27
John
This is the itch I needed to scratch. Like I have, we have these machine tools. We have grinders here. Like, can we, can we grind glass? Can I use a UR robot? Cause they were the guys that on our local astronomy club, I know we're Phil's 10 minutes into circling the parking lot right now, but, ah The guys that grind their own glass talk about how they'll spend, you know, hundreds of hours just repeating a motion.
00:54:50
John
um And I'm like, we got it. We got I'm currently on track to own three URs in the next four weeks, um one of which will be the kind of extra lab robot to test with.
00:54:57
johngrimsmo
Nice.
00:55:01
John
So it's like I set up UR robot to sand or grind to move a lapping thing around.
00:55:05
johngrimsmo
Yeah.
00:55:07
John
That's easy.
00:55:07
johngrimsmo
Yeah.
00:55:07
John
That's kind of fun. and
00:55:08
johngrimsmo
Between Adam and extra minty Matt. Um, do you follow him?
00:55:13
John
I'm writing it down.
00:55:13
johngrimsmo
Oh, extra minty Matt is the optician of all opticians. um He's a super nerd. He's friends with Adam. I've been following him for a long time on lapping.
00:55:19
John
Okay.
00:55:21
johngrimsmo
He typically laps glass, but I've learned a lot about lapping in general and optics from him.
00:55:24
John
Okay. Yeah.
00:55:25
johngrimsmo
Just a great Instagram account to follow. He's really zany and funny and cool.
00:55:27
John
Okay. Awesome. Awesome.
00:55:33
John
Apologies if I hugged the conversation there.
00:55:35
johngrimsmo
No, that's great. Great.
00:55:39
johngrimsmo
Yeah, Leif and I looked into making that DIY telescope and just moved on. But um it would be really cool.
00:55:43
John
you
00:55:45
johngrimsmo
Because he like he loves space.
00:55:45
John
Yeah. um
00:55:47
johngrimsmo
He's like super interested space. Yeah. All
00:55:50
John
I'm literally going to Google when we hang up.
00:55:52
johngrimsmo
man.
00:55:54
John
i
00:55:54
johngrimsmo
That's about all I got.
00:55:56
John
Okay. I'll see you next week.
00:55:57
johngrimsmo
Sounds good. Bye.
00:55:58
John
Take care.