Hail to the signal switcher! On this episode, Parker wraps up his prep work for the Extra-Life Charity stream and Stephen discusses switching signals.
Why is estimating a projects completion time feel like it takes more work then the actual project? Estimating Project Time, the quest of management.
Our Spider-sense is tingling... OH that is actually a man-made zombie spider crawling up my leg. Who thought that would be a good idea? WHO!?!
Parker is an Electrical Engineer with backgrounds in Embedded System Design and Digital Signal Processing. He got his start in 2005 by hacking Nintendo consoles into portable gaming units. The following year he designed and produced an Atari 2600 video mod to allow the Atari to display a crisp, RF fuzz free picture on newer TVs. Over a thousand Atari video mods where produced by Parker from 2006 to 2011 and the mod is still made by other enthusiasts in the Atari community.
In 2006, Parker enrolled at The University of Texas at Austin as a Petroleum Engineer. After realizing electronics was his passion he switched majors in 2007 to Electrical and Computer Engineering. Following his previous background in making the Atari 2600 video mod, Parker decided to take more board layout classes and circuit design classes. Other areas of study include robotics, microcontroller theory and design, FPGA development with VHDL and Verilog, and image and signal processing with DSPs. In 2010, Parker won a Ti sponsored Launchpad programming and design contest that was held by the IEEE CS chapter at the University. Parker graduated with a BS in Electrical and Computer Engineering in the Spring of 2012.
In the Summer of 2012, Parker was hired on as an Electrical Engineer at Dynamic Perception to design and prototype new electronic products. Here, Parker learned about full product development cycles and honed his board layout skills. Seeing the difficulties in managing operations and FCC/CE compliance testing, Parker thought there had to be a better way for small electronic companies to get their product out in customer's hands.
Parker also runs the blog, longhornengineer.com, where he posts his personal projects, technical guides, and appnotes about board layout design and components.
Stephen Kraig began his electronics career by building musical oriented circuits in 2003. Stephen is an avid guitar player and, in his down time, manufactures audio electronics including guitar amplifiers, pedals, and pro audio gear. Stephen graduated with a BS in Electrical Engineering from Texas A&M University.
Special thanks to whixr over at Tymkrs for the intro and outro!
Hey, all you macro fibers out there you want some pizza? The answer is yes. And who are we kidding? We all like pizza stop by the Microsoft website and that is Mac fab.com on Pi Day that is March 14 or 314 Pi Day you can take our pop quiz. There should be like a little banner that pops up see pop quiz for a chance for some pizza pie. That is macro.com on Pi Day. Welcome to the macro engineering podcast a weekly show about all things engineering, DIY projects, manufacturing, industry news and tool testing. Where your host electrical engineers Parker, Dolman, and Steven Greg, this is episode 319. So before we get started Feedspot I don't know if that's a reputable website or whatever, but has ranked this podcast Yes, this one. Number one for Best 10 Electrical Engineering podcasts. I want to thank everyone for listening and sharing this podcast with others. We are just getting started. I think
the next goal is number one for best engineering podcast.
Who's number one on that? You're already toppling next?
slowly climb the ladder one rung at a time. Yeah, one
listener at a time. Yeah, thanks so much everyone for supporting this podcast and making it as awesome as it is.
It's exciting. Cheers, everyone. Cheers, everyone.
And then one more thing is I don't think it's live yet as the live stream. But should we live? By the time this podcast episode comes out? Is Mac fabs automatic uploader? And our platform supports key CAD six now. That's good.
You know, Okay, question, just because it's been a while since I've used the platform with key CAD. I did years ago. But what all does that mean? You have key CAD six support, does that mean you can drop your key CAD files directly on
your KiCad PCB file, you just drop that right into the platform, and it uploads and does the bill of materials and everything. That's rad.
That's super nice. And KiCad? Six is the most recent,
the most recent one. Yes. Very cool.
So as of right now, in terms of the Mac fan platform, KitKat and Eagle both have those capabilities, where you just drop in your files, and it just scrubs and extracts everything, right? Yeah.
And then everything else is either ODB plus plus raw Gerber's. And then dip Trace has their ASCII file. Yeah. Which works. I don't know if it's updated or not. But used to work probably still does. If it's not version of whatever the case it was.
You know, what I've noticed with dip trace is the ASCII stuff, never changes never changes. That's always the same. So I bet you know, it still works. And you can do ODB from dip trace. Yes, you can do all the all the, I guess more labor intensive ways, but the guaranteed ways. Yeah,
Odv is my favorite way of uploading stuff to the platform, I should say. Because it it definitely works the best I would say Yeah, and it's a more complete package out of everything else. So highly recommended. I can't wait to like Eagle maybe actually outport that. Maybe someday.
Sounds like a ULP thing.
One Crazy you'll be it'd be a monster you Oh. All right, so our first topic today is testing ratchet wrenches. Now this is electrical engineering podcast, as evidenced by our number one ranking. And this is not really electrical engineering. At first glance.
Yeah, there's some electrical here. Yeah.
So how long it was like two months ago, I think like two months ago, I broke i ratchet right so ratchet wrenches like think of like a socket, not a socket wrench, but like a regular crescent wrench. But it has basically the end of it ratchets his around in a circle so you can quickly loosen or tighten a fastener in a low profile manner got a socket wrench has like you know, the socket plus like the ratchet mechanism that's thick and I broke it had a Harbor Freight Pittsburgh Pro I think it's what they brand branded. And I've actually broken a couple at that point but this one I broke a half inch which is the one I use like the most. It's a
fairly beefy one.
Yeah half inch ratchet wrench racket I Basie the Paul stopped working in it just like sprung apart that.
So what did it just like Free Rotate after they just freely
rotates and like parts fall off of it. So I was like, Okay, I'll probably how I view it is if I break a tool or something. It's not the point where I need to just go buy the better version of that tool, right? Because those are wrenches I bought when I was like in college, so they lasted 1012 years at this point. So I wasn't like sad that they broke because I got my use out of them. Right? Yeah, it's just one of those time to finally actually buy something nicer. And so I started looking around online, I'm like, there's a lot of choices out there. Oh, what different brands, brands of the other brands have like different kinds, and different specifications and that kind of stuff. So I'm like, You know what I want to buy like one of everything and figure out what set I want to buy because I was gonna buy a whole new sets of these ratchet wrenches. And so I first came up with qualifications, right, because I'm like, okay, it has to have these features for it to be in this tech contestant. Yeah, contestants. So it has to be a ratcheting wrench. self explanatory, but it has to be in there. Right? Your your spec sheets gotta have the first qualifier is, it's got to be ratchet wrench nonreversible style, which some? Excuse me, some ratchet wrenches have a little clicker on the back. So you can reverse the direction.
Right, this is this is flip it over to flip direction. Yeah, this
is you flip the ratchet o or the wrench over to reverse it. And the reason is, is because for me the entire point of a ratchet wrench is its low profile illness, when those little clickers that the reverse direction actually take up extra thickness.
You know, I can't stand it when when you get a ratchet wrench, and it has an angled head on it. Because it like defeats the entire purpose. Yes, so zero degree
offset. Yep, and the head. So it's a flat,
just a straight bar.
Yeah. And then where you're going to have to come in half inch hex. Because that's where I'm testing everything. That was the one I broke that made me start down this path. And so I'm like has to be half in checks because there's a lot of brands out there found they only make it metric. They don't make any SAE or imperial specifications.
Basically what you're saying is the half inch represents the entire set,
yet entire set, because this thing is we'll get into a little bit later. But when you go online and look at people testing wrenches, their test like let's say AV E, very popular YouTube channel. He tests his like big wrenches and stuff, or like ratchet wrench big ratchet wrenches and like big guy because he's into like mining equipment. Yeah, mining in an oil and gas and stuff like that. Whereas, like most mechanics, don't use stuff that big. And so like half inch hex is like, kind of what I use a lot of or like 10 millimeter or, or in that size range. So half inch hex, and then it could be 12 or six point drive. I don't use a lot 12 Point stuff. So it can be 12 or six point.
Yeah, it did you notice that? Do some brands have a choice to over six? Or do you just kind of get 12 or six
some brands have a choice of 12 and six. Some of most of them are you just get what you get. Most of them are 12 is very few that are just six only. So so far, I've collected 20 ratchet wrenches that fit this mold.
You know, real quick, one thing I noticed that's not in your qualifications is how many Poles the Ratchet has?
Correct? Because that's part of the tests. Okay. Yeah. So I collected 20 Different wrenches and they range from prices $45 for a wrench to two for $8
So 42 for $8 ollars That's, you know, I like I like in how you started right there with the highest number first because I was like, Oh $45 You know, the most expensive one is going to be $200? No, two for eight to $8. Nice. Okay, so you bought a pack of two? Are you gonna test both of those? Because that would be funny. They have very different
Yeah, that's the only one I actually have dual. Twin of. So yeah, we should see if they match. So typically, I'm 21 ratchet wrenches.
Okay, yeah, you gotta have a sample size here.
I am going to, I'm going to make a blog on my website that's like these all the ones I'll have with the model numbers now, because the some of them are like GearWrench. They make a 72 and a 92. Tooth, Paul, and so I have both of them. Okay, okay. So like, when I found a brand that would fit the specifications, I got everything, they would fit those specifications from that brand. Like, I think I have like two matte COEs. And they're like, almost identical. I think, is it MakerBot? Or is it or one of them is like, the only difference is the color. But I did get I did get it because it met the specification. You know, it'd be really funny if there's like a significant difference between colors. It's chrome or black? Well, honestly, the block wrench looks really cool.
Yeah. Is it anodized or something? It shouldn't be anodising.
I don't know yet. Because I haven't actually really opened up the packages, because one of the things I'm testing is also like the packaging that came in
just as you're going all the way here.
Anyways, I got 20 of them now cool. Or 21 If we caught the twin, so I'm going to make a list. Hopefully before this podcast comes out, so people can go look at the list. And then maybe suggest ones I've missed that hit those qualifications.
Man, we all need to like you need to release the list. And then we place bets on the winner who's going to be the best. That'll be fun.
So what are we the things I want to test the thickness of the wrench so like how, how low profile is the wrench cuz that's, that's the most important thing with these ratcheting wrenches for me is how, like, you know, thing of a spot does it fit. Because even though I work on Jeeps, there's sometimes you got to get a wrench in that you can't get a wrench in like a normal wrench. Like sometimes you have to use a clawfoot style or whatever. The worst is like clawfoot with like a weeble wobble extension, and like three O's that game and you're like, yeah, those a lot of fun. But the whole point is compact wrenches. So another thing is, how does the tool feel to use? So I'm going to actually like use, like, tried turning some bolts with them? Probably not on a car, I'll probably just tighten up some bolts on like, the device and like to see like, how well does the mechanism move? Like, is it smooth?
Do all the teeth feel the same?
Yeah, there's Oh, there's Yeah, exactly. Or like, one of the big things is, some brands are wrenches have like sharp edges from the stamping. And it's not like sharp, where to cut you. But it's like they they press on us, the surface area that hits your hand is smaller. And so it will push into your hand on a sharper edge. Whereas like I have some really old craftsmen that are that have like a more rounded edge autumn. And so they don't, they don't hurt your hand when you really push into them. So that's that's part of that as well. Like, how, when you really have to reef on a on a on a bolt? How much does that hurt your hand? Right? Yeah. And then the other thing is max torque to either failure or cam outs. That's the big test because it's basically when will these fail another thing in there, I guess we need to put in before is like how many degrees of arc the more teeth it has, the less degrees you need to swing it before it clicks over. Right? Yeah. catches.
And that can be can that's really convenient when you have
that's very tiny areas. Very tight areas. Yeah. So any other tests that people can think of? We can add in here. I don't really want to do like I've been watching a lot of like, people testing wrenches and stuff on you to to get ideas of tests. And I think just doing a max torque to either failure or cam out either or basically, is going to be that's what I'm looking for, like, how much can you force? Can this wrench take until it just fails?
You know, one number that's coming to mind that I think could be fun is I'll get so this is a number that I guess would create would would dictate its usefulness not necessarily just based on how much force can apply, because a max max torque doesn't necessarily mean, like, if something has less max torque, in terms of like brakes sooner, that doesn't necessarily mean it's bad. That just means that it failed sooner. Because like, most of the time, you're not going to take these things to that No, right? So that's how about this ratio of its thickness to its max torque. And that gives you that gives you like, maybe something that is thin, which is really useful, but breaks earlier, you can you can create, like a, like a gray ratio based on that, you know,
a thickness, basically a thickness to torque ratio, and see if that actually even like, plots out correctly.
Yeah, see if they see if that shows anything interesting, because it might be like, Yeah, this thing breaks earlier, but it's thin, so it gets a better rating.
Yeah, I'm writing that down.
Yeah, that data might just show garbage. Who knows. Yeah.
It's probably actually just gonna look at a scatterplot.
I mean, yeah, yeah. Yeah. But you know, depending on your needs, it might tell you something better.
Yeah. Yeah. Or, and also the method of failure. You know, it could be it fails by cam out versus, you know, breaking the Paul basically, because I don't think we're going to be breaking like the housing around the mechanism, the ratcheting mechanism, we're probably just going to either break the Paul teeth out, or it's going to come out and so I've actually got so to test all this stuff. The first thing I needed to figure out is like, how much pork am I going to need to apply to these actually break them? Yeah, you need a system capable of at least that much. Yeah, so I got a another Harbor Freight one on my door that actually still worked around the same I think you got like the 13 millimeter, which is the metric version of half inch, basically, half inch plus half inch plus it's, it's we're half inch bolts that are little rusted. And I put it in my vise. So I get I stabilize it. And then I took well, before this, I actually took a half inch hex head, grade eight bolts, which is like five sixteenths 18. So five sixteenths, three major diameter 18 thread, threads per inch. And I toured, I tightened it up. And then I put the wrench on it and then try to tighten it more and see that I can get the wrench to fail. I was able to by hand, I was putting almost all my body weight into it. But I was able to actually share the grade eight bolts. And then I looked it up. And those bolts are supposed to be torqued to about 30 foot pounds or things like 28 foot pounds is like before you start to go into yield. So I was like, a me, I probably put those bolt immediately into yield.
I was about six, because let's just say that that wrenches, let's just pretend it's a foot long. And you put one Parker foot on it. Yeah. And 230 foot pounds. Yeah, like, Oops. Yeah.
So I immediately sheared the bolt. And I'm like, okay, so we got to figure out a different way to actually load these. And this is the thing. I think what happens is, over time the Paul wears, and then it takes less force to break it. I think that's what happens. It's gonna be really hard to test. But I think just going for a max is going to be good enough. I
would say, well, and everything's new, which you're going off of it
and everything's new. I can't I can't wear it. 10 years by using it, right.
I guess you could just hook it up to a motor and run it on stuff.
Yeah, I guess you could, but we're not doing that.
That would be awful.
And so then I'm like, okay, I can't make this fail by a bolt that it's supposed to tighten technically, right? And so I'm like, Okay, how much forced how much pork do I need to apply to this to make it break. So I, that's when I put it in my vise. And I took a half inch allen sockets basically, and put it into my torque wrench, I put it first I put it in my three eighths torque wrench, which maxes out to 100. So I put the clicker to 100, put it in, and I got the clip before the wrench failed. And I'm like, okay, it clearly is more than 100. And then I got my half inch torque wrench, which goes up to 250 foot pounds. And I got the shear. I got the the I basically got it to basically rip the Paul out of the teeth. And so how it failed, though was interesting is it actually tilted the whole mechanism out. Which disengage the Pol, basically, because now it doesn't have that so and actually now that wrench is like frozen, it doesn't like you can't even turn it anymore. And so that's actually an interesting thing is one thing we should put in the test is how much degrees of deflection does the mechanism have?
So what do you mean by
so when when you have a, let's say you're able to like? So the mechanism that rotates inside the housing can wiggle until Twain. Right? Well, no, it can wiggle on the same plane as the wrench itself. Oh, I see what you're getting at. Yeah, yeah. And so that's how my other one failed is like because you're turning it and you're never going to be turning it perfectly on access. Right, right. Right. And then you turn it on slight axis and you put a little bit too much force and kind of just pops out. So just maybe being able to measure how much plays in there. Maybe the more expensive ratchets have less play maybe ideas,
huh? Yeah. Okay. So if you have it, if you have it engaged with the head of whatever bolt you're talking about, and then a way to measure its angle of deflection. past that. That would be that would be a nice thing to know. Because I guess the thing is, like, is having a lower angle, necessarily a better thing? I'm not sure.
I don't know, either. But it'd be fun to actually measure that and see if it results in less racing brake torque.
Yeah, yeah. Like it would be cool to correlate, like, do the ones that have higher angles fail sooner?
Yeah, fail sooner. So how are we going to be able to apply between 100 to 250 foot pounds to these wrenches, so clearly, to be able to, so we have to first apply the force like to it because using a five sixteenths 18 bolts with a half inch head, everything great eight doesn't work at all, like because you just share it. So I started looking at basically like, how to actually put that much force into it. And it clearly though a three eighths inch drive, a half inch allen socket, is able to do no problem. So you need to buy those off the shelf, which is great. Now we have a part that we can apply the torque to the wrench that's replaceable, right. And so I was actually thinking about us is so now though is we have to be able to couple that force to something else. Well, it's got on the back of it's got a square drive. And so I was like, okay, what can I buy that square drive, right? You can get square drive. Bolts, so I'm like, Okay, well get a bolt that's a bigger thread that can handle the torque. And basically, we have to go over like, has to be bigger than half inch with 13 to like half inch 13 has to be bigger than that. Like that's half inch 13 is like, like 200 some odd foot pounds, I think is like 230 When I looked it up, which is probably barely enough. I like to go bigger. So these I don't have to worry about the the thing. I'm measuring it going into yield right.
Oh defeats the whole purpose.
Yeah defeats the purpose. So I'm like okay, there's two different paths we can take with this. We can take a three quarter inch bolt and machine the head to hex the right size, which wouldn't be that hard to do. Okay. Or, we take the half inch allen socket onto a three core Motor Drive, and then put that into a three quarter drive square bolts. I think that's the way I'm going to go because it's like parts, I don't have the custom machine I can just like get them from McMaster. Now the half inch socket, the three quarter inch drive is like, it's like a $50 socket, though.
Well, hang on it perhaps is there a simpler way of doing? Well, I guess what we haven't discussed yet is, is the force being applied to the ratchet, or is the force being applied rotationally to some
force, well, I'm going to be moving the ratchet, the ratchet wrench and we move in the wrench. And that's going to be applying a force on the bolt or the screw mechanism, basically, because basically, what I'm going to what I'm going to end up measuring is the clamping force, well, then I'm going to have two plates with basically load cells in between them, and then a screw in between them that clamp that all together. So the torque you apply results into a clamping force between these two plates. And that's what we're going to measure.
Gotcha, gotcha you could you not? Could you not just buy some sort of hex and then just weld it to a plate? And then just crank on that.
So how do you so you'd have to measure the force that you're applying to the wrench at that point, right. And the problem with doing that is it depends on the angle. Yeah, yeah. Whereas if we measure the clamping force, we take out basically the the force vector.
Yeah, I see what you're getting at. Yeah, yeah.
And I thought about doing like a hydraulic puck. And then measuring, like the pressure off that using, like a pressure transducer, might still do that by kind of like the load cell idea, too. Because I can use like for load cells, and then they're already like, threaded and bossed correctly, and I could just like, bolt them through like a plate, right? And just like everything will line up. Yeah, it seems like and then I know, we're welding something
is an easier way to guarantee that you get the torque, but it's harder to measure the torque.
Yes, yeah. And then we'll just use like, a really well, I was going to use as whatever the screw drive is, is I'm going to use Teflon and bases. So that No, nothing's lost into the screw drive or minimal loss, and basically, clean it between each test and then reapply it and put it back in, which should get any various variations out of the tests are most of them. But yeah, the big thing is, I think I'm gonna go with the half inch allen socket, with a three quarter inch square drive on the back. And then that goes into a, a three quarter inch square drive bolt, which is going to be like three quarter 11 I think is what it is threads. Something like that. Beefy, and then that's going to be I think that's like something like 400 or 500 foot pounds or something like that something stupid number, which is gonna be plenty to and then on basically on the backplate screen is where I would weld a nut to the back part of the contraption. Fixed Weezer squeezer. I don't, I don't, I'm not going to suspect any, like super crazy, catastrophic fails, like the head exploding or whatever. But
if it does, it'll probably explode and move like an eighth of an inch, right? It's going to like explode all over and throw something.
Yeah, it wouldn't be nice to automate the part we're applying the force, it'd be nice to automate it. I don't know how yet. I want to do that. It would be nice to do that though. Like because it needs to be able to apply 200 foot pounds to that wrench. So 400 pounds, because that wrench is roughly six inches long. So 400 pounds to the end of it. I don't know exactly how to do that. Because I don't want it to be a lot. But some of the problems I find is watching a lot of the torque tests is online is they use like a pneumatic powered hydraulic system. A pusher. Yeah, and pushes it or pulls it or whatever, which is fine, because it works great. Because you can just hold a pneumatic down and then it goes click click click as as the hydraulic pressure goes up. The problem is the clicks. It's kind of a discrete addition of hydraulic pressure. And I'd like to be more of a smoother curve which is like One thing is, like, putting like a big lever on it and just pulling on it. It's actually kind of smooth compared to that. But I don't know a way to, like, apply that kind of torque evenly. It doesn't have to be super repeatable. I just want to be smooth application of it.
Right? Right. So I guess what's what's coming to mind is some kind of contraption that translates rotational into like a pressure. So like, a screw that that has something that pushes on it, and you just rotate the screw. And it continually applies pressure to the end of that.
Yeah, I wonder if you could have a basically a, a, almost like a linear actuator, because there's a screw driven to
something with like a worm gear that has a crazy ratio and just crazy pushing. Yeah, it's like, because you don't need to go fast with this. You just want it to be smooth. Yeah,
I want to be smooth so that I can. Because what I'm what I'm basically looking for is I'm going to is the impulse of when the torque fails on the ratchet, right?
Yeah, yeah. Yeah. Are you going to try to plot this?
Yeah, I'm gonna plot them. Okay, cool.
Cool. Yeah, something I, whenever there's, like, whenever it's when you're looking for smooth like that, I'm thinking electrical, to apply the force in a mechanical.
If I had a hydraulic pump, electric hydraulic pump, I would definitely do hydraulic. I don't have that. And I didn't look them up. And they're pretty pricy. Whereas like, oh, you can do a hand pump. But then you get back to more discrete editions whenever you stop. Yep. You have a discrete step there.
Yeah, the whole motion of whatever pressure arm you're pressing, that is fairly smooth.
Yes, yeah. Because that part's smooth. It's just when you release through, do another pump, you actually have bleed down to there. Right? So it'll be it'll, it'll still spike. Well, you'll see stair stepping, I like to not to have that. So I got. So there's two things I need to figure out is, can I get a linear actuator style system to push that hard push or pull about 400 pounds 500, let's do 500 pounds, 500 pounds. Because if you think about the lever arm, it's only roughly six inches long, because because there are different links. And that's also thing is probably going to have it hooked into like the end of the other, like rat, other end of the ratchet. So the great thing about measuring the clamping force is we don't have to care about where this force is being applied to the wrench, as long as it's into the wrench.
Yeah, I mean, you want it to be 90 as possible to it. But that's
not going to be the most effective this. But we are not measuring based off that. So that's why we don't have to worry about it.
Well, that's just the thing. As soon as the tooth catches, as long as you're relatively close to 90, you're going to be the most efficient in terms of applying your force. Yeah,
applying the torque. Yeah. And then for the clamping load, I need to calculate the clamping load that's basically going to be between these two plates that like basically go at 250 foot pounds of torque, this big three quarter inch bolt to 250 foot pounds, how much of the clamping load is going to be and I did some napkin math and it's stupid, how much clamping force is in there. So I'm like, Okay, I gotta figure out some load cells that are gonna handle that. Yeah, and I'll probably go with for load cells. And then I'm gonna build a little circuit board that will handle the ADC front end of those, and then it's probably gonna be an Arduino shield. I'll make this plug into an Angeleno. And yeah, quick and easy spit serial onto my computer to record it. Yeah, we need some high speed footage of it.
Do you have a high speed camera?
I get one. Yeah, I'm not worried about it.
Well, I mean, honestly, from a safety perspective, you hope that the high speed camera footage is boring. Yeah. I mean, honestly, like if you could get yourself like a three quarter inch chunk of plexiglass or something like that that you could put around
I'm just gonna wear my safety glasses and maybe a face shield. Might I thought it was gonna be like that, but even the cheap wrench like cheap Harbor Freight wrench. I tested to even get close to these whatever numbers I needed. Know which 102 50 foot rounds like it didn't They explode or anything now it could explode, which is why I'm gonna wear safety glasses and stuff. But it's also like, I'm not going to be like, my face is not gonna be right on it either.
Oh, no, you're gonna be you're gonna be a distance away from
this, which I guess why I want to be able to apply the force smoothly and away from me.
Right? Well, I guess one of the other failure modes in this would be an extreme failure mode is, if you're if, let's say you have, like you mentioned, whatever linear actuator or whatever you're pressing on the other end of that. If one of those teeth on the other end of the ratchet breaks, those could launch right. Yeah, but that would mean the casting of the failed wouldn't would fail. And that's probably not good either. Yeah, well,
it's gonna measure. One thing I forgot to mention the measuring is I'm going to measure the the actual like, teeth, the teeth, not teeth teeth. Hex point, the hex point, maybe in the in the ratchet mechanism, just see what the number is. Because like, technically, the closer to half inch, it can be the technically the better it is right?
The better it is. I mean, but at the same time, so there's a balance between like how easily it like, goes on to the head of the bolt versus how effective it is. Yes, yeah. Because I mean, some slop is necessary. Right, but too much slop is terrible. Have you had a press fit socket? Yeah. Right. You'd be hammering it on.
You have to hammer it off. Right?
You know, actually. Okay, if you measured the span of the of the teeth, get an average of that. What I'd like to know, just from a design perspective, what do most people think the fat should be? Yeah. Like, what if I were to design a ratchet? Is it 10 1000s? Is it 30,000? I don't know. Like, how much is a good gut feel for that? Maybe we find out that all of them are within like, 5000s of each other? Yeah. Actually
within like machining tolerance. Yeah, maybe maybe we find that out. And it's like, oh, well, that's a good bet. I bet you there is a tolerance for it. And we should see through actually within that tolerance. I mean, is there though, like, yeah, there's a there's a specification for the specification for Hex heads. Yeah.
Well holds can be no whatever hex heads, but for the ratchet, is there a standard?
No, no. So you could take the tolerance at it, and see, like, how close are things to that tolerance? Like that number,
right. But what I'm saying is the designer of the ratchet, they have to take the tolerance of the socket head or the hex head, and then add something to it. I'm asking, what did they add? Yeah.
What did they add?
I don't know.
I did. Honestly, that right. There is the hardest part about engineering is like, some people just know what they add what it should be. I don't know that number. I wonder, I wonder if some of these really cheap ones I have are actually like 13 millimeter ones. Those Those guy has a stamp on them? Yeah. So because it's only what half a millimeter more.
So what is half an inch in in? metric is 12. Point 7.5 4.54? I don't think so. Right? Yeah.
What's that? 3.37. Yeah. 12.7. Like I said, yeah, so that's, that's a little over. It's point three millimeters away from 13.
This, this is us freedom units, guys trying to play metric games.
And actually, it was, so I put all the brands and all the model numbers, and I'll put the prices and actually where I bought them from. When I was, I was really disappointed that there's not a lot of Japanese manufacturers and because most Japanese manufacturers make them only in metric. And the Advent text so I mean, go figure. Yeah, but I was like, I really wanted to try it. Same thing with German brands. I only found like two German brands that actually made half inch hex head ratchets that fit all those specifications or qualifications, I should say. It was kind of disappointing. I guess.
I'm going to add one little wrinkle to your plan here. Okay. If, if you have the time, let's let's just put it this way. If you take these and you get you get failures on all of them, maybe take the best one and the worst one, break them apart. And then under magnification, look at the broken steel and see if You can see any difference between Oh, yeah, best and the worst. So like, does one have like really fine crystalline structure? Does the other one have like schmutz inside this deal? Who knows?
I'm going to bet you that the two for $8 ones are going to punch well above their weight. You know, and that'll probably make people both mad and sad. But I would not be surprised. It's like, it's like when Dave Jones tested the Harbor Freight multimeter. It's like a $5. And sometimes you get it for he was so pissed. He was really pissed off, because how good it was.
I remember he put he puts his like, voltage standard, and it was like 10 volts or something like that. And that meter read 10.00. Like, straight on. And that's not necessarily the most difficult thing, but I know, but he was expecting it to be off awful. And it actually
ended up being not bad. Yeah, I really don't think that what those meters are the leads, which, which
we've discussed multiple times, you could buy really good leads on eight bucks. Yeah. And they have needle tips on them. They're awesome. And oh, that, that silicone. The leads themselves are made of that real flexible silicone. It's nice. I'm excited about this, dude, this will be fun.
So next week, I'm gonna have a design and probably ordered the parts to build the tester. Nice. And then I'll have I'll have my calculations on like, how much clamping force is going to have to handle and a lot. It's gonna be a lot. It's, it's something like 200,000 pounds. Yeah, there's a reason why like your car, if you when you properly torque your your lug nuts on your car, your wheels do not fly off. Even though how much energy like calculate how much energy your tires have, like rotating at 80 miles an hour? And how heavy because they're like, What 40 To 60 pounds is what a wheel weighs? Like how much rotational energy is in there? Why don't you just fly off, because you have five lug nuts torque to 100 foot pounds. And they are there's a lot of clamping force holding that thing onto your hub. Because it's not your your studs. And your lug nuts aren't actually holding your wheels on. It's the clamping force that holds it on. So it's not a Yeah, it's because like, when you think when you calculate it out, you're like, Oh, I'm only talking five lug nuts to so total your 500 foot pounds of torque, right? But you look at the clamping force and some like Slyke astronomical number of like safety factor.
Well, they put it where it counts, right? Oh, yeah. That's also why like, I pucker a little bit when I get on the freeway and I see a guy with a donut go in at five. Oh, god, did you put that on?
Or I always, doughnuts don't really scare me. It's when I see someone who's got like either a really out of balance wheel or their suspension shot on that one corner. And so the wheel is like going up and down like six inches. Like how do you not feel that? You probably do and just don't care?
Yeah, or you're behind that guy. And you're like, I can clearly see your realtor hires. And your front tires. Your cars so out of alignment.
Crab walking down on the street. Yeah, yeah. And still going at? Yeah, yeah. Okay. I think we have time for one more topic, Steven.
Yeah, yeah. Let's, uh, you know, we've been talking for probably four or five weeks about, you know, we were gonna mention Python. So let's just rip through that one real fast with the Python one. Yeah, yeah, it'll be quick. So it's, it's kind of funny because it's now kind of ancient to me. I haven't actually been touched by THON and a bunch of weeks, because I accomplish the goals I was going for. And Parker and I've discussed have discussed and kind of joked about this a while ago. If you want an engineer to do a lot of work, give them a reoccurring task, give them something that happens on a regular basis, or specifically, like on a scheduled basis. We've always said an engineer will spend 20 hours of work to save one hour of work, as long as that one hour of work happens regularly.
Yeah, if it happens multiple times, your engineers will do it.
Yeah, like if something. If you have to do the same task over and over, automate it, find a way to automate that task, unless that task is somehow unique every time automated.
A. A good engineer is number one On Thoreau. Number two
lazy, yes, engineers hate to work, except for when they're finding a solution to work, then they will work extremely hard, like, like, it's super, super hard. And so, okay, I had some some data gathering that I knew know I needed to get done. And I could have done this data gathering manually. But we have a management meeting every week where we present our, our numbers, and I present the numbers for a variety of departments at work. And I'm in charge of some people who they themselves need to present numbers. And so I'm looking at how are we getting these numbers, and it's just the method through which we would need to receive these numbers. It's just absolutely garbage. And we would have to do it once a week. So we would literally be spending hours of work between multiple departments and heads of departments getting numbers, that's like, Guys, we can just create something that automatically does this and hands us the numbers. And yes, it might take half a week, or even a full week worth of worth of work to in order to get this but then we never have to do it again. So a handful of weeks ago, I ended up kind of digging into Python, and creating a script that that would do this all for me. And here's the thing, I'm not a coder, like I've said this a bazillion times, I don't write, let me put it this way I write code. Most of the time when it when it affects bare metal, like that's my kind of code I love see on microcontrollers. I'll do that Oh, you like you like knowing that that bit flips? 100% I don't like the few levels of extraction give me one or two levels of abstraction. Not like 400 levels. But But I was like, Okay, I know that this data is gathered doable. I know, this is something I can do. And I know Python can do it. So I was like, Okay, I'll bite the bullet, I'll learn Python. And the one thing that that I think is like a real testament to Python itself is the fact that a guy like me that I get programming, but I'm not good at it, I can go dive into Python, with the idea of this is what I know what my outcome should be. I know what my inputs are, I don't know how to connect those dots. And in a few days, I can connect those dots and not be banging my head on a table like that's, that's where Python really excels, in my opinion. So this script I have, it accesses an API that goes out to our database, where we store all of our part numbers, all of our build IDs, all of our production information, it goes out and it gathers all this data, it parses all of it. And then it creates an Excel sheet that is already pre formatted and hands it to all of our departments such that we can open it, gather our numbers, and just present it each week. So the whole goal was I wanted myself and my department managers to spend only a few minutes each week just reviewing the numbers before the meeting such that we can go and present it and say, Here's how well our departments are doing kind of thing. And, and with Python, there's there's two, there's two plugins that I want to kind of showcase that really made this available or possible. And the first one is requests, which they call it HTTP for humans. And it makes it makes going out to a web address, asking for something and getting information back, it makes it unbelievably simple. Like I make something
good. That module I use practically every script I write,
it's so good. It's so good. Because you just basically say get, you give API keys, you give it web addresses, and it just craps out data to you. Which then at that point, it's just a matter of manipulation. And just asking, or Python to parse through what you need to. So your request is fantastic. And then I found another module called Open pi Excel, open py XL. And that is a whole suite of commands for creating, adjusting modifying Excel spreadsheets. And in fact, what's funny is, the amount of work I spent going and getting the data is probably 5% of all the work that I did. The other 95% was just wrangling this to create the spreadsheet in a way that my department manager
would be happy with through the JSON data and the right way.
And just things like oh, you know, I wanted in this cell and I want it centered and I want this one to be bold and I want that arose. You're
actually in formatting. Here's the thing. This is where he started doing PDF
documents. Oh, PDFs are fun. Yeah. What No, no. Okay, so here's the thing. I've found with people is. And this is not a dig towards people, I think this is actually a good thing. But if you if you present, somebody who isn't regularly using just like strings of data, if you present them that their eyes roll back in the, into the back of their head, and it means nothing to them, if they can't, can't just give a list of numbers to somebody, and they do it to, you could probably do that with an electrical engineer or many other engineers. And they'd be like, Okay, fine, they'd be pissed off, and but they know what to do with it. But like to somebody else who is not that kind of tech savvy, they're just gonna just forget about it. So I know, I had to present this data weekly to my department managers in a way that they would be happy to use it, and they would run with it. So I went through the trouble of like, creating a beautiful Excel spreadsheet.
Or just doing that automatically, like go out to the API, get the data, parse it, make it pretty hand it to people, that's the whole path. So you know what the next step is, with that, figure out what they are doing on their end to make it presentable. And just automate that part.
You know, it goes even, we've learned how to boil our departments down into a single number. And here's the API's sort of, sort of, okay, I'm not a fan of management meetings, they 99% of the time, they're a waste of time. Yeah. And so basically, what I did at work, and I'm saying this in a very positive light, I learned how to become the person who runs the meeting. And then I changed the meeting. And the way we do it, such that they're as short as possible. So basically, what I did was I was like, Okay, we're getting rid of all the fluff, each department says, they show up and they say, I've either reached my numbers, or I've achieved my numbers, or I don't, and then they give a reason why they didn't if they need to, and then we move on to the next person. So I've taken a meeting that was traditionally like an hour long, and I've turned it into like 1520 minutes. Because I've, I've the rules for what what makes the meeting start and stop this, right?
See, because I don't want to mirrors should, engineers should dictate all meetings.
And I laugh about it, because like, in my mind, I've actually made it more efficient. And just a better meeting, because we're not just like talking about gut feel about what what's happening in the meeting. Like we have rock solid things that we talked about. And, and the meeting will go long if we need to have it go long. But most of the time, we now have goals that we say, as long as we reach these goals, the meeting doesn't need to go longer. And so yeah, I think that kind of wraps up to what we were saying earlier, it's like, engineers are lazy, and they will spend a lot of work to make sure that they don't work in a way. So you know, hands up, I'm totally one of those guys. But but you know, now having dug into Python in this way, I already have a ton of ideas in terms of what we what we can do next, because we have another system where we track failures, and RMAs and things like that. And I know they have the ability for API calls, and being able to ingest data and modify things. And so you know, in an effort to even further make our meetings efficient, I want to be able to report quality data automatically without me having to go and scrub a bunch of stuff on an online spreadsheet, and just say like, Okay, show me what failures were done this week, show me, you know, how we repaired them and where we're at and our total numbers and things like that, just just give me the numbers automatically. The only thing that sucks about this Python script, and I actually had to build a cooldown timer into my script, my API that I go out, if you if you have more than 30 calls to the API in a minute, it will shut you down for five minutes. So you have to wait five minutes to spit out another 30 calls. And what's annoying is, depending on what you're asking for, from the API, a call could mean a massive slew of data, or it could mean a single item of data. So I had to create this this cooldown timer where it would keep track of how many calls it went out. And if it reached a certain number, it would just pause for a minute such that it could so
basically like how you have a certain calls per minute that it's trying to keep track of
Korea Yeah. So
that's a normal way of doing
it but but but here's here's an example of where it's it's goofy. I could I could send them an API call that says, show me all of our active builds right now. And that could be pages, pages and pages of data. All right? Well, or one, but we have many builds going on at once. But if I were to say like, Hey, what capacitor is in this location on this board, that also counts as an API call. So I can either get like this massive data set or an individual item, both of them count as the same thing. And so sometimes I have to ask for individual items. And sometimes I have to ask for the whole list. So I just have to keep track of that. So what sucks is to get my data every week? It's about 30 minutes of just the database. Yeah, you're right. Well, it's 30 minutes, because like, I do 30 calls. And then I have to wait a minute and 30 calls and wait a minute. And so I set it up on a window scheduler, that so it happens at like five o'clock in the morning, on Thursday, such that everyone shows up to work, and there's this new data set that happens.
Do you own that database? Or is it in the cloud?
It's in the cloud? That's why
Yeah, that's called up and be like, Hey, guys, I need this rate limit increased. I don't you know, I
don't care about like, if it runs overnight, it doesn't. It doesn't matter to me. But but that is annoying, where it's like 30 calls, you can blow through 30 calls in no time flat. Yeah. In fact, developing this code was annoying, because I had to keep track in my mind, just like, oh, you know, I have to rerun my script. How many times have I done it? Because like, I'm not keeping track of like a variable between run between calls or between runs? Yeah, yeah, exactly. Because there was a handful of times where I tripped it, and it's like, oh, well, time to go get some coffee and wait five minutes. So that was that was fun. And I've had I've actually had managers in the past, utilize what we've talked about where it's like, oh, we're gonna we need a bunch of work from an engineer, we'll just make them do the same thing over and over. Or we'll just schedule work for them on a regular basis, and then they get pissed off, and they solve the whole problem.
Yeah. Now, for a suggestion for people out there that want to get into this kind of like Python automation stuff. There's a really good book called automate the boring stuff. It's like a Python introductory thing for beginners. And it's automate the boring stuff.com. It's a really good book. I think there's an e book too, that you can get. I suggested. And then, then I have one thing I want to talk about here is commenting code. Oh, yeah. What is useful comments, for maintaining the code later down the road? You're putting in a comment that is this. This line does two plus two equals four is not good commenting. I would say anything that has to refer to external documentation, like, like, let's say, you just paste it in, like this crazy formula, maybe have where you found that like, the URL in your comments, or whatever, or what that function does. So I like to do is actually function level commenting. So like, if I write a function, I go, this function does paragraph of stuff. Right? Right. Right, then I go example. Give it this. And I'm like, this is the input. And then this is the output. So that way, you can, you can technically kind of like a mini tests for it. So you can be like, there's a bug, does it actually go through its test? Example correctly? That kind of stuff? Because commenting line by because I had one professor in college that want us to comment every line and I'm like, That is such a waste of time.
Well, okay, if
you don't need you don't need a comment. Hilton does add dition
Yeah, right. Right. That's, that's okay. So that's my point. If it is a built in function to the standard language, that probably doesn't like what you're saying like x is equal to two plus two. Like, you don't need to comment that unless there's something unique about it. But like, when I was using that, open pi Excel, like a lot of times I would write for loops that would go through and format groups of cells. I'll write a thing above that saying, this will take cells A through whatever, blah, blah, blah, and make them all bold. And so like, the chunk of code underneath that it just gives a quick snippet of what's happening there. That I think that's plenty enough, but you don't have to go line by line.
That's plenty and actually useful commenting code.
Yeah, yeah. This next section centers all the text, you know, like that.
That's exactly, yeah. And you could go as much as like using really good. Variable names. It's, in my opinion better for variable names be longer to explain what they are than to be shorter. That's just my opinion, though. I have some variable names that are like, three or four words, sometimes that's like, that's what you need to properly explain it. So when you open that thing up in three years, because it's broken now, what this is variable do again? Yeah,
actually, you know, it's funny. So I'm probably stepping way outside my boundaries here. But if a variable gets sent to a function, or returns from a function, I usually have a fairly long name for that, or I have a very explicit name for it. But if it's used inside the function, like an increment in a in a loop or something like that, it's usually pretty small.
Yeah, that's kind of normal. Like, yeah, if it's a loop variable, you might name it. C and T for count. Right? Sounds like that's totally normal. Yeah.
Yeah. There's probably wouldn't send cn T to a variable or a phone.
There's a lot of different schools of thoughts from schools of thoughts on that that are way beyond us in this podcasts. Exactly. Yeah. Like camel case, or like our like our Global's always capitalized? Or is it the first sort of Yep, camel case or under? Some people don't like underscores in them. I use underscores all the time.
So that's just a Yeah. No underscores all the time for sure. Yeah. underscores between two different words.
No. That's gonna wrap up this podcast. Yeah, I think that's it. Yeah. So that was the macro engineering podcast and we were your hosts, Parker, Dolman.
And Steven Craig. Later everyone.
Thank you, yes, you our listener for downloading our podcast. If you have a cool idea, project or topic or you have a ratcheting wrench that I should test. Let's do it. And I know Tweet us at macro lab at Longhorn engineer or at analog EEG, or emails at podcasts@macro.com. Also, check out our Slack channel. You can find it at Mac fab.com/slack. And I still don't have a URL shortener slash shortcut for the Twitch stream thing that happens at six o'clock on Tuesdays. That's six o'clock central time in the United States. So you have to go to twitch.tv/mcwrap or follow Steven and I on Twitter, as we tweet it when we send it out, and we also post the link up in our Slack channel. I think that's it everyone. Have a good weekend.
Why is estimating a projects completion time feel like it takes more work then the actual project? Estimating Project Time, the quest of management.
Our Spider-sense is tingling... OH that is actually a man-made zombie spider crawling up my leg. Who thought that would be a good idea? WHO!?!
Hail to the signal switcher! On this episode, Parker wraps up his prep work for the Extra-Life Charity stream and Stephen discusses switching signals.