- Battery Charger debacle update
- Two batteries recovered, One failed
- Amazon Sidewalk draws electronics from among TI, Silicon Labs and Semtech
- Amazon Sidewalk, a new shared wireless network for IoT consumer device
- DXF to Board outlines?
- Eagles DXF importer doesn’t work all too well
- AutoDesk Fusion seems to import just fine
- Connectors on schematics
- Rfdave on slack channel asked a question about this
- Am I the only person who hates a schematic consisting of 1 connector, 2 IC’s, 4 bypass caps with all of the interconnects being done with net names, and no interconnecting wires?
- Currently working on a design that has multiple boards with cable interconnects
- How do you like to show this?
- Reference Designator?
- Rfdave on slack channel asked a question about this
- Potentiometer board – one board to rule them all
- Using slots to make panel mount “solder lug” potentiometers into pcb mount
- Not a lot of values in each taper is available in a through hole version
- This means a lot less wiring is required.
- EDA schematic and grounding
- Using slots to make panel mount “solder lug” potentiometers into pcb mount
Special thanks to whixr over at Tymkrs for the intro and outro!
About The Hosts
Parker Dillmann is MacroFab's Co-Founder, and Lead ECE with backgrounds in Embedded System Design, and Digital Signal Processing. He got his start in 2005 by hacking Nintendo consoles into portable gaming units. He also runs the blog, longhornengineer.com, where he posts his personal projects, technical guides, and appnotes about board layout design and components. Parker graduated with a BS in Electrical and Computer Engineering from the University of Texas.
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.
Host 2 00:00
Oh you want to start saving? I did yeah.
Host 1 00:13
Yeah, I do. Okay, thanks.
Host 2 00:15
Host 1 00:18
Welcome to the macro fab engineering podcast. We
Host 2 00:21
Are your host, Stephen Craig and Parker Dolman. This is episode 243. So I have an update on the battery debacle from last week. Just remind us what happened. So what happened last well, week and a half ago at this point was my I have a bunch of batteries for my cars, and the battery tender that I have hooked them all up all to all of them decided to hit the hit the bucket, and it took the batteries with it. And basically what I can, what I can figure out is one of the cells inside of the battery tender, took a dump and drained all the batteries. I was able to recover to have the batteries fully leased in terms of you know, they fully charged up and you can hit him with a starter motor and they crank engines over perfectly fine. But the main one, the one that killed everything. That one is no longer with us.
Host 1 01:28
Was that the $400 one? Yeah, that
Host 2 01:30
Was the $400 one. Oh, yeah, the good thing is I did buy that one used. So I only paid $200 for it, which is like still that's still you know, not cheap battery. So and I went today, and because I ordered it last week, the replacement battery and it showed up today once go pick it up. And yeah, it's completely fine now. But yeah, that battery has become a core deposit in the sky.
Host 1 01:59
Josh, please, please play us some sad music here.
Host 1 02:02
This music was removed by artificial intelligence for copyright infringement. Morons.
Host 2 02:08
So yeah, I did get a spare. That company ended up sending me a replacement tender. I don't think I'm going to use it because I don't really trust this brand anymore.
Host 1 02:21
The battery Butler. Yeah, the battery battery puller.
Host 2 02:26
Say I probably won't use the replacement because the one I got to replace the one that went went is a lot better unit. You can see Steven in the background. There's the box that the new old one came in. I'm thinking about actually like opening it up and see if it has the same kind of like poor solder quality to it. I think he showed it.
Host 1 02:50
I mean, what it's a paperweight for you now, right?
Host 2 02:53
Yeah, I'm thinking about actually bringing it basic. Bring it down the Galveston and then let my dad use it because he's got a lot of 12 volt batteries for like boats and stuff.
Host 1 03:03
Oh, and the the golf cart. Yeah. How many batteries are in the golf cart? Is
Host 2 03:09
It to know it's 36 volts with six volt batteries. So six batteries in it?
Host 1 03:16
Oh, wow. Okay, I didn't know ran on six. Was that standard for golf cards?
Host 2 03:20
Yes, that's standard. So in Golf Cart world, you have six, eight and 12 volt batteries. So a lot of the lower voltage ones. Let's say low voltage 36 volt. So in golf carts, 36 volts and 48 volts, and now 72 volts or like common battery voltages or banks for for golf carts. And so my parents have a 36 volts, and we're going to upgrade it to 48 volts. This winter. We're going to put 412 volt batteries in it. Okay, lithium,
Host 1 03:54
Is there some kind of like, like master controller? Or can you just like overcharge
Host 2 03:59
It? So yeah, the stock ECM or Yeah, ECM, the electrical motor controller or EMC, I should say, actually, you have to replace that with a new one. But that's it. Usually, a lot of times the motors, you have to check your motor recourse but most of them are rated to a lot higher voltage than what the golf cart, use them at. So you can put more voltage in them, it's fine. Basically, the big thing is you limits the you can you can run more voltage, but you limit the current. And so basically your total power output to the motors got to be the same. got it got it. But usually, I think the golf cart motor that my parents have, it's rated at 96 volts, so you can't exceed 96 volts but you can put whatever voltage you want up to that and then you just limit the current is make sure the total wattage is the same and you're fine.
Host 1 04:53
I could just see it. Now Parker's mom has got a bottle of wine in her hand and she's popping wheelies as she goes down the street and
Host 2 05:01
That is I really wanted to go to 72 volts because it would it could do burnouts at that point.
Host 1 05:10
Does it actually have pneumatic tires or are they just hard plastic?
Host 2 05:13
It's pneumatic. Okay. Yeah, they got it's got real tires on it. Yeah. You can put
Host 1 05:18
Forklift tires on it and then really feel out.
Host 2 05:24
Host 1 05:25
Don't Don't Don't tell anyone at Walmart. But when I worked there many years ago, the parking lot used to get wet when it would rain and we would totally peel out in the forklift.
Host 2 05:35
We did that at fab once. Did we? Yeah. That was when we rented that big forklifts.
Host 1 05:43
Oh, yeah. Yeah. And everyone got to drop. Yeah, we
Host 2 05:46
All got the drive in the back parking lot. Right. Don't tell the rental company. No, no. Yeah. And then it rained that afternoon. And then we all drove it out there and the donuts. That was a that was the biggest forklift ever driven.
Host 1 06:01
We had to get a Heller 1703 or what? Yeah, it was 1703. I think we have to get that off of off of a flatbed. And it was just like how we're doing this?
Host 2 06:11
Well, this just that forklift was ginormous. Because I remember Oh, man getting the when we were when we bought the micronic the first one. And we were missing that old fab. Oh, yeah. Yeah. Then we did. So it came on a flatbed truck. Yeah. And how that shop was set up is it was the floor was elevated above grade. So you had a dock high loading dock, but you also had a ramp? Well, we couldn't go through the dark high, because the it was too too wide. The micronic was too wide for the my chronic box was too wide. And so it wouldn't fit through the bay doors. And so the only way was to basically back it up into the loading dock area, the ramp. And, but we couldn't do that because the place was built out with offices there. And so like you only had like 10 feet, the turn whatever you got up in there. So you couldn't like back you couldn't like put it on a trailer, so to speak and back it in there. Yeah. Which was like what I would have been maybe it would be so much easier. But anyways, so we we got it off the flatbed and put it onto another flatbed truck that was smaller. But it was a it was a flat tow truck. It was a flat tow truck because it had a tilting bed. Right and so so because we had to get it halfway up the ramp, and then tilt the bed a bit. And then as we were backing it up, tilt the bed more and more, because like it was like ceiling clearance or there was some reason we were ridiculous, right? Because we had ceiling clearance issues. So you had to like, tilt the button, then we got it all the way in. Right. Yeah. And then we had another forklift in there. That was definitely too small to pick this thing up. 100%. And because you'd pick it up, because you had to have extensions on the forks, which you're not supposed to use and
Host 1 08:11
Yeah, OSHA does not look kindly upon fork extensions. Yeah.
Host 2 08:14
But so yeah, that fork extensions on this forklift. And you picked it up and they were bowed like crazy. And then we slowly back that thing off. And then the, I remember we
Host 1 08:26
Had to lift this $200,000 pick and place machine that was brand new. We did lift it almost to the maximum of the forklift, such that the flatbed truck could get out and tip its its bed. And so I remember, all of us are sitting there. Our butts are so puckered. We couldn't cut steel with them. Like everyone was just like, oh my god, are we going to drop this $200,000 thick and so
Host 2 08:54
The next time we accepted a big crate like that. From my chronic it was the it was the other pickup place and we just hired a company that unloaded. It was probably 10 times easier. Oh, yeah. They showed up with the right equipment and then did it immediately. I think the Heller was the last thing we unloaded ourselves.
Host 1 09:15
Yeah, I rent I was the one who had to rent that forklift. And like the problem was like, I had to tell the forklift company what we were picking up with it and they're like, we're not letting you get anything other than this forklift. I was like, fine. So we got to enormous there was.
Host 2 09:31
But yeah, the thing is, though, is that it lifted it up. No problem. We moved everything, put it back down. And that was it. Yeah, it was like super easy. Yeah, I was that first one. It was. We just, we, all I remember was it was one of our neighbors in in the, in the in the warehouse area complex. They had a forklift and we're like, hey, you know, we have to lift this thing up and move it and we gave him the weight and like, you think you're Cliff could do it. And he said, Sure. It was like on the backs of what that forklift could do. Yeah, for sure was there was not a safety margin on that forklift for,
Host 1 10:11
You know, it's funny. We don't have a forklift at work. But we've had many, many situations where we have needed a forklift to move our equipment around. And we're so lucky because we have some neighbors, you know, a few few buildings down that does have a forklift. I cannot tell you how many times I've walked over with a case of beer and been like, Hey, can we get you to lift something real quick? And they're like, Yeah, sure.
Host 2 10:37
But didn't you have the when you got that big forklift? Didn't you have to drive it down the road? Or how did I get
Host 1 10:43
Actually, they so they delivered that? They delivered that on a and, and our shipping manager at the time was the only one. Well, I've driven a forklift plenty, but I was like, I It's been a while. And he was the only one who had or he had the most recent experience for driving a forklift. So it's like, this is all on you, champ. Luckily, he did it like no problem. That was Colbert, but no, that was That was Bob. That was that was. Yeah, that was Bobby He did it. Huh. Yeah. Yeah. Even Colbert was like Yeah, those were
Host 2 11:24
Good times. Yeah. Rent renting heavy equipments. Always a blast. Yeah,
Host 1 11:29
Yeah. Me too, because like, I rented that forklift. And I rented it for I think, the minimum amount because we rented this forklift because we had to get one box off of one truck once, like that was that was the whole thing. I rented this forklift, and I rented it for I think the minimum amount of time was a day and a half or something like that. Yeah, usually, it's like 36 hours. It's like 36 hours. So I rented it out when I was done. I called them I'm like, we're done. You can pick it up. And they left it in our parking lot for like, five days. Like, Yeah, cuz they just picked it up whenever they wanted to pick it up.
Host 2 12:03
It's free storage. Yeah, it just stayed there until the next person needed it.
Host 1 12:07
Probably. I bet you they picked it up and just, it probably got rained on and everything to outdoor equipment. Yeah,
Host 2 12:13
We remember that. That scissor lift we had. Did you ever drive that scissor lift around?
Host 1 12:19
I don't. I have driven a scissor lift. But I don't remember what the fab.
Host 2 12:24
It was when we were doing the final build out of the new I say new location. We've been there for like four years now. They it was the contractors had it. So I don't remember that. Yeah, they just left it outside for like couple weeks.
Host 1 12:43
Host 2 12:44
Yeah. Not enough.
Host 1 12:47
Yeah, we've wasted 12 minutes of your time. regaling you with nostalgia
Host 2 12:50
Here. Yeah, about forklifts.
Host 1 12:54
So one of your batteries died or to your batteries died but
Host 2 12:57
No. No to live to enable our government to have them one died. Yeah, I think it was because two of the batteries are just deep cycles. They're not starting. And they have they have higher internal resistance normally then starting battery would. And so they might have been safe from whatever happened to that charger. Because those things we pulled the board out looked at it. There's nothing wrong with the board. So something other something else happened to it, maybe that microcontroller that's key description is what was it again? It wasn't value saving or something like that, but it was what was it? I don't remember. At least I'm gonna have to look this up.
Host 1 13:45
So I realized I haven't really looked into this. How many amps does a starter pull from a cold start? Like a few 100
Host 2 13:56
Yes, few 100. Usually like for a medium size V eight engine, the starters the rater that they're not rated at that rotation they've heard they have a stall rating. So if you stalled it what it would pull which is basically like your instantaneous like when you click it because it's it's what you stole it it's just the resistance of the coil. Yeah, but the radii that the that amperage. It's because basically when you turn the key it starts at stall and then moves
Host 1 14:27
Right. So yeah, as soon as you turn the key that's what is drawing effect. Yeah.
Host 2 14:32
But then there's a continuous winds rotating which is a lot less but you have to hit you got to get that hit that that amperage hit. And for the wagon, which is AMC 360 engine, which doesn't really matter. It's they're all about the same. It's 300 amps up stall. So you get basically a 300 amp instantaneous hits and then it probably is like 90 to 100 Continue rotate it. Oh, yes, that's right. The microcontrollers key feature is cost effective. So I don't want to trust my new $350 battery to a cost effective microcontroller. That makes sure it doesn't die again. So yeah. Fun. What do you been up to Steven?
Host 3 15:27
You know, actually, so earlier today, was it today? I think it is. It was today. Yeah. Yeah. Yeah. Yeah. Earlier today, someone on the Slack channel RF, Dave posted a question that he posted it. And then I started thinking about it. And it was going through my head all day. And I was like, You know what, I want to talk about that on podcasts? Because I'm curious to know, your opinion. And, you know, I feel like if we were doing a live cash this time, this would be a good conversation to have with people. So our Dave posted this question, am I the only person who hates us chromatic Kunis consisting of one connector to ICs, four bypass caps, with all of the interconnects being done with net names and no interconnecting wires?
Host 2 16:13
I feel personally attacked. Because that's like my schematic.
Host 3 16:18
Well, here's the thing. Okay, you and I have talked about this topic multiple, multiple times. So this is nothing new for the podcast. But what I'm what I'm kind of curious about is like, is there a standard, or what's the best way to do this, because there's an easy way. And there's kind of a really shitty way to do it in terms of being able to read the schematic, but it's even easier because you don't have to type in that names. You could just spaghetti all the connectors together. But I'm kind of curious about this situation, because I'm actually designing a board right now that has multiple connectors that connects different boards together. And even at different locations on different boards, it connects them together. And I sort of, I have my way that I like to do it. And it's a kind of a hybrid between names and wires at the same time. So I think I agree with RF Dave like, yeah, if there's a simple schematic where you can show the connector on his on the schematic as like one item?
Host 1 17:22
That's sort of the best way if you can get away with that. But I don't know, what are your thoughts on that
Host 2 17:28
Point? So I draw mind with the connector is one part, like a symbol? It depends on how complicated is the board is whether or not it's actually connected to things or is just connected to nets? I think that there's not there's not a standard at all with this, what you really have to do is, who's the schematic for you? Because you're the only one dealing with the hardware? Or is it a team or a company. And if that's the case, and you have to come up with a standard that everyone agrees on, and there's there's a threshold, but the thresholds not like a touchdown line where like the ball just has touch it and then you get a touchdown, right? It's, it's this big gray scale of, of what how complex something gets, because like something will be super complicated. But then you only have one little segment of the circuit that you can draw out. And like like, let's say, an op amp with a feedback circuit, and it connects to a sensor. You can draw that all as one little thing and when you look at it, you don't need names, right? Because you can see the feedback loop and the bypass cap is there and all that good stuff. But then you go okay, now I have this ginormous 800 pin FPGA that's also in the design. Yeah, that is connected to like 483 backplane connectors, right, right. Or pins in a backplane connector. It's like drawing, like if given drawing out all those little lines wouldn't help you.
Host 1 19:05
Yeah, it doesn't, it doesn't make any sense. Especially if you're the end result of or if you're the person reading the schematic and all you get as a PDF, where you can't like search by net names or click on a net name like it's virtually worthless. You're sitting there with your finger trying to follow lines and
Host 2 19:21
Exactly so it's not like it's not like in a an EDA tool when you you can just like right click on net, and then just get like or an eagle it's like you just show right and then it highlights it as a different color. So you can easily see it on your screen and trace it because like even even in super complicated things, drawing stuff out with the line or with the with the netline doesn't really help you. So it's you got to figure out where that that scale is that where whatever is right complicated enough to where you have this switch over and then might be the moment that that line or that line goes off the page to somewhere else. And so you had, you would have to scroll in your schematic, because like you use different different pages for your schematics. Yeah. You know, a lot of people do that. I don't. But I know a lot of people do that. So it could be, if you have to go to a different page, then yes, you have to you have to go nuts, right? You just have to. But for like me, I view it as if I have to pan somewhere else to see where a trace goes. That's pretty much has to be a net now. And a virtual trace, right? I don't know what's a good word word for that. Yeah, that that green or? Yeah, and that name. But yeah, I started doing well, yeah. But what I've started doing is I made a in Eagle, at least you can, you can do it by net name, where it just puts the name down. But you can also make it like a tag. So it like it draws a bounding box over the around the text. And so it looks like a test point flag. And so that makes a lot easier to read that kind of stuff. Yeah, yeah. I guess I should post some, some schematic images of like, simple stuff, first complex stuff I've done. But I've actually, you know, when you start working at McAfee ages, they just go, I started doing my schematics more like you because you complain about how shitty mine were? And yes, it's a shitty, they were shitty, no hard to follow. They were hard to follow. That's not necessarily shitty yesterday.
Host 1 21:47
Yeah, actually, you know, it's interesting. So at WMD, we've talked about this a handful of times. In fact, we were talking about this today, because we're training up some new people in terms of doing electronic diagnostics. And, and some of our legacy products, stuff designed six, seven years ago, were done where there were no net names, every connection has a connection. And so if you look at it, you pull it up, and it's zoomed way far out, and there's just wires going everywhere. And, you know, that's that's just how the, you know, the designer did it at that time. And there's, there's actually, I wouldn't necessarily say there's anything wrong with doing it that way. It's just, especially if you're the designer, and you interface with it all the time, and you're the only one interfacing with it most of the time. There's not necessarily something wrong with that. But it is a nice nod to other people, as soon as other people have to start looking at it to group it. And so I've started doing two different things, actually. So So almost all of our products now are two boards, we have one board that interfaces with a panel, and we call that typically a like our mechanical board. And that's where all of our potentiometers and jacks go. And then it has connectors that connect to another board. And that's our main board that has most of our SMT stuff on it. And so I've started drawing one page, that on the right side of the page, I have whatever my female connectors are, and then on a different schematic page. On the left side, I have all the male connectors that that are interface directly with that. And so one of you, I can't remember which one off the top of my head, what the male or the female connector is called J x and then the number. And then the female is J y, and then the number so if you see J x one and JY one, that's the male and female that connect together, and J x two and JY. Two are the male and female that connect together for connector, one, two, or however many connectors you have. And that's our inner connect boards. And those are on my schematics that sort of like defaults now. Because they're there, there's, they're always going to be there. Because we're pretty much always due to layer, two boards backups now. And then for those connectors, I'm doing net names, for the most part. Because in general, anyone who needs to see a schematic has access to EDA tools in our building, we didn't we're not making PDFs. If I were to do PDFs, I would probably handle it differently because just like looking at at you know, pin 15 of this connector is this name and then you have to search the schematic for, you know, physically searched pieces of paper for it that's kind of garbage. So I would handle it differently if we didn't have access to digital schematics. Now one of the things that I've done recently is instead of creating a schematic symbol for a connector, basically a rectangle with however many connections represents however many pins, I've actually created some connectors where each pin is its own Simple, it's so like, say, if I have a 10 pin connector, I'll have J 1.1, J 1.2 J, one dot all the way through. And then I can connect those pins wherever I want them to. And the way I kind of liked that is because some of my analog circuits make a lot of sense to draw them in a certain way. So you can see what this schematic is doing. But they don't make a lot of sense to be all lined up perfectly like a connector, a connector. And so you get to choose Do you want your analog side to look really pretty and look the way people think it should look? Or do you want the connector to look really pretty? Well, in my opinion, I think the analog Trumps in that way, because it makes the analog circuit easier to read. And if you need to trace out the connector, which the only time you would need to trace it out is if there's like a diagnostic problem. And in that case, you're going to be digging a lot deeper into the schematic. So I feel like in you know, it's kind of a toss up whichever way works best in whatever situation.
Host 2 26:01
And you could do you could do both, really is you can you can draw your pins like that. And then you can make a virtual part. That is what the connector is right in real life, the symbol, and then all the nets that are supposed to be connected to it. So you can totally have both, and that actually might be the best way to do that. Yeah, I
Host 1 26:21
Think I think hybrid is, I think one of the most powerful tools in Eda. Eda world is the ability to use multisection components, instead of having to draw everything as one block with all of its pins, like being able to break them all out, makes schematics so much easier to read, especially things like op amps, like take a take a standard dual op amp package, you're gonna have two op amp packages, and then you'll have two pins that represent positive and negative power. I always draw those as three separate things off and one op amp two and then the two power pins as its own little chunk, because I typically have a page on my schematic that is dedicated to power. So I take those two power pins, and I put those on the power page sucks that if you ever want to see what voltage, whatever op amp on my circuit is run at, go to the power page, look up its reference designator. And you can see Oh, this one is run on plus minus 12 volts or whatever. And what that does is it cleans up your op amps. So you don't have to show the power pins on the actual op amp package, which is annoying as hell, because it just makes things cluttered. So like if you got a circuit that uses one op amp, sure, fine. But like, my most recent design has like a ton of op amps and would make my pages just very, very large. And on top of that the way a standard op amp package is drawn, you know you have the triangle Yeah. And, and almost almost every single time you have a resistor that's fed back unless you're doing a comparators symbol. So if you have the plus minus symbols, you're going to have to have net ports that show plus minus and then if you put bypass caps, those are in there too. And then you have to draw a feedback resistor that goes way around all of those things. It just makes everything super ugly and harder to read.
Host 2 28:17
Or, or you could have been like me and flip the part and not realize that the power and ground were also flipped.
Host 1 28:24
I remember that that offense got pretty hot. Yeah, they got
Host 2 28:27
It got about two watts of power through it.
Host 1 28:32
So yeah, no, thanks for that RF. Dave. I honestly us that question. I was thinking about it all day. Because literally this last week I was I was kind of wrestling with that because I have some projects at work that I'm doing a lot of connectors on. And then I have a handful of projects at at, at home the personal projects, that in fact the stuff I've been talking about with my my DIY Pogo pin connector. I was I was laying that out this week, and figuring out okay, I've got this Pogo pin connector, do I want to represent it on my schematic as one thing? And then I have to draw lines to it? Or do I want to to be net names that I've just connect to it? Or do I want to have every pin be its own schematic symbol and connect them wherever they go? Conveniently and I actually opted for that last one, because I like the way to represent this particular thing that it that it connects to in my schematic, I like the way that's drawn. And I don't want to have to put weird net names all over the place. So I've represented each pin with its own item. And then I just connected those directly to what they connect to
Host 2 29:39
The Yeah, I think we should. It's been a while since we've shared schematics and stuff of what they look like. So we should do that. Yeah. Yeah. So I got this a news article today about Amazon sidewalk. And so I shared it with Stephen and basically what it's about Out is items on the title is Amazon sidewalk draws electronics from TI silicon lands and syntek, which is like, basically it's like an article saying Amazon sidewalk is using, like all these parts. And that's about what the article is about. Not very interesting, in my opinion, because like, yeah, like how many different manufacturers and are in one of your products like eight, nine? Like, yeah, you're going to use a lot of different manufacturers because you have to write to get all the features you want. Anyways, I was like, what is what really interested me though, was what was Amazon sidewalk? Because I'm like, that my mind immediately went to like solar panel roadway,
Host 1 30:40
You know, it's funny, you posted that to me today. And without even clicking on the link. I was like, Oh, this is a solar sidewalk I'm reading.
Host 2 30:49
And so I clicked on it. And I'm like, Okay, there's no information here. What this actually is, besides the tagline, Amazon sidewalk, a new shared wireless network for IoT consumer devices, and I'm like, Oh, we've talked about this kind of stuff and had people on the podcast, talking about different kind of IoT mesh networks and that kind of stuff. So
Host 1 31:09
I feel like it's been a while since we've talked about the Internet of turds like it. It's really been a while, right
Host 2 31:15
Has been a while. But I was like, Okay, what, what's the infrastructure behind this? That kind of stuff. So I started looking into it. And I'm like, it actually has nothing to do with sidewalks. That's just the name. It's just a Cubase,
Host 1 31:29
That they came up with you. Yeah, basically, it
Host 2 31:31
Is a low power sub gigahertz, net mesh network that they are going to implement into other devices that they currently have. And so like, let's say the ring, which is there, that's the doorbell that like spies on you on your front door. So it would eventually have it doesn't yet, I don't think but it will have another radio in it that's designed to broadcast out into the streets. And so that you basically create a mesh network eventually, in neighborhoods and everywhere else. And that way, other devices that would normally only work inside your house, because your IoT network is in your house, it would work elsewhere, as well, because you have all these other things, broadcasting out so everything becomes a repeater. Yeah, basically, yeah. More devices become repeaters. And since they're outside, like, let's say in the ring, for example, it's broadcasting out and doesn't have anything in front of it anyway. So your distance into your into the street is bigger, which is actually a pretty interesting idea of handling the the IoT network problem of like, basically, the moment you leave your house, your IoT devices useless. Unless it's cellular connected, and then you know, now you're talking to high power. And so you, you eat your batteries up. So very interesting. And see what how this goes.
Host 1 33:03
I, what's coming to mind is called What's the second Christopher Nolan Batman movie? You know, where, where they put like, the cell phones down, and they're all like, ping each other. And then Batman has like cell phone vision. And he that's like, this is, I don't know, sort of like that.
Host 2 33:25
You'll have seen technology based off that trying to figure out room, what's in a room based off reflections of waves, which is what that is. It's just they did it with cell phones. And it doesn't make any sense
Host 1 33:42
At all, not only four gig cell phone,
Host 2 33:45
I don't think a cell phone is kit has enough channels to figure out all the reflections in your room. Because that's what you would need, you need a antenna that can handle that many kinds of reflections. And then calculate all debts. Bananas, anyways. Very interesting idea. And those one video out there of like being able to find your dog, if your dog got out, because your dog has a is currently you know, it's it has an IoT collar. Right. And so it would it could tell you what house it's by.
Host 1 34:19
Which is interesting that Yeah, that's cool. You could be I mean, assuming you live in a in a densely enough populated area that also has enough of these things to create a mesh. Yeah. But I
Host 2 34:31
Think the I their idea is integrating this into other other products, so it doesn't matter. So like if you bought an Alexa, and would have this already.
Host 1 34:44
Host 2 34:46
I mean, that's cool. Yeah. It's just an interesting idea that that instead of building their own infrastructure, they're like, well, let's just toss it into every single consumer device already. It doesn't matter.
Host 1 35:00
You know they Well, right? I guess the question really comes up like, I don't know, it starts to bring up some some concerns in my head where it's like, okay, you purchase this thing, and you're expecting it to be purely for you. But now you are like servicing everyone else. And that's fine. But do you know you're doing that.
Host 2 35:21
And there was a big thing with Comcast slash XFINITY, where their boxes would broadcast. They accent because XFINITY he's got a similar idea where you can if you're, if you have XFINITY, you can find the XFINITY hotspot and just log into it. I think 18 T does the same thing. And all those guys sold all that dudes, there's like but expending he was like, or slash Comcast was like the first company to be like, oh, yeah, let's put that on our home modem or home modems as well. And so like, you actually have to go in onto the XFINITY website and opt out of it. That say, don't use my home box as a hotspot for your free Wi Fi. Now, I don't think it counts against your broadband, it like it classifies the packets as something else. So like, it doesn't, if you have a broadband cap or whatever, it doesn't affect you. But I don't know if Amazon is able to do that. But, you know, no, it was just an interesting idea of, yeah, pretty instead of trying to build out a network and then getting people to buy in on something like this, like, let's just put in everything we make no matter what. Yeah, and doesn't matter.
Host 1 36:34
I mean, it's a cool idea. I like I said, it just kind of brings up some concerns because like there's traffic going through your item then that you're not aware of, and you didn't purchase it for that. So what what kind of controls and safeties? Do you have? You know, who knows what kind of like weird communication is going through your ring doorbell terminal, and then the FBI is hitting you up being like, we need to gather this information. You know,
Host 2 37:02
You got got a bitcoin miner going through your ring ring. Doorbell now. Yeah. The there was no side thought of, but I can't remember anymore. Oh, well. Cool. Cool.
Host 3 37:20
So I got I got something fun that I've thrown together. I've done something similar to this a handful of times. But it's always, it's always fun to convert parts that are not traditionally intended for through hole into through hole applications. And so a great example is potentiometers.
Host 1 37:41
If you've ever really spent much time searching for potentiometers, you can bang your head on a table for a long time, trying to find what will actually work for your application, especially if you need a variety of potentiometers there's always the option of if your volume is enough, you can just purchase whatever you want, or you get them custom made. But if you fall underneath that threshold, then it gets difficult, quick, sticky. Yeah. So a lot of the big players will stock potentiometers that are more generic, shall we say. So things that are like panel mount and solder lugs. Like if you go to Mauser right now there's a pretty decent amount of potentiometers available that are exactly that panel, mount and solder lug. And if you want to get like the entire catalog of potentiometers. So all the values and all the tapers, then you're pretty much stuck with that style of connection, panel mount and solder lug. Which sucks because that means soldering wires, which sucks because soldering wires, and labor. Yeah, lots and lots and lots of labor. And if you look at this exact same potentiometer in the through hole version, they have it but they don't have the entire catalog, you're not going to get all the tapers, and you're not going to get all the values. And so when designing like a suite of projects or products, it's nice to know what's available, especially for prototyping and for small manufacturing runs before going into the big stuff. Well, so I've been using these are V 24 af potentiometers for a decade now. These are like just your Joe Schmo potentiometers and they have the nice solder lugs on them. And I've finally set off to just say like, you know what, I want to convert these into a through hole version thing. So I'm holding this up to the camera now we'll take some pictures and send it to everyone or put it on the show notes. But I made a potentiometer board that just interfaces with these pots. So these pots have I don't know like a an interesting leg on them because it's like an island It's an island Yeah, and there's like a circle with a with a wall. Weird curve on it. So effectively, I just mirrored that and made slots in a PCB such that an actual potentiometer can slide in. But there's plenty of extra room and a cutout for the body of the potentiometer. So now whenever I mount these into a chassis, I can put a row of potentiometers. Or even if I want to do two rows of potentiometers, and have them all interface directly with this potentiometer board,
Host 2 40:29
You know, what makes I'm seeing where you're going with this? Yeah, is, if you ever had to disassemble it, you only have to unplug connectors. Exactly, instead of how to de solder wires on potentiometers.
Host 1 40:41
Bingo. Now, here's the thing on top of that, the way I had previously had this product design is that I had a JST connector on a main PCB, and it would go out to three separate wires that will connect up to whatever PCB so every potentiometer had one JST connector and three wires, not necessarily, that's not the worst thing, but it is a lot of wiring. And if you ever have potentiometers, that, you know, share functions or connect to each other. You're you're doing extra connections and extra wires that you could do on a PCB. So I've taken this board that I'm holding up that has six potentiometers on it. Now, each one with its own JST connector on it, I can actually redesign this board recently. So six potentiometers now only needs two JST connectors because the potentiometers share functions on them. And I put some, I guess, like ground and power, and then honestly, it's in out and ground. And all the pots do all their things on it. And I have a handful of passive components that go on there also. So it just makes wiring and everything so much easier. And now I don't have to worry about I don't have to worry about oh, is this potentiometer going to be available? Well, I've got a board that just interfaces with your, your being the one that's available that's intended for panel mount and wire, but my board interfaces with it. And the cool thing is, is eventually if I get to the point where I do would ever want to go with a more custom potentiometer I've already proven this board out so I just replaced the footprints with whatever works with whatever pot I'm buying and Bob's your uncle at that point. So yeah, kind of fun. I like I like doing these kinds of it's fun to do the mechanical side of how do you make a part that's not meant to be through hole? How do you turn it into a three wall part?
Host 2 42:44
Yeah, I think we just we talked about the, like doing the exact opposite. couple podcasts ago where we're talking about potentiometer that we were turning a through hole into SMT.
Host 1 42:55
Right? Yeah, cuz cuz legs had to be been bent out.
Host 2 42:59
Yeah. And by the way, that actually worked really well. I made a little tool that would slide over the legs, and then you would just so it was like Ay, ay. Ay, it was just like a butter knife, basically, that just had a slot in the end, and the slot was big enough to fit over the whole legs. And, and it would the legs would bottom out in the slot and you would just bend them. Oh, nice. Great. Yeah,
Host 1 43:22
Easy. So, there is one one thing I want to bring up about this, this extra board. So I've been designing this particular product, which now has 12345 separate PCBs that go into this product, but I'm designing them all in one schematic. So the way dip trace works is if you design one schematic and you go over to your PCB, it thinks that all of those parts is one board, right? And so I create all the boards and I put them in a array such that I can break them apart with mouse bites or V scores, or however I do it. The biggest issue shows up is like the way dip trace handles ground is it thinks, oh, all of these things are ground. Well, now I have separate boards. So the question is, have you ever seen an EDA tool that handles separate grounds? Well, because I've never seen one.
Host 2 44:23
We've talked about this before? The answer's no. Yeah. Right.
Host 3 44:27
And I'm just I'm curious to anyone out in the ether out there. If you've ever found one where you can be like, this is a separate board, but like I'm designing it together and I want it to have its own ground.
Host 2 44:39
Because what you were saying is basically because this goes back to the net thing. You can just call it ground one and ground two and ground three. Yeah. Because it doesn't matter. I mean to the EDA tool. Ground isn't a thing. Ground is just g and d as a string that's attached to this net. Right, right. Yeah, it's brain it doesn't matter. Right.
Host 1 45:01
I wish it did.
Host 2 45:03
Ah, see what you're saying,
Host 1 45:04
I wish I wish it knew that these were the same. Like, I wish that new ground on this board and ground on that board is the same, but they don't have to be connected. But they're the same thing.
Host 2 45:17
So you wouldn't need to have a virtual name. And then you just thought you'd have to have two layers of names basically.
Host 1 45:26
Yeah, effectively. And I've added, I'm adding complexity to this, because on one of my JST connectors, I'm sending ground to the separate board. But I'm also sending a chassis connection to this separate board. And that chassis connection doesn't connect to anything, but it's connected to the chassis. But ground is also connected to the chassis. And so I'm adding lots of confusion in my EDA tool. And it goes back to what I was talking about earlier is like, how do you make the schematic easy to read? The schematic should be the easiest thing to read. But I EDA tools do not seem well equipped to handle what I'm going for here. I want separate grounds that are actually the same ground. And I want chassis ground that's connected at one spot. Like I'm adding all the confusion in one project here, you know, yeah, it
Host 2 46:15
Might be you. How I would do it. Is I would call all those grounds separate names. Yeah. So like you like you were doing with analog ground Digital Ground. Right. So you would do that. So you'd have a chassis ground? And then I would make a part. Right? That is like an SMD. Pad. That's it. Yeah. And then you could connect multiple pins to that part. And it would effectively be you know, that's like your star ground. Right.
Host 1 46:45
Right. Right. Yeah. And that's sort of how I've already done it. Like, and because like that, because that's all my only option.
Host 2 46:54
Yeah, but on the schematic of easy to read. That's the
Host 1 46:57
Good. Yeah. So So I for the board for any separate board, other than what I'm calling the main one. The main one has the the official thing called G N D on it. Yeah. And then all the other boards have the name of the board underscore GN D. So like they're still met, but the EDA tool doesn't think that they are exactly connected. Yeah. But at some point, you have to be smarter than the EDA tool.
Host 2 47:26
I mean, I hope so. I don't
Host 1 47:28
Know most of the time I'm thanking the EDA tool for saving my bacon, you know? Because if you if you left this stuff up to me, it wouldn't work.
Host 2 47:39
Alright, one last topic for today. Yeah, we'll see how long this one goes. Because we're already at 47 minutes, we don't have a problem talking. Now we don't have a problem talking is. So dx F drawings to board outlines, okay. Generally, this isn't a big deal. But I've run into some DX F outputs from Adobe Illustrator, which is not a proper CAD tool. And so it draws things really weird in its exports. So like I was given a DX f from a customer. And I'm always trying to get the important to Eagle and Eagle was not having any of it, because it would import it and it'd be like missing half the half the stuff on it. So I haven't fully done this yet, because I got the file today. But the I was able to import the DSS successfully into Autodesk fusion. So I'm probably going to do is go into Fusion, and then export that as a PCB outline that eagle can ingest. But what else could someone try if you were to say not using eagle and Autodesk because I'm lucky that those those programs talk to each other? Right? But what else could you do?
Host 1 49:04
So I ran into this exact problem just the other day. And the reason I ran into it is because dip trace does not allow you to create donut pads. So a pad that is just a ring that has like a void in the middle. It has no problem making through hole pads but think of a doughnut SMD pad. So
Host 2 49:28
I had to trick you you want you want the annular ring but not the whole.
Host 1 49:32
Exactly, exactly. So so how do you how do you create that what I did was I actually created a DX F from CorelDRAW where I drew the donut, I brought that into Fusion 360 And then I use Fusion 360 To export a DX f from a DX F and then dip trace was fine with ingesting that. It's interesting because I've found that EDA tools tend to not be as happy with taking a A CAD from an art program. But if you take fusion 360 to re export CAD from a CAD file, that seems to work just fine.
Host 2 50:12
Because that's the same thing I'm having is I'm trying to import from an art package, dx F. So I bet you how our packages output arcs, because you're probably running into arcs issues, because that's because I can get straight lines and Eagle, but I can't get the arcs. So I'm going to bet you, however, aren't packages, define arcs is not happy in a normal engineering CAD tool. And Autodesk just has some magic under the hood that just couldn't do it. Right.
Host 1 50:44
Yeah, it gets both sides of both. Yeah. Actually, we had we had a customer that had an unbelievably insane board. It was actually not very big. It was maybe like three inches by two inches. So not particularly large. But the board. If you turned it sideways, it looked like how do I describe this, like, if you're standing in front of the Notre Dame Cathedral, it looked like the profile of the cathedral. And the silkscreen was like this ridiculous silkscreen file. And this and it's funny, because this board looks like a cathedral with like, the steeples and all the Gothic art on it and stuff like that. And it was a power supply PCB that connected to a different board like this, this was incredible. Like most most PCBs, if you look at them, they're rectangular, or maybe they have some angles on them. If you looked at the board outline, let's say an average PCB might have between, say, four and 20 vertices or points on them, right? This guy's PCB had like 800 points. It looked really, really awesome. But it also sucked because he gave us that board. And we had to put PCB rails on it such that it would fit on to our conveyor belt. And it was like, Oh, my God, I had to do it by hand. Like I had to go in and like edit points, everything. But it looks really cool in the end of the day. I think that was a dF dx F input.
Host 2 52:24
Yeah, I think that's gonna wrap up this episode. So I know we said last week, we're gonna try stream this one. This episode. There's a hurricane in Houston right now. So I could not Well, I guess I could have tried. We didn't really have any problems during this whole podcast, but I didn't really I've been having issues all day. So I really didn't want to tempt fate. But if next week is nice and clean, we did test it. There's a little audio glitch. We have to work out but it seemed to work fine. So yeah, we'll give it a shot next week. Hang
Host 1 52:58
Tight, we'll actually do it some time. Yes,
Host 2 53:02
This would have been a great episode for it too. Because how many like board design stuff we did?
Host 1 53:09
Well, please, six. Talk about that stuff in Slack. And know that if you put stuff in Slack, it might end up on the podcast because you put it on there and then I think about it all day long. No,
Host 2 53:20
It's not it. It's when it's likely
Host 1 53:23
Yeah, it's actually quite likely. So you might get called out on the podcast.
Host 2 53:30
So that was the Mac Feb engineering podcast. We're your hosts Parker dome and
Host 1 53:34
Steven Craig. Later everyone take it easy.
Transcribed by https://otter.ai