From A Current Perspective

Welcome to the macro fab engineering podcast. We are your host, Stephen Craig and Parker Dolman. And this is episode 115.

Okay, so before we start, we got all the events that are happening at macro fab. We have the Twitter chat. This Friday, April 13, at 1pm. Central. Whoo, Friday 13th. I'm actually going to be in Milwaukee doing it. So I'm going to the Midwest gaming classic. That's the pinball festival or involve video games, etc.

Just like everything. Yeah. So

if you're in Milwaukee this weekend, stop by MDC and, you know, tweet me at Longhorn engineer with no O's. And, yeah, I'll be there. I think it's also cipher con is like, right next door, what's that? It's some cipher con. I think it's like a codeine or like, crypto conference.

So like the real nerds go there. Sure. I'm looking it up real quick on my computer phone.

And so the next meetup is next week? Because we moved up from the last Wednesday of the month to the third Wednesday of the month. And I can't remember the 18th. I think

I'm looking it up currently, because I also do not Yeah, the 18th Wednesday the 18th.

Yeah, Wednesday, the 18th of this month, I will be giving a talk about the key parts of Arduino, which is basically we'll go over what what is an Arduino? The hardware? And then if you use it on your project, how do you take that hardware and port it over into your own design? What's important? What's not?

Oh, well, yeah, if you take all the different blocks, like which ones do you need to keep? Yeah, yeah. Because it can get a little bit confusing.

Yeah, especially since they have really weird power stuff on there. You don't need most of it for a single application. It's designed to be really flexible. So if you just take the flexibility out of it, then you lower your bill material cost

well, and and when they switched over to the what the 16 you for handling USB, correct. That is confusing, because it's not a drag and drop component anymore. Yeah, you have to program it, right. You got to program it. So it'd be cool to talk about like, okay, so if you need USB, what's the easier way of doing it? Correct? Yeah,

or least expensive, right. And then trade German will give a talk about taking off shelf enclosures, and customizing them silkscreening stuff like that. He was on trade. You might recognize his name. He was on the map episode 10.

So he's an OG and episode 3838. Yeah, where he

returned. So before we have him on that he'll have another return on the podcast, probably eventually, because it's been over 100 episodes since he's been on

episode 38. I think it was either, right? right before or right after the Icarus train after the Chris trophy, right that where he did like, he flew for like 10 bazillion miles. It has parasail or whatever. And around the world universal? Yeah. And I know he's done like 10 More since we recorded episode 38. So he'll probably come back on and talk about that again. Yeah,

that'd be pretty cool. Yeah. And then we have something new, we have the Houston hardware, happy hour, h four, we're gonna call it and that's the first Thursday of each month. So the first one is going to be May 3. And we are going to do it at a venue called slowpokes here in Houston. Because it's really close to the fab. And it is got a lot of open area. It's coffee and beer, coffee in beer and food. So it's something for everyone. Basically, I'll probably go there and get a beer and then drink coffee, while hopefully people bring their hacks and stuff so we can talk about their projects. Yeah, tell me a little bit less. That the hardware meetup here at the fab is more of like, engineering and like networking and a little bit educational, educational, where I'm hoping that the hardware Happy Hour is more fun and more relaxed and more of like show intelli.

Yeah, and like trading in a way like trading what you know. Yeah, yeah, yeah. And so that's the first Thursday of each month. What What time does that start? 6pm 6pm?

All right, cool. So Steven, you've been getting on to this week.

So I Gosh, Last Episode Two episodes ago, I can't even remember now. I was talking about some some voltage controlled filters. Last week. Last week. Yeah. Last week. And I wanted to week, we've been doing a burger. And I've been doing a lot of like extra stuff. So I'm all like, confused now. But yeah, last week, last week, so I wanted to talk about a chip that I've been using in those filters. It's actually really cool. It's called the LM 13 700.

Yeah, that's that dual operational transconductance amplifier with linearised diodes and buffers built in.

Oh look, someone did There's homework before I just copied pasted on the sheet.

Oh yeah, ti filter to

di there's, there's a handful of other guys. But yeah, Ti is kind of one of the main guys. And you can get it right from Mouser. So it's actually a really cool little widget. It's an OTA operational transconductance amplifier, which is not an op amp.

What does transconductance Ah, see, that's a really so if you break it down, trans is going across a country or across a long way?

Sure, well, yeah, it's it's the change, okay, of conductance. Okay, I conductance changes in a way. So this this little device is it looks. So it gets confusing, because it looks like an op amp, when you draw the symbol, it actually looks like an op amp, where if you took scissors and clipped off the nose of the op amp, so you kind of draw a trapezoid okay for it, but it still has like a plus and minus input. But it has a third input, which you actually inject a current not a voltage into that third input. And you basically have control over the gain of that block with that third input. So an op amp has to, you know, a positive or a negative input terminal.

So it's so that current inputs, kind of like a BJT, than what's a current controlled amplifier.

So actually, okay, so if you peel the hood of of an OTA, the front end of it is almost identical to the front end of an actual op amp, okay. However, the front end of an op amp is to, to transistors or to like employing elements in a differential pair. And the amount of current that flows through them helps dictate how much gain the whole system has. an OTA has that exact same input, but it allows you to, with another pin control how much current goes through the front end. So a regular op amp has, you know, gazillions of game it like, like a crappy op amp will have like 200,000 times gain something stupid like that. And an op amp just applies that game to the difference of its input terminals. Correct. So if you have one volt difference on an op amp, it's trying to make it be 200,000 volts on the output, which obviously won't work your rail, right. So it's an OTA and operational transconductance amplifier does the exact same thing. However, it spits out a current that is proportional to the change or the difference in its input terminals. And that third gain terminal will change the kind of the gain of the current. So transconductance is, is it's measured in units of MH O's, which is ohms spelled backwards, okay. And it's how much current versus how much voltage. So like a traditional voltage gain is how many volts per volt, this is how much milliamps per volt, okay, in this case is how transconductance is is handled. So you put a difference of voltage, say, put one volt difference in there, that might equal one milliamp on the output. Gotcha. So it doesn't work like an op amp, because you don't apply feedback in the same way. All the same things don't work in the same way. However, what's nice about it is that third pin that controls the gain, you can make all kinds of things that are variable with it, and control them easily with just a current. And that current can be generated easily by just putting a resistor on that pin. And then you apply a voltage, that resistor, and you basically get a current into that pin. So all said and done, you can make Voltage Controlled amplifiers, you can make Voltage Controlled filters, you can make devices that act like a variable resistor, you can actually even string two of them together and make a variable resistor where one side doesn't have to be tied to ground, it can be anywhere floating, it could be floating. So technically, you could actually make a virtual resistor that is in the feedback path of a regular op amp, and you could control the gain of the regular op amp. Regardless, this thing is just allows a lot of control. And what's cool about is on the output, all you do is you slap a capacitor on the output, and you have a low pass filter. Yep. And based on the value of the capacitor and a handful of other things, you can control the low pass filter where the cutoff is. And if you need higher order filters, just string these blocks together. Every time you add one of these blocks, you get another order another order. So I thought I just might showcase that that little chip. It's confusing at first, because most most of the time you don't learn about an OTA you know. So actually, just as a quick little side note, an OTA is sort of like half of an op amp. It's like an op amp that doesn't have the voltage output tacked on to it. Like they cut the op amp. Before it gets to that thing out the nose off, they cut the nose off of it and just give you voltage. So I've been using those to make a handful of filters. They work really great. I was you I'm actually able to make some variable all pass filters, which I don't know if you've ever used an all pass filter, all pass filters are pretty clear

ohm resistor.

Yeah, but it's but here's the thing about an allpass filter. That's, that's weird. It doesn't attenuate your signal at all. What it does is it just shifts the phase. So if you need a signal, where, like, say high frequencies are delayed, but low frequencies or not, you can, you can do that with an allpass filter, but you can shift it also. So consider this, if you have like a low pass filter, you know, and it's cut off frequency, it's supposed to have 45 degrees of phase shift, yeah, and further out, or to go all the way to 90 degrees. Well, if you want to do like algebraic adding or subtracting to that signal with another signal, the phase has to match. So you can take a low pass filter, and an all pass filter, pass your signal through them, and then you can do any kind of algebra you want. Because at every frequency point, they will be in phase with each other, and you'll get real results. So this is what US analog guys mess around with.

What keeps you up at night? Excellent. Yeah,

I like this stuff, I think phases weird phases is like, because a lot of times, you know, when you look at filters online and stuff, you just see a magnitude response. You're like, oh, this was great, I can reduce this, this blah, blah, blah, yeah. And that works. Until you forget about phase, and then your your whole circuit goes crazy. And at the same time, like if you're doing you know filtering, you do have to worry eventually about it have you know, phasing to the point where it's 360 degrees out of phase, and then you start oscillating somewhere else in your circuit, and you don't you didn't realize it. So you kinda have to pay attention to that. Regardless, the LM 13 700 is an

OTA. So what's the best way for people to learn more about OTAs and stuff?

You know? So I think the funny thing is OTAs are so kind of funky, and you're not taught them in school, unless you take like a really high level class, go download the datasheet for the Texas Instrument, lm 13 700. It is just it has pages and pages of example, circuits of things you can build with it. And I think they put all those examples in it, because they're like, nobody knows what this thing does. Yeah, nobody would know how to actually do it. Because it's funky.

Some gray beard wrote all those examples. Yeah, one of them over the years. And the thing

is like, so Okay, great example, let's say you had a left and right channel stereo, and you wanted to control the, the volume on them from a microcontroller, and a really cheap way. That's actually not as easy as you think. But with one lm 13 700, it has to have these OTAs inside of it, you can control both left and right channel at the same time, and your they'll track very well. And you could just put a PWM signal right into that control pin, you know, with a little bit of filtering, and you have, you know, you have left and right channel volume control from a microcontroller a for like a couple, not even a couple probably like $2 worth of parts maybe. So, you know, that kind of stuff is really, really, it's useful for that. But the thing about it is like most of us engineers don't really think in terms of current. We think that current is not byproduct. Yeah, current feels like a byproduct. Yeah, exactly. But if you have a device where it they don't tell you what the output voltage is, they only tell you what the output current is. And then you put a capacitor on the output. Like, think about that. It you kind of short circuit in your head, you're like, oh, wait, how do I calculate this? Yeah, you have to go back to your equations for a capacitor. Yeah. And if you remember the s domain and doing Laplace transforms, do a Laplace exactly neither, neither did I, before I start playing with this. But again, no, if you look at a Laplace transform, it'll give you all the information. Because it doesn't make sense. Like, if, if I were to tell you like hook a inductor to a current source, like, you'd have to sit back and be like, oh, gosh, what I do, but if I, if I said, you know, I put five volts across an inductor, you'd be like, Okay, I know what it does. I know it does. Yeah, yeah. So you kind of have to rearrange your mind. But it's a good exercise. That'd be interesting. Class.

Electrical Engineering from a current perspective,

which is funny, because it's actually all supposed to be from a current person. Yeah. Because current amperage is the unit that is like the primary defined unit, but it's the one that we're like, that's kind of the bad one, you know, like, we put five volts across the resistor, therefore current flows. No, you have to think you're supposed to think of it as Oh, I'm flowing current through a resistive element. Therefore, a potential develops. Yeah, that's the way you're supposed to be taught. But, and it's different in analog and digital because digital guys, it's zero and one and those zero ones are not current, their voltage. Current is sort of your enemy in digital circuits. You don't want things to flow. Yeah. So

yeah, well, that's why everything's like high impedance input or or, you know, high Z and stuff like

that. Everything. Everything behaves nicely when you do that. But yeah, when when you start getting into the need for things like Voltage Controlled filters, you have to start thinking in other terms.

Yep. So cool. Yeah, that's that. And I heard you fix something.

I did a actually I fixed something earlier today and it was a little bit scary. Scary just because I've never done it. And I've been told I shouldn't do it. But I did it. I actually recharge the refrigerant in my

old 10 fingers. I do have all the fingers and all the hairs on my

head and like I'm not a weird green color or anything like that. Although I could have been you never know. Like so. So yeah, recharge, recharge the refrigerant in my kegerator. I've had a broken Kegerator for years now. And it was just out of Freon, right. Yeah. But like I went through like all the steps to try to figure out because I because everyone always told me you can't do that. And I even people at my local hardware store were like you have to have a license in order to even buy Freon? Well, apparently you don't. If you go to amazon.com They don't care. They'll say anything. Once you buy like our 34 A, it's 134 dash a Yeah, that's let's like that. Usually what goes in cars and

refrigerate, you can just buy 130 for a at Autozone.

Which, which is interesting because I went to Home Depot and I went to a local hardware store and both of them were like, Do you have a license? Like no, I just probably just tell it to everyone. Well, they probably looked at me and they're like this, this guy has no clue what's

out in the parking lot and just stab it with a knife.

It's so hot. We just pour our 134 on us. Yeah, just to cool off. Yeah, no, no. So like, I had to go by like a needle piercing valve and

get back down because that was like hard to get to Amazon.

Actually, what's funny is okay, so at hardware stores, you can find refrigerant, but you can't find the needle piercing valves. And here's what's so goofy about it is Amazon has the needle piercing valve as an add on item.

You know, like buy that Freon stuff.

It's like was add on. You can buy anything as long as it's $25 you can get a needle valve for like three bucks. And I don't know like so apparently enough people buy them that Amazon's like this is a sponsored item or whatnot

Amazon's choice for destroying the ozone.

That's right. Yeah. freons pretty nasty stuff is pretty toxic.

I think our 134 A is it is not that bad. I think it's

one of his lower on the bad scale.

Yeah, one of the ones that used before are 12 I think one of them's just straight up propane. I think that's our 12 Regardless, or 134.

A I recall is acetone also, I think

yeah, well you want anything that you can if you design your system, right for compressive heat management, but you're pumping, right? You just want something that phase changes really well. Right, which any kind of ethanol or any kind of highly backward

design circuit a system around and

one of the some of the original ones used system a Okay, is that is that acetone? But

any freeze? No,

no, no, no. Why did I can't remember. Oh, come up later.

Not acetone. Okay. No.

Oh, using the space station right now for the same thing. It's like one of the best ones.

Yeah, but it can be corrosive, right. Yes, it is. Well, I mean, depending on the situation, it'll be corrosive. Yeah. Yeah. I know what you're talking about. I can't it does not come into mind. But regardless, I did read the can of r1 34. A and it it had a lot of precautions about not breathing it. And it was very bleak. And it said things like you will die. Touches can, which is funny and my listeners are gonna, they're all gonna laugh at this but so I tightened everything down really well, but it was still like, I don't know if I've got it good enough. So I opened the can up a little bit. There's a valve at the top of the can to my feed hose. And the pressure went up, but I heard a little bit of a hissing sound like so it's like is that the hissing of the Freon going into the compressor? Or is that something else? What did you find it you were looking at now

are 290 is it's just straight up propane.

Okay. 297 Okay. But yeah, so I did hear a little bit of hissing and I corrected everything and tighten everything ammonia, ammonia, that's it. Yeah. Yeah. In fact, um, so the in Dubai, there is a indoor skiing in Dubai, which is like that's one of the hottest places on earth. And they created like a place that will snow indoors and you can like take a ski lift up and go skiing in Dubai. There's a lot of money there. But regardless, the entire like, arena that it's in is blanketed in ammonia. And that's how they that's how they keep bit cold and they and it will snow in Dubai inside this like tank, look it up. It's super dense crazy. There's a whole documentary about it. And yeah, they're actually the documentary is really great because they go through like the inspection of the welds. And apparently they got like 75% of the way done. And whoever was contracted through the welding did a horrible job. So they had to go back and re weld everything, which that's because, you know, even out in the middle of the desert, you have an ammonia leak. That's a bad thing. Yeah, you don't want that happen.

How much ammonia we need to cool down entire building, like a ginormous warehouse that you can ski in.

Okay, so this this, this whole ski arena looks like kind of, it looks like a giant radiator. That's like tipped up at an angle. So like, like I said, the whole thing is surrounded in ammonia tubing. And things. Look it up. It's weird. Okay, so. So yeah, that's, that's what I've done. I've analog filters and free on free. Do you been up to?

So I've been, you know, working more on knowledge base kind of got that thing all worked out by people bored of me talking about that. So I'm in kind of working on the next set of articles, because I'm still waiting for the deck stuff to show up. That should be here next two weeks. I think give me a week to write article and then that will be all wrapped up. So that'd be cool. But the next one I'm going to work on is USB type C articles. I did we were talking about this a couple years ago, but it kind

of sorta right when USB type C was coming out?

Yeah. But so I'm gonna actually write those articles now, basically, because now that's kind of like my new position here at Makram is content writing. So I'm like, Cool. Now I can actually get through this entire, you know, three year backlog. You know, Article what

I like about USB type C now, it's, it's still well, and I think it will be for a while. It's like newfangled, in a way. So if you go look at I was at Fry's the other day, and I was looking at something. And on the box, it had like one of those big

like now seeing on TV with USB type C and

one of those big bubbles that has all the spikes over and it's like now with USB type C like, and I know in a couple years, that's gonna be like old hat, but still, like, it's one of those things where it's like, ooh, it's got C on it. Sorry, go ahead, you're gonna write an article.

So the article is gonna be about converting your USB 2.0 projects that have like micro USB, or whatever to type C. Yeah.

Are you going to reference that footprint that you made for? Yep. Cool.

Yeah. So it's going to be that kind of stuff is I'm going to bring that up in, probably do a exercise in, you know, here's a design that you can download that has type has Type B micro on it, and we're going to convert it to a Type C, and what extra stuff you need or stuff like that, what considerations you need.

Do you have a kind of like a chip in mind that you want to use for that?

You can pretty much use any USB 2.0? Chip? Yeah. So we'll probably just use like an ft 230. X or something.

Okay, so you're not going to necessarily include the power handling side of

it? No, no, this is just if you want to implement your old project. Okay. Yeah. Yes. Be type. See. So you at least can use the new cable. Yeah. Power delivery stuff is a completely different beast, and one that I've only read like, tech docs on not actually done. And that's probably more than an article, correct? Yeah, that's gonna take like, I need to read stuff for like a month and actually start experimenting and stuff.

I want to see 100 Watts go through that. That cable that

cable? Yeah, you got that you have a special cable for it. A special see? No, it's a special power delivery cable that's actually got circuitry so that the device, or the two devices know that cable can support 100 watts.

Ah, yeah. So how do you prevent? How do you prevent both not trying to deliver 100 Watts to each other? They talk to

each other first before turning stuff on. Okay. Yeah. Yeah. So that's why like, if your phone is completely dead, any charge be plugged in type C into it, it might take a while before it starts fast charging, because that's a power up enough to so it can actually talk back.

Gotcha. So yeah, it would charge over the lower voltage or the lower current. Yeah, just spring the life.

Yes, spring life. Cool. And then I started looking at my website, because I'm kind of like in that weird state where like, I have a lot of projects that haven't finished over like the past, like, five years, you know, and so so looking at my website and figuring out what one like what kind of personal project Can I finish, or start working towards and so I'm going to do that Gameboy VGA finish that up, because I got the proof of concept. We actually did the proof of concept for The Ben Heck Show when we built this ginormous seven to one scale. So it's like, you know, three feet tall Gameboy or whatever. four foot tall game board actually. Yeah. And that's, yeah, that's Yeah, the thing is that's running like a really early version of the code. That's like hard coded for that reason. pollution and all that stuff. And so what I want to do is take it. So first I'm gonna, I'm gonna design a board, because right now it's a dev board. So it's like in wires everywhere. So I want to take that designer board that can go into the Gameboy, which is a VGA port hanging out of it. And then start developing kind of like an a graphics engine, not really engine, but like a configurable driver, configurable driver, so you can select what resolution you want to output, that kind of stuff. So you can output to, you know, 320 by 200 resolution, or you can go all the way up to 16. But 1600 by 1200, PJ, which is like the highest.

How are you going to make that selectable? Is it going to be on the PCB? Or is it going to be software? And so, yeah,

what's gonna do is, when you turn it on, I'm just gonna make it so if you hold a and b, at the same time, when you turn on, instead of reading from the Gameboy data port, it will just go into the, they'll just have a menu, and you can just select which one, oh,

it'll save that to memory, and then you just reset and it goes back to the processor.

Yeah, yeah. Okay, so be able to sniff the button presses and stuff. Yeah. I'm actually I'm thinking about having the button presses go into the FPGA and then pass through, because that will only take like a couple nanoseconds to happen. But that way, I can actually interrupt it, or put my own signal in to the Gameboy. So like, if I want to run a external like NAS controller. I can do that with that.

Okay, Okay, gotcha. You can break the path and put whatever you want in Yeah, Xbox on a Gameboy

if you wanted to. Yeah, that's cool. But then I eventually want to do HDMI, but I got to learn HDMI. So that's,

well, you've done it on the pin. Heck, right. Well, you didn't, I guess,

yeah, that's kind of implemented on the Raspberry Pi. And I just had to route that I didn't have to actually build a driver and spit out the signals correctly. Yeah. So you got to go learn all the standards. And there's actually some really good HDMI stuff on FPGAs are fun. I think it's are fun. Anyways, I can post link, it has like all the stuff like how to like demystifies, HDMI, and how it works. And they have like a, like a dev board that you can get the hazards like an HDMI port and FPGA on it. And you can walk through the code and make it work and stuff. Like

just spit stuff to the screen. Yeah,

yeah, that's, I'm gonna try that.

Because Because HDMI, it's it's both like a hardware thing. But it's also like a handful, a handful of protocols that go

yeah, you basically are setting up data streams, right? It's almost like a spy. It's kind of like, a more like i square C, where you serialize a whole bunch of data at once, within like four channels.

What's the data rate on HD? Fast? It's really fast, because it's spinning both audio and video. Yep. At high resolution, so it's got to be ripping, but it has a ton of lines, right? It's not like one data line that shows us it's four channel 14. Okay, yeah.

What are some communication? other random stuff clocks? And, yeah. Yeah, so yeah, that's where I'm gonna start working on so hopefully next week, my at least my goal for next podcast is I want to get all my old crap out, compile my old project and see, first of all compiles still, oh, geez, you should video that, see how many errors it pops up for some, yeah, get that running, get it running on the old hardware. I hopefully actually, hopefully, I get that by tonight. That's my goal. And then start routing the board. Now the first draft of the board, so I'm going to pop on like a cyclone for FPGA cuz I already have that design Eagle have all the support stuff for that already designs, I actually cannot be available yet. And that's part still available. It's only like 10 bucks. Put in a decent because it currently runs off of a, a just a voltage divider bridge to build the VGA. And like, I'll just put it in a proper DAC and just drive the DAC you know spy so that's actually gonna take some time to figure out how to do that. Yeah, cuz I've never done that before and then actually implementing the will the first part is going to be basically abstracting all the math out of the the code to build all the timing signals based off of a chart now because if you select one resolution, it's got whole different timings for each one. And so you have to put all those in and then make the bases state machine work that way. Wow. Because originally it was only designed for like 800 by 600 resolution.

You know, you know what job I would hate to be the guy who has to like draw all of the timing diagrams for data sheets, data sheets, you know all the ones where like, it has like all the definitions of like, this falling rate you get to Yeah, like THC. Like every single little thing. Like, Oh, that sounds terrible. Like, yeah, but the person who does it loves it. Probably yeah, his life lives between like two lines like zero and high. And like all of those little nuances in between there. Because like, I've seen some data sheets where they'll devote an entire page, top to bottom, where like, up at the top is your clock, and then all your signals are like, oh, yeah, sounds like teach their own.

Cool, so yeah, hopefully next week, I have an update on that. And we're gonna talk about next week, then, I guess. More filters, right.

Ah, we'll see. We will see there's a lot of things I filters and cold beer because I've got a kegerator now. I don't know what we'll see.

Alright, so on to the RFO. Rapid Fire opinion. Hmm. So Amazon is doing in home delivery now.

What they deliver from room to room in your house? Is the guy who walks in, here's your pizza. Alexa,

bring me my pizza, or give me a cold beer from the fridge.

guy who lives in your fridge?

No, no, no, no. So when instead of leaving the package on your front porch, they can access your house. Well, that's creepy and put the package inside. So you have to give delivery instructions like exactly. Well. So what it is, is a lot of these smart locks. They work off like I guess nearfield you know, NFC communication, possibly or RFID. And so the Amazon livery people will just unlock your house with that code. And then they put the package inside includes the door.

Let fewer stolen packages.

Yeah. I think column porch pirates porch

says I really? Oh, that's great. I haven't heard that before.

I think this is like the evolution of their delivery service. Because it makes sense. I think it's a little creepy, though. A lot creepy. I'd rather than just like put the package on my back porch.

How do you sign up for that?

I think you just opt in.

But like opting in requires like, some extra information, you

have to have a that they support, like eight locks different manufacturers. And so you had to say I had this lock, and then and then the code and whatever it is. Uh huh. That's, I don't know what kind of code you get for a Smart Lock. Because a lot of them don't even have like, they just have like a box on the front of your house. There's no like, keypad or anything. It's just a flat box. Right. And so I guess it's all wireless.

So I like working with if you opt into that. And then let's say I don't know, something happens and it doesn't work. Do they leave it on your front porch? Or do they take the package back with them? Or we leave on the front porch? So it just defaults to the normal way? Yeah. It's um, it's creepy. I don't think I'm going to be signing up for that anytime soon. But you know, they probably have some kind of like, I don't know, Star Trek scanner thing that the UPS guy carries around. And it is a phone. Yeah, it's just a phone. But it but it can, I guess it can activate whatever it can configure itself. Yeah, that's cool.

So the next one is the FTC says warranty void if sticker removed is not valid. This one is crazy. So basically, the FCC is saying people can repair their products, which is really cool. companies can't put repair restrictions basically on their products unless they provide parts and services for free. So if you charge for parts and charge to fix stuff, like if you break your screen or whatever. A third party, fixer, I guess, is a repair person or repair company isn't allowed to fix your phone without word bunkin your warranty? Basically, right? I'm kind of like in the middle of this, because I'm not sure it's really good to be able to fix your stuff without the warranty. But it's like, it's that mechanics old mechanics thing where it's like, it's $100 an hour if you let me fix it $200 an hour if you if I let you watch me while I fix it. And it's $300 an hour if you try to fix it before I got to it.

Right. And it's that and it's $400 an hour if you talk while I fix it. But what I'm most curious about on this, okay, great FTC, the Federal Trade Commission says that that's not valid. So what the manufacturer just goes, I don't care. It's still valid. Well, they can actually. But they can prosecute them. Yeah, they get prosecuted on them and levy fines. Wow. Yeah.

So this is a great thing. It's just one of those. They create this gray area like because it says for third party. Are you a third party fixing your laptop?

No, clinically, yeah, no, because your end user? Ah, well, you're not third party. How would you be third party? If you handed it to me, then? Well, you would be a third party because you're not the end user. You're not the end. You're not the person who purchased it for the use of the product. So So, in terms of that, technically, yes, although I'm sure there's all kinds of like mumbo jumbo that's in all the documentation that says that you have to be certified or licensed or whatever, or known as a repair technician. Yeah,

it's one of those weird areas where like, let's say, your laptop, your USB port blows up, because you plugged something in that was bad, right? But it only blew up like the front end. So like the power delivery part. So it like blew the poly fuse the kingdom come, okay. And you go in there, and you wire a jumper across the poly fuse, and basically pumping five volts to your port, so it fixes the port. Okay. comes later you plug in another bunk device that basically now entirely blows up your laptop, because your five volt lines not protected? Should the manufacturer be liable for that now?

Of course not.

Technically, you didn't void the warranty?

You you will know I think because you if you did not apply a fix that returns it to the original operating functionality of the device. So the answer would probably be no in that case, especially because in your example, you got rid of

safety. So bodgers Don't count. Yeah, aluminum foil

over a fuse is not a fix

or putting a 22 LR bullet. And you know what, finally?

Yes, that's a showcase next,

yeah. So yeah, so if you had asked me, I guess, there has to be some, you know, boundaries on what constitutes fixing it. Ah, I I've seen some pretty nasty fixes them like I've done a handful.

I mean, you fix the decorator? Yeah, yeah. But yeah. Which technically, I think I wasn't supposed to be allowed to do. But I did it. And I did it in a way that like, that's the way a normal technician would do it. It's just, you know, I didn't know how to until I actually just did it. But But I think what they're getting at with this is talking about the hole. Do you have to have it repaired by that company? Or can it just be repaired by anyone else? Correct. Yeah. And, you know, maybe this is also

well, okay, I know where they're getting that. It's just one of those it starts turning into when you start going to what I just said, because technically, you didn't void the warranty. Unless you say, you have to get it repaired by a official third party or you have to get it repaired to original specifications. Yeah, then I could see that being fine.

Well, I'm sure there's plenty of legal terms about modifications. And if you do anything outside of,

oh, my jeep has no warranty on it now. Oh, zero years?

Yeah, nobody would warranty that thing. I mean, it's well done. It's just nobody would warranty because it's so, so different. But at the same time, this has to do with the sticker, you know, and so if you remove that stickers, you're going to void according to the company's Yes. But according to the FTC, I guess no, no. Yeah.

So the warranty it no matter what's happened inside of it, which is why I'm like, Yeah,

seems like an overreach of power to me. You know, maybe, because I mean, that that means that means you could pour, pour Coke in your laptop, you know, mess things up, royally, take off the sticker and still hand it to them be like you, you owe me a new one. You know, like, that seems a little nasty. I wonder why they chose to do this.

Well, it's actually has always been like this. Now they're actually starting to slow

down. Okay. Now, one of the things and I actually did some research when I was manufacturing amps as my main job. I put a sticker on the amps. And I put a big warning on there. And that was not because I didn't want people to modify them. I could actually care less if someone wanted to, but there's lethal voltages inside. So I use that as an indicator to see if someone had been inside of one of these amps. And to protect

weighing. Do you need to be even more careful when you open up that panel?

Right, because technically, you don't know what's in there. I mean, I inside of the laptop, there's probably not a couple 100 volts but there could be for whatever gonna be a ferocious ferret in there. So you know, you're right. You're right. Yeah, you don't know.

amps like rattling?

Yeah, yeah. I have had a cockroach in an amp explode before I can see. It shorted the high voltage. So yes, you you have no idea. And so I don't know. So a lot of times those warranty stickers mean more than just that.

So we have a last topic. I found this on Reddit and it was a person asking about avoiding cite entry into SMT pads. And why you should avoid that because there's not a lot of stuff on this. And especially when you Got a lot of open source projects like traces just go wherever into pads and stuff usually brought up some samples from like SparkFun and Adafruit and stuff like that, where like they do it. Why is this bad practice?

So side entry, as in a trace comes into the side and has to make an immediate 90 degree to go into the component?

Well, not just that, but like, let's say you have a, a Oh, six or three resistor, and you have the instead of coming, you know, upwards and down, it's kind of hard to explain without drawing, but instead of coming directly in from the top, from north south, let's say the part was zero degrees. From top and bottom, you came at the side, right, right. Yeah. And the main thing to worry about that stuff is, is thermal pulling during reflow, is why you don't do it.

Right, because if you cause as a trace, expands and contracts in the oven, it will no dress on there. Well, no,

it pulls heat away, because that traces copper, so pull heat away from that side of the of the pad pad, and will actually rotate your part from under the the surface tension to the solder will actually rotate the part. So in low volume prototypes, largest parts, like over, oh, 603, probably not a big deal at all,

I was just about to say my gut feel is like on a big part with a lot of mass. Yeah, in comparison to the trace, it's probably not going to do much.

But when you start getting 204 to smaller stuff, and really when you start doing high density stuff, so you have parts right next to each other. So if one did tilt, it could short out to the part next to it and bridge, then you really have to start worrying about this kind of stuff, and bout kind of almost balancing the thermal load on your pads.

And that's really interesting, actually. So this kind of reminds me something I bet we've talked about this before. So if you're following those rules, let's say you had a string of, oh 402 caps together, you wouldn't, you wouldn't put a trace just straight across all the pads, if they were all if they already be connected together, you would make like little humps in between

where you would go basically you draw one that went all the way across and then you drop down,

right to them, right, you would drop down such that they all get pulled in the long axis as opposed to the shorter axis. Which is funny, because technically, that's the correct way to do any two adjacent pads. You see this a lot in real fine pitch components where, yeah, if you have two pads on on a microcontroller, say, and they're connected together, some people just put it right from pad to pad. And that's super annoying, it's actually better practice to go out of the pad, loop around and come back into the next pad, even though it seems ridiculous because you don't need

to. Because usually what happens in that case, let's say you have t q if P which is a point four millimeter pitch, yeah, and you have you put that trace in the middle, there's not gonna be any solder masks that will cover that either. And so what you'll get is will look like a short, even though they're supposed to be connected. When you start doing QC work, you'll see a short and basically be like, Oh, we have to remove the short and you know, causes quality issues and stuff like that. Right?

Yeah, it's, it's, it's a QC nightmare, because the only way they can tell if that's okay, is if they look at your design file. Yeah. And at the same time, even with your design files, that's still a failure, QC failure. So in order for a QC person to pass that they have to forfeit whatever rule that they're doing. And if you're getting something manufactured, a lot of times if you fail at QC, for some reason like that, they'll make you sign a piece of paper that says, I know this is a failure, and I'm still okay with accepting it. So it just comes up the mix.

Yeah. And it's not just the bridging either. Well, because that that traces not as solder mask on, it will actually get solder on it and bridge. So you're actually wicking. Paste away from your, your actual part that needs to be soldered. So you might have lower amount of solder on your foot, and what's the heel of your of your lead. And so, I mean, that's not good either.

Well, and it impacts if, if that part ever has to be reworked again, ever, you know, you want you want basically every pad on a component to have the same equal amount of solder such that if it's hit with hot air it all reflows at the same time, and you can pull a component off, but in that case, you'd have to hammer that one little piece a lot harder. And so let's say you had a DQ a fee that had a bunch of bridge legs on one side of the component, you'd have to hammer that one side of the component with a lot of hot air just to be able to lift it off. It's not good practice and that can damage the component and the board and right and the board. You can Yeah, you can lift traces and things. And what's great about it is the fix is free. It doesn't require anything more and just get in the habit of looping out and come Going back to the pad next to it. So, but it's cool to actually put something like a reasoning behind why because at first if you if you just look at electrically, it doesn't matter that it doesn't matter and actually

doing it that way. If you go like let's say we had a whole row for twos, and let's say that was a ground strip, that's actually better noise wise to do all the way across the Pats.

Yeah, you get like it's lower parasitics. Yeah.

But, yeah, it's not good for manufacturability. And so you gain a slight amount of parasitic conductance and an impedance

well, okay, but let's, let's put it this way, if your circuit is so sensitive, that it requires you to do that, then your design probably needs some updates. You know, like, if that is the linchpin on that makes your circuit work. You're in for a butterfly flaps, it messes up when someone farts in the other room. It stops working.

Oh, cool. So I'll post some pictures of like what we were talking about there, because there's not a lot of I haven't been able to find IPC document about this yet, either. Because I know this is a thing. So I'm reading a lot of, like manufacturer papers on buy. I haven't found an IPC document yet. Maybe I'm not searching for the right.

I know, I've seen the one that I was talking about, about jumping to pads on a Goldwing component. I know I've seen that before, but I don't I don't remember what number it is. There's so many. They have like over 100 documents that are just your standard operating documents. Yeah, you know, if you were to get like every document, it would be a whole bookshelf full of standards. It's ridiculous.

So cool. We'll wrap up this episode of podcast talking about standards. Yeah, so that was yeah, we need a standard way to end these shows.

How about how about we do this? That was the engineering macro fab engineering podcast. We were your host Steven Greg and Parker

Dohmen. Cater easy Thank you. Yes, you are listener for downloading our show if you have a cool idea, project or topic, or if you found that IPC document that we are trying to think of Tweet us at macro fab on Twitter, or email us at podcast at macro comm if you're not subscribed to the podcast yet, click that subscribe button. That way you get the latest net episode right when it releases and please review us wherever you listen as it helps the show stay visible and helps new listeners find us