Are you ready to Tango?

Welcome to the macro crab engineering podcast. We are your hosts, Parker, Dolman. And Steven Craig. This is episode 156. So to kick it off this week, that wagon train module, I actually plugged it into the wagon over the weekend and blown behold, it works.

Yay. Like no issues just fires right up.

Yeah, it just beeps like crazy now. Now it's like, I should just remove it because it's annoying.

I'm just about to say like, Wait, why did you I still haven't,

you know, broken down and made a schematic for it yet. But at least it like works inside there now. So I'm gonna pull it out and draw up a schematic for it. And I kind of want to know how it's because it's got some digital logic in there. That's probably playing. Like if these switches are high. And these are low. Play this chime. That's probably what is using those those a NOR gate

shirt. You sure it's not just like an analog implementation that just is all like timing and stuff.

Has that but I'm talking about like, what circuits for the oscillator to trigger?

You but it but doesn't it play the exact same pitch every time?

No, it has that it has that continuous tone, and then it has the bong, bong bong they're the same pitch. So there's probably one oscillator that is this, that's the output of the speaker. That's one continuous tone. And then that has like an enable line that's also on a trigger, either continuously on or it goes up and down.

I you know, with the age of that thing, I would not be surprised if it's that's an analog implementation, where it doesn't have a digital thing going on. If it has like a VCA and voltage controlled amplifier, and it's it's just envelopes and based off of what pins are high or low. I mean, it's it's, in other words, I'm thinking like an analog implementation of a digital thing. In other words, not like a processor that's making decisions.

Yeah. So I'll have to print it out. Today. So or after this podcast, so cool. Yeah. And then I've been working. I don't know if we talked about this last time, I think I had it on my list. And we ended up skipping it. It's the USB type C article I've been working on. Yeah. You mentioned that in the Slack channel last week. Yeah, the I have an implementation for doing USB 2.0 with a Type C connector to like an ft 230 x, which is basically, you know, you are a USB UART bridge. And so I finished that implementation. And I created a design blocking Eagle, so anyone can kind of just grab it and just plop it down. And it has everything routed out already to. It's like It's like the maker block for USB. Yes. Yeah. So it's it's a building Lego block, you can say for a USB interface for to talk to a microcontroller. The only thing I don't know I want to do yet is if I want to put ESD protection on the Configure pins, the CC pins because they're both tied together at the connector, and then those go through a 5.1k resistor to ground. And so that's, that's like a lot of resistance there. It's like, does it need a ESD? TVMs.

I mean, what what for though?

Well, an ESD event would hit the connector, it could hit that CC pin. And then now you've got only 5.1 kilo ohms to ground impeding you at that point.

So so if there's like a 10 kilovolt spike or something like

that, yeah, is there enough? I don't know if there's enough energy and an ESD event to like, basically, shunt through that resistor and you know, effect your ground plane basically.

Got to go back to calculating jewels. Yeah,

I guess I need to figure that out. If someone in Slack knows, let me know.

I mean, it wouldn't it can't hurt, right. It just can't hurt adds bomb cost. And so

yeah, it's adds a little bit of tiny bomb cost. And I can't use like this is like my idea for this article is like I have a USB 2.0 design that uses a micro or mini USB connector. I want to go to type C what is the minimal I need to do to do it. Now? Well, if you say minimal

then don't so but you do have ESD protection on the data lines, right?

Yes, there is there's ESD protection there. I've what Chip is that actually?

Is it is it one of those

six pin jobbies Yeah, it's a little six pin guy that actually the cool thing is it has a clamp on the voltage line. It's not as low as I'd like but at least it prevents a mostly catastrophic, you know, voltage event overvoltage event. Sure. I am looking I don't have it ready right now.

You know, I think I've even implemented back in my macro fat days, I used your design block. It wasn't a design block at the time, you just kind of gave me your USB circuit and I just implemented in in a handful of things. It works great. It was nice because it was just like, okay, someone's done the work for me drop it in

the ESD part I like to use for USB 2.0 is IP 4234 cz six comma 125. I think the common Wente five is like the package designation. But yeah. That That part's gotten me through lots of CE and FCC testing.

Oh, yeah. Just slap that on the front. Actually, you know, that's a that's an interesting thing for your design. You Your does will this design pass?

Sure. What else is connected to you? Right? Well, yeah, yeah, but

this like design block? Is it is it? You know, it's, I mean, obviously, you probably designed it with that in mind,

correct? Yeah, it has the ESD protection and the clamps on ground and the power rail coming in. That's the only thing is like, is a 5.1 kilo ohm resistor one robust enough and to is it going to snug enough of the current off of like an ESD event if it hit that pin I might just have to put a standalone you know little TBS ESD resistor, diode there. We'll see if anyone argues for it on Slack. And the cool thing with this this layout I did I put like test points everywhere. So you can actually like test stuff. That's cool. You can easily delete them. Basically just delete them in your off the schematic and it will remove them from the the layout,

it'll automatically move from them from the design block.

Well, yeah, when you when you drop it into your your design.

So when you actually click the Delete key on the schematic side, does it disappear on the PCB side? Yes. Oh, that's cool. Yeah. I mean, that sounds dangerous. But that's cool.

I can be the best is when you like, accidentally delete something and put it back in. If you put a new part back in there, it's gonna be in a completely different spot on your board now.

Yeah, that's true. Yeah. So it's funny, I was trolling the or lurking on the Detroit forums earlier, and I was looking at some feature requests. And one of the things that dip trace does when you when you're done with your schematic, you have your schematic like exactly perfect, all your package definitions and everything and you're ready to hop over the wall and go to layout. You can, there's just a keystroke, and it takes you right into the layout and load of the program. And what it does is it just gives you a blank screen, and it just craps out all your components. And that's what Eagle does, yeah. And it just, it just literally spits them everywhere. And the thing is that one of the fundamental differences between dip trace and Eagle is dip eagle has the to the schematic and the layout linked, whereas dip trace, they're not linked you, you link. So you create

a netlist and that barfs into your layout tool,

right. And then the schematic is is nothing after that, well, I shouldn't say nothing but but if you ever wanted to link them again, it's a manual operation that you have to do every time. So you can get them out of sync if you choose to. Or you can bring them back into sync if you choose to. But so the thing is, like if you select something in your schematic, it will not select in your layout. Gotcha. So the thing is like, and one of the guys on the forum, what he was bringing up as a kind of an interesting thing, he makes these really large scale designs with that have multiple pages. And what he wants to do is be able to say like, I want to select all the components on a single page in my schematic and group those and have them somewhere on the on the layout. And it was funny because this is so I don't know, such a hacky workaround, somebody suggested that he has like a safe copy of his schematic. And what he could do is just delete everything but the page he wants on his schematic and then load that into his PCB. Oh, and then we'll just dump it and then it's only those components and then lock those components and then undo changes on the schematic side and do that page by page. Bring in just those components and lock them and I don't know that sounds like such a terrible way

that sounds really a terrible workflow.

Yeah, it's it's awful.

So I actually just came up with a cool idea. Because that's what when you when you cook, make PCB When you go to the board editor and Eagle it just barfs all the parts tries to like organize it organizes them by when you put them on the schematic. Okay, and so like the first part you put on schematic is going to be like at 00 location and it just goes straight down. Right? Oh, geez has this weird like, wait,

wait, what if you have like 3000 components, it'll just be a line of 3000.

Couple of more or less.

That's, that's terrible. And

I actually thought it was really cool idea is what if you, you could do that operation, but then it put all the parts how they are in your schematic. Zoom, make it make it look like us, given the relative positions? Yeah, the relative positions are how they showed on the schematic. Now, that doesn't help you like in routing or anything, but at least like, because I like to have the schematic open on one page as I'm doing the routing on the other page. And so I can like, Okay, I'm this this op amp need, R 13, R eight and C 14. I go ahead and group all those. That's the first thing I do is like, group all the parts together, like Yeah, that's what I need to be together, these bypass capacitors, go to these parts. And be nice if, when eagle will just automatically would do that. Yeah, yeah. Or any EDA tool would automatically do that.

Yeah. Okay. So let me ask you this. This is this is something that I'm actually really anal about. And I like doing it. Let's say I put a bypass cap on an op amp. So the way I designed op amps, and this is the default in dip trace, you have the you have the schematic symbol for the op amp, however, many op amps are in a package. And then there's, there's another schematic symbol that is power, v plus and V minus. So that that can be placed anywhere on your schematic, it's not directly connected to the map, and it makes for things to be a whole hell of a lot cleaner. When you do that, that way, you don't have to search for the one schematic symbol that has power and ground like you can, if you see like you too, is two op amps and a power and ground thing. Yeah. So the thing is, here's the part where I get a no, if I if I put a bypass or two bypass caps on the power, and it's like C eight and C 50. And those are connected to that op amp. When I lay it out, I put C A and C 50. On that op amp, you know, even though they're bypass, they could go on any of the other bypass ones, it doesn't matter, I make sure that my layout matches my schematic. Exactly. And that's never been a problem where like, you know, I didn't know which one was which. Because if you look at a bypass cap, typically it's connected to some power rail and ground. So it doesn't matter. If that bypass cap was on a different op amp, it's still the function is still the same, but I make sure that the layout is exactly like the schematic.

Yeah, I do the same thing. I'll make sure if I'll I grew up my bypass caps with the chip on the schematic. Yeah. And so I'm like, Okay, this one gets this one. And it's not for like, it's just makes it easier when you're reading the schematic later, like, a couple years later, and you're like, Oh, what is C? 12? It's by this part. Right, right.

Yeah, cuz because yeah, if you had if you had a bypass cap that on the schematic was connected to an op amp, but then you laid it out where that bypass cap was on a processor somewhere else. Once again, that could function but it'd be really confusing. Yes,

I try to avoid that as much as just minimizing it for you too.

But that's all part of like, when you barf all the components in and you're doing all the grouping of sub circuits and things like that. That's all part of that.

Yeah. I wonder if I could write a better I could write a ULP that does that.

Well, how would it how would the ULP know what to group together?

I'm thinking it doesn't like proximity just looks at it. Just look. No, it looks at the schematic and just pulls the coordinates for the parts. And then replicates that in the board layout?

Oh, you're going back to your idea where it's just like, yeah, yeah, yeah. Put down because then

I'm already grouping stuff, how they should be on the schematic already. Yeah. And so I'm, like, just replicate that on the layouts that I have, at least, you know, the first hour of my board layout is already done.

Well, okay. i That sounds really useful. But what happens when you have multiple schematic symbols that are in a single package?

Oh, yeah, like an op amp, like a quad? Op amp that's got Yeah, five, multiple XY components. You just have to have have a priority. It just picks the first one. Yeah. Make it easy. You can't make it too complicated. It's just a grouping algorithm, right. It actually splits all the pins out into four different parts and spreads them all over the board.

Exploded diagrams of ICS start going all over the place. Oh, that's a great idea. Actually, you know, so I would love to hear. I personally am not super aware how other EDA tools handle that exact same situation when you hop the wall from schematic to layout. So if anyone wants to jump on the Slack channel and let us know, like, Does every EDA tool out there just go. But when you hop that wall and just shoot the components everywhere, or are there like better grouping algorithms? I don't know. I'd love to hear about that. Yeah,

I think it would be nice to like, if you can, like pick a certain section of your schematic and be like, I want to route the USB UART section today. And so it just picks those parts for you. Oh, yeah,

yeah, right. Right. Well, and eagles sort of does that right?

No, I'm saying it just barfs it a lot into the let's just like the trace.

Well, no, but I mean, but but Eagle at least, like if you select five components in the schematic, those will be selected in the layout, right? Oh, yeah, that's true. Yeah, well, in dip rates, I have to go and be like, are 17 and then flip over to the layout and be like, where's our 17 and then move it to a place and be like, that's where our 17 is going to be not like it's, it's super manual and clunky. And I heard one of the guys, I think it was on the forum, one of the dip trace, one of the one of their people mentioned, like, there's a way to pull up a list where it shows a list of every component, and you can just select a component from the list. But that's just a list of components, like the list isn't grouped, it's just a different way of selecting a component like it doesn't help.

When I have that situation, Eagle, I use the I think people forget this thing exists. And eagle is it's just the command line that it has. So you can type in show C 13. And it will just highlight it. And that's super convenient.

Well, and in DipTrace, you just do Ctrl F, just do a find find find c 13. And it'll it'll, it'll center the screen on that component. And you can choose the magnification level. So it'll like center and zoom in on the component. But that's still like, you still have to do like a find command on every component.

I guess. So. Yeah. Yeah, whatever. Maybe by next week, I'll have this up. And so we can like experiment with it. Sounds like a cool little project. Easy. Easy, easy win right there. Low hanging fruit. Yes. So Steven, you are resurrecting your ribbon mic?

Yeah, yeah. So well, last week we had talked about, I was working on my graphic equalizer. And actually funny enough, right after we finished recording the podcast, I started feeling kind of rough. And then the next two days were hell, I ended up getting food poisoning. And I was among the dead for a couple of days. Like it's taken me a couple of days to recover from that because it was rough. So I didn't really do much work on the on the equalizer like soldering didn't sound that fun when when I had food posts. Well, you can

barely move. Yeah,

exactly. So So I was, I was actually in my basement kind of looking through some stuff. I think it was Friday. And I found that old ribbon mic that we had mentioned a few podcasts ago, the the one that I had, that you 3d printed, the shell for and I built up this ribbon mic. And it are always kind of been my intent to take that to the next level. And so I started playing around just with some new ideas and designs on that. And I kind of want to redo what you and I had created because that was like a that was a really fast and easy fun project that didn't take very much effort and we got a win out of it. You know?

Yeah. And we had a really good little podcast recorded with that Mike.

Yeah, yeah, it actually surprisingly worked. Okay, for just being something we 3d printed, super glued together. And I put a sock over it as a windscreen.

Yeah. And in the transformer was just Gator clipped onto it.

Yep, yep. Yeah, and we just Gator clipped into an XLR cable.

That was episode 110 By the way, oh, dangling transformers.

Oh, and we use the magnets that were part of a the mic at to just stick like we just stuck it to the mics. Yep. And regarded that way. So okay, so here's what I want to do in in ribbon mic version 2.0 Because we've we've proven the concept what I want to do is ditch the 3d printed frame and I want to make the actual frame out of out of PCB material. Do I stack up of PCB material and A lot of the reasons why I want to do that is I want to embed the transformer into the design as opposed to just having it hanging out, I don't want long leads from the transformer, I want the transformer to either be mounted directly to the PCB and the in the leads from the transformer or like soldered, you know two millimeters from the PCB, or I want to get cute with it and have like, if you think of doing laminations of PCBs like four or five fr for PCBs thick, do a silhouette cut out of the transformer and stick it in on a different axis. Yeah, so kind of like fits inside. So we can get an even lower profile and make a really tight clamshell. Exactly. Well, and the thing is, if we, if we do it that way, where we do PCB laminations, then I can potentially do some kind of interesting shielding work and have all the signals run on the internal laminations basically making multi layer PCBs out of multiple PCBs. Yeah, in a way. So I would love to just have the signals go directly to the transformer, and have have all the pads and everything that, you know, sandwich the the actual ribbon sensing element have that all be on fr four. So effectively, all we're doing is just swapping out, you know, our plastic for fr four, but giving it some conductivity, to allow it to work that way. But I also want to try because when we mentioned on or when we were talking on the on that podcast, whatever episode that was one of the one of the issues was getting the correct tension of the ribbon. And the ribbon is like, I think I bought three micron thick aluminum, it's not like something where you can just like manhandle it, put it in there and pull it taut, you know, like, you do that and it falls apart it like disintegrates. So one of the one of the ideas that we were mentioning on that podcast was having a system where you can squish it in the PCB. And then by using thumb screws, you can pull the PCB and apply like fine tension and fine tune the Guess how much it's being pulled. And that will almost guaranteed have effects on the frequency response to the mic. So I would love to like fine tune that and see like, what where our high end roll off is? Or maybe we can get a different kind of response with it and things just

sick a speaker at it and then sweep it, basically.

That's it? Yeah, that's a good idea. I hadn't thought about that. Yeah, do like a speaker, you know, with a specific output a specific distance away from it on axis. And then yeah, watch it on a scope and see what's happening. Yeah, that's a good idea. Yeah, that would be that would be fun. Although, you'd have to make sure that the speaker is outputting, a flat frequency response? Yeah,

you'd have to basically see. Because if you had that, you'd have to, you'd have to have the frequency response to the speaker and your amplifier. Right? Well, and then and then basically, record in with your mic, and then compare the two. And the difference is what your mic is doing.

Exactly. And you know, so I, I had a project I was doing a long time ago where I did something pretty similar to that. So Amazon, you can buy reference microphones on Amazon for actually pretty cheap, less than 100 bucks. And the reference microphones come calibrated with a known frequency response. And they're more for acoustic testing than they are for recording. Because they don't particularly sound well. In other words, they just don't have any character, they're super flat. So what you could do is put one of those right next to your mic, and then use that to cancel out the error from what the amp is. So if the amp is rolling off, you can tell with the reference microphone, cancel that from your, what your measurements are from the mic, and then just get the mic. Yep. So that would be that would be kind of fun. So you know, one of the other things I want to play around with, so that, you know, the ribbon itself is a really, really low impedance. It's just a chunk of aluminum. Aluminum. Yeah, it's like point O two ohms. You know, that's your source impedance. But the voltage that comes off of it is tiny, like 2030 micro volts is what you get, you know, yelling into this thing. So it's not particularly strong. And that's why there's a transformer in there to kind of you know, give it a fighting chance that boosts it up a little bit and the transformer also works as impedance matching, you know, you so you can match down to that point oh two, or whatever the impedance is when the things actually vibrating who knows that? But one of the things I was thinking of is I would love to try ditching the transformer entirely. and putting a instrumentation amplifier on there. And basically going with like a, you know, basically an instant instrumentation amplifier that has like a Tara ohm. Input impedance is like absolutely insane. And then applying 1000 times gain over, you know, two or three stages of op amps. After that, it would have to be AC coupled through this. So there's be a little bit of funkiness there. But if you apply 1000 times gain on a 20, or 30 micro volt signal, then you're closer to mic level that, you know, like the mics we're talking into right now. So maybe I could get away with embedding a instrumentation amplifier in this fr for material, and then making this ribbon mic work off of phantom power, like, like a lot of I like this idea a lot. So it could be it could be pretty fun. So either have a fully passive version that has the transformer embedded in the PCB or have an instrumentation amplifier in there. So it runs on phantom power. So I don't know what in the in the the other benefit of the instrumentation amplifier is you can have a super low output impedance. So you could drive whatever I mean, the thing is, it's going to drive an XLR cable anyway, so it doesn't matter too much. But it this would be kind of fun. And honestly, like, probably not going to be that difficult to you know, design that up. It's going to be more of like, the mechanical stackup is is more difficult than the other stuff I was talking about. So

it's Ryan's we have Brian bench off is doing a badge a PCB badge for like Def Con this year. Yeah. And he's Korean what's called the Oreo PCB stackup, which is kind of what you're thinking of

is two blackboards with a whiteboard in between.

Sure, but it's actually it's one PCB with the traces on it, you put your parts on it, and then you have one that's just all cut outs. Yeah, that's like the sandwich that you're talking about. And then the new solder one on top of that. And he actually had a pedal design on Hackaday that use that.

Okay, did he Okay, so but the question is, did he get really, really goofy with it? And via stitch around all of the pockets and make like, little shield tanks for everything?

I don't think he did.

Basically, like, you know, if you've ever watched, like the signal analyzer tear downs that Dave Jones does, where he, you know, he opens up this case, and you don't even see the board, because there's a giant milled aluminum block in stuff. Yeah. Like, it's akin to that.

Right? Yeah. So I posted the link in our notes that says Oreo stack up.

Let me take a look at this. Oh, yeah, that's, that's effectively what I'm talking about doing here. Embed the transformer, but on a different axis, or embed the instrumentation amplifier in there. And what you know, what would be what would be really cool is to find a right angle, XLR connector, that this could solder on the end of the PCB. And then what I would love to do is make a two, a two side or a clamshell enclosure design and have the two enclosures 3d printed and then they just come together around this stackup. Yep. So it would be really cool to 3d printed such that the base was solid, but you could have mesh, 3d printed mesh for the where the ribbon is itself.

Oh, yeah. Yeah. I'm wondering how fine you would have to make that mesh.

Not very fun. It doesn't have like, if you look at old ribbon mics, they're just like, a steel that has a bunch of holes punched in it. It doesn't it doesn't have to be super. Yeah. So 3d. Yeah, 3d print that and when you put it all together, take two pieces of medium density foam, put them over the ribbon itself, and then sandwich the two clamshells together. And then maybe a nut on the bottom or something like that, that holds it all together or even glue it together. So you you get the two pieces of foam act as a little bit of a windscreen. Yeah. Which which it needs because if you if you do a p or A t into it, you can break the ribbon. But like I don't know, like this a sort of comes together in my mind. It's like an all in one little package. That's pretty easy to put together. Honestly.

Yeah. I'm looking forward to to this project.

Yeah, and of course, I want to I want a fusion 360 And I want to I want to actually make it. In fact, I have in the notes. One thing I haven't done with you yet or with anyone is go from fusion 360 to 3d printing. Yeah. Which you know, there's nothing particularly special about that, I mean, cuz it's just file formats, but I would love to design it all in Fusion 360 send it to you have you 3d print it? That'd be cool. Should work. And will you dry brush the screen silver? So it looks like metal? It wasn't spray paint. Okay, that works. Yeah.

You know one thing we were talking about was this on that one episode when we recorded this with this mic. Yeah, it was the PCM nine a 2912. A, which is this old one? Microphone USB package?

Oh, yeah. Forgot about that.

Yeah. So this would be cool to do a one that's USB powered to make that work?

Oh, you could totally do it entirely USB powered. Yeah. Because even if you gain up the signal 1000 times you still like three millivolts? You know? Yeah, you could run this all it Okay, so it'd be really cool to have USB or phantom XLR option here. Like you could pick on one mic? In one mic. Yeah. So you pick what you want. If you want it to be like studio, plug your XLR in. But if you want to take it on the go, and you just have your laptop, you know, do that?

I think we should do this. Yeah, well, I have to have a a session of just like coming up with like the board idea.

Hey, you know, like when we did my, my my micro tracer, and we did the code thing, the live stream, the code thing? Let's live stream design of this. Yeah,

that sounds good. Yeah, that'd be a bunch of fun. Hi, right. And that downside do not forget,

we can talk about I mean, it might be more than one live stream, there's a bit to talk about, but mainly the mechanical design and like kind of like layout of how it all needs to work. Because one of the things I didn't pay attention to at all with the first one was, the length of the ribbon was sort of fixed, but the length of the magnets was just like whatever I could find, yeah, the thickness of the entire thing. So there's a ratio of the, the length of the channel that the ribbon sits in versus the thickness of everything. And you want that ratio to be something decent, because if it's off, then you get this. The I can't remember exactly how it is. But the the wavelength of certain frequencies doesn't vibrate the ribbon properly. So it doesn't resonate. Yeah, we want to get that ratio. Correct. And I obviously there's probably a lot of wiggle room because we just threw one together and it worked. And it worked. But let's let's actually try to get something decent.

Cool. Cool. All right, onto the RFO.

Yeah. So I found a good one, this week that I'm a bit excited about. So the MIDI Association announces MIDI 2.0 dot dot dot for prototyping.

So what are they been doing for like the past 30 years?

Obviously, MIDI 1.0 Yeah, but it's been done. MIDI MIDI has been around for at since 83. Yeah, I think it was 83 is when MIDI was first implemented. And so at NAMM 2019, which is happening January 24, through the 27th. The MIDI association is which? Gosh, what's it called? MIDI something Association. And it's MMA and I've totally think it's the mixed martial arts of electronics. But, yeah, so they're announcing MIDI to

transistors, finding op amps and microcontrollers. Oh, yeah.

Yeah, like rolling around on the ground until someone gives up. Yes. I probably offended somebody by saying that.

What MMA fighters?

Oh, no, like UFC guys like they're the UFC guys. Yeah, they're not gonna be listened to this podcast, I'll probably not know if you

if you, you can watch whatever you want to watch though.

Okay, so MIDI, MIDI 2.0 is the next logical, maybe not fully logical movement in the MIDI world. And the reason why I'm kind of saying it that way, is because, like, MIDI is sort of antiquated and not sort of it is totally antiquated. It's slow, it's old, it's clunky. It doesn't have a high resolution. It's basically a just eight n one codes that, you know, flow down whatever wire you're sending between two devices. And it's super great. It's super robust, but it's just like, you know, we could have, we could have a much bigger protocol, and I think MIDI 2.0 is trying to become that. However, the thing that I find really interesting is that MIDI 2.0 is promising to be 100% Fully backwards compatible with MIDI 1.0, which means that it's going to be crippled by all the things that make MIDI 1.0 kind of rough. So I'm really interested to see what is going to come of this and how it can kind of like, become something bigger. Because there's other audio style protocols are actually technically as a control protocol. It's not an audio because you're not actually sending audio. It's a control protocol. And MIDI 2.0 is supposed to kind of like, extend beyond the whole like keyboard or foot controller world and go more into like your computer dot control, being able to access a lot of, you know, other codes and commands and things, it's supposed to also give you greater resolution than 127. For everything setups, with the thing is, though, it's still supposed to run at the baud rate of 31, to 50. So it's kind of slow. And it's still supposed to be able to handle eight on one. So what I'm really interested is like, is it going to have like a completely separate mode where it'll like, upon startup, it'll ask like, Hey, are you ready to MIDI 2.0? And if so, then then like, starts blasting about like, because that would be kind of cool. Oh, like it has basically little handshake at the beginning? Yeah, if something if something like waves back, it's like, alright, let's let's tango, and then it just starts ripping like actual data, you know, that would be that would be pretty cool. Israel, I, you can actually get information on the, the new protocol, if you go to Gosh, it's escaping me. It's the it's like the official MIDI website, you can sign up and make, you know, register to get information. And you have to tell them like, Oh, I'm a hobbyist, or I'm a designer or whatever. And they have some cryptic PDFs about like, what's happening? And the whole thing is it they're not releasing the standard. What they're saying is they're entering prototyping phase, which means they're gathering a bunch of information, everyone's kind of starting to begin to argue and agree,

like, here's a white paper on, like, what we think the standard should be.

Exactly, exactly. So none of this means that anything is set in stone, they're just beginning the process, I'm sure it's going to take a very long time, because everyone out there is going to want to, you know, put their two cents into it. And the big guys like Roland and all those guys are going to say like this better work with all of our current offerings and the last 30 years of offerings. Well, that's the thing is like, Oh, I don't want to find that gray beard design this device 30 years ago, they get him to implement something, make it work on 3.0. Exactly, yeah. And MIDI 1.0 tried very hard to allow for expandability because it has like predefined set codes for most of its stuff. But it also allowed for banks of codes that were user definable, or they were definable by, you know, whatever designer was out there. So, you know, Roland might have their own specific code at, you know, whatever address they have. And, you know, it allowed for that, but MIDI 2.0 seems to be more focusing in that realm where it's like, Okay, we're gonna give everyone their own codes. And, and we're, we're finally, in an age where all of our stuff can actually talk and not have to have standardized codes that can ask if this code is is good or not, you know? Yeah,

actually, I'm looking at the it's midi.org.

I should have guessed. And

someone on there, how 2.0 is gonna be backwards compatible. And yeah, basically, they have something called MIDI capability. Inquiry. Yeah. MIDI CI. Yeah. And so it goes and talks to the device. And it's like, if either device says, No, I can't do that, or just straight up doesn't respond, because it doesn't know what's going on. Right. Your host. MIDI controller just does. 1.0

Yeah, it pulls the crutches out, and it starts going real slow. Yes. So and that's what I'm saying. Like, I hope they implement that, obviously, this is all prototyping, you know, I'm sure there's, there's a general amount of politics that go behind this kind of stuff. So you know, whoever complains the most, it might not even that might not even work. But it would be really cool to see that like, you know, if two devices are like, Hey, we can do this, let's do it. And then you got all kinds of cool stuff, because then you can really like do like true automation via MIDI, through your computer, have your computer like talking to multiple devices, controlling all kinds of aspects on them, like talk about like some really cool power there. Whereas like, with old MIDI, it was like, well, good luck. You'd probably have to program a whole bunch of stuff and it's going to be not fast enough to automate live, you know, that's why a lot of a lot of stuff has like custom protocols over USB, but the don't really work because everyone has their own custom, like protocols and drivers and things like and then it gets, like what happens one year later when that product is obsolete? Well, they come up with a new product with its own USB driver and its own protocol. And it doesn't work work too well. So I'm rooting for this, I hope it'll turn out pretty well.

Well, it sounds like, MIDI has stood the test of time because people still use it. And it's still very popular and it hasn't really changed in 30 years. And so yeah, having a big change up like this, and hopefully they do a good a good job with 2.0. So it lasts another x years.

Well, and in my realm, in the synth realm, I would love to see synthesizer start using MIDI 2.0 such that you can go MIDI to voltage, and then you can take this MIDI 2.0 and use it to control your entire synthesizer rack and do it at high speed. So you can you know, instead of just like sending a command, that's like play keyboard, G five, you know, like that's old MIDI. Now it can be like, I'm sending you all the codes to play this and modulate it this way and do it that way. And I'm talking to eight of you guys at the same time. That would be really cool.

I wouldn't be cool. Okay, so the next topic is solder tension, or salt, solder surface tension and why you should care

about solder tension sounds like a, like a nerdy engineering metal band

of solder tension yet

soldered the liquid solder tension experiment. I'm sorry, go ahead.

That's great. So the current source, aka Derrick, he released a really cool video of basically like, what is surface tension? And why is it important with it for a PCB assembly?

It saves your butt after?

Yeah, that's about half the time. Derek was on the map episode 103. So go check out that episode. But that's a really cool video. And Derek does a really good job. Like explaining what surface tension is and how it actually works. And at the very end of it, he has a example of how changing a part footprint can affect how it reflows

Oh, wow, that's really good. Yeah. So I'm kind of breezing through the video right now. So he does different, different footprints and shows

at the very end. Yeah, he's talking about like, a complicated module footprint,

where his Oh, it's like an ESP. Yeah, how

the I think it's a Silicon Labs part. Okay. But it's, it's how the stock footprint from the manufacturer actually doesn't work really well. And mass manufacturing. Ah, neat. Basically, you need to move some trade some, some of the pads around to get better surface tension action, so to speak.

Okay, so you know, this, this kind of harkens back to our chat that we had with embedded FM, where they were like, why wouldn't you just download somebody's library and just use it? It works, right, like, well, here's a great example as to like, maybe not, you know,

yeah, even even the manufacturer's recommended footprint. wasn't good. Like that footprint would be foreign, probably for hand soldering, but for automated assembly, you need to move these pads more away to prevent the part from shifting under reflow.

Right. Right. Right. Well, and and at the same time, like part shifting under reflow. That's what I was. Meaning when I said to save your bacon earlier.

In this case, it ruins your bacon.

Yeah. Yeah. When you're, yeah, when your hand soldering and you're placing components. Don't spend all of your time trying to get the component perfectly aligned before you've float the base. Yeah. That is really cool. Um, okay, I'm totally gonna have to watch that. Yep. Okay, that's just one quick thing. They should they should play that in college. This video should Oh yeah. taught in college guaranteed. Yep. Okay, cool. Next,

I have a choice of grippers helps dual arm robot pickup objects faster than ever. And this is the new Dex net 4.0 Multi Tool picking robot. It says dual arm and you're like, oh, yeah, totally makes sense. Humans have two arms. But you got to think this is more like a It's got two different tools. So it's got a it's got a grabber and it's got a little suction cup to pick up stuff. The problem with picking robots has always been weird shaped objects. And robots aren't very good at computers aren't very good at figuring out that kind of problem. Humans, we just live, oh, I know how to pick up that soda can by just kind of looking at it. And your hand just kind of grabs, it's like it's some some cerebral thing going on in our brain to make that work. And computers just can't do that. Like, even this is like the best picking robot ever. And it can only do about 300 picks an hour at 95% success rate. Whereas a human does double that at 100%. Unless, unless they're tired. Unless they're tired. That's one thing the robot won't get tired. Yeah, it'll just keep going. It just keeps going. But the cool thing is how they're getting this high percentage of success rate. It's not just, you know, a suction cup and, and a grabber. It's, it's learning things by simulation. So it looks at everything, it has a 3d camera, I guess it does depth mapping, and scans the objects to figure out how it's basically it takes one of its tools and simulation and tries to pick it up, you know, 1000 times or 100,000 times, and it goes, Okay, this is probably the best way to pick it up. And then goes and picks it up. And they're doing weird things with the physics model, though, in those simulations, because you're 3d sensing isn't perfect, because there's some fluctuation of error there. And even just like driving the arms, you're going to have overshoot and undershoot with your mechanisms and stuff. And so they're actually adding a little bit of randomness to the physics to basically mask up sensor defects. They're adding fuzziness to it. Yeah, they're actually adding a little bit of fuzz and it kind of masks the physics problem. That's funny. So I think that's a really cool idea. I'd like to see, I bet you like Amazon is gonna buy this thing.

Amazon's gonna buy this company. I like they broke things out into grabbing difficulty levels. Yeah. How hard it is. And I love it. Because like, in the hardest difficulty level difficulty level four. There's one of those honey bears, you know, like the bear that's full of honey. That's a that's a level for difficulty.

Yep. And then also, it's like a action figure in a plastic box. Yeah, a whole Cogan or something like

that, actually, you know, I bet you that these things are all in difficulty for not necessarily because of their shape because of what it has to see.

That's actually one of the things is surfaces that are easily manipulable. Under pressure, or stuff that's transparent. The system doesn't work too well on.

Well, they have a loofa. And they have a glove. Those are both probably awful to look at, you know. Yeah. And there were cameras up. Yeah. I mean, if you're trying to do edge detection on a loop, like what? Crazy. That's cool, though.

That's a cool little robot. Going on to the next one. The basics of USB battery charging a survival guide by Maxim. I'm only going to cover just a couple things in here is a really cool app note, if you haven't done any kind of lithium battery charging over USB, and actually they go into nickel metal batteries to it's a little bit outdated on the USB end, because as it does cover USB 3.0 but doesn't cover the Type C connector or anything past battery charging specification. 1.11. It's like 2.0. Now 1.1 was like the last one before type C came around. But yeah, so if you're doing a USB 2.0 device, this is like all the app this is like the Bible for like doing lithium battery charging. That's great.

Yeah, cuz that's super prevalent.

Yeah, I'm hoping either I haven't found it yet. Or they haven't done one yet for battery charging 2.0. So that would be nice. Because that's like getting your 100 watts of power.

Yeah, it's, that's the serious guy. And then USB, USB 3.0. They have, like one paragraph on that.

It's like, Yeah, this is kind of new when this came out.

Yeah. What would you does that? Is there a date on this? I don't see it. Okay. But no, that's great. Yeah, this is like four or five pages and plenty of December 9 2010. Oh, okay. Yeah, this is this

is old, but it covers any kind of USB 2.0 charging for lithium batteries. This covers all of it. Yeah. And

it's still good. Yeah. This is a good good app note here filled with a bunch of diagrams with maximum parts in it. Yeah, course. Go figure.

I bet your half of these are obsolete now.

Yeah, but there's probably some better version of it. Probably. Cool. So there's one more quick topic, that it's a fun topic to bring up. So, over the weekend, Mauser did a slight change to their website. And I wouldn't be surprised if a handful if not many of you have experienced this also. So Mauser for a while, had a kind of a static website, they didn't really make much changes, but they did a handful of graphical updates to things. And so I came in to work on Monday, and I was having to look for some parts. I was like, oh, Mauser kinda looks like Digi key now. So like they their old site. And what I'm talking about with this is their like filtering tool, like when you go to passive, like resistors, or something like that. And up at the top, it has all the banks of information where it's like, select your resistance, select your tolerance, your package, blah, blah, blah, all that stuff. It used to just be a string of boxes that had, you know, all of the stuff you wanted. And it still is, but now they added a whole bunch of graphics to it effectively. So it looks more slick. But it's just spaced out. Yeah, the question is, is it better?

No, because so I'm on I'm on mouser.com/passive components slash resistors. And so I see, and I'm on a 4k monitor. So everything's really wide. I see resistance. And I have to actually have to scroll over to select the case code.

Yeah, yeah. So one of the things that really stood out about this whole graphical change is the fact that everything is kind of bigger. Now, they made all the selection boxes larger. And even when you actually drill down, and you see a list of the components you want. All the list of components are much wider, not wider, I'm sorry, taller now. And actually, you know, what's, what's kind of interesting is, it sort of reminds me a little bit of the macro fat bomb, the macro fat bomb looks kind of similar to this. So maybe Mazzer stealing from you guys. But the macro fat bomb is there to show you your parts, it's not there to for you to search for parts. So there's very different intent behind each one of the bombs. And the thing that kind of is grinding my gears right now is it feels like they made it a lot more difficult to quickly get to your part. Because now that everything is so much bigger, I have to scroll around to get to the package. You can't use your mouse will scroll over there. Exactly. So previously, I didn't have to do much sliding across the screen. Like, I could pick my, my tolerance, my resistance, and my package code, which that's like the first things that I drilled to. And I could do that all just from the page that I was at now I'm not capable of doing that I have to use their little like slider bar and slide over to all my stuff. It's like, what, why? Why was this necessary? And like I said, it sort of feels like Digi key now because Digi key was always bigger and clunkier in my opinion than Mauser.

Yeah, and DigiKey used to have this kind of problem. But they fixed it recently, they had this like, checkbox slider thing that's called filter options. And they have stacked in scrolling mode. Guess what mode you always use stacked? Because you can see everything. Yeah, it doesn't stack doodle vertically. Yeah, it will it stacks all these filters up into what will fit on your screen.

Yeah. Which great you know, someone was actually thinking about, like the, the experience on there and it kind of sucks because like now Mauser has made the experience harder. It's not, I shouldn't say harder. It's not like it's any more difficult. In terms of like, I can still get to the thing I want. I just have to do it in more clicks now. Whereas previously, there was nothing wrong with the graphics. It all worked out fine. And I was could do it. You know, I don't want to sound like an old an old man being like, I want my stuff back. But it's at the same time. It's like, why like what was wrong with it? Like what executive came into Mauser and is like, I'm gonna do a good job and change all this stuff up. You know, it's like what it no go back?

Because this is the thing is it looks slick. Yeah, like I really like how they're doing this colors and stuff, but because of this scrolling thing, it feels like it's unfinished.

Yeah, it's just it's clunky. They just added clunk, but made edge. You know, they made it slick.

It's like an Apple device now,

I you know, I shoot, if it was an Apple device you they give you one resistor that's all like you can you can pick one. That's it and it costs $1,000. But, but no, no, like, I wish there was a button that was like Mauser classic, just, I mean, it's Tuesday. It's been out for a day and a half. And I already want to Mazda classic button.

Kind of like the Reddit Reddit classic.

Yeah, yeah. There's, there's so many threads on Reddit that talk about like, how the mobile app is like, incredibly superior to the website now.

Yeah. Well, the good thing is read it. So has the classic, you just click classic, and you get the classic stuff. Right now. Don't look at mousers mobile on a desktop computer. I bet your disease is just as bad. Do they have a option to flip over to that? They don't.

Also, you know, it's funny.

A huge assembly. You're looking at it. And you're like, this is totally optimized for like vertical scrolling. Oh, yeah,

for sure. But like I'm not griping about this. I think it's funny. You I mean, last week, if you went to the Products page on Mazda, you go to Mauser and you click on Products. It shows everything categorized like circuit protection connectors, electromechanical, it used to have a picture of like, next to passives. It just showed a capacitor next to connectors that showed some kind of connector, and then they ditched all those. So now, when you go to products, it's just text, which was like, Why did you get rid of the picture? The pictures like the first like, it should gives you like a quick overview of that entire genre of components that you're looking at, you know,

now, but they do have pictures. In the when you actually drill down one pass, like if you go to passive components. Yeah, there's a visual representation of the parts now. You click, like, if you go types of resistors, you can do list or visual which.

Okay, yeah, I see that now. I'm looking at it. That's cool. Oh, wow. Wait, you wait, you could search based off of datasheet? That's weird. Oh, look at that. Or search? Search off of images? Hmm,

how do you how's this data sheet

work? I don't know. This is Oh, it has a nut part number

in datasheet. Let's go to

the data sheets, and then click List. And it just shows like lists and lists of data sheets. That's interesting.

I would have to play around with that. That looks really interesting. Yeah. Oh, I think this is of like, if you knew what you really wanted. Yeah. Like if you wanted to stick with the single family, this is the way to go.

Yeah, yeah. And yeah, I see that. That's, that's convenient. That's nice. I like that. If you know where you want to be, you don't have to like backdoor it by searching for a specific component and then finding that data sheet, like the global data sheet. That's that's kind of nice.

Now I don't think the image result searching for passive components is all that useful. But I just found a really cool looking

part. So okay, I also the mouse wheel it it does work for scrolling left and right but only if you click in the right area. Oh on that filter. Well, but not entirely on that filter in the right area within the filter. Like there's there's a gray box that you have to be within to be able to why

school mail seen man?

Yeah, like I

don't know. We had my magic All right, I guess that's our read. Yeah,

I think that'd be the end of this podcast.

So that was the macro fab engineering podcast. We were your host Stephen, Greg. And Parker Dolman take it easy everyone see you later

Thank you. Yes, you our listener for downloading our show. If you have a cool idea, project topic or interface you want to complain about? Let Stephen and I know Tweet us at Mac fab at Longhorn engineer or at analog E and G or email us at podcast at Mac feb.com. Also check out our Slack channel. If you're not subscribed to the podcast yet, click that subscribe button. That way you get the latest 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