Come celebrate the third anniversary of the MEP as Parker and Stephen discuss a potential video game.
Stephen and Parker discuss where debouncing a switch should be done in Hardware or Software.
Calculations
12.01V from power supply
11.98V at input of circuit
60mA current draw so
0.03/0.06 = 0.5ohms resistance
But that is across two wires so 250mOhms per wire or 0.083mohm per foot
33 AWG wire matches this but that is fairly thin.
Figure 1: Non working Space Echo RE-201
Figure 2: A NuTube 6P1 straight from KORG.
Figure 3: The constant current source Stephen is using for his NuTubes. The equation for the constant current is I ~ Vgs/Rs.
Figure 4: Jig of Destiny 3D render.
Special thanks to whixr over at Tymkrs for the intro and outro!
Hello and welcome to the macro fab engineering podcast. We are your hosts, Parker, Dolman and Steven Greg. So over on Josh's rack, the, the stereo racks here, they're 19 inch, like regular, I guess server racks.
We love Josh's rack
he's got this space echo and that we've talked about this a couple times, I don't think on the podcast, but beforehand, unfortunately, it's broken. But uh, Josh will actually let us take it apart and fix it. And so maybe we should do, like maybe a live a live podcast of us fixing in order to record us fixing this thing. I'm sure, but the cool thing is, Josh actually has a digital plugin on his computer that will make a sound like what the space echo does. And guess we're gonna be a little goofy here and just basically say some sci fi nonsense and have it take what it would sound like. So just give us a sign Josh when it's running. Alright, so he says it's on now. We can't actually hear what is sounding like, but uh
so this is actually perfect. This is this is a double blind test of space echo. Echo Monster Jam.
Spec. Oh, was that was that thumbs up for Josh? Okay. Is it okay, cool. Cool. Cool. Are we still recording? Okay. So, Steven So, Steven, what are you been working on this week?
Well, we got we finally got our shipment of new tubes in Yeah, the
I knew to start with the MMU.
So we if they can't, who was cork
Korg? Korg. Makes them? Well,
yes. Cork? I guess they originated nor takeI nor
Taki. And why is it called Neo tube? Neo? Yeah, it's a Japanese tube. Um, good view. Well, do you like Neo Tokyo
and all that stuff? It's just like some kind of strange anime or something like that? Well, no, it's just
like, all in like, you know, futuristic stuff. It's like Neo New York and Neo, whatever. So it's kind of like new New York. Honestly,
the whole like, new craze crap was like, mid to late 90s. So they're pretty far but
Bellet slightly different.
Well, I'm thinking of like the the new metal with Linkin Park. And
oh, like that was? Or like future. I'm a New New New York. But was that spelled and you know, is it's any W? Okay? Yes. It's two news.
Regardless, it's not a good name for that. Well, okay, the the official name is six P one. And I assume they just felt that that was kind of boring. And no one would know what it is.
It's not a really long part number. With tubes, you
don't need to because you're not competing against other port numbers.
It's not like a 16 character, part number for like a capacitor,
or ICS that just get insane. Yeah.
So yeah, again, back on topic slightly. So what have you found so far with these new tubes? Um, okay,
so we got we we got nine individual new tubes and one evaluation board. And the it looked like what we got was some engineering samples, because some of the things look like they've been kind of tampered with or partially used. Yeah, parsley is one of the one of the little daughter boards that connects to the evaluation board had clearly been soldered before. And unsoldered the pads look a little bit
rough. Yeah, the look a little rough. And there's actually a flux residue on the Hassel finish. Yes, I thought it was hassle finish. But yeah, it's got flux residue on it.
So it's kind of annoying, you would expect that, you know, with an eval board, you can just take one of these units and plug it right into the eval board. No, you have to solder it to a daughterboard and then that daughterboard plugs in. I don't know why they chose that. Now, I don't know either, but whatever
and better because that the new tubes don't actually have 100 mil spacing. Depends Yeah, and I bet you'd that was cheaper than putting in a blossom header, custom header or or headers in the right spots, so to speak. Yeah. Also, it wouldn't be an Arduino compliant header. A jab at the 50 mil spacing on Arduino. Right? For those that don't know.
So well. And the thing is, it's it's In a VFD package, and it only has pins on one side, so they probably didn't want it flapping around just being held by. Yeah, one side of it. I get it, it just seems a little odd. You know,
I haven't actually seen a VFD that size, I guess I need to look at nor talkies site and see if they make something that small. They are a lot smaller than I thought there would be. Yeah, not thickness wise, but x y wise. They're a lot smaller.
11175 by about point six, three.
Yeah. It reminds me of a, like a 6507. CPU chip out of an Atari. Yes. About that. About that, right. sighs Yeah. That's right. It's a large
chip. And Parker had an actual VFD on his desk. Yeah. What size is that?
That thing? Is a like an inch and a half by almost six inches, seven inches? Yeah, it's a giant 40 by four character display actually have it right here in this box? Yeah.
So I, I looked at the internals of both of these. And they're identical. Absolutely. Except, okay, so So on the if you if you look through the glass, at the very top, there's a tiny wire that crosses over where the displays are. And that's, that's your heater or your your filament filament, and then the port that actually glows would be that the anode. And the only difference in this tube is that there's a mesh screen in between the two. That's it.
So yeah, I'm still very interesting how they turn that VFD technology into more of a tube, a transistor based technology, I guess,
well, they didn't have to reinvent the wheel, or really, now that I look at it, they probably didn't even have to retool their machines very much, because they just installed one extra thing. And it's done.
Yeah.
And you got something here about a 100 mega ohm current source.
So these tubes are not very good. Their specifications are pretty bad. And they have really high output impedance. I found in a book that I have constant current source that has a an output impedance of 100, mega ohm all the way up to about 10 kilohertz. And then it drops to like 10 mega ohm all the way up to 20 kilohertz.
So and you said, this is a chart? Sorry, what was this? This is a chart. It's a it's actually in a book that I have. What's that book?
It's called? Designing Hi, Fi preamps by Merlin Blencowe.
Is there an edition for that, or first first edition? Yeah, signed by the author,
I wish. I love it, guys, books, his books are awesome. In fact, they just released in February. Cool. So I'll be posting some information on that circuit. The cool thing about it is if you use one of these things with a tube like we have here, you actually negate effectively, all of your resistance for your, your anode. And the gain of the circuit that you have reduces down to the inherent gain of the tube. Oh, that's cool. So these tubes have a gain of 14 times. And in a typical circuit with just resistors and caps, you'd be lucky to get about eight out of that. If you use one of these constant current forces, you can get about 13.9 times almost 100% 100%. Yeah. Cool. So yeah, I'll be posting the schematic for that. And in some info on is that
gonna be done by tomorrow morning? Yeah, cool. That way, it will be in the blog post for this podcast needs. And then I've been working on what we've codenamed the jig of destiny, we've always had to have cool names for the stuff we work on at macro fab. And essentially, it's a jig or fixture that will hold our giant 16 by 16 prototype panels. So for those that don't know, at macro fab, we have very low volume pricing for our prototype runs. And how we do that is we basically batch all the prototypes into one giant panel, basically, because that way, you can spread out the tooling cost over the entire year. Everyone shares, everyone shares tooling costs instead of having to, you know, hit someone with a $200 tuning charge, you spread that over, you know, 10 or 12 people. And that's how you get the prototypes really low. The secret sauce, the secret sauce. So anyways, the problem with these giant panels is when they go through reflow, they tend to bow a lot, like you're talking like four inches of deflection. Yeah. And so what we typically do is, that's normally for single sided boards, it's not a problem because you run it on the mesh conveyor in the reflow oven. So it's actually the whole board is supported, right. But the problem is when you do two sided assembly, you have to support the board completely in the air to prevent the parts basically being ripped off in when you when you heat the backside up the liquid assayed again Hmm. And so what this jig does is allow it allow support in the middle during that process. Yeah. And I actually made a prototype out of open beam, which is a 10 millimeter by 10 millimeter profile extrusion kind of like at 20, our Misumi or a bunch of other adult directorship. Yeah, adult Erector Sets, bloods a very small version of it. And I was able to make a fixture out of that, and ran through the reflow. And it worked great. Awesome. And so I basically I've made a version of it that's a lot more precise. Basically, the entire bottom is actually going to be milled out of 3/16 aluminum. And the top plates will be made at 16th inch aluminum, which is a typical thickness of a PCB. Yeah. And because of everything will be machined on one, one plane, so to speak, the tooling holes will be in the correct spots, I pretty much had to eyeball it with my open beam version. Right? I was surprised that everything fit really well. Well, in
the border around our PCBs, we're putting holes, yes, this this panel aligns on the jig? Yeah, it's really easy to install.
Basically, the entire panel sits into these tuning holes. And then there's pins that are adjustable. In between that you can slide them around and and lock them in place. And so those hold up the middle of the boards,
right, right, because there's cutouts in between all of the boards, so we need to have adjustable standoffs.
Yes. And I hopefully in the next week or two, I'll get the machine parts back. And I'll be able to put them together and see if it works. And it works. We'll order like power 1210 12 jigs for that. And then I'm going to work on a version that's adjustable. This one's just for 16 by 16. And then, but we have other panel sizes for higher volume projects and stuff.
And isn't the goal for this to be able to go through all of our machines? Yeah. So
it has to be able to go to the My 500 paste jeder has to go through my 200 Picking place and it has to go through reflow. Right. And so that way, you the first operation in assembly is the put it into the fixture.
Yeah. It's a good way to indicate, hey, this thing is ready to go. Yep, we put it on the jig and you know where it is? Yeah, yeah.
It's really going to help out with with. It's also great when, if you have parts on the backside, and you have to do like hand placements on a board, it actually raises the board up and it keeps the board level. Yeah, on your desk. So I mean, it's gonna be pretty cool. They operation seems seem to really like the open beam version with its slightly out of tolerance pins and stuff. And they actually really liked it. Yeah. So hopefully, this version, the new version works really good. Yeah. Cool. Works really well, I should say. And then I got, I think on the last podcast of the podcast before that I talked about this fan controller. I'm designing for my jeep. Yeah. And actually got it today. Cool. Ended up on my desk today. Since I work at macro fab, so I don't have to have shipping.
It's kind of nice. Yeah, it's
kind of, you know, I wonder if we can get the developers to put we can have custom shipping locations. And so you can type in like Parker's desk or or Stephens desk. I think the developers would look like, no. Anyways, I got it today. So far, all I've done was powered up and took about 60 milliamps at 12 volts, which is about what I was expecting. And, but the interesting thing I noticed was I was measuring the voltages on the rails, and I noticed a point 03 volt drop from the power supply to the choke that have on on the power input. First, I thought it was the power choke problem. Yeah. And then we ran some numbers. So explain the numbers that were in?
Yeah, yes, it was, uh, we started taking some measurements. We wanted to see what is this point? Oh, three. So 30 millivolt. Drop? Yes. Where's that coming from? So our power supply indicated that it was spitting out 12.01? Yes. And we actually measured that at the terminal, and we measured it with our
multimeter and it was 1201 1201.
But then if you if you look at the input to your board, it's reading 11.98. Yes. So start putting two and two together, you got 60 milliamps worth a draw. The only thing in between the two sections that we measured was wires, the wires that went there, and these were banana wires, right?
Yeah, Bana wires inside his banana. Yeah, once it's been in the power supply and the other side are the little Test clips.
Yeah, the little like, springy finger clip clips, not not the alligators. No, but the little spring clips which are awesome. I love those things.
When they're brand new, they're awesome.
Yeah, they're well, you gotta you gotta be a little ginger with. Yeah, when that when that clip
wears out that just aren't way.
Right. So we got we got 60 milliamps and a 30 millivolt drop across those wires, which indicates a half an ohm of resistance, yes, that we're seeing in between there. But that's across two wires. So each wire would be 250 million ohms, or the resistance. And looking at these wires, they're somewhere in the 1314 gauge range.
I think so. Yeah.
And and so if you if you look at how many ohms per foot or ohms per inch that has, it sort of doesn't seem to work out. Because these are brought three feet, dude, yeah, two to three feet long, somewhere in that range. And you'd be looking more in the range of a couple of Milla ohms worth of resistance for the whole cable. But we're seeing 250 Yep. So we came to a conclusion in there. Well, actually, if you if you look at it, in terms of what gauge wire would be required in order to actually be 250 milli ohms. It would be 33 gauge wire every really and know that that's like magnetic wire and a transformer. That's and that's clearly not it. So it has to be something else. Yes. So we're assuming that it has to do with the actual clip itself?
Yeah, the actual clip the spring or the contact for the clip onto the test point or some
some grime in there. So so it creeps in a little bit, you know, if you start pulling a bit more, you can actually drop enough voltage to make a difference.
Yeah. It's something to think about when you're doing tests for your devices and stuff. Because we even saw this with basically crappy USB ports. I think we talked about this last week, did we? I can't remember we did. But we were doing a lot of tests with some USB ports and pulling, you know, more than basically close to the max of USB 2.0 power, which is a half an amp. Yeah, getting close to that and seeing how much does a USB port dipped down? And it depends on the computer. Like the front, the front ports on my really old desktop. They can barely even supply 100 milliamps. Yeah, before you start dipping below the standard, which is plus minus a quarter of a volt. Which is actually a lot by the way. Yeah, yeah, plus minus a quarter volt. So you can be 4.75 volts or 5.25 volts within spec. Yeah, it's pretty wide. And actually the interesting thing about USB 2.0 voltage spec, it doesn't have a ripple spec. Really. So your ripple can be a half volt.
Jeez, yeah. Well, honestly, I guess they're probably just making it easy for people to do it cheaply.
Yeah, cheap stuff. And well, that just means that your supply can be cheap. But your device has to be robust enough to be able to handle a half volt ripple on its on its supply line. Yeah, that's a lot. Yeah, that's a lot of rejection you have to do.
So I guess I guess the moral of the story is here here is don't just assume a wire is is just, you know, 00 ohms. In fact, I have an interesting story about that I was I was working on an amp a while back and I, I had the speaker pulled out from it. And instead of, you know, soldering the speaker terminals back because this was one of those kinds of amps. I just alligator clipped it in. And I noticed that my bias in my output section was going crazy, it was all whacked out. And it took me forever to figure out the issue and I eventually got to the point where I measured the resistance of the alligator clips. And they were about an ohm and a half each 303 ohms but it was a four ohm speaker. So my entire output is doubled effectively now and the output section of the amp did not like that. So do not and I actually what's interesting is I took apart those alligator clips I actually clip the end and you the from the size of the jacket, the diameter of the jacket, you would expect a pretty significant amount of copper in there. No, there was like a an angel hair strand of copper. So it had a have
like a have like 10 kilovolts of isolation on the on the jacket. Hmm,
yeah, and a gnome and a half of resistance. It was worthless. I threw all of them away.
You know, I wonder if we should take those test clips and actually cut them. Sure. I mean they're made by was it Electra something? It's a it's a company that makes supposedly good. Basically power clips and stuff you can buy on Amazon which is what we did.
I would, we should go look to see if they have a data sheet. And if they call out the contact and spring resistance, yeah, that'd be fun. Yeah,
which I'll report back next podcast on that
stuff. Cool.
And so we'll be going into the RFO section sounds good. And some we got a lot more content. There was a lot more interesting things this week. One of my favorite articles was bolt.io. They had this cool article about well, they called demystifying hardware jargon. And Josh might actually like this one.
We have 15 pages of jargon manufacturing and some of its business some of its hardware or hardware.
acronyms and jargon. Yeah. So I had Stephen go through it. Yeah. And pick out things that he that caught his eye that were interesting. So so first of all, on Josh's litany, and by the way he's
the the third page the very first entry on the third page is cogs. I love the word cogs cogs. COGS is great, it's nothing special, and reminds me of the Jetsons. Yeah, and it stands for cost of goods sold. And basically, it's the bill of materials of Bill of Materials. Yes, it's the at the end of the day, add everything up labor parts, the air conditioning, add it all up. What is the absolute and dollar that you spend that you're called?
And so bull IO says it's the COGS is a bomb plus, which is what you just said. includes labor, freight, logistics, customs duties, blah, blah, blah, yeah, stuff that you don't think about. Right? Taxes,
the it's, it's the end of the day, if you know your cog it's, it's a it's a it's a, you can have a sigh of relief because that means you did all of your homework Exactly. Everything,
everything. And actually on the same page. I'm going to enter it says a POC, which is proof of concepts. I don't think that's a real acronym. I've never heard that before. POC
actually reading through this there's a lot of things that we say that I mean, they they break them down into a you know, acronyms and things like that. And I was like, I've never heard any of these said as an acronym.
Same thing with EV T engineering validation test. I've never heard that before. Yeah, I would say that something you probably would use in like FCC CE testing. But I've done that before, and I've never heard a VT. Or maybe
it's written on the documents or anything or something like that. Maybe it's possible.
What's your next one? Let me let me sift through. Oh, wait, next page. Next page. NP mass production. Oh, really? That's not an acronym?
No, they gotta have three to be an acronym.
It's not the Gumina macro fab rule. Oh,
here's one that I actually didn't know. Okay, gift box. That's an acronym. No, it's not an acronym. It's under Gosh. What's the section? It's under? Where pages a filament. Oh, sh i got supply chain and logistics. Okay, a gift box. Apparently a gift box is just the thing that your your product comes in.
So just what your customer gets at the end,
they know it's the box. So when Amazon sends you something, it's the box they send it in. So but they but the jargon is gift box.
So gift box final individual box a customer receives gift boxes, mugs, gift boxes, mostly commonly referred to retail packaging, but can also refer to generic gift boxes from things like Amazon worryfree Oh, no, it's not it's not the actual box. Because they're saying Amazon's worryfree packaging. So like that's that's that's like you can check mark a gift wrap option no cuz
I mean look at look at the look at the image they have up at the top of the page. It's not a gift wrapping. It's it's a GoPro in its box and box, right? So it's interesting, whatever it comes in. I mean, I guess like you wouldn't call let's see you got an Arduino in that like official Arduino box. That's a gift box. That's the gift box,
not the not USB box that came from Adafruit in right, gotcha. Yeah. Interesting. I didn't know that. Yeah, I didn't know that. That was a new one.
I don't think I need to know that. Oh, and then on page 10 of 15 this is still logistics LTL LTL less than truckload less than truckload it basically means if you don't have enough to ship a full truckload, you have less than truckload
lots on obvious.
Yeah. So yeah, some of these are a little bit. I was really hoping
for FTL what? Faster Than Light
Some of these are ridiculous.
But this is actually a very good article to read. Yeah, for those starting out there, it gets a lot of jargon. Some of the jargon is, as we've shown is a little silly. Yeah, I went and picked the ones that are just like really? Yeah. But some are pretty good. Like GM. General Motors. General gets injured. No gross margin. Makes sense. We've only picked that like maybe five and this list is 15 pages long. Yeah. They could have just dropped out those five that are just complete BS. Yeah. And it was one of them. No BS. Maybe
that would be funny if it did any stand out to you.
Uh, not? Not really. I want to see what your opinion was. Yeah. So yeah, I think I think they could drop a couple of them that were low BS and it would been a fine article regardless.
Yeah. They probably just sat around and brainstormed a ton of and just
not gonna look on the whiteboard. What can we make an acronym? Oh, wait, that's what we do at macro FEM engineering
web whiteboard.
That Warner Brothers
ridiculous. Oh,
yeah. I think I talked about that one. Um, oh, yeah. EP, what does EP stand for?
Extended Play?
I knew you would say that. No, engineering prototype, according to them?
Is it really harder to say engineering protocol? Like, are you saving time and effort and money by shortening that to EP?
No, because I bet you if you if you email someone, I Oh, yeah, I guess got that EP from blah, blah, blah. They'd be like, what? Because they probably would actually probably think EP extended play vinyl. And then you'd have to send another email like, oh, yeah, engineering. You know? Prototype. Yeah. Engineering product. I didn't forgot what. Exactly. And so it actually wastes more time. In that regards. It explains most of these unless, unless, unless in your signature in your emails, you link this article. So they have to go and read it. So it's like it's like going to a What's that website? Urban Dictionary to find a new lingo that kids are using, right? Yeah, yeah, it's really the same thing. What's the lingo the engineers are used?
I went to Urban Dictionary today to read what their what their definition of tubular was.
Oh, yeah. tweeted that today. Yeah.
So So I think I'm going to be talking to our shipping manager. Now all about LTL every one of our shipments is an LTL now
not when we got the the machines in. We have actually three trucks bring it for my my daddy deal
right there. Yeah. What was that? BTL bigger than truckload?
Like loads to what? A TTL. Three truckloads? Three truckloads. Yeah, I TTL. OR logic, yeah, or logic. Anyways, good article. We'll put the link in the blog post. Um, I'll let I'll let you do this next one. Okay. Yeah.
I have. Okay, so actually, let me preface this. We talk a lot about data sheets and we talk a lot about connectors. Yeah, and how bad they are. Because there's a lot to say. Yeah. And I have a datasheet PSA for for everyone who is willing to listen here. So connectors that have more than one part that is inherent to the connector.
So like, like pins that have to go into the connector that you have to solder to the wires or crimp to the wires, you
have to crimp the wires and plug them in there. Or if you have like a little holder that clips into it or anything
even farther is like the male and the female connector. What actually does fit into that connector, right? Yeah.
So So here's the PSA, if, if it requires more than one item, put the damn information on the datasheet. The part number, the part number, give me even a picture would be good. I was looking for a connector. Just Just the other day for a customer. The customer gave me information on what he wanted, but it wasn't directly specific. And I found what was necessary, but I found the male and female shell of the connector, but it needed a clip in part and it needed a socket and it needed a pin. But it didn't tell me which one got which clip which one got what what pin which one got what soccer, the datasheet was just here's the dimensions of this plastic shell like, Thanks, great. I literally had to go to Google and do a Google Image search and find people who had these and figure it out from what parts you actually needed. And finally, I was like, okay, so this one takes the pin, this one takes the clip. That's what takes the socket. It was terrible. Put the information on the datasheet. The best
ones are, or the best data sheets for this kind of stuff, or when they do a, a family data sheet for that connector. And then you get all that information. Yeah. Please do that connector accompanies
a little note that just says, this part meets with this partner. Yes. Oh, my gosh, it's amazing. And uses
these conductors? Yeah. Yeah. How often will we talk about connectors, every other podcast?
Pretty much. That actually shows that we deal with connectors pretty often a lot. Yeah.
Or then Amazon earlier this week, released a, an IoT programmable button for developers. And actually, they are trying to get some of these for macro fabrics. We do a lot of IoT stuff. Reluctantly, sometimes.
That's another thing we talk about a lot. Yeah. Anyways,
so um, I think it was last year or the year before that Amazon released these buttons that you can put on, on stuff, I guess. And you can press them and it would automatically order stuff for you, like toilet paper, like toilet paper. So you'd have a button next to your toilet. And if you're on toilet paper, you'd press the button to put an order in for toilet paper? And I'm sure
Amazon Prime Now you'd have to wait an hour. Yeah, when
I wonder if they could come in. If you can make like, if you press that button, like it printed out a key for them, so they can come in your house and hand it to you too.
Can it be called the oshit button? We need to make this.
Well, we can now because they released a generic Amazon IoT button now.
And what do you just tie a product number to it? No, you
you actually get the program. It's okay. So you can make it do anything actually, it's basically a Wi Fi enabled button. Okay, and you can make it you can press the button and you can make it you know, brew a pot of coffee if your coffee pot had Wi Fi. Hmm, etc, etc. Yeah. So it's really cool. I haven't haven't looked too much into it. I'd saw that today. And like, oh, yeah, I want to talk about that. So I'm gonna see how much they cost how much a development platform costs for it. See if we can get some at Matt crab. Yeah, that's cool. Really cool idea. I really like it. It's one of those. I bet you I actually kind of want to get one and tear it apart. I bet you someone's already done a tear down. I want to see what kind of hardware is in there. Cuz I bet you it's probably like an ARM Cortex M zero. Which is probably actually way overpowered. For what this thing does a button and sends it over Wi Fi.
Well, okay, so once we've, once we see how much it costs, then we can make a judgment on what's inside. Because if it's like five bucks,
oh, it might have somehow they got a Wi Fi stack that fit in like a tiny eight. Eight bit.
MCU. Yeah, yes. Something something crazy.
Yeah, maybe. Well, they also Amazon's really good at the scale of costs. They're very good at reducing cost by scale. So is that is that a jargon word and bolt? Probably scale cost scale of cost SOC
soc. System on Chip? Yes. As much.
Anyways, I think it's cool. I'll save them and get some for macro fab. Awesome. And this is going to be on you.
Yeah. So I saw on Hackaday there's a there's a cool article about the monster 6502. Basically, the 6502 was a CPU in the gosh, I guess it was late.
80s? No, ah, no. early, mid to late 70s. Mid to late 70s. It was the the cut down version to 65 07. Yeah. Was in the Atari 2600. Yeah, the difference between those two chips was this. The O seven had less address lines. That was a smaller package.
Okay. Okay. So this processor was in a lot of it was in the Apple two, and the Commodore 64. A version of it was in the original Nintendo. Yes. It's a modified version hold decade later. Right. Right. Well, it was tried and true.
Yes. At that point. The it's actually interesting back in the day they would. CPU technology didn't develop that quickly. Back then. Like the 6502 architecture was used for a long time. Without any significant upgrades. They did change from Na style semiconductor to c MOS. And that was a big leap. That's a big leap because you you drop power a lot on that leap. Basically the difference in that is is nos is basically n channel only right transistors. So you almost have double the amount stuff, or see moss is both and type and p type, right?
And see most of smaller, yes.
But anyways, the interesting thing about this, this monster is it's all na style. So it's all n type. Yep. Or NP ns.
And even more interesting, a gentleman has gone out and made a discrete based board. That is a one to one copy of the 60 50211. With I'm sorry, a few small
changes. I'm actually surprised how tiny it is. It is 12 inches by 15 inch. Yeah, it's a lot smaller than I thought, oh,
it would it be? Yeah, yeah. Well, it has 4000 individual, a little bit over 4000 individual components on it, like 3900 of those or something. Something in that range are transistors and so 2933904 is probably something like that, what I'm thinking Jelly Bean transistor. Yeah. And they were they were talking about cost of this thing, because that's obviously the first thing that came to my mind. It's like, oh, my gosh, really? And he said, It's above 1000 and below 5000.
That's a true maker right there. Yeah, yeah. Well, you're not too worried about the cost, you still have a ballpark?
Actually, what I think is probably even more true to the maker thing. Somebody has asked him if it comes in kit form. And of course kit form kit form. If you think about a transistor, three legs, and then resistors, two legs, you have you know, 10,000 plus pins to solder on something like this. And his response was great. It was anything can come in kit form. If you try hard enough, if you try hard enough, anything is good for him. Because it's all surface mount though, right? Yep. Surface Mount Dual Side. It's actually only four layer board.
Ground power signal signal. Yep. On the outsides.
And apparently he wrote some kind of scripts to help with the the layout of you know, probably just like Step and Repeat kind of script. Yeah. Things like that. That would make sense. Can you do that Neagle actually has a ULP?
Yeah. There's a couple of yuppies out there that you can step and repeat nets and stuff.
Okay. Yeah. I'm sure. Yeah, he probably did something like that. Yeah. So yeah, it's cool. I mean, just to be able to, I've always geeked out on those things, being able to see it like that. Yeah. It's
a pretty cool project. Yeah. And then there was a 6502. Emulator you showed me today.
Yeah. So that's actually been around for a bit longer. In fact, this emulator, this website has a graphical representation of the dye of a 6502. And you can press play and watch it run through its commands and see every single line in that.
Yeah. And that's a visual 650 two.org. Yeah, I believe that's it. Yeah, I got right here. Okay. Yeah,
yeah. And we'll post the link up. Super cool if you nerd out on this stuff. And, and on top of that, like it shows all the layers and you can peel them away. So if you want to look at just the metal layer, if you want to look at just the poly silicon layer, you can see each individual layer it's, it's just a total nerd fest.
Now. You were you were I can't remember which layer you're looking at. But it was like, pulsating. Like back and forth. It was almost like, like your list like trance music background.
So yeah, it would make a perfect background to a electro dance beat.
Or like a rave or something like you put it up on a wall. Yeah. And really,
you could put that up there. No one would know what it is. Except you're you're the one guy in the audience. He's like, Yeah, that's a 6502.
Yeah. And the last RFO it's not really a news article or anything, but I saw a interesting comments, or post on the Arduino subreddit. And it was hardware or software D bouncing for switches. What do you prefer?
I'm honest opinion. Yes, both?
Both. I was gonna say both, but which one? If you could pick one, we do software. You know, it's really funny. We're both hardware guys. And we picked software. Because it requires it's cheaper.
It's cheaper and it's more direct. You don't have to deal with tolerance.
Yeah, you know, tolerances and we basically have to add two components to each switch. Yeah, that a cap and a basically a low pass filter cap and a resistor. Like port what would be like 100 ohms. And probably a point one microfarad. That part make a pretty good low pass filter.
Well, okay, so you can always shotgun it, you can, you can do the math and find your time constants and do all that kind of crap. The thing that's great about software is if it doesn't work, you type in another time. Press reprogram and give it another shot. Instead of having to pull out a breadboard and stuff. I get lazy when it comes to that kind of stuff. I'd rather do software. Yeah. And I've had better success with software D bouncing than
than hardware. Yeah, I've actually had same luck that too. It's also cheaper. Yeah. Well, the pins when you say cheaper because it does consume more cycles. But yeah, I guess you say consumes more CPU cycles, but it takes the same amount of time constant time. Because that hard work like the Herbert debounce is basically going to make a an RC circuit with a roll off. Yeah. And so it's going to have a instead of having a sharp transition that bounces back, you can have this nice, I'm gonna say rounded square wave. So you're going to have that strong square wave. But in that same round square very time, you're basically going to sample your pin twice to see if it's still high. So takes the same amount of physical time. It just takes more CPU time. Well, and it doesn't really, it only takes like one more x, it only takes a comparison on your CPU
on your comparison, but you have to write and run a delay function.
Well, if you do it that way, which is the worst way. So well, how I like to do software D bouncing is you record an old value. Yeah. And then when you run it again, you run like an interrupt for it. So you're not waiting for a time period. So you interrupt happens again, and you read your IO again. And then compare that to the old value. If a if a if you're saying one is a switch has been put pressed. Sure. So it's been pressed for both of those time periods. That's a press. That's that's a valid. And so it takes less computational time. It just takes one more interrupt cycle. Sure.
It depends on if you want to just pull in your main loop or if you want to have an interrupt, snagging things. Yeah, it one's easier ones harder. One takes more time one takes less whatever.
The my main thing is with doing hardware D balancing is it just takes more physical layout on your board. And yeah, I wonder if a 10k point one microfarad would be a good RC circuit. Because that's like a part those are two parts you use everywhere. On your on like a digital layout. Mm hmm. And those would work for a debounce circuit? Of course, which because then that way you're not adding any real cost to your board. Yeah. I different SKU numbers, whatever.
Well, check.
I think that would probably work. It probably really slow roll off, though. Comparison.
Yeah. Well, it would it would be worth experimenting, especially if it was like let's say this was going into a medical device, and it had to work 100% You'd have to do both. Yeah, well, I would absolutely do both in a situation like that. But I would, I would take the time and actually measured it measure it saved my waveforms, you know, all that kind of stuff. It just don't just take a lot of time.
It takes a lot of time. software's easier for that stuff. Oh, yeah. And that is the end of the RFO section. Is there anything you want to add to this podcast? Steven? No, I
think that's I think that's pretty good. This is episode
Got till next time. Take it easy.
Come celebrate the third anniversary of the MEP as Parker and Stephen discuss a potential video game.