Stephen and Parker chat about updates to the SSPS and FX Dev board.
Podcast Transcript
Unknown Speaker
Hello, and welcome to the macrophage engineering podcast where your host Parkstone and Steven Craig. Awesome. This is episode 28. If I recall, that's 28. That's what my sheet says.
Host 3
You have to look at the sheet.
Unknown Speaker
So yeah, we we don't have a guest this week. No. Um, so we've been basically hard at work at Mac fab. Working on some projects. Yeah. Getting
Host 3
back to getting the project moving again. Yeah, finally. Yeah. So, got some got some cool breakthroughs this week. The SSPs. I think we talked about having the analog board done a few weeks ago, I finally got it all hooked up, you got a whole prototype testbed running. So got the board, got a whole cap stack got some beefy rectifiers. And the big power transformer running,
Unknown Speaker
and you made a mess of our bench.
Host 3
Hey, that's what it takes, man. But But yeah, so got the got the whole rig. So it's a kind of, not kind of it simulates a full single channel SSPs right now. So big power transformers got 230 volt taps at 13 amps each rectify that up to 43 ish volts. And then on my analog board, I got two high voltage regulators that hammer that down to 35 volts, which is the maximum for the output op amp. And then I got a handful of low voltage supplies for, you know, powering op amps and things like that. And then all the digital controls and got that up and running this weekend. I was actually outputting some stuff.
Unknown Speaker
Yeah. And you had a prop those stick basically simulating the front panel sending the data.
Host 3
Right, right. So literally, the the output is just controlled by a single I squared C line. We have to their max chips. I can't remember the the actual number. We'll have to post that. But the it's to those days, right? Yes, 16 bit data as one of them. When you write a number zero to 65 535. That controls zero to up to 35 volts. So you got 16 bits for the positive, and then you can write to the other DAC, and you get zero to negative 35. So it's I think we argued about this a while back, it's 32 bits across the whole 70 volt
Unknown Speaker
line, correct? Yeah, right. Yeah, it does work that way. Right? Yeah.
Host 3
So I wrote up a quick little spin code to have it just output a ramp wave basically just writes. It doesn't write every value on through the 16 bit, it writes every 1,000th value. And I was able to get a full 30 It was 33 volts because there's some calibration but a full 33 volt peak to peak ramp wave. Nice sawtooth. Yep. Yeah, sawtooth and I think it was 10 Hertz. And I wasn't trying to time it. It's just however the processor hammered it
Unknown Speaker
out. Yeah. And you had Did you have a load on it? No, no,
Host 3
I haven't done a load testing yet. Because we don't have heat sinks. WaterBlock or any heatsinking whatsoever.
Unknown Speaker
We should probably just bolt a big piece of aluminum to it for the for now
Host 3
with insulators. Yep. Yeah. So the I was I was pleasantly surprised after some mishaps with putting caps on backwards.
Unknown Speaker
Yeah, that was scary. Yeah. So you know, these ginormous capacitors. And so this is the first time Stephen was turning it on. So we had the rectifiers basically with a hacked up cord that went to the wall suicide cable, suicide cable, he plugged it in, and immediately all the lights in our engineering area dimmed in there was a CAPS unplugged it and he had him wired up backwards. And it wasn't, you know, like dusty even. Why did you do that? Well, these caps are actually positively marked on the stripe,
Host 3
but they have a big solid stripe down the positive side of the count. No. Yeah, this was a dumb thing. Because surrounding that positive stripe is a whole bunch of cool saw signs. Yes. But I looked at it. I was like, oh, it's got a big stripe on it. That's got to be negative.
Unknown Speaker
chalk that up to you did it before lunch. Yeah. So Stephens not allowed to play with high voltage before lunch now.
Host 3
But flip the cats and the funny thing was, or the cool thing I should say is first time I fired up the board. Everything did
Unknown Speaker
what it was supposed to do. And I'm actually surprised you didn't blow up those caps either.
Host 3
They look beefy. And I did not have it plugged in very long. I heard some noise was like oh shit. Yeah, to pull out the lug?
Unknown Speaker
Well, then the real thing is we put the meter on on the cap. On the caps. Yeah. And we were actually we're measuring the voltage and the voltage didn't get up to like 10 volts.
Host 3
Well, it went 10. And then it started to drop. Yep. And I was like, Oh, this is bad. This is not good
Unknown Speaker
now, but we flipped them around, plugged back in and got up to was it 4040? Something? Yeah, 4040. And they stay charged for a long time, even with your bleeder resistors.
Host 3
We have eight caps, and I have a bleeder resistor on every cap. The value of the bleeder resistor is 22k. So 22 divided by four whatever that ends up being. I don't I'm an engineer. I can't do math in my head. But, but yeah, that acrossed point oh, four fair adds. Takes a long time to discharge up. And I didn't want to put lower yet because I we only have quarter watt resistors. Yeah, I would just burn them just from discharging. Yeah,
Unknown Speaker
I think we need to get a discharge screwdriver that you use for working on on TVs? Yeah. Basically, it's got like a, I think a one meg resistor, and you just touch it, the stuff and it discharges through, I actually
Host 3
have something similar to that for my, my amp work. It's, it's actually a really big cardboard tube with a with a probe at the end of it. And it's got a 220k wire around inside of it. And the other side is like a really beefy alligator clip. And you just hook that to chassis and touch whatever you need.
Unknown Speaker
Interesting. Yeah, yeah, just bring that in now work. But
Host 3
the only thing that sucks about that is 220k it would still take to discharge.
Unknown Speaker
So we need something like 1k or less or less, and then a beef enough, you know, beef enough resistor to handle that heat dump. Yeah.
Host 3
When when we get to finalizing everything, we'll probably want to get like five watt resistors and put them across the caps. And you know, something reasonable, like 500? ohms? Oh, no,
Unknown Speaker
no, all we can do is actually in the design, we should put a relay that that you can enable to discharge the caps. Oh, that's a good idea. That way it's already built in. So when you open it up, you type in Hey, I'm going to be taking this apart. Yeah. And then you punch in, you know, the command for discharging caps. It the relay turns off the the input supply. Yeah, to the caps. And then it turns on this relay that dumps everything through big resistor. Yeah. And Bam, done. Dig it. So this is not simple at all,
Host 3
no. And I came across one issue that we're going to have to address, okay, and it comes with turning off the power supply. So, on my board, I have 235 volts plus or minus regulators. But then I have a plus minus 15, I have a five volt and a three volt 3.3 volt. For the digital stuff. Well, the 3.3 volt, all those low voltage stuff are actually on a separate tap on the transit transformer, yep. When you unplug it from the wall, the high voltage regulators still have a whole hell of a lot of charge on them. But the low voltage stuff drops out real fast. What ends up happening is you get a giant spike to the input of the op amps. And the output of the op amp even if you hadn't driven to ground, it still shoots up like 20 volts. Yeah, that would that would kill whatever you have attached to the load. So we need to think of a couple of ways to fix it. And it could be something as simple as the output just goes through a relay. And when you click it off, when you kill power, it opens that relay.
Unknown Speaker
Yeah, make it that that relay is powered by the low power side.
Host 3
Right, right. So it kills that. There's there's some other stuff we could do with pulling the ground.
Unknown Speaker
Really, we're going to have the outputs connected to a relay. So we just use that relay.
Host 3
Yeah, it will just we'll just have to make sure they're timed. Right.
Unknown Speaker
Yeah, yeah. Um, and it might be where we beef up the capacitance on the on the low voltage side too, so that it lives longer if like, let's say you like the power went out to the shop. Yeah, yeah. Cuz you have to, you know, pay attention that too. Yeah, yeah,
Host 3
yeah, we could we could certainly do that. We could also technically do like a soft power.
Unknown Speaker
Oh, yeah. The the front panel is going to be soft powered.
Host 3
Okay, well then in that case, we can we can have it control that's with intelligence. Yep. Yeah, yeah, that might be the
Unknown Speaker
more we can do is actually put a battery backup on the front panel. Yeah, so it so that if it loses power, it can detect that and actually properly shut stuff down. I mean, you put like a big 18 650 cell on it that it'll can last hours. Because why not? Why not? We have the space for it. Yeah, I like it. Yeah, cool idea. Okay, so Eric's the board update on that. The
Host 3
enclosure finally came in and it looks awesome. We've been waiting weeks for this enclosed Yeah, when I was kind of sweating bullets because I had to pick a color for it. And we wanted to use macro fab red. But when it comes down to picking color color, yeah, like, there's just gazillions of colors to pick from. And so I was literally sitting there with my computer screen when I was defining it and just trying to find the best one. And I think I landed it. I think I think it looks really good. It's, it's nice. And the board fits really great. As strong as hell. And I'm digging it.
Unknown Speaker
Yeah, it looks awesome. Yeah, I think the only big problem is the it's actually fits a little too well. Yeah, yeah. It's a little too tight. But that's okay. That's, that's, you know, loosen up the tolerance a bit.
Host 3
Yeah. Cuz the powder coating when they when they fired it, it actually pulled the the, the wings in because it was a sheet metal. I think it was like 19 gauge or something like that. And it kind of shrunk the dimensions just just enough to make it fit. Really nice. Yep. And really nice, meaning it needs some more room. Yeah.
Unknown Speaker
So I guess we need to get the magnets and the rest of the hardware. And then you have the next revision of the FX dev board, which is the final one. Yep. That has everything lined up perfectly.
Host 3
And the footprints are right for the breadboard. So this time we tested that yet. No.
Unknown Speaker
Okay. So because it has no tomorrow morning.
Host 3
Yeah, well, I'm confident this time.
Unknown Speaker
Confident. You heard it here. First. You'll check out you'll check out Twitter tomorrow morning. If is like curse words on Stephens Twitter accounts
Host 3
because of crappy data sheets.
Unknown Speaker
Oh, that sheets? Terrible.
Host 3
Yeah. So that's moving along. I'm really unhappy with the enclosure.
Unknown Speaker
Yeah. So I've been doing just basically bookkeeping this entire week. And last week, updating Eagle footprints. We've made some changes to our what our footprints look like, and how this the symbols are. And so I've been updating all that stuff, updating all documentation. We are updating. Dip trace next week. Yep, we should be getting our CAD footprints soon. Mm hmm. Dustin holiday, he's been on the podcast twice already. He's working on that. And that should be ready by next week. So hopefully next week, we can push out, you know, the update DTrace. And then actually have KiCad libraries.
Host 3
These are gonna be all the libraries, they're also going to be DRC and Cam rules. Yep.
Unknown Speaker
So you'll be able to download, basically templates with dip trace, and then files for KiCad. And make sure that you when you upload your boards to macro fab, you actually can build them.
Host 3
Right, right. And they're kind of drag and drop, where if you use the templates, or the files or whatnot, you know, it'll fly through macro fab. Well,
Unknown Speaker
exactly. So, you know, that's all that bookkeeping, that's been basically the bane of my existence the past two weeks, trying to keep tabs on all that stuff. Yeah. And that was Josh's ringtone. He said, sorry. It's okay. I'm Nick tomorrow. So Friday, we'll be releasing pin headers as health parts. roadmap. Whoo. Yeah,
Host 3
we've we've kind of been talking about that for a long while.
Unknown Speaker
Long time, but we finally worked out how it's gonna work. Yep. How it's gonna work on mainly in the back end on operation side, like, how are they going to handle all these pin headers?
Host 3
Right, right. How do you how do you manage it? Because it's, it's what is it two all the way up to 40. Pin one up to 40? We'll for a one pin. Yes. One pin. Awesome. So one 240. How do you handle inventory? That's an
Unknown Speaker
inventory for that. And so we got we got it all worked out. Yeah. That should be going live tomorrow. Awesome. Hopefully the about probably about two o'clock, three o'clock PM. It'll be live on the house Part page. And we'll make a Twitter probably post about it. Yeah. Because we're always should make a blog post about it.
Host 3
Oh, yeah. Whoa, we'll hit social media.
Unknown Speaker
Speaking of that, um, you're going to have an update for the SSPs on the blog tomorrow, right. That's
Host 3
right. Yeah. So I wrote up a so we've already had to design blog post for the SSPs. And I finished up the third, which should go live tomorrow. Yep. And it kind of covers what we talked about earlier. All the all the testing and bunch of images of my hacked up prototype job. Yeah. And you can see how dangerous it is.
Unknown Speaker
It's actually a really cool article. It's got a lot of good information on there. Also about testing and that kind of stuff. Yeah. Yeah, so we'll have two articles or two blog posts out next, tomorrow. Mm. Well, three, including this podcast, right? Yeah. So it's gonna be a big day. Probably make Abel, who's our marketing person like insanely crazy? Oh, he's
Host 3
gonna pull us here. Yeah, he's like, why are you doing three one
Unknown Speaker
day? I'm like, because we need to. Uh, let's see. Oh, then earlier this week, I had a blog article come out about how to do attributes in Eagle. So go check that out. It's, you know, most Eagle veterans would only know how this stuff works. But you know, some people don't know. Yeah. And the cool thing is, with the attributes is, if you put NPN and populate into your Eagle part attributes, we actually pull that in at macro fab. And so your manufacturer, part number, get pulled in, or the Populate field will get pulled in populate can be a 04 Non populate, or a one four populate. So you have like a part that you you want in your schematic, but you don't want it to be placed on your board, you put a zero for that field, right. And the Manufacturer Part Numbers will be good for basically, it's the to make the readability of your schematic better. Yeah. Because let's say you're using a specific cap, because it has a, an ESR rating, you know, an a series resistance that you actually need to have. So like, let's say it's for switcher, okay, power switcher. When you look at your schematic, you don't want to look at a part number for the Value field, right? Because you have no idea what that most time with that part number actually means unless you've memorized everything, yeah, memorize every single part. So use your put like 10 microfarad, right? Well put that as your value field. And then if you in the attribute field for eagled, you can put that actual part number there, right, and when you upload it to macro fab, it'll pull that NPN field out, and bam, automatically grab that that part number, put it in populated up. Good go. Neat, neat.
Host 3
Excel, I like I like the whole populate thing. Something I've done in the past. And a lot of a lot of people do this, it's really convenient is to just populate a whole bunch of like, oh 8050, ohm resistors, put them all over your board, when you're prototyping, but make them non pop, such that if you need it, you can just plug something in there or jump across it. And that makes that makes life a lot easier. And on macro fab we we have the ability on the user interface to select a part as non pop. But now you can do that in your EDA tool. Yes. Which is even more convenient. Yes, it is.
Unknown Speaker
And so we will be working on the dip trace libraries that come out next week. And the key CAD stuff will have these kind of attributes. Yeah, now they won't be automatically pulled in yet into macro fab yet until we get that back end stuff working. Right. But it's a good
Host 3
start with both those others, we can do that. We just have to kind of define how we do it. Yes, exactly. It's a bit easier with Eagle.
Unknown Speaker
Yes. Yeah. And then on Episode 26, which was two episodes ago, I talked about having like a hack together. USB type C. Yep. Working. Well, I finally actually got my board built. And it works. So I have a type C connector that actually uses the standard manufacturing practices. So just the standard DRC rules for macro fab. It talks over an FTT 30x FTDI, USB 2.0 USB bridge chip, and then to a parallax propeller. And it all works.
Host 3
And it worked on the first go round right worked on the first go,
Unknown Speaker
No, zero rework. That's That's awesome. No green wires. It's pretty awesome.
Host 3
Unknown Speaker
You know, I'm not bad. But yes, it's, it's pretty rare. Yeah, for a board that for a first time go round. That's rare. Yeah. So it works. I'm pretty I'm pretty happy with it. Awesome. And that's because she, so I'm going to have a blog post in the future. It's gonna be a couple months on the road about type C, and all that other good stuff. Yeah, you've written a lot about that, aren't you? Yeah. It's almost like an e book we're going to be putting out for that. For Type C, obviously, like, why, why is it important for makers and engineers start using this connector, even if they don't need USB 3.0 Because the great thing about it is you can use 2.0 on it. So it's backwards compatible,
Host 3
because as d plus d minus lines, yep. Yeah. You just make it like the Tome of type C or something, you know, like, like, you know, like old dungeon dragons books, you know? Yeah, a player's guide to type C.
Unknown Speaker
Oh, like that. Engineers guide to type C chapter one. Are we making ebook? No, not ebook I audiobook.
Host 3
Oh God, that would be terrible. But But if Morgan Freeman did it, I would
Unknown Speaker
be whole thing all his voice, you can just listen to anything he says. Okay, I think that's up for all the projects we've been working on. We kind of rambled a bit. But yeah, that's certainly what we do RFO RFO. So there's some research into a new type of battery. And say dissolving battery when they say transient electronics, which I guess is code word for bad or electronics. I only get used once, or they are short term, like a transit event. Sure. I guess that's the only that it makes sense. It makes sense. Yeah. It like dissolves like,
Host 3
well, or something like Yeah.
Unknown Speaker
So it's a lithium ion battery battery that provides 2.5 volts. And this is what the article says, yeah, it can power a bad laptop for 15 minutes. This 2.5 volt battery. So what is
Host 3
2.5 volt dissolving battery? can power a laptop for 15 minutes. That's what the article said they claim
Unknown Speaker
and the size of this battery is five millimeters by six millimeters by one millimeter. So I did some quick, dirty math. So this math might be wrong.
Host 3
You have to warn everyone. Yeah. Cuz we've been rolling
Unknown Speaker
in, in the past and been called out on it. So first, I went like I typed into Google, typical laptop batteries. sighs Yeah. And usually it's 60 Watt hour is the typical size. Okay, so just taking notes. So took 60 divided by four get 15 Watt hours, right? 15 minutes is a quarter of an hour, right? And so our battery is 30 millimeters cubed, or point oh, three centimeters cubed, which gives us 500 Watt hours per centimeter cubed, which is enormous energy density. I don't know what energy density that comes close to. But I'm like, Okay, let me do some math on a typical battery, like an 18 650. lithium ion battery. Yeah, that's got a watt hour rating of point six, four watt hours per centimeter cube. three orders of magnitude No, four orders of magnitude less. So
Host 3
so they're claiming the world with this battery?
Unknown Speaker
Yeah. Well, I think the article probably has something messed up with what their size or the fact that that laptop is, you know,
Host 3
well, so maybe the instead of saying 15 minutes, it was supposed to be 15 milliseconds, no seconds. It's also kind of funky. Because there's not a laptop that runs on 2.5 volts. No. So they were clearly talking about power.
Unknown Speaker
Yeah, it's probably multiples of these can power for 15 minutes. It's kind of silly when it especially since the article was on a, like electronic engineering website. Let me see which one that was on. It was on electronics weekly. So it's like, where is all that is clearly like a copy paste job. from some other it's an aggregate. Yeah, it was suffering from some other. You know, when it got press release press kit stuff. And the person who wrote this article wasn't even paying attention, or even did the simple math well, and
Host 3
those kinds of numbers. A hit home for people who are not inclined to actually understand what the numbers are. They look at it. They're like, Oh, my gosh, I can put this liquid all over my my laptop and it runs for 15 minutes. Oh, yeah. Now that sounds like a I'm calling BS on it. I'm calling BS on those stats. But okay, so here's the one thing where it might be different. If you don't think about it in terms of watt hours, but think of it in terms of total joules that it has. If you compare those two numbers, the amount of joules that are in an 18. What is 8650? Yep, compared to this thing? They might have comparable jewels, but this dissolving battery can't provide it quickly enough. That could be an option. Yeah. But that's a really far stretch if you ask me. Now,
Unknown Speaker
I would agree. Yeah. And so apparently, so what makes us batteries special is it dissolves? So it's for you no one time use devices? Yeah. Because apparently it's it's kind of polyvinyl alcohol based like enclosure. And when it gets attached with when When water hits it, it kind of just dissolves. And this is what they say is that the power casing swells and the electrodes are broken apart, causing it to dissolve except for some nanoparticles, which they don't mention ever again. So what happens to the nanoparticles
Host 3
Okay, wait, wait, wait, wait. So you get this thing that they like, I'm assuming it's kind of like blow sticky where two materials come together and a reaction happens. Yep. Okay, great. How do you extract the energy from it? Well,
Unknown Speaker
it's got, it's got, um, the anode and cathode on the outside.
Host 3
Okay, but you but you're gonna have, you're gonna have reactions going throughout the whole thing. So you're clearly not gathering all the energy. No,
Unknown Speaker
no, it's like a typical battery, but the materials that's made out of breakdown and contact of water. Weird, except, except for nanoparticles, which it doesn't say what? It could be cancer.
Host 3
That's just leftover crap.
Unknown Speaker
It just is. The batteries dissolve into cancer? Was what would happen if you were in California? Right, right. And then I was thinking, you know how humid it is here in Houston. These batteries would probably dissolve. You just had them sitting on your table.
Host 3
Oh, yeah. Guaranteed. You know what happened? Yeah, cuz I mean, we got we've got gallons of water to float around us right now. Yep. Wow. So
Unknown Speaker
yeah, it's a BS BS. Wait till we see some practical applications. Yeah. And then this one's not really a news article. But a cool project I saw. It was a we like talking about Pokemon, and especially pokemon go on the show. And IoT devices. And it's combines both. It's a pokey ball that wiggles When Pokemon are near for. For the Pokemon Go app.
Host 3
Unknown Speaker
this guy, the hardware side is the way do you have
Host 3
like a belt like in the show where you have a couple of these on your belt and your belt starts to wiggle?
Unknown Speaker
That'd be awesome. But no, video doesn't show that he shows the pokey ball just sitting there on his table wiggling. Okay. hardware wise, it's pretty simple. It's a 3d printed pokey ball. It's got a server with a big weight. And so when the server moves, it makes it you know, wiggle around. Yeah. And then and then uses a photon, which is by particle IoT, the the photons are really popular IoT platform and server and bunch of other stuff. Yeah, and then apparently, Pokemon Go has an API. Do they really? Yes, they do. Hmm. So he uses that API to basically figure out where it's at. Yeah. I think usually Wi Fi. So thing probably uses the Wi Fi location. And then and then you know, walks up hey, is there any Pokemon near me? And it says yes, then it wiggles? Well, it's nifty. Yeah, it's pretty cool project.
Host 3
I actually Okay, so I've got a complaint about Pokemon Go. Okay. I think I told you about this the other week, but it's worth telling everyone. Pokemon Go has messed up my YouTube watching schedule. Oh, yeah. So okay, I get this. When I'm driving into work, I like to just listen things on my phone or even at home, you know, watching watching videos and and I pretty regularly watch the Eevee blog.
Unknown Speaker
It's probably a pretty popular thing with people who want listened to our show.
Host 3
Yeah, I'm pretty sure our audience knows that. So for not for a short period of time. For years, I've been able to go to youtube on my phone and type in three letters. Evie, and the very first search that pops up is Evie blog. Now when I type in the Eevee, it pulls up the Pokemon Eevee as number one, or is that a Pokemon? Yeah, it's a it's a it's a Pokemon. And not only the first search, but the first like four searches are all. So now, because of Pokemon Go I have to type in Evie B to get to the
Unknown Speaker
core. One more keystroke to do you think um, YouTube's algorithm would know better?
Host 3
It should know me because my searches on Eevee blog and Pokemon vastly outweigh each other. Yeah. Although I have for poker,
Unknown Speaker
I think we call it with the Pokemon hardware project.
Host 3
Unknown Speaker
got to think of something, I guess well, and it has to be IoT. You ask me IoT.
Host 3
It has to be IoT and it's got to be really stupid. Yep.
Unknown Speaker
So we'll think of it I can't think of anything right now. Can you?
Host 3
Actually I mean, like, a lot of the cool projects are already being made like GPS hacks, the cool poker decks. Little Box. Yeah. All that wiggles bocce ball that Wiggles. We'll have to come up with something.
Unknown Speaker
I wonder, I guess I look into API more but it'd be really cool if you could play it with the pokey ball that Wiggles? Yeah, so it like like making sure it vibrates when Pokemon near And then you throw it. So the text up you through it. Yeah. And then it tries to capture it. I think wiggle does
Host 3
actually know a guy already created that. It was on Hackaday like two weeks ago. Yeah, he had, gosh, I can't remember how he did it. But he had an actual poker ball that he throws, and his phone sends to him throwing that, and it completed the action also. Oh, that's cool. Yeah. So actually, along that lines, what could be cool is if you had like a, like a capsule or a pod or something, and you could throw it anywhere, and it would pop up a pokey stop, or you could like plant a gym somewhere. Now, you'd have to somehow put that into their system, but that would be cool.
Unknown Speaker
Yeah, very cool. Yeah, I guess we'll look into it some more. Yeah. And that's gonna wrap up the our first section, right? Yeah. Cool. And that was the macro fab engineering podcast. We are your hosts, Parker Dolan
Host 3
and Steven Craig better guys. Take it easy.
