- Stephen talks about the Stupid Simple Power Supply (SSPS). He designed a OPA541 breakout board for testing. See figure 1.
- OPA541 Datasheet : Street price of $21.88.
- Parker brings up that blinky LEDs are the most important part of a project. He is designing the control panel and chassis for the SSPS.
- FTDI is again at war with counterfeit parts.
- Microchip buys out Atmel for $3.8 Billion. The consolidation of semiconductors continues.
- France building a huge solar roadway. This again?
Special thanks to whixr over at Tymkrs for the intro theme!
About The Hosts
Parker Dillmann is MacroFab's Co-Founder, and Lead ECE with backgrounds in Embedded System Design, and Digital Signal Processing. He got his start in 2005 by hacking Nintendo consoles into portable gaming units. He also runs the blog, longhornengineer.com, where he posts his personal projects, technical guides, and appnotes about board layout design and components. Parker graduated with a BS in Electrical and Computer Engineering from the University of Texas.
Stephen Kraig began his electronics career by building musical oriented circuits in 2003. Stephen is an avid guitar player and, in his down time, manufactures audio electronics including guitar amplifiers, pedals, and pro audio gear. Stephen graduated with a BS in Electrical Engineering from Texas A&M University.
Host 1 00:10
Hello, and welcome to the macro fab engineering podcast. I'm your host, Parker, Dolman.
Host 3 00:14
And I'm Steven Gregg.
Host 1 00:16
And so this week we will be talking about our stupid simple was it simple, stupid,
Host 3 00:20
Super stupid power? Super
Host 1 00:22
Stupid, powerful supply. Yeah. And so once you first introduce our, what we want to do with this power supply, so Okay,
Host 3 00:32
Let's, let's take a step back and look at kind of the way that Parker and I kind of envision things. So we sit next to each other at work, and we just pass ideas by each other all day long. And, you know, I would say something like 90% of ideas are just absolutely stupid, which is fantastic. I love that. And but we we've been wanting a power supply for our bench. And of course, when you start talking about a power supply, you start you know, putting together what you want the the requirements, yes, Spec wise. Yeah, the specs. And of course, it just keeps growing and getting bigger and having more features and doing more and more ridiculous stuff. So we ended up creating kind of a beast of a power supply, at least in our heads. It's an hour trying to set forth and actually make something
Host 1 01:19
That makes something that at least meets most of the specs. We do have a regal three, what 380 Whatever it is, yeah, I can it's a was a three channel 30 volt max. Yeah. 15 amp.
Host 3 01:30
Yeah, the one that has only I can't remember the dial indicator option on the
Host 1 01:35
Front. Yeah, it makes it look like a like a speedometer. In a car. He presses a button. It's a little silly, but it's a really good power squat.
Host 3 01:43
Yeah, we've been
Host 1 01:43
We've been real happy with it so far. But the problem is it only goes up to 30 volts. I think it's five amps per channel.
Host 3 01:50
You know, it's three amps for for both the 30 volts, and then it's five amps for the five volt. That's right. Yeah, that's right, which is pretty typical. Actually,
Host 1 01:58
That's a it's a pretty beefy subline it weighs. It weighs enough to bend the shelf it's on. But anyway, so we want to basically make a more powerful power spline and also be able to do kind of like high powered signals as well with it. basically be able to output low voltage, AC and that kind of stuff as well.
Host 3 02:21
Yeah. So one of the ideas that that came up is more like what if what if you have a dc coupled power amplifier? Effectively, basically, so So what if you can basically output a DC with an AC wave riding on top at 10 amps or something? So just just an absolute beast, and,
Host 1 02:42
And so we, well actually, you found this really cool op amp? It's how did you find this op amp? Well, okay,
Host 3 02:49
So So occasionally what I'll do is I'll just get on Mouser Digi key or one of the big guys, and I'll just go look for the most ridiculous thing I can find. Like, go to the resistors and go filter everything like what's the biggest resistor, or most expensive? Yeah, almost expensive, like just for fun, just kind of thought experiment kind of stuff. So I did that with op amps one day, I was like, what's what's the biggest, beefiest op amp I can find? And I stumbled across the OP a 541. Made by Bernie Brown. And this thing
Host 1 03:20
I think it's also I think T ball says T at the bottom. Oh my bed? Yeah, yeah. It says bird Brown from
Host 3 03:28
Texas and Texas Instrument I guess. I guess Bert Brown did the design and Texas Instrument bought it out or something
Host 1 03:32
Or Texas bought beurre. Brown, I mean, could typically do that. Yeah.
Host 3 03:37
Burger and actually has a bunch of really good chips they they're pretty well known for for their quality. So I was looking through this op amp and what basically, it's almost an all in one package. It's the the the tail end of an amplifier, I guess you could say. The thing that's crazy about it is it works just like a regular op amp. But its output capacity is just obscene. It has rails that you can, you can take all the way up to 40 volts. In fact, the supply voltage, negative to positive is 80 volts on this thing, and its output is up to 10 amps, I believe, with a power dissipation of 125 watts in a 220 wide package. Or up to
Host 1 04:25
221 125 Watt. Internal Yeah,
Host 3 04:29
Yeah, yeah. So it's it started scratching my head, it seems like we might just be able to use this chip and get almost everything you want out of it. Because like I said, it works like a regular op amp. So you can just apply feedback and get gain right right there. But yeah, so just putting a signal into the input. There you go. You got you have your your capability of doing high current high power output almost directly from a micro
Host 1 04:55
Yeah, and actually reading the the feature list The best thing about this is industry standard pinout. I, is there an industry standard print out pinout for op amps? Or is it just Burb rounds industries?
Host 3 05:12
You know, when when you look at like a single and dual package op amps, they are there. There isn't industry standard yet. But, but for this to 220 11 pin pack, and I'm sure there's not a standard. Yeah,
Host 1 05:27
Especially we're 40 watt 125 Watts. Yeah. And speaking 100 Franklin watts. We're gonna have to sync that somehow. And either it's gonna be big copper heat sinks, or we could just water cooler. So the Zucker
Host 3 05:41
Water cooler be fun. Yeah. So yeah, I'm actually I designed up a kind of a breakout board that has two of these packages on it. So hopefully we can pump 250 watts out of this little breakout board.
Host 1 05:54
Things can be size of a playing card.
Host 3 05:57
Yeah. Yeah. It's gonna be it's gonna be monstrous.
Host 1 06:01
And let's see. Applications motor driver, servo amplifier Synchro excitation. I have no idea what that is. Do you know what that is? My best
Host 3 06:13
Guess is it's some kind of motor coil, that it that it requires a ton of power.
Host 1 06:19
Yeah. Synchros. Well, Synchros remind me of the gears and transmissions. Yeah, but I don't know what that means in terms of op amps. And then audio amplifiers, I guess you can probably use this as an all in one audio amplifier. But yeah,
Host 3 06:33
Just capacitively coupled with the output. And then just honestly, I bet you could just make a power supply to supply the 40 volts to this. And plug, plug it honestly your your audio input from whatever source, apply the right amount of feedback. And you basically have a single chip 100 Watt, 125 Watt, you know, monster here.
Host 1 06:56
So and then, of course, number five, and this is programmable power supply. So I betcha someone's used this before can make a silly stupid power squat.
Host 3 07:07
And actually in the in the datasheet, they, they have examples of putting these op amps in parallel. For increased current output, just in case you know, you need more
Host 1 07:17
Like for these, we're going to supply it with 240 through 40 volts from the from the utility closet. So here in the States, you don't have to 40 Most outlets, you have to get to have another line.
Host 3 07:33
Jump across the lines. Yeah. So
Host 1 07:35
You have to have, you basically have to have an outer phase 120 volts, right. And then you get to 40. When you combine those two, well, on opposite sides.
Host 3 07:42
So we haven't talked about the one downfall or the one downfall that I can think of right now. In singles. This chip cost almost $23 What does
Host 1 07:51
Everything Yeah, it like the only thing it lacks? I mean, I haven't checked the datasheet completely, but it doesn't look like it comes with a kitchen sink.
Host 3 08:03
Host 1 08:04
Um, voltage. So if you do five amp continuous then so 10 amp peak
Host 3 08:10
To 10 amp peak, but what in one of the applications I was looking up for this when they were saying on the on the the parallel operation you put to these in parallel? They could do 15 amp continuous? So I'm not sure how that how does that work? But I don't know, I'm kind of tempted to use usually what what
Host 1 08:31
Is the rate the peak is, is what it can do in a transient for a short period of time, because it's way over its thermal junction rating.
Host 3 08:40
Yeah, right. Right. Right. And there's a short period of time the the temperature is not going to rise to a point where it gets damaged.
Host 1 08:46
Yeah, I mean, it says 125 degrees C operating. Wow, that's really hot.
Host 3 08:53
I think we're gonna have to have a monster heatsink on these things. Yeah,
Host 1 08:55
I'm thinking water cooler with with dual 120 fit 120 millimeter fans and giant heater core coolest guy. You got to turn this thing on and then the power will dim and the whole building.
Host 3 09:09
Well, the thing that's great about it is we can just program up a micro with a DAC on the output and and slap that right in front of this guy and then you have a full on controllable power supply effect, you know would be kind of cool, is it a digital pot in the feedback loop. So you can not only control the input, you can also control the gain of the whole thing. So you can you can reduce the gain and get higher accuracy at lower. And then when you need to go up to you know, say 40 volts at a couple amps. You can pump it up that way. Remember pretty cool. Yeah. So, Porter, you got you got some ideas for the front plate on this thing, right?
Host 1 09:50
Yeah, so the whole power supply is going to fit into a four year rack case. And so to make it look cool, because that's the most important thing about engineering. Well, first of all look cool and blinking LEDs. Oh, yeah.
Host 3 10:03
Who cares about functionality? Yeah, who
Host 1 10:04
Cares about the functionality? I mean? So the whole idea is to make it all led up on the front end, basically, you know, old school power meters. LED readouts basically make it look like the interior of the DeLorean from Back to the Future. Awesome. Yeah, it'd be pretty sweet. If I couldn't make neon tubes work. It would totally had neon tubes.
Host 3 10:26
Talk about adding noise to your your output though. Yeah.
Host 1 10:29
Oh, just isolate that part. That's all it's all for show. When you turn it on. It's just like the border. It's just the border glows neon. You want to use cold cathode?
Host 3 10:39
Rope? Well, okay. Yeah, good to have you. Okay. So yes, you can isolate. You can isolate those those things. But try plugging both of those into the same power strip. Even if you're isolated, you're going to get garbage on the Ask for sensitive stuff. They neon tube spits so much garbage back online.
Host 1 11:00
It's it's not. I actually have an old Budweiser neon sign. And when you plug the thing into the shop, you can hear it's actually only the far force to start to go. It doesn't flicker, but it does not like it definitely does not like it. Yeah,
Host 3 11:16
LEDs are quite a bit quieter. Yeah.
Host 1 11:19
And so we're going to be probably the microcontroller will choose the control it is probably the parallax propeller, because one price is an object. And two. I actually really liked programming spin, which is, I find a lot of engineers seem that's to be really weird, but you know, whatever.
Host 3 11:38
If you like it, you like it. And you like
Host 1 11:39
It. You like it. I like this tool, the tools for so cool.
Host 3 11:45
Host 1 11:46
Oh, that's so spacey design. I copy and paste that.
Host 3 11:50
Yeah, if you have a backlog of code, then it makes life a lot. Exactly. Cool.
Host 1 11:56
So I think that's going to do it for what we've been working on. Yeah. Let's go move on to the the rapid fire questions section. So recently, Microchip is buying out Outmail. And so this is comes after C comes after Intel buying out Altera and this is a huge deal. It's like $3.6 billion, something like that, which is That's an insane amount of money. And in fact that I think it's almost all cash. Which first of all, I didn't know microchip had that much cash on hand. I
Host 3 12:35
Guess they're bigger than we think. Yeah, I
Host 1 12:37
Guess so. So what do you think about this merger?
Host 3 12:41
Me personally? Thumbs up on this. I started my whole micro learning experience with pics. So I I say thumbs up. It's great in my opinion.
Host 1 12:53
What do you think about the what do you think's gonna happen with the toolsets losing on version will keep them separate
Host 3 12:58
Or personally I would like to see a merge because both sides of the of the realms are so big. And there's so many fanboys on both sides. I'd love to see them combined together to make something better than each one of them individually.
Host 1 13:13
That makes sense. I'm just interested when that will be able to buy up pick eight Mega 328 P. Kinda like AVR with like the peripherals and microchip Yeah. Okay, so got that. FTDI gate 2.0. So apparently they're messing around with drivers again, working devices.
Host 3 13:34
Well, it actually in this one, they're not necessarily bricking devices. They're spitting back garbage. On on the counterfeit chips.
Host 1 13:43
Yeah, counterfeit chip found or something like that's what's spitting out the serial bus.
Host 3 13:48
This one's really tough. I was actually reading some stuff about it today, the whole bricking of devices. That doesn't seem like a good move by by FTDI that just completely shoving people out of the market just doesn't seem like a smooth move. And they they got with Windows and updated their driver. So that doesn't happen anymore. But now they're spitting out garbage information and effectively breaking
Host 1 14:14
Them. Yeah, you basically break any device that actually requires that serial communication.
Host 3 14:19
Right. Right. So yeah, you break it again. Now, is that is that really the way to go? I mean, it's hard because they're trying to protect their investment. They spend a lot of time on it. Somebody is stealing it. And in fact, you know, they write cease and desist letters. That company disappears and a new one pops up the next day doing the exact same things actually the same people. Yeah, it's a tough situation. I you know, honestly, I can't blame them for protecting their investment. Yeah,
Host 1 14:46
I think they should. I mean, I think it's well within their right to do this. Yeah. Now, is it the right way to actually respond? That's probably not my decision. But I would like to see them do something else like it If you plug in a counterfeit device, basically it says spitting out garbage code which could do whoever knows what, to that device, it actually just says, you know, it just doesn't work. Or it pops up with a window or something, some notification on the, on the window side to say, Hey, this is a counterfeit device, please contact your manufacturer to blah, blah, blah, blah, blah.
Host 3 15:25
Yeah, yeah. But when it comes down to it, that's, that's, that's wonderful and lovely. But what if somebody buys some consumer gear and they plug it into their computer? And it says that? I mean, are they really going to go that far? It's kind of difficult to look at it that way. You know, some Joe Schmoe just buying?
Host 1 15:42
Well, yeah, so Joe Schmo buys a piece of equipment, plugs it in, and it doesn't work, they return it. And I guess assuming, you know, most people will be returning these devices, and then that gets shoved up the supply chain. Yeah, because it won't be best buy on the look, it will be the supplier to Best Buy. And then that person or that manufacturer will go up the supply chain and figure out what actually happened. That's technically what's supposed to happen. Well, okay.
Host 3 16:08
I think I have a solution for this. You're ready for this?
Host 1 16:12
Yeah, I'm ready for this one.
Host 3 16:13
Don't buy crap on eBay or Alibaba. Buy from reputable sources that fixes the problem.
Host 1 16:21
Yeah, it does. Unless the reputable source gets tainted, which has happened before. Yeah, I think Digi key got tend to tainted way at the I got a bit. I think it was like one chip, I have to look it up. That was a long time ago.
Host 3 16:36
Did you keep buying stuff off of eBay? One?
Host 1 16:38
And one surprise me if it was a return? Yeah, I got it and get lost in the mix and then got into someone's order.
Host 3 16:46
What I see what I see is just a giant box like a huge like, no more like a dump truck filled with chips. And the guy who has the one bag chip kind of drops it in. It's like oh, crap.
Host 1 17:00
And just whistles and walks away. Yeah. And walks away. Exactly. No one. Oh, no, no, no, no. Okay, and then this, this will not go away, either. Solar roadways are making a comeback in France. They're going to build one and it's gonna be really, really long. I can't remember the exact distance but it was several kilometers which is less than a mile. I think like two kilometers a mile. Right.
Host 3 17:24
Don't make me do math. I'm an engineer, man.
Host 1 17:29
Where's your TI calculator? Haha, cheese. Anyways, um, so they were built a solar roadway? Last I heard they built one in somewhere up in Scandinavia area.
Host 3 17:43
Yeah. And that was a bike way
Host 1 17:45
Bike way and it broke. Yeah. Um,
Host 3 17:49
So let's put a lot heavier items on it that go faster and are more disruptive. Exactly. That'll work.
Host 1 17:55
Um, you know, I think in a perfect ideal world, this works. Problem is. I mean, everyone that like even drives on a freeway, you always see a car that's broken down on the side of the road or a wreck. Yeah, yeah. The moment a car wrecks on this thing. It's gonna just no metal on that surface. Yeah, gone. Yeah, it's, it's
Host 3 18:16
Done. completely done. Yeah. So you know, okay, so here's the thing that that I think is funny about about these kinds of solar projects, because so everyone gets caught up on efficiency. Solar panels are X amount efficient. And you can do the academics. And you can write down all the formulas and stuff and actually figure out efficiency. But if you think about it, the efficiency is, you know, what percentage of light actually hitting a solar panel becomes power? Exactly. Right. So 35% efficiency is, is pretty good for a solar, right. But that's assuming that you started by producing the light or you started with a certain amount of light. Really, they're more than 100% efficient, if you think about the fact that you didn't have the light before. You have that power before and now you have more power. The way we have to think about efficiency is how difficult is it to purchase to make to install to maintain, right, that's where you have to start looking at your efficiency as opposed to just like lights hitting it and power comes out. Exactly. So when you start looking at that your efficiency is just in the toilet.
Host 1 19:32
And I'm also wondering as a you know, most time like how often you see someone cleaning a road or street sweeper not super off. Yeah, and here, here and here in Houston. We've never seen D I think I've seen like a couple downtown. Like after like a big like Mardi Gras Yeah, right. Um, I mean, I've done done solar panel work up in Oklahoma for remote radio Do stuff, you pretty much have to clean up every month to keep the dust off from now, because if the dust in just like a week's worth of dust knocks the efficiency down by half. Now you're talking about a road that you have to clean every week. Right? So you have to add that cost into.
Host 3 20:17
Yeah, yeah, absolutely. Well, and above and beyond that, you have to think about the economic impact of actually creating these solar cells to put in there. I mean, it's, it's not it's not a trivial thing to go and harvest silicon and create the panels and, and do all that. I mean, so yeah, great. You make $13 a year on a mile of solar panels. I guarantee you, that's a pretty negative return on what it took to actually put it in there. You know? So, do you have a solution for this?
Host 1 20:49
Don't put them on the road and put them off to the side of the road?
Host 3 20:54
Okay, or, you
Host 1 20:55
Know, why can't you just put them over the road? Because then it keeps rain off of them? Yeah, um, I think roads will get dirtier, faster, which might,
Host 3 21:04
Because because no rain on them. Well, and these are in these are in areas of the world that are fairly north, I'm sure they will get snuffed feet of snow
Host 1 21:13
Or snow, but then get snowed on anyways, and you didn't have to plow them. And then you had to be even more careful about plowing them. Because if you dig that, because I mean, I mean, if we don't have it down here, of course, but if you go up north, you'll see in the concrete where like a snow plow has dug into the asphalt and ripped up a chunk. Yeah, well, if you
Host 3 21:33
Think about it, you got to coat these things with something that is hard, yet cheap and easy to apply. And here, it has to be incredibly optically clear to get any kind of efficient scratch resistance.
Host 1 21:46
Right, right. impact resistant, and it's like the golden window that you have to design.
Host 3 21:53
Yeah. Well, and what's funny about these, this kind of project is every variable that goes into it makes it worse. There's no variable that makes it better, right? I mean, you just put these things down hope that power comes out, but everything is against you
Host 1 22:08
On everything is against again, put them above the roads, then sure keeps the snow off keeps the rain off.
Host 3 22:15
Maybe we they design little robots that go and clean them every day.
Host 1 22:18
Something like little rainbows. Yeah, something like that. I got it. What's that? Puts instead of rubber tires, sponge tires, and they active and clean the road as they everyone drives?
Host 3 22:30
That sounds like an absolutely horrible
Host 1 22:34
Ride would be terrible. Yeah, no,
Host 3 22:36
Absolutely terrible. And your efficiency in your car goes down to nothing.
Host 1 22:41
Because imagine that you have to have like soap applicators on your tires.
Host 3 22:45
Yeah, they slowly just drip on your front and clean the road. Yeah, perfect entire something to put them then you leave a film of soap on the solar panels and it goes to crap. Then
Host 1 22:56
Half the cars are just water. Half of them are soap.
Host 3 22:59
We make vendors that are a squeegee. Yes, fender squeegees fenders going to perfect. Next product done. We're making this so I guess I guess the main idea is that we don't necessarily approve. It just doesn't seem like a feasible project. It seems like Okay, great. I love these kinds of experiments where we're like, let's try to do something better. Let's think of ways to make this better. But why don't we Why don't we focus our time and our effort and our money into research to making solar panels even that much more efficient, that much more shatterproof? That much more. You have have a longer lifespan. I just seems like a better use of funding.
Host 1 23:43
Yeah, exactly. Well, I think that will do it for this podcast. So I'm your host, Parker Dolan. And I'm Steven Craig and catch us next time on the macro fab engineering podcast. Take it easy
Transcribed by https://otter.ai