Embedded - 217: 10000 Pounds of Pressure

Episode Date: September 29, 2017

Bob Skala of Interactive Instruments spoke with us about very large servo motors, wind tunnels, and staying current in tech.  Hydraulic Press YouTube channel (and our favorite video) The Wright Brothers by David McCullough Other good tech podcasts included The Amp Hour and HamRadio 360 WorkBench Chris talked about getting into WSPR in 197: Smell the Transistor but we first talked about it in 76: Entropy is For Wimps. The new WSPR mode he mentioned is called FT8 (google it). And a note from Bob: Below is a link to a type of servo system that tries to simplify the interface to be more like a stepper.  It integrates the driver and motor into a single package so you can treat it like a stepper with digital step and direction or serial commands.  You get the smoothness, speed, accuracy and low power (when idle) of a servo but the servo motor, driver, and cabling are integrated into one magic box.  You add a DC supply and simple control signals and you are all set.  They came out with this to replace stepper motors.  I haven’t used one yet but I hope to at some point. https://www.teknic.com/products/clearpath-brushless-dc-servo-motors/

Welcome to Embedded. I'm Elysia White, here with Christopher White. We've sort of stumbled through explaining servos, those little motors in hobby aircraft. Apparently they come in extra large, and you shouldn't arm-wrestle them. Our guest this week is Bob Scala. Hi, Bob. Thanks for joining us. Hi. Thank you. Thanks for having me. Could you tell us a bit about yourself? Well, I graduated in 79 and went to work at Smith Corona, a typewriter manufacturer. actuar and we worked developing typewriters and doing test equipment and so forth for

Smith Corona for a number of years. Then I had an opportunity to start a company with a business partner of mine. And so I became president and co-owner of Interactive Instruments back in 93 and we develop a line of products and we use we've used stepper motors but we primarily use a lot of servo motors and other motors to in our products and use embedded controls as well okay of course we're going to have many more questions about all of that. But first, lightning round. This is where we ask you short questions and we want you to answer with short answers. But that doesn't always work.

All right. Favorite movie or book that you encountered for the first time in the last year? The Wright Brothers by David McCullough. Oh, yeah. It was a good book. It's on my list. Yeah. Hardware, software, or firmware?

Everything. Sorry. Would you rather be in charge of making a product or help someone make their product vision successful? Making a product or help someone make their product vision successful? Making a product. 8051 or STM32 F-series with Cortex-M something? The STM32. Roger Waters or David Gilmour?

Roger Waters. Favorite type of wave? Electrical, I guess. Favorite encoder? The quadrature. All right. That was very lightning. That was pretty good.

We can retire it now. Someone's done it correctly. Okay. So my understanding of a servo is that it is a small motor, usually fairly cheap, not very powerful, not much torque. And the big advantage is that I can tell it to go to a specific position. And it goes to that position. And from the outside, I don't have to deal with encoders. And it's usually pretty exact.

You work with servos that have some of those characteristics. Yes. And ones that have far more. Mm-hmm. So tell us about your servos. Well, servos come in a wide range. There's AC servos, DC servos. DC servos is what you're talking about, the small servos that can use for a quick mechanical motion.

You want to go to a certain position and hang out. But then there are also AC servos. These are larger. They could be more powerful and so forth. And the control is a little bit more complex because you aren't using a DC system. You're using alternating current. You usually use three phases to feed into the servo motor. So it adds a lot of complexity, but you get a lot of rewards such as high torque, positional

accuracy, velocity accuracy, and so forth. How much torque are we talking about? Well, numbers. Inch pounds, foot pounds, you can get up to 700, 800 inch pounds, foot pounds. I tell you the truth, I don't have the exact numbers in front of me, but they can be enough to, as in our application, we've made machines that can push up to 10,000 and 20,000 pounds. That's enough, yeah.

It seems like enough. It. That's enough, yeah. It seems like enough. It seems like a lot, yeah. Okay, so you said that the DC control method, which seems like it's really related to PWM waves, is different than an AC control method. Mm-hmm. How? Well, at DC, you can use a pulse width modulation to control the power into the motor, and you use a resistive feedback or some kind of encoder feedback. A lot of times,

the low-cost ones are just a rotary potentiometer, And it's constantly comparing the position that you're commanding to the position that the motor's in, and it just changes the pulse width modulation width up or down to maintain that position in a closed loop. Okay. And from my outside processor, this is all just send it PWM and it does everything it needs to do. Exactly. Yeah. Inside there's a little control that basically determines, you know, depending on your commanded PWM signal, what PWM it'll send into the motor to control the position. And all the tuning for the feedback loop is all done for you?

Or is that something you have to... Yes, that's all done for you. Usually in the small ones, it's all integrated into the box. I mean, that's the beauty of them is that they're super easy to use. I mean, once you give it a PWM and you get used to the frequency it needs, you just say, go here, go there, and it does. And if you're smart, you maybe try to build some ramps in there. But overall, I'm not the only one just jerking them around

to go from one position to another. Right, right. So the AC ones, you said require more complex input. I think you said sign commutation right yeah they use uh uh three phases uh usually uh we don't design our own servo drive we purchase those but uh there's a number of manufacturers that are coming out with them and they they come out uh the they're getting less and less expensive as we go on the last 10 years have made a big improvement in reducing the cost and adding more features and improving the performance but essentially

you can either serial command a position or a velocity or a torque depending on how you set up the drive and the the motor will do whatever it needs to to maintain that. So you can either use serial, Ethernet, RS-232, or you can use step and direction. That's a popular one as well where you use streams of pulses, which tells it the velocity, and then a direction line, which gives it, you want to go counterclockwise or clockwise. There's other protocols as well, but those are two of the common ones.

So it's still very easy to use, even if it's doing something more complex inside the motor. Yes, very easy. Once it's programmed, what you have to do is you have to program it. And every mechanical system behaves differently due to the mass and inertia and so forth. So you have to program it and every mechanical system behaves differently due to the mass and inertia and so forth so you have to tune them and that requires going in there and adjusting a number of constants there's position constants current constants and so forth sometimes they'll have an auto tune feature if you can turn that on and if it works, you're golden. But many times you turn it on and the thing just shakes all over the desk. Is this like

tuning a PID loop where you're trying to do one thing at a time and then you get the proportional control settled and then you start tweaking other things? Yes. You'll have a proportional position loop and uh you have sometimes pi and on that loop sometimes d uh and you also have a current loop and uh there's there's a few other loops inside there they can be extremely complex depending on on the sophisticated the controller is. And so you are mostly buying these and then integrating them into other devices? Yes, yes. We make a 5,000-pound and a 10,000-pound materials tester,

which basically takes the rotational torque from the servo motor and puts it into a ball screw, which is kind of like a nut and bolt combination, but uses ball bearings instead to make it more efficient. A screw is typically not very efficient, and so there's a lot of energy lost in it. So if you use a ball bearing nut, you can improve the efficiency. And what that does is it translates the rotational torque into linear motion. And I should be very excited about all of this mechanical stuff, but I need to interrupt you in order to ask you what is the coolest thing you've ever crushed?

See, I don't do a lot of crushing. The customers do that. I mean, what does an Oreo cookie look like after 10,000 pounds of pressure? Flat. I'll have to check on that. I don't know. I haven't. Are you aware of the Hydraulic Press YouTube channel? Oh, yes.

Oh, my goodness. That's a fake of mine. Yeah. Oh, yes. Goodness. That's a fake of mine. We mainly do tension, and the reason is most of our customers pull parts apart. They'll take a piece of material from their boiler or their pipe that's developing cracks, and they pull it apart, and they monitor how fast the cracks grow over time. Oh, interesting. In our case, we primarily do tension, but we've had some customers that do compression.

An Oreo cookie pulled apart by 10,000. I don't think that would be that exciting. That isn't that exciting. That's just frosting. I could do that with my hands. It's going to hold a lot of weight on YouTube. What kind of applications are people working on with this? I mean, tension cracks, yeah, but why? I mean, what are they doing?

Well, say in a nuclear reactor, if you have cracks. That escalated quickly. It didn't. Say in a nuclear reactor. We have a lot of customers that test parts, pipes, and so forth in a nuclear reactor. And as the nuclear reactors age, they have cracks. And they know about the cracks, but they need to know how fast they grow because they may want to squeeze, instead of 20 years, they want to go 30, maybe 40 years on the same part. So by studying the crack and how fast the crack grows over time, they can do things to slow the crack rate down by adding various chemicals and changing temperatures and changing the pH in the water and so forth.

And so that's a popular use for our product, but also oil pipes. Oil pipes are going to, you know, they're going to pump various oils at times. So it's going to put a lot of pressure on the pipe and then they're going to release the flow. And cracks in oil pipes, they need to monitor how fast those cracks grow so they know when it's time to swap them out. Is it a realistic test to put it under a huge strain like that? I mean, these cracks develop because the world isn't steady and aligned because temperature shift and chemical shift.

But putting 10,000 pounds of pressure or tension on it, it seems like that doesn't scale. It's like accelerated aging testing in a way, right? Yeah, it's more of an aging test where what you do is you have instruments on there that can measure the depth of the crack and they actually have computer programs that knowing the depth of the crack, they want to change the intensity. They call it a K factor around the crack which allows you to control the rate of the crack growth. So you can increase and decrease this K factor, which is called a stress intensity factor, and they can control the rate of growth of the crack.

Sometimes you want to run a test where you'll keep running at a decreasing K factor until the crack stops. Now you know what load you can run that particular boiler at or pipe at and still not grow the crack. Of course, the materials you use to build the tester, the screw and the ball bearings all have to be able to withstand all of this without falling apart themselves. Yes, yeah. They run for days, months, and sometimes years on end, unattended. So we should just make the nuclear reactors out of those.

Yeah, yeah. So what part of this do you do, Bob? I'm the programmer. I do the board layouts, board assemblies, firmware development. When you're part of a small business like we are, we only have nine people or eight people full time. You do what you can to get the job done. So you're kind of a jack of all trades, master of none. And do you enjoy working for yourself? Sometimes I say my boss is a jerk,

because I am hard to work for. But do you enjoy being in charge? Yes, being in charge is when you're when you're doing your own project is a lot of fun. It's more challenging when you're running a company and doing your project at the same time. But yes, I do enjoy my job. Okay, we should go back to servos for a bit, although I have more job-related questions. We should go back to servos and steppers. How are they different? I mean, steppers are a lot harder to use. Well, they're simpler to control, but they have some nasty side effects because they are simple.

I mean, a stepper motor is kind of like a servo motor, but instead of having just maybe six or eight poles, poles being windings around the motor, you may have a number of poles on a stepper. So you have to advance the phases, because in a stepper motor, it'll have typically two phases. You have to advance the commutation quickly to get the same rotational speed as a servo. So steppers are typically great for low-speed, high-torque applications, whereas servo is great for high-torque applications at pretty much any speed. So just the basic design between a stepper and a servo,

you have to kind of pick your applications. Is there a position and repeatability difference? I mean, I used to think, oh, stepper is what I would choose if I wanted to go to this position exactly within, you know, the commutation resolution and hold it. Whereas servos i always felt like i kind of over here a servo can uh can be both uh servo uh it depends on your your position feedback sometimes they use uh hall effect sensors which are just placed around the

rotation you may have just a handful of them placed around so you you only have a uh you know you know maybe a five degree or two degree difference between uh you know position control and in a hall effect sensor where but they can also use uh more sophisticated controls there's there's things called resolvers which is it's an analog way of uh measuring your position by using a uh a, it's kind of like a coil. And due to the rotation of a magnetic field inside that coil, you can tell the rotational position. But, you know, it's analog, so there's noise involved with that. You can also step up to the next level, which is digital encoders.

And digital encoders can be extremely accurate. We've got one right now where we can get the servo motor position divided into two to the 17th power. It's at 131,000 counts per revolution. So it's very fine. So it depends on your feedback. So I guess I always thought that servos had steppers inside of them, hiding the complexity of the stepper drive. Is that not true at all? Well, they're wound differently. The windings are completely different, where you typically have

two-phase windings in a stepper motor, in a servo motor it's a three-phase there's fewer poles in a servo motor uh but uh uh but the uh stepper uh is just more of a it's it's more for low speed applications you can't get very high speed in a stepper just because of the number of poles you may have to go through. You may have to advance through 24 different waveform configurations on a stepper motor to rotate in one rotation on a stepper. But a servo motor, with the three-phase, you're pumping a lot more power into the motor over time. But you're not getting as accurate a position in the motor itself, and so you have to use the encoder in order to get a better accuracy.

Yes, that's correct yes yeah servo by itself without an encoder is is more of a it's just an ac motor if you don't have control or feedback uh uh it's it's very difficult to tame them and we've said the word commutation a few times. Do you want to try to explain that? Well, commutation is kind of the advancement of the magnetic or a computer for a servo or a stepper motor. So commutation is basically the control of the magnetic field that surrounds the magnets, which are usually in the rotor. And if I have this right, if I have a stepper motor with, say, six windings, that means that in order to get, so if I'm looking at it like there's a clock and I want it to go around, and there's 12 points on a clock,

and I have six windings, so every other hour gets a winding sort of. And if I want to go from 12 o'clock to one o'clock, I have to energize the one o'clock at just the right way. And if I'm doing it right, I'm also energizing the 12, energizing the two o'clock in just the right way. And if I'm doing it right, I'm also energizing the 12 o'clock to push the motor. So there's pushing it and pulling it. And you're going through multiple stages all the way around the clock face so that you're constantly changing what you're doing. And you don't just want to jerk it so that it's pushed and pulled like boom. You want to go in a sine wave so that it's pushed and pulled like boom, you want to go in a sine wave

so that it's gently pushed or pulled, and then it's more strongly pushed or pulled, and these sines overlap so that even as you're going from 12 to 2 to 4, you're trying to get a smooth motion between those so that you don't burn your motor up. Yeah, that would be current control, where if you had each phase was under current control, you can actually microstep the stepper motor. Is that what you're thinking about? I think it is, yes. I mean, I have done this sine commutation thing a few times,

but usually only with somebody who's already done it. Yeah, use sine and cosine waves. And sine and cosine will create a circle for you. And you pump the current into the two windings, one sine, one with cosine. And depending on what angle you pick up from your sine and cosine table, you can control the position. It's a lot smoother than what they call full-step or half-step stepper motor commutation. But the big problem, though, with stepper motors is if you overdrive it, you can actually lose your position. It'll backspin. And because the stepper motor is stable, when you put a certain current into the two windings,

it's stable at multiple positions because of the multiple poles that it has. And so in your code, you're going, oh, I'm going forward. Oh, wait a minute. The encoder just said I didn't go forward. I went backwards because I did my commutation incorrectly. Yeah, most steppers, they don't use encoders. But if you do have an encoder, at least you can find out where you actually ended up and try to get back there. We had an encoder on the test unit because we needed to know how bad my software was. But yeah, if you don't have, you know,

two or three times the torque that you need, you will definitely need an encoder because it's very easy to overdrive them. You mentioned you get to do everything. And as you go from firmware to software to hardware, you end up having control over the whole system. And so you get to design every piece of it. And there's some benefits to that.

What do you think the benefits are? Well, it's kind of self-satisfying to be able to come up with an idea and carry it all the way to a product. That's a big benefit. But it's getting more and more difficult today. As you know, the specs are getting extremely complex. I mean, I remember back in the 8051 days where if you had 20 or 30 pages to describe the chip, that was a lot. Now, I was looking at the uh beagle bone

chip the microcontroller and that uh and it's 4700 pages it's it's more and more difficult for a single person to take on these projects that's for sure that is one of the big disadvantages and then there's the part where when you do take it on, you don't necessarily have anybody to talk to about it. Yeah. Yeah. Yeah. You got to use Google a lot or listen to a lot of podcasts like Embedded FM. If somebody wanted to build their own high torque system, where would you say they should start?

Should they buy a servo? Should they design their own motor brushes? You can buy linear actuators that will have a servo motor on a screw, a long screw with a slide. And you can buy your own components and do your own software development to control the servo motor. Buy your own servo drive.

You match a servo drive depending on the current capabilities of the motor. You need enough current to be able to uh to control the motor efficiently and uh and to its peak power and uh and then um depending on the inputs to the the uh servo drive you could use uh serial communication or uh step and direction commands to to control your own, make your own. But it's hard to make something yourself today because it requires some precision machining, and that's what we have. We have a lot of three, four, and five axes milling machines and so forth. So the advantage of having our mechanical designers

come up with a product to our exact specifications and produce a product like that. It's challenging to do that if you want to do it in your basement. It does seem challenging. What kind of power do your strain systems take? All of it. It's 110 volts. We like 110 because it doesn't

require special power. So they may take five, seven, eight amps or so at 110 volts.

A kilowatt or so then? Yeah, a few kilowatts, yeah. Cool. I mean, the nice thing

about these servo motors, they're very efficient because they're all current controlled. It's kind of like a switching power supply. So it's very efficient. And so you can get a lot of power relatively efficient with these drives today. Do you have any sort of health monitoring of the motor? I imagine it gets pretty hot. So are there temperature sensors and is there any sort of cooling or is it all passive?

The servo drive will monitor the temperature in the motor. It notices that if it's putting too much current in there and it knows that it's over spec, it'll actually try to back off. And if it can't get any farther farther it just shuts down completely actually that's a good point what other health warnings do you have for not the motor but for the humans don't put your finger in there any things like that uh yeah we uh These are used in labs. It's not something you'd have out on the bench for a common public to be using. So they're usually installed in a piece of equipment back in a lab someplace, turned on and left.

So very little human intervention. So you don't need the sign that says, do not lick machine. No, but I mean, I think that that's in the manual. Do not stick the remaining hand in. Your company also makes wind tunnels. How is that related to Servos? Well, it isn't. We started the company making wind tunnels, and that's what got us going.

And we had an interest to make a wind tunnel. And so myself and my business partner and one other person helped design this product. And we took it to shows, and it started to become a hit. A lot of people were interested in this. Nobody had anything like this. It's a desktop wind tunnel with an AC motor, one horsepower AC motor, and lift and drag sensors where we measure to within a thousandth of a pound the lift and drag forces inside the tunnel. And a pitot static tube that controls the wind speed. We brought it to a show down in Charlotte and North Carolina and had a lot of interest.

And so that's kind of what got our company started. And now, I mean, are they still popular? Your company started a while ago. Yes, we've been selling them. This is our 25th year. And we've been selling them right along since the beginning. We sell a few internationally. We just had one that's going to be delivered into Germany shortly. But we've had companies, primarily they're for high schools and junior high schools for STEM training. But we've also had companies as well purchase them. We've had a company that Mack Truck, they wanted to buy one so they could test their uh aerodynamic packages on the on the trucks and so you take a little itty bitty mac truck like the size of a matchbox car truck and then you put it into the desktop wind tunnel i just can imagine doing this and

ending up at the end of the story with like those trolls that have the fuzzy heads in the wind tunnel driving a little car. I mean, but people are using these to make plans. They're the prototype phase things for companies. Is that right? Right, right. What do the students do with them? They make airfoils. That's where you learn a lot.

I mean, there's two things that they typically test. There's cars, which kids love. Yes. CO2 cars, which is a project where they make these cars and they put a CO2 cartridge in them and they mount them to a wire and they tap the back end of the CO2 cartridge and it shoots down the wire. And so aerodynamics is everything for something like that. So they put their cars in there and they can measure cross-sectional area drag and so forth. But where it's really exciting is when you put an airfoil in there and you can measure lift and drag and the efficiency of the airfoil, which is lift divided by drag. So lift is larger and drag is less.

You have a larger number, so it makes the airfoil more efficient. And so you can measure uh lift over drag coefficient and airfoils are like wings and they usually i mean i guess a wind tunnel would have been very useful because as far as i remember about airfoils and wings is there's something about the bernoulli principle and nobody knows how they actually stay up i mean that's that that, right? That's state-of-the-art science? Yeah. I think there are some people who take issue with that, but that's fine. I think it's money, but I'm not quite sure.

Those airplanes are pretty expensive. As long as we keep believing that they can fly, they will. But as soon as the first person stops. Is there smoke in your wind tunnel? i mean i see that for big ones yeah uh it's difficult to to put smoke in in a classroom so we don't have a product that actually generates smoke i mean we use it for more for the uh in our photos there for the visual effect to show you that wind is actually blowing because you know wind tunnel sitting on a bench looks like one that's turned off so uh so we use smoke in our in our uh our photos but uh and some schools come up with uh with their own ideas we had a

a student that used dry ice uh down in texas and uh he wanted to study icing on airfoils. And so he used dry ice to create the fog effect. But, you know, stuff like that doesn't, unless you're under strict supervision, it could be dangerous. How hard is this to scale up to, I mean, I don't want you to scale up your wind tunnel, but how hard is it to take the results of a small-scale test and extrapolate to,

okay, so we put a truck in here, but it's a toy truck or a small model. Are the results directly applicable? Because, I mean, air density changes, or can you just say, okay, we're going to scale the wind speed appropriately and everything will work? Ours is more for demonstration purposes. We're not looking for analytical results. The businesses that buy it, they realize, we tell them, this isn't a calibrated piece of equipment that you want to just run down and

start taking notes off of. It's relative because we need to keep the cost down. Schools, as it is, have a difficult time budgeting a product that's an expensive wind tunnel. So keeping the cost down makes a big difference. And for the schools, they are doing small models, so that's perfect. Yeah, yeah. They want to see how the kids make an airfoil. Which one's more efficient? Which one has better lift?

Which one has more drag? Just to give them the idea of what's going on. And we were talking about STM32s, you and I, before the show. Is that in the wind tunnel or in the other, the strain, not sensor? The wind tunnel started off with a little 8051. And this was back in 93 or so. So 8051s were very popular. It was a nice little chip for that but as as time goes on it was more and more difficult to uh find these through hole components and then

the price started going up so i chose a um you know it was probably about four years ago or so i had this it was on my desk it was a maple mini and uh he was an astm or stm 32 103 chip and it says wow this is nice you know the the a51 had a long life and you know i'm sure people still use it today but i i wanted to choose another processor that would uh continue on so so i chose this this maple mini it's a little board or old 40-pin through-hole device, and I can still deal with through-hole circuit boards and so forth. And I designed this wind tunnel control around it, but then shortly after, the Maple Mini went obsolete.

So, yeah, I switched to the STM32 microcontroller, and it's a nice controller. It does a lot of things for very low cost. Very reliable. It's good to hear. I mean, we talk a lot about how nothing works. It's all broken. It's horrible.

You have to test your code because it's for sure not going to work. And yet I remember the software that's broken you don't have to wonder as much anymore yeah yeah I mean the hardware is just one nice little package and everything's all integrated in there you have your uh your analog to digital converters, PWM converters, spy controllers. Onboard flash and onboard RAM. Exactly. You don't have to worry about your traces going out to your little 64K chip that stored your program and make sure all those lines are are working just works it's a mental shift to go from something that has worked for you for years and years to something that's new and different how do you get over the hesitancy of my God, it's all new? Well, you've got to accept challenges.

Challenge yourself. When I went from that little 8051 to the STM32, it was a big leap, but you just prototype it and test it and play with it. And over time, once you build your confidence level, then you see it's time to go into production. And you just kind of take these challenges one step at a time. And don't be afraid to challenge yourself. Challenge is good.

I mean, failure, you have failures every once in a while. I mean, anyone who's been in the business long enough will know failures. That's how you learn. You learn from your failures. And been in the business long enough. May I ask you how long you have been in the business? Sure. I'm 60 years old, just turned 60 this year. So I spent 16 years working at Smith Corona and another 25 years working here. Do you ever worry about getting bored? If you're bored, you're just not challenging yourself. You just got to go out and find something else that interests you.

But you run a business where you have to produce these systems. Aren't they the same systems over and over again yeah yeah that that uh but you're always changing them uh you know a good project is never completely done you're always just adding to it or refining it making it uh faster better smaller more cost effective. So even though we just have a fixed number of products, we can always keep tweaking them. How do you stay current?

That's not easy. When you're the only person in the department, podcasts are great. Podcasts have been a big help for me and i listen to podcasts and as i'm driving into work and uh says wow yeah i gotta remember that i gotta google that when i get in there and and i've picked up a number of great ideas uh that way well i know you listen to our show so i'm not fishing for compliments but which other which other podcasts do you listen to

uh amp hour is a good show oh the enemies show i don't know and i also listen to i'm into ham radio as well and i listen to a number of uh ham radio shows there's one called ham radio 360 workbench, and they get into projects. I've gotten a number of great ideas from them. One of the things Bob said to me was about how your interest in Whisper gave some to him. Oh, yes. Well, now I feel bad because I haven't done it in a while. I need to get back on that. I got my amateur license back in August of last year and did my tech in general in one shot.

And then I was kind of thinking, well, what else can I do with this? And then you came on and mentioned Whisper. I said, yeah, it sounds like a fun one. So I went and ordered it and wired it up. And I've got contacts all over the place. I just had one not too long ago that went out to Switzerland. And that little quarter watt transmitter can actually do pretty well. Where are you based? Because I don't think it's near Switzerland. No, we're just outside of Switzerland. We're in upstate New York, outside of Albany, north of Albany. Yeah, I think I've gotten contacts up in

near Edmonton. We're in California. And I think I got to Florida once, which I was amazed at because I didn't even have a good antenna. I just threw a wire over a tree. An antenna does make a big difference. I just got a new antenna, and it doubled my area. So Whisper is this, not contest, but this... It's a mode of, it's a digital mode.

It's a digital transmission mode where you have a little thing that squawks at like a quarter watt, so not a lot of power. And then you try to put it in the right place with the right antenna and good impedance matching in all of your components. And then you let it just shout its little heart out for a bit and go online and look to see where somebody heard your little widget. So it's ham radio without any social contact. Right. For the introvert. But it's amazing that with a quarter watt, you can get that far. And this isn't one of the ones that bounces off the ionosphere, is it? It isn't the frequency that's...

It's in a lot of bands, so it can. Oh, okay. Yeah, it can bounce off some of those long things. Yeah, the one out to Switzerland was, I think, a little over 3,800 miles. So it's got to bounce off as something. And the information you send is your location and your call sign and not much else. That's pretty much it. And your power level.

And your power level, yeah. And you also send it like... And that takes two minutes. With lots of error correcting. Yes. There is a new mode, and it's extremely low repeat rate. With lots of error correcting. Yes. There is a new mode, and I forget the name of it. I'll have to look it up and put it in the show notes. But there's a new mode that somebody's come up with that's supposedly much faster and equally robust.

So, interested in looking at that. My wife will be glad to hear that I'm buying another box then. I'm hoping the boxes are upgradable too. Okay, that's great. I'll have to check into that. But yeah, I like it because it kind of reminded me of, I don't know, Voyager or something. You know, the space probes out there and they're not transmitting more than a couple of watts now, if that. And they're billions of miles, billions of miles away.

Of course, they have, you know, Arecibo and gigantic dishes to receive them. Oh, and some of the people who receive your whisper contacts probably have very large antennas. I remember we looked at one of them just snooping on Google Maps. And yeah, you could see those antennas. From space. Yes. Yeah. Yeah, mine's small.

Mine's just in the garage. Chris hung an antenna from one of our trees. It was a little odd. It worked fine. It worked fine, yeah, because you tied it to a rock and threw it over a branch. Yeah, slingshots, drones.

That's not slingshots. Slingshots are not a good idea. We did that urgent care visit. From personal experience. But you can use drones. Some people use drones to lift the wire. We have too many trees to do the drone. I mean, yeah.

I have drones, but I haven't done that. I know where that would end up. With the drone in the tree with the antenna? Yes. And with the antenna? Yes. Yeah. And with the antenna wrapped around one of the props. Okay. One more question about your age, and then I will get off this topic. It's something that kind of stresses me out because as I'm 40-ish, I don't—I love engineering.

I love what I do, but I talk to people who keep telling me that it's hard and you, you, you lose the love for it. And I hate hearing that. And I, I, somebody said to me not too long ago about how you have to go into management because you can't keep up with the technology. And I just, no, I don't want that. Was that person a manager?

Of course that person was a manager. Yeah, I got told that by a director at a company, too. It's time for you to go into management because there's nowhere else for you to go. Well, okay, I guess I'll go somewhere else. Yeah, there is no up from here. You're going to be an engineer doing the things you did three years out of college

for the rest of your life i i don't that's not a happy path for me no no i think it sounds like you have a passion for it obviously you do and you know i don't think you're going to go down that management uh trail uh happily uh just stay with engineering i mean engineering you can I've known engineers that are, you know, 16, 17, 80 years old that keep going at it. It's just, it's more of a, you know, just follow your passion. I like that. I like to hear that, of course. And I admit, I'm a much better engineer than I was in my 20s or 30s.

Oh, yeah. Yeah, you grow. And you build your yeah. Yeah, you grow and you build your toolbox and the toolbox grows and projects become more and more challenging. And even the tools that get old, like my knowledge of 8051 and compilers of that era, it's never really wasted. It just gets built upon. Yeah. Yeah.

Yeah, I mean, controllers are pretty much the same. You know, it's just which compiler do you use and if you use IDEs, which IDE do you use? But, yeah, IDEs today are a lot easier. I remember back in the earlier days when you were dealing with Intel boxes, doing 8051 programming, and it was a real challenge using big eight-inch floppy disks and typing it all in, and all in assembly language, of course. And you push a button and hope it works, and if it doesn't, you't, you roll up your sleeves. There are some things that have gotten a lot better.

Really a lot better. Except the not working. That still happens a lot. Yeah. But you have better tools, though. Debugging tools are a lot better. And it takes a little bit more time to find out, but it just kind of increases your level of, your confidence level when you find those really complex bugs and it takes you days and weeks and sometimes months

to find. I remember having this one bug, I was searching and searching and searching trying to find out why wasn't this feature working, as the manual said. And so I had to consult the forums, and I put this message out in the forums, and I didn't get a response. I said, oh, boy, that's not good. About two or three months later, I finally get this response from a guy. He says, yeah, I don't know if you're still looking into this, but the manual's wrong.

Don't pay any attention to that do it this way instead and sure enough it worked great yep i've been there with data sheet yeah oh yeah actually in my case it was there was one data sheet that was correct and one that was not correct and the one that was not correct was just a different package of the same part but it was a timing diagram that was wrong. So I don't know how they got the timing diagram wrong between just a different packaging spec. But yeah, I was very angry.

I tried everything. Didn't work. Gave up? Did you ever find a solution? Yeah. Well, the other package data sheet was correct, so I just did what it said. All right.

You had mentioned working on typewriters. And I remember the printers in the 80s, they were either Daisy Wheel or IBM, I think, had this ball thing. The ball, yeah. And I am still amazed that that could be controlled accurately and quickly. Can you go through how those things, how the control systems on those worked, if you know a little bit? Because I'd just love to hear that. I don't know about the IBM ball. We did study it.

There were mechanical people that studied it because it was a remarkable piece of equipment. But that was all mechanical. I don't believe there was anything electronic in there. Wow. When you push the button, it just dialed up a bunch of cams and it positioned the ball and then the ball went smacking

against the paper. That's insane. Yeah. That was a beautiful piece of equipment. But the daisy wheels, those were electronic? Yep, those were all electronic. You used stepper motors. And the beauty with the stepper motors, you design a daisy that stops at all these positions along the stepper motor. And so you just dial in your position and whack the hammer into the pedal.

The pedals are all flexible plastic, and it makes the impression on the paper. Why did we go to electric typewriters? I mean, we had manual ones that worked terribly. You answered your own question. Yeah, my question wasn't really going to go off the rails, was it? I mean, we had manual ones that worked terribly. You answered your own question. Yeah, my question wasn't really going to go off the rails, wasn't it? I think electric typewriters started becoming popular in the late 50s, early 60s, right? The electrics, yeah, it took a lot less force to, it was a lot easier.

The impression control was a lot better. Where manual, your pinky doesn't push as well as your index finger, so you had different impressions. So the nice thing about electric is you just push a lever and it flipped a cam and the proper hammer went up against the paper. It's funny how we've, in so many things, exchanged mechanical genius and complexity for software complexity because i mean printers now well it's you know it's all software and

controlling lasers and things uh on toner drums and the same thing with cars right we had in the internal combustion engine is full of thousands and thousands of parts it's incredibly complicated mechanical machine and we're rapidly replacing it with an electric motor and a computer yeah yeah sort of saddened away times march forward and you worked on some of the electric typewriters that had spell correct and the little buffers, right? Yes. It was a big deal when you had a typewriter and you just hit a single button and it would automatically know what letter you hit previous and it would just automatically erase it.

Or we had a feature on there called word erase. You hit that and it would just erase the last word that you typed. That was a popular feature. It was magical. Yeah. Somebody who's not a great typist and an even worse speller, it was magical. Yeah, yeah.

It was a nice feature. It was very popular. All right. I'm not sure we have more questions, except I want to ask all of the... Do you have any times when you realized, oh, this is what I want to be when I grow up? I mean, maybe it's in the start of your career, but maybe it's halfway through it, too. Well, I don't know.

I just take each day as they come. You have a general idea of what you want to do. I had an interest in running a business myself for years. It's got a lot of upsides to it. I think that was the general direction I took, but I didn't have any plans, anything laid out how to get there. When the opportunity came, I took, but I didn't have any plans, you know, anything laid out how to get there. And when the opportunity came, I took it. That's actually good advice. Christopher, do you have any more questions? No, typewriter stuff was what I wanted to do. Then I think, Bob, we're going to wrap up and ask you, do you have any thoughts you'd

like to leave us with? Well, as I mentioned before, don't be afraid to challenge yourself. Reach out and try something new. And that's how you grow, by trying things that are just not something you've expected to try. Just experiment. Our guest has been Bob Scala, president and co-owner of Interactive Instruments. Thank you for being with us, Bob.

Thank you. Thank you also to Christopher for producing and co-hosting. And of course, thank you for listening. My quote this week comes from C.S. Lewis. No book is really worth reading at the age of 10, which is not equally and often far more worth reading at the age of 50 and beyond. Embedded is an independently produced radio show that focuses on the many aspects of engineering. Thank you.