2018, Chapter 7
Velocity vs. Sanity
Hey all, if you’re not into some navel-gazing with a side of engineering, perhaps best to skip this chapter. After all, a logical alternate title could be “3D Printing (and other Technologies), Part 2: Full Crazy Mode.”
Because here’s the problem: today’s R&D tech means engineers don’t get any rest at all.
Lull? There’s never a lull, not when you can get parts on the same day.
Pause? There’s never a pause, not when you can get boards in three days.
Slowdown? Why slow down, when you know everyone else has the accelerator matted?
Sigh. Let me explain. In the near past, when we did sheet metal designs, we could count on prototypes taking 2-3 weeks to turn around. (“Two weeks, ha!” my inner voice says. “More like 6 or 8 weeks!”) But, the point is, the design went away—disappeared—and there was nothing we could do about it but wait.
Wait. As in, relax.
Because it was out of our hands. So sit back, kick the feet up, walk over to Pocock Brewing and have a beer. (No kidding, there is a nice brewery on the other side of our parking lot. Snarky guys can say, “That explains plenty, yep it does.”)
But the main thing was: there was nothing we could do to get further on the design. Not until we got the prototype back. Then we could see if it fit, if it worked as well as we wanted, and if we hadn’t missed anything. Then we could make a few tweaks and wash, rinse, repeat until it was right. And in every cycle, there’d be time to rest and recharge.
Today? Ah hell no.
Just print that sucker and have it the same day.
See if it works.
Make changes right now.
Run the next version overnight.
Wash, rinse, repeat.
No time for rest.
No time for beer.
“Wait a sec, you can 3D print metal?” a sharp-eyed reader asks.
No. Not with our current printer. But even though it may not be an actual metal part, but it’s definitely good enough to see if it works and fits. Heck, our mid-level 3D printer prints in some fairly exotic stuff, including high-stiffness composite glass-reinforced plastic and resins that simulate rubber, ABS, and polypropylene. If you want to experiment, you can even do ceramic. Affordable 3D printers now print in carbon fiber/Nylon 6 blends, or do continuous reinforcement with fiberglass or CF. Less affordable 3D printers print metal bound in plastic and sinter and shrink with accuracy good enough to hold M3 threads (!). And that only begins to scratch the surface of what’s available. Hell, I can walk across the parking lot and turn right to Scicon rather than to Pocock, and they can print and finish pretty much anything we want, in any size, in a few days.
The point is: no matter the part, you don’t get a rest anymore. It just goes and goes. You can’t stop. You can’t grab a beer. And that’s just the start, because lots of other things are moving a lot faster than the old days, as well. There’s no time to relax.
“Well, if it’s that sucky, why bother?” a smart person asks. “Why not do it the old way?”
Great question. Let’s talk about that.
The Payoffs: Better Products and Dopamine Hits
It’s simple: rapid iteration can result in better products. More than that, it can result in better, more ambitiousproducts.
Being able to play with mechanical designs in real time means we are willing to take bigger chances—on products that work better, look better, offer more features or higher functionality—and we have a better chance those more ambitious products will work right the first run out.
Because, let’s face it, if you’re looking at 6-8 week turns for mechanical stuff, you really have only two logical strategies, if you ever want to introduce a product:
1. Make it realllllllllllllly simple.
2. Get reallllllllllly lucky.
#2 isn’t really a strategy. But #2 is pretty much what we had to do when we went to the “fastenerless” slide-lock chassis on Jotunheim. That was pretty ambitious back then. The stuff we have coming down the line makes that look like baby stuff.
And yeah, you can try to skip the physical and do 2D mocks and 3D assemblies and all that wonderful stuff, but it frequently doesn’t substitute for actually holding the part, or snapping the real thing together, while imagining what an owner might be thinking, if they had to do that. It also doesn’t simulate the look, the heft, or the physical presence of the product.
So, you engage in modern R&D tech because it results in better products, which gives you an edge. That simple.
But you also do it for another reason: because youlikeit.
Because every part that comes off the 3D printer is another dopamine hit.
I made this,you think. This is cool.
And even if it isn’t cool, you do it again. And get another hit. Because it’s really cool to be able to make all sorts of stuff. You can make light pipes and brackets and insert knobs and you can even make turntable pulleys when your consultant is being vague about the interaction of turntable and belt, and if the pulley is wrong you can make a simple change and run it again, and test it again, until it runs exactly the way you want it to. You can bypass your slow metal vendors and run chassis for fit testing, heck, you can have the thing tweaked and perfect before they ever see it, and you can figure out a half-dozen optimizations you would never have seen if you had to wait weeks for a prototype.
And every time you get another dopamine hit.
Because, let’s face it. For people who design things, for people who engineer things, this is a whole heckuva lot of fun.
And that’s great.
Until you realize there’s no rest.
Speed Galore: Fast PCBs, Comparison Curves, and In-House CNC
And it doesn’t end at 3D printing. Oh no. The acceleration continues across the line.
A couple of years ago, we’d order prototype PCBs from the same manufacturer that does our production runs. We were cheap, so they’d take 20 days. Smaller stuff could be turned in about 10 days from a different prototype house, so sometimes we did that.
But still, the same principle applied: you’d send out the files, and they’d go away for 2-3 weeks. Sigh. Sit back, kick the feet up, go for a beer. You know, rest and recharge.
Now, we’re primarily using a prototype PCB supplier that delivers in 3-5 days. Which means it’s entirely possibly to get boards back the same workweek. There’s nothing quite like seeing a bunch of 450-component boards show up on Friday, just before the weekend.
Because you know they’ll be there Monday, waiting to be assembled.
Because, you know, you maybe kinda should come in on the weekend and get a head start…
“So why do you use a vendor who’s so fast, if it’s a problem?” another smart person asks. “It shouldn’t be a problem to find someone slower, right?”
Well, yeah. We could go back to our slow production PCB supplier.
But these guys are actually less expensive, as well as wayyyy faster.
And, again, it means better products.Thanks to this new fast, low-cost source, we’ve done more PCBs in the first quarter of this year than in the last two years. We’ve been able to tweak and iterate at a much higher rate than ever before. And it is resulting in better, more ambitious products. Heck, it’s even allowed us to purpose-build some engineering test sleds with custom designs that allow for rapid iteration of plug-in boards (on products never intended to be upgraded), which allows us to explore many more options than we have in the past.
So yeah, it pays off.
But, again, there’s no rest.
And it goes beyond PCBs. With the new Audio Precision APx555, we have a test and measurement tool that is far, far more capable than the Stanfords, but even more importantly, it also offers a much better way to automate tests, store results, and compare curves.
Sounds basic? Yeah, even the Avermetrics do that. Stanfords, not so much. One of the reasons we haven’t published reams and reams of test data is that it’s so difficult to get out of a Stanford. They don’t have a lot of standardized tests. They don’t have a good way to store and compare results. It’s frustrating that the fastest way to document stuff on the Stanfords is to take a picture of the screen with a smartphone.
Now, the Stanfords are great bench machines, wonderful for early development. But now we can define custom measurements, save the results, make changes to a design, and immediately come back and compare.
Again, the ability to rapidly iterate goes through the roof.
And again, the ability to rest and recharge goes through the floor.
And there’s more. We’re finally bringing the in-house CNC on-line. This allows us to make actual chassis parts—not just plastic simulations. Will it fit? Yep, we can answer that definitively. Will it get to hot? Again, answers are as easy as making a part. Is it durable enough? Well, heck, mill it and find out.
And when it’s that easy, again, we have an opportunity to tweak and perfect in near-real-time.
And again, the velocity increases. More speed! More tweaking! More dopamine hits! Heck, you could do this 24/7, because 3D CAD usually means more work, not less, because, well, you know, you really should do assemblies and integration with PCBs and connectors and stuff isn’t trivial and you may want to do both 2D views and 3D assemblies and you know, you could probably have the logo milled or stamped into that chassis, and you know, having a rounded feature there would be so much nicer and really, when you come to think about it, the whole way we do chassis is kinda old and we should probably start over again and see what a modern take can do for us, and really, maybe we should look at a whole new approach…
…and you create even more work for yourself, since, you know, there’s no time to slow down.
I’ll Rest When I’m Dead, Damnit!
I’m sure there are some of you out there, arms crossed, frown highlighted in the blue-white glow of the screen, wondering, Why do you want to slow down? Are you getting lazy?
Well, I’ve always been lazy, so there you go. But that’s a deflection. The real reason you want to slow down is simple: getting too close to something means losing perspective.
Stop. Go back. Read that again.
And then sit back and think about it. No. Seriously. Turn off your damn phone, close the screen, and think about it. To get perspective, you needthe big picture. The long view. If you’re too close, you can’t have perspective.
By definition.
Because, instead of running the 38threvision of that bracket to add rounded corners to make it look cooler, maybe you need to step back and ask, “Why am I making that bracket at all?”
Or, you might need to ask, “Hey, will that bracket work for all the different kinds of boards we want to use it in?”
Or, it might be smarter to wait for some bigger parts and assemble the whole thing, just to see if it works.
Maybe that’s why I’ve thrown away a couple of runs at updating our most expensive analog product. Maybe I was too close. Maybe I didn’t take the time to sit back, relax, and wonder why I wasn’t addressing all the shortcomings, only the easy ones. Maybe that’s why I got bit on the butt with Vidar last year, when I found that you couldn’t assemble the product at all, and I had to go to a clip-based design for the output stage (ironically, this was one of the smartest things we’ve done…you’ll see why eventually.) Maybe that’s why this year has seemed busier than any other, even though engineering hasn’t really been working on any more products than usual.
(Well, and the fact we’re growing up on the production side, and making some big changes under the hood—that does take up a reasonable amount of time.)
But, you get my point: you run so fast, you don’t see things that might trip you up.
So, sometimes it’s better to slow down, sit back, zoom out, and take in the big picture. And make some decisions based on that, rather than the micro-detail you’re currently immersed in.
Yeah. Let the 3D printer be idle for a bit. Ignore the boards. Go have a beer.
“So, you’re saying ‘screw off, and that’ll make things better?’” some might ask.
Well, yeah.Maybe.
Here’s the thing: there should be a balance. I can’t tell you what balance works best for you. Running full-out with the fast-turn tools we have right now doesn’t seem to promote the best balance, at least for me. But everyone has to answer that question for themselves.
“So should we worry? Are you getting burned out?” you ask.
LOL. No, I’m still having the time of my life.
But life has seemed, er, kinda busy. And I didn’t really know why. At least not until I tore that design apart for a third time, and wondered why I hadn’t caught this earlier. And as I tore it apart, I realized I was sitting next to the 3D printer, watching it iterate yet another design, and I saw the piles of boards on the same table, and suddenly everything clicked: things seemed busy because everything moved so fast that I couldn’t get ground under my feet.
At that moment, I walked away. And I spent a weekend just thinking. No, not putting my feet up and drinking, but just sitting there with nothing more than a mechanical pencil, writing down what the challenges of the design were and how we’d addressed them, and the pros and cons of continuing down the path I’d defined, or starting over.
And things got a lot clearer.
Mike’s working on some new stuff; I did the same thing there, going through pros and cons, what-ifs, making simple sketches—far away from 3D CAD or layout software. The result is we both have a much clearer view of the pros and cons of what might come to be.
And, as a result, we’ve been working more collaboratively—Mike and Dave can play their own game of what-if, and be better informed as to the ramifications. That way, there’s no “oh schiit” moments when a design is presented like a fait accompli.
And, I think, it will result in better products.
How to Hit the Brakes?
“Okay, okay,” some are saying. “Maybe you have a point. Maybe I need to pull back a bit. But how the hell do you do it? It’s so easy to keep going and going and going…”
Sigh.
I wish I had the one answer that would work for everyone. But I don’t.
I do believe one of the keys is in forcing yourself to turn schiit off. Put the phone in Do Not Disturb mode—all the time. Turn off the automatic email checking—have your system check email only when you ask it to. Disable all notifications from all apps. Unsubscribe from all your threads—yes, even this one. Because each of those things is another interruption, another thing to get the adrenaline spiking, another thing to keep you on edge, another thing that makes the constant grind seem completely normal.
And I also believe that another key is in setting limits. Don’t check your email after you’ve gone home. Respond to phone calls only twice a day—one block of time in the morning, once in the afternoon. Or just one block, once a day. Don’t schedule meetings whenever possible—they’ll warp your entire schedule, and something will usually come up right before them. Don’t stick your head into social media while you’re getting dinner ready. Make things work on your schedule, not on the constantly-on, always-accelerating timeline that will eat your life.
“Well, I can’t set limits like you, I don’t own the company,” you say.
Fine. Sure. Yeah. And you can be like Mark, who looked as healthy as a friggin athlete (he actually was, when you get right down to it), and who ran at the pace of the company he worked for…and then had a heart attack last year, then went in for open-heart surgery for a quadruple bypass. Mark’s only a couple years older than me.
Even if you can’t put the brakes on, set the limits you can.
Seriously.
(Mark’s doing well now, by the way. He also works on his own terms.)
Or, conversely, you might say, “I have to work as hard as I possibly can, I run the company.”
Yep. And I worked like that for years. Decades, actually. And it can be a lot of fun, and it can actually work out pretty well, if you’re of a certain constitution. That’s a choice you can make. Just make sure you don’t get a reputation for being fast and sloppy. Or taking on too much. If that’s your choice, have fun with it, and I wish you great success.
And, in striving for balance, I also believe that you need to find something you love as much as what you do—but is absolutely and completely different.
For me, I’ve discovered that there is an insane range of hiking trails literally right behind our house (literally—I can walk up the hill from our backyard and hit one of them, less than 200 feet as the crow flies). So instead of waking up and going to do PCB layout—or, worse, marinating in the forums, where there can be soooooooo much negativity—but that’s a whole ‘nother thing, ain’t it?), I get up and go hiking for an hour or so. Some of it is quite strenuous and some is kinda pants-schiiting, but you know what? I can’t think about engineering there.
And, more than that—I frequently come back with a clarity that I didn’t have when I started. Problems are more easily worked through, new ideas are more frequent and ambitious.
Is it a panacea? Nope.
Hell, it’s not even my only panacea. I go stand-up paddleboarding with my wife, I try to find a new brewery to go to every weekend, I re-visit old breweries, I cook and experiment with recipes (also with the wife, now—she never cooked before, but she loves doing some of the more tedious stuff, like prepping peppers for grinding your own chili powder (try it, seriously), I work on some of our old cars, and I go off-roading with Lisa and her Jeep (she’s the off-roader, not me).
Even more importantly, it isn’t necessarily your panacea at all.
You might be more into gardening, or competitive swimming, or hell, even just more work. I don’t know you. I can’t tell you what will work.
But I can say, for me, that current engineering technology lets me keep up a pace I could only imagine in the past—and that pace is so fast I need to make myself take breaks from it.
And, in finding ways to take breaks, I think we’ll end up with better products.
We’ll see how right I am.
Depending on whether my tubes get here in time. A Coaster amp is a circuit board shaped like a coaster that is a hybrid headphone amplifier.
If you build the amp, then you have to figure out how to mount it in a cabinet or such. I am working on several options.
