How to guide a rocket using servos

Is there anybody on the planet who hasn’t heard of Estes solid fuel rockets? When I was a kid, I loved going to a local park to fire off my model rockets. Then, just as today (and for as long as I can remember), if I wanted something I had to earn the money to buy it myself. This detail becomes germane in a moment.

However, first, let me mention I’d received an Estes rocket starter set for Christmas (one very similar to the photo below). This, to include the rocket, and parachute, plus launch pad, complete with launch controller and wiring to reach the launch pad (a safe distance from the rocket, of course).

– Estes Alpha rocket and starter set – image #1 use awaiting Estes’ permission
Gray line delineating sections breaks within a blog entry

Assembly was a doodle, the only tricky bit being that of carefully aligning the balsa wood fins when gluing them onto the cardboard body. Basically, if the fins weren’t perfect, then the rocket would spin (not too bad an outcome), or fly off course (very bad outcome). Until I got good at it, this was somewhat painful for this young modeler because some rockets got lost before I developed my craft.

Similarly, the nose cone was also balsa, but it had already been turned on a lathe to the correct size and shape and thus, only needed me to screw an eye into the underside for attaching the parachute – plus painting. Easy peasy even for an inexperienced 12 y/o.

Fortunately, these days, it’s much easier because with an entry level rocket like the Alpha shown above, you now get an injection molded set of fins and nose cone. Makes for a really good present for kids, believe me. Best part, nothing needs painting.

Anyway, when I was a kid these bits were wood so afterward, I painted my rocket with yellow spray paint I found in the basement. It turned out nice but, predictably, heavier than need be because I didn’t know about primer so I used many coats of paint to hide the wood grain.

Fortunately, I didn’t paint the joint where the cone fit else I’d have ended up with a lawn dart as the ejection charge wouldn’t have separated the cone reliably. Anyway, this is one of my earlier memories of gluing bits together to make something, e.g. modeling.

Gray line delineating sections breaks within a blog entry
Launch day

So come the big day, my mom drove me to the park where, almost anticlimactically, I immediately experienced success as I only had to unwind the lead for the fire control, and after sliding the rocket down the launch pad rail, connect alligator clips to the starter/igniter leads hanging outside the engine. Then with everything ready, I performed an overly dramatic countdown, pressed the button, and it went off without a hitch – except for a spiral of sorts (an imperfect fin attachment).

But this didn’t bother me as it whooshed into the sky, and that’s all I wanted. Then at apogee, with a puff of smoke, the nose cone separated, the parachute unfurled, and it drifted down to landing a few hundred feet away as I held my breath. That’s when I discovered that I also wanted recovery – actually, at that instant, even more so than launch!

Perfect day, too, almost no wind. Honestly? I was besides myself with joy. And I was hooked! Then, after I fired the other two engines (recovering the rocket undamaged each time), that was that. On the very short ride home I almost couldn’t believe it. My excitement was off the charts!

Give a gold star to Estes . . . I’ll forevermore have a soft spot in my heart for them!

Anyway, then as now, engines are sold separately (so I was fortunate mom had the foresight to buy me a pack along with the rocket set). Point being, if I wanted more engines (and I did), then I had to buy them with my own money. Allow me a brief birdwalk.

Gray line delineating sections breaks within a blog entry
Back in the day

So back then, earning money meant I worked pumping gas at a corner Gulf station after school. I actually started off washing windshields, along with checking tires and oil while Mr. Wiggins (the owner), pumped the gas and made change. Reflecting back, I’m pretty sure he didn’t actually *need* me, so why he gave me a job is unknown. But before long I proved an apt student and soon was pumping gas, collecting the money and making change.

I was proud to be entrusted with the responsibility. And for his part, he could focus on whatever car was in his two-bay garage for repair, thereafter, only getting involved if I needed him. Like maybe to occasionally break a $20, or some such (rare because gas was 31.9¢/gallon for regular and premium, a nickle more, was 36.9¢/gal).

I also delivered the morning edition of the Birmingham News (rising before the crack of dawn to first fold them). My bicycle was key to both jobs as it was my principal means of transportation. So along with mowing lawns in summer, and raking leaves in fall, pumping gas and delivering papers were my most reliable sources of pocket money.

Also, while I was encouraged to save (half, which I conscientiously did), nobody really said anything about how I spent my money. And I spent a fair bit on rocket engines (hadn’t really discovered girls, yet, but that’s another story). And, of course, all this before I discovered motorcycles, model airplanes, and cars, but I digress.

So after work on Fridays, I scurried to the bank to cash my check only to then haul butt (again on my bicycle) to the hobby shop before they closed. Why? To buy a couple, three packs of A-motors (and occasionally a new rocket). Did this for maybe two years, or so. This, until one day I saw an RC model airplane fly and it was game, set, match as far as me and model rockets went, but again, I digress.

Anyway, in the intervening 50+ years, I’ve only again thought about model rockets, once. This, only briefly, as I perused an idle thought to do with firing rockets off a model helicopter thinking because there’s no recoil it would be easy to stay on the target for follow-up shots (so multiple rockets). But no sooner had the thought crossed my pea brain, than I simultaneously visualized cops surrounding the place – weapons drawn – and, ‘Self, you’ve had better ideas!‘, crossed my consciousness and I promptly ditched the whole hare-brained scheme! Note to Fed-watchdog, let me reiterate, I only *thought* about it, I promise!

Gray line delineating sections breaks within a blog entry
Basics of rocketry

So in case you know less than me about model rockets, they’re powered by solid fuel engines (or motors, both terms are acceptable). These are ignited when you close the circuit of a battery wired to an igniter, which you had previously shoved into the backside of the engine, where there’s a hole sized to accept it snugly – won’t do to have it falling out, eh?

These days you get a plug to secure it in place but I don’t recall such a thing so I took great care not to jostle it once installed, and especially when connecting the alligator clips.

So the engine slips into the body with a light friction fit and stops against a ring in the body. After the liftoff, there’s a delay charge, then an ejection charge that pops the nose cone to deploy the parachute.

This all seems simple *but* it’s finely engineered and manufactured to close tolerances, else the nose cone sticks (bad juju).

Here’s a link to a YouTube video (brief, just 1:30) of how all this works.

– Estes solid-fuel rocket engine – image #2 use awaiting Estes’ permission
Gray line delineating sections breaks within a blog entry

So when the igniter circuit is closed, the tip glows hot *briefly*, and this lights off the solid fuel and whoosh, the rocket is launched. After the ejection charge goes off the model floats back down so you may recover the rocket, repack the parachute, slip in another engine and igniter, rinse repeat for as long as you have money – and don’t lose the rocket!

As you might imagine, I soon had a favorite, which I’d also painted bright yellow. This was 1969-1970 and I’d grown aware of television-news in large part due to the manned NASA missions. SpaceX does this for today’s kids to the tune of 100-ish flights per year, thank you Elon Musk!

Anyway, back then I was glued to the television flying right along side these pioneers – if only in my MItty-esque imagination! And today? As it happens, I just stepped back inside from watching SpaceX launch another mission – this time in-space cell-phone towers for T-mobile. For me, this never gets old! And I’m fortunate KSC (Kennedy Space Center) is just 30 miles east of me so I get a next thing to front row seat.

Note; there’s an interesting website I like to visit called arstechnica and I follow the launch schedule through them. Sign up for their Rocket Report newsletter, if you’re interested.

Anyway, I probably launched and recovered those simple yellow rockets hundreds of times (had they known, Estes shareholders would have loved me). Eventually they flew better as I learned shiny paint jobs decreased performance due to weight. That, and I got better at aligning fins when gluing them in place. Important because once it’s off the rail the flight path and recovery depend on how well the fins were attached, the wind, and the altitude. But everything is a learning process, right?

Sometimes a broken fin on landing brought the day to an early end (before super glue), but often I experienced a successful half dozen launch and recovery missions on any given Sunday after church. I was in hog heaven, believe me!

Gray line delineating sections breaks within a blog entry
Regarding solid fuel engines

Estes engines are sold in impulse units of N-sec (Newton-seconds) and the A-engines (classification of size to do with thrust), range from 1.26-2.5N-sec before the next size larger, the B-engines. Note; there are also 1/4A and 1/2A-engines for smaller rockets (or lower launch altitudes), but I only did A-engines except for one E-engine adventure that ended in disaster because it went so I high I never saw it again.

Anyway, the B-engines step up from 1.26-2.5N-sec and range from 2.51-5N-sec, and this business of steps goes all the way up through C, D, E, before you get to the solid-fuel monsters, the 80N-sec F-engines. This engine chart explains better.

Gray line delineating sections breaks within a blog entry
– Estes engine chart – image #3 use awaiting Estes’ permission
Gray line delineating sections breaks within a blog entry

Naturally enough, there are various classes of A-engines (maybe a half-dozen, or so) as well as B-engines, etc. each tailored to the mass of the rocket.

Me? I bought the A-engines (maybe A8-3 but memory is hazy). This, because I had quickly learned from my one E-engine model, the higher powered rockets (B’s, C’s, D’s, etc.) cost more than I could afford *and* went much higher than I could reliably recover. In fact, after the big E-powered model, I lost a few higher powered A’s also before I figured this out (word to the wise).

This next chart, also swiped from the Estes website shows the typical thrust curve.

– A8 engine thrust curve – image #4 use awaiting Estes’ permission

Anyway, the A-engines were my ‘thang’ and this thrust curve is the data for one of these. Basically, not quite 1/4 second of strong *whoosh* after which thrust drops off precipitously. The model attaining an altitude, depending on weight, of maybe 1000 feet but for Yellowbird 1, it attained maybe 500′ due to being overweight, and later ones, realistically maybe 800′. There were probably 8-10 Yellowbird iterations all told (each painted the same canary yellow).

Gray line delineating sections breaks within a blog entry
Fast forward 50 years

So is anyone surprised at learning many folks getting their feet wet with Estes starter rockets, like guys who begin RC model airplanes with an E-flite Apprentice (a trainer) and progress to 200cc powered giant scale model aircraft, will similarly transition to advanced rocketry? Turns out rocketeers develop and take their sport considerably further, too? How far? A LOT further!

So this askJOHN began like most, with communication with a modeler raising a question. And this is the story of what happened after I innocently asked what it was he had in mind for our servos. So back and forth we go and this ultimately led to the purchase of three DS205BLHV brushless mini servos. Amongst our very finest.

Me? I’m thinking you might be interested in the rest of the story (as Paul Harvey used to say). And believe me, it’s a good story, so here goes. We’ll begin at the best place, the beginning, his initial message . . .

For a few years I have had a project for an enhanced stability system for our high power rockets. See attached pic. This uses an inertial control unit driving two sets of forward mounted canards (Pitch and Yaw) and a single roll-control fin at the rear. The entire intent is to keep the thing pointed vertically in order to minimize recovery dispersion.
Gray line delineating sections breaks within a blog entry
Big time rockets

So you’ve probably twigged to the fact this askJOHN is to do with rockets. Well let me tell you, I like to think there’s nothing a modeler will say which takes me by surprise, but this fellow’s query left me momentarily speechless. Basically he needed guidance servos for the canards (pitch and yaw) for steering, plus a 3rd servo for roll-control for his rocket. Yes, on board guidance system to control his rocket using servos! Why? So it won’t come down 3 miles away! Like how cool is that?

My first instinct was to double check his email address. This because while our defense contractors have a different channel of communication, occasionally project engineers reach out informally through the same email system everybody else uses. But no, this was definitely a civilian customer. How large a rocket?

This is him with his Kraken 2 . . . bit more than an Estes starter rocket, agreed?

– Wayne Robinson at Lake Louise, Alaska with his Kraken 2 ready for launch

Note: Wayne is a proud member of the Alaska Northstars Rocketry Club, National Association of Rocketry NAR, section 726, Anchorage AK

Gray line delineating sections breaks within a blog entry

So the rest of the story is large accomplished with pictures. Why pictures? For the same reason the saying a picture is worth a thousand words originally came about! But first, a bit more background, a preface if you will, as offered up by Wayne to help educate me, and which I feel is perfect for everyone;

Preface
'Almost all of the machines flown in amateur rocketry are unguided, fin stabilized, dumb rockets. They are pointed up on the launch rail and released with high hopes. Once clear of the rail their path is subject to the vagaries of fin alignment, motor thrust vector misalignment, atmospheric wind and turbulence, etc. These often result in significant dispersion in the flight paths and large possible recovery areas. The problem gets exponentially worse with multi stage high altitude flights.

The Kraken project is an attempt to reduce recovery area diameter by keeping the rocket pointed “UP” until apogee and drogue parachute deployment. An inertial navigation unit controls three servos in order to minimize roll and to control pitch and yaw to keep the nose pointed at zenith. This is often called a guidance system, but more properly would be considered a stability system as it has no target other than UP.

Kraken 2 has flown successfully a number of times as a single stage to altitudes between 10,000 and 15,000 feet, but its real purpose is to act as a stable booster for an unguided upper stage to about 50,000 feet. That remains to be done.'

. . . back to the story. So Wayne has built a rocket stabilized to a track a vertical path by an internal gyro based autopilot driving forward placed pitch/yaw canards and a rear mounted roll control fin through three ProModeler DS205BLHV servos. And he’s reached out to us because the hobby grade servos he’s using are failing. This is exactly why so many RC modelers (as well as defense contractors and industry) reach out to us, also.

– ProModeler DS205BLHV – an all-alloy servo, with steel gears, & brushless motor – it’s a jewel
Gray line delineating sections breaks within a blog entry

And it’s also important to note, this article is a compendium of two years+ of email ‘conversations’, some very brief like when he wrote to say . . . Three flights to date and zero issues with the ProModeler DS205BLHV servos.

Adding a few days later when I asked a few questions regarding what he had been using and how they failed; ‘I had been having unexplained intermittent failures on the previous servos. Have had zero issues with yours. This is the PCB of the servos (not ProModeler) that have been dying on me.

– Unsupported PCB components of this hobby grade import have failed during use
Gray line delineating sections breaks within a blog entry

So our customer had zeroed in our DS205BLHV servos to replace those he’d been using. Would they work to loads exceeding 14Gs? Honestly? This is always an open question with applications of extreme performance and there’s really only one way to find out, empirically. Or put another way, let’s try it and see! So he ordered a set of three (pitch, yaw, and roll).

Background on construction and pricing

We offer servos in five fundamental sizes ranging from our largest, the quarter-class massing 3/4 of a pound, through standard, mini, micro, and our smallest, the submicro-class, which are about the size of the tip of your thumb. All are assembled with ten machine-thread Allen-head bolts, which instead of threading into plastic, always thread into metal (the aluminum center case). Six bolts secure the transmission section with four of these boxing-in the output shaft bearings resulting in a more rigid assembly than hobby-grade servos typically offer.

– Offering 5 sizes ranging from quarter (largest), to standard, mini, micro, and sub-micro!
Gray line delineating sections breaks within a blog entry

So the foundation of our servos is the center case section. This is the structural member to which everything attaches, gears, motor, potentiometer – everything! And it must be rigid to transmit the forces accurately. Anyway, while we build some servos within an all-alloy case, others feature hybrid construction (polymer top and bottom along with the alloy center case).

Importantly, 100% have cooling fins machined into the the center case. This, to better shed of heat, which is always the enemy of electronics because when they’re working hard, they get hot.

However, unlike people who sweat to cool off, and dogs which pant, servos – like air cooled engines – must dissipate heat through convection. And the best way to improve convection is by increasing the surface area.

And this is the point of the cooling fins, to help the servo shed heat more effectively! Anyway, we refer to the center case section as the porcupine because of all the bolts used for attachments.

– ProModeler servos (except those with exposed motors) have tiny o-rings beneath the heads
Gray line delineating sections breaks within a blog entry

Why are cooling fins such a big deal? Allow me another birdwalk, please. And note; as usual, some will consider this educational, while others will consider it propaganda. The truth, as always, lays somewhere in the middle, so the Page Down-key is your friend.

First, do you remember the formula for Watts form high school physics? It’s W=V*A (Watts is Volts X Amps), so basically, a servo consuming 5A at 8V means it’s a 40W device (5*8=40). Ever touched a 40W incandescent light bulb? Burned the crap out of you, right?

Major point being, the cooling fins are *not* there to look trick, e.g. be pretty. They are there to reduce heat because heat is detrimental to the lifespan and performance of the electronics. Thus, cooling fins on your servos are 100% about shedding more heat to improve reliability and have nothing to do with looks and/or decoration, e.g. the coolness factor that appeals to unsophisticated consumers.

If you’re curious how they are made, keep reading and if not, Page-Down to skip ahead.

Gray line delineating sections breaks within a blog entry
How cooling fins are actually made

Cooling fins are considered a feature and they are created with a CNC-machine called a mill. When machining something, anything that requires a tool change, or reposition of the work within the fixture, is called a feature). Not a feature like air conditioning is a feature of a car, but feature in engineering terms.

Anyway, the process of machining each features takes time. Some add considerably to the time each individual case spends within the machine cell (a CNC mill, in this instance). Cooling fins are that kind of feature.

This is because these very thin cooling fins are actually hogged individually of 6061-T6 aircraft aluminum using a tool called a slitting saw. The mill and the slitting saw work on a solid billet of aluminum in *exactly* the manner of a great artist chiseling away everything that doesn’t belong from a solid block of marble before ending up with the details of a finished statue, understand? But, of course, there’s more to the story.

Strazza’s Veiled Lady © Barnsley Museums / Cooper Gallery Collection
Gray line delineating sections breaks within a blog entry
Time is money

We’ve all heard the phrase, *time is money*, right? Well, it takes time to cut the cooling fins, (more time than *not* cutting cooling fins – and MUCH more time than using an injection molded plastic center case). But that it taes more time to mill a case with fins actually means is there’s less time in a day before the CNC-mill can complete the workpiece (the center case) and begin work on the next case.

This also means adding more time for the men servicing the mill (more time waiting on the part to be made). And of course, since men are paid on an hourly basis, then this also means paying more in labor cost for this added time for creating the cooling fins.

So there’s added machine-time, plus the extra labor-time, plus added tooling costs (they get dull and have to be replaced more frequently). But it cascades from there once you think about it even a little bit.

This is because, for example, if you need 1000 cases per month (just plucking a number out of the air, I could use 10,000, or a million), then because you need a certain number of machines and men to make 1000 case/month, then because it takes longer to make a case with fins versus one without, then the number of cases you can make from each machine is reduced to maybe 800. So to make the same 1000 cases/month this means you either need more mills (plus more man/hours) and/or another shift, and/or a larger shop in which to have more mills working to make them!

So while it seems simple to add the fins feature to the case, and it is, it costs more time and money to do it. And in a competitive market with customers who don’t know the difference, consumer electronic companies offer servos to these uneducated customers (by definition those who don’t understand the cooling fins are actually pretty darn important) that don’t have fins because it’s cheaper to do this. Yes, it really is all about the Benjamins!

Anyway, hang in with me, I’m going somewhere with this.

Gray line delineating sections breaks within a blog entry
Cooling fins equal greater reliability

So all ProModeler servos have cooling fins because our primary customers are the engineers and scientists at government and defense contractors. Means we’re not actually a consumer products company at all. Surprised? This is why you probably learned about us through word of mouth instead of full page magazine adverts or paid field reps because unlike consumer companies offering hobby-grade products, marketing is a nearly non-existent afterthought.

Anyway, because fins add to product reliability, then the obvious conclusion to be drawn (the corollary if you will) is . . . hobby-grade servos missing cooling fins aren’t as serious about reliability as ProModeler. Interestingly, once they figure it out, modelers become just as demanding as our engineers so consider yourself informed.

Recapping; engineers and scientist care about cooling fins because they’re dealing with missions where the price difference with consumer-grade servos is a rounding error. Remember, they’re not using servos for fun like modelers (flying a model after work, or on Sunday after church), but because there are lives on the line. Big difference, agreed?

How much importance ‘you’ once gave to your servos having cooling fins is beyond our ken but I bet it’s more important, now. Call it propaganda if you will, but this is the kind of information about what’s a big deal to the project engineers for whom we produce our product we feel modelers should also be informed about.

Are you curious why competing servos leave cooling fins off their product? Let me briefly explain the added complications the consumer electronics folks deal with, beginning with distribution, or reach.

Gray line delineating sections breaks within a blog entry
Cost cascades due to distribution

Product price is actually determined not just by what something costs to make but also by how many wet their beak touching the product (but not contributing to development or construction) before *you* actually lay hands on it (these are folks termed middlemen).

For example, with hobby-grade servos, after someone produces it, these middlemen usually include an;

  1. importer making 15% for his trouble, and
  2. distributor taking 25% for his effort, and
  3. hobby dealer earning 40% for his part.

All of these middlemen are the players who make up the product distribution chain holding their hand out for their cut before you buy the servo.

Gray line delineating sections breaks within a blog entry

Thing is, these middlemen-costs are being paid for by *you*, are you surprised? What’s more, grade school math informs you a significant portion of what you pay is actually going to someone who didn’t do diddly toward producing the servo. That’s right, they just touch it to earn their cut. Stands to reason the better deal is always one, which eliminates middlemen, agreed?

So the fundamental reason hobby grade servos don’t have cooling fins isn’t that they are stupid unaware about how cooling fins make for a more reliable servo, it’s because the middlemen of the distribution chain are sucking up a huge chunk of the budget. But it doesn’t matter to them because the average hobbyists isn’t an engineer. They don’t know the difference, or as it’s often phrased, ‘Ignorance is bliss’.

Bottom line? Product costing isn’t rocket science and the reason hobby grade manufacturers can’t put servos in your hands, which are feature and price-competitive with ProModeler isn’t that we perform magic, but due them choosing high volume sales, which result from using the established chain of product distribution.

And while we know we’d sell waaaaay more ProModeler servos if you could walk into any of 3000 hobby shops to buy them, we’d have to raise our prices to account for the margin the middlemen need. We’re happy enough with our direct business model, instead, and you, the modeler, get the better price for it.

Gray line delineating sections breaks within a blog entry

Anyway, this particular rocketeer selected mini-class servos, which are right in the middle of the 5 size-range. But not just any of the half-dozen minis we produce, he ordered three of our very finest, the DS205BLHV.

Note; we actually make 8 mini-class servos but 2 are specialized products. For example, one offers 180° vs. 60° rotation, (typically for retracting landing gear or operating a feature like a sliding canopy), plus one has its neutral at 760us vs. 1520us (for a helicopter’s tail rotor, where the servo is routed through a gyroscope before the receiver – if plugged into a receiver it does nothing because it’s outside the receiver’s control range). For his purpose, these six were the ones of interest.

– Within the mini-class, we produce 6 servos priced from $50-100 a pop!
Gray line delineating sections breaks within a blog entry

So why did he select our finest mini-class servo? Basically, and without putting it in so many words, because, a) the Kraken2 project is important to him, and b) his time traveling to test fire (launch) meant going as far as 150 miles from home, then c) risking everything on saving a few bucks, was deemed silly. A less charitable view is, as spoken by Hanks, in the role of Forrest Gump, when he said, ‘Stupid is as stupid does!’.

Figuring it out isn’t difficult and thus, distinguishing the best is also kind of how there’s a difference in sound when you shut a Kia’s driver’s door versus slamming that of a Mercedes (you know which is which, even with your eyes closed).

Gray line delineating sections breaks within a blog entry
God is in the details, Ludwig Mies van der Rohe

So what makes ProModeler servos better than competing designs? We already mentioned the cooling fins. Another part of the equation is insisting every single part of the product is pick of the litter. Put another way, since the whole is more than the sum of the parts, better parts result in better servos.

Think of it as paying attention to the details. Like installing hardpoints in the 6061-T6 alloy case to reinforce the bores. In the DS205BLHV, these hardpoints are tiny bits of precision Swiss turned 303 stainless steel. They’re knurled and pressed in place.

Close up photo of steel bushings turned and knurled on Swiss lathe for the purpose of reinforcing a ProModeler alloy case section.
– Super high precision Swiss turning produces the 303 stainless steel inserts
Gray line delineating sections breaks within a blog entry

They exist to reinforce the bore in which the steel gear shafts are fitted because aluminum is lightweight but actually pretty soft stuff. So basically, they’re like armored points in the case.

Why do hardpoints matter?

To preclude accelerated wear under load because otherwise, the much softer aluminum gives way to the steel gear shaft under load. This is when the perfectly round bore into which the shaft is fitted becomes egg-shaped because of the pressure exerted by the shaft when the servo is loaded and unloaded. Remember Newton’s Third Law a.k.a. Action & Reaction.

His 3rd law states that for every action (force of the servo on a control surface) there is an equal and opposite reaction (shafts being forced against the aluminum of the case). So when the servo is working to force the control surface to move, it’s pushing against the case and Newton said, when object A exerts a force on object B, object B also exerts an equal and opposite force on object A. Aluminum being soft is fairly easy to distort. And distortion is when the round hole is hammered into an egg-shape by the stainless steel gear shaft!

End result? After heavy loading, the un-reinforced case begins to fail. The round bores go egg shaped. The previously precise gear mesh goes to Hell resulting in accelerated gear wear and ultimately, gear failure. Worse, if you just replace the gears without replacing the case with a new one, then the gears quickly fail again! Put another way, good money after bad! But when you do the math of gears plus new case the servo turns out to be a throwaway. Modeler never considered that, professional engineers always do.

So the best servos, ones like ours, always reinforce the case! Hobby grade servos? No because there’s no money for that when you also have to fund the distribution chain. So this next photo shows where the inserts are pressed into the case for one of our servos for reinforcement. And all of our servos have something similar, even the hybrid case servos.

And it goes without saying this costs more to do but for the savvy consumer, this is one of *the* most important reasons for coming to ProModeler for your servos.

– When the inserts are pressed into the soft aluminum, it’s almost like armor plating the case
Gray line delineating sections breaks within a blog entry

If you found this image interesting, then if you have time, review this article to learn more;

Another thing making ProModeler servos better is that because of whom our most important clients are, we certify servos to several MIL-STDS. For example, one of the things that recommended the DS205BLHV, especially for use in a rocket is that instead of the usual 3, this servo meets five MIL-STDS.

MIL-STD-810G-Part 16
  • Shock – Test Method 516.6
  • Vibration – Test Method 514.6
  • Acceleration – Test Method 513.6
  • Altitude <70,000’ – Test Method 500.6
  • High Temperature – Test Method 501.5

Beyond a motor supported within the alloy case for better cooling, meeting the first 3 test methods is aided by the heavy application of monkey snot to support the surface mounted components.

– When touching the finned aluminum case, the motor is cooled better during hard use
Gray line delineating sections breaks within a blog entry

FYI, monkey snot is assembly’s loving term for potting compound. Why? Simple, because it’s a stinky and sticky mess to apply. Gets on everything and once on your hands, you now need lacquer thinner to help get it off. Seriously, it’s a pain in the hind end but this aerospace building technique is found in every military grade electronic component in the air and in space for good reason. It works.

So basically, a liberal application of potting compound reinforces the components mounted to the PCB itself, thus combining the rigidity of the motor in the frame, with the PCB soldered to the motor endbell and supported by the lead exiting the case, all working together to ensure meeting test methods 516.6, 514.6, and 514.5 – so mechanical reinforcement via a liquid that solidifies once in place!

– The potting compound on the PCB flows into all the nooks and crannies and then solidifies
Gray line delineating sections breaks within a blog entry

When combining potting compound, along with a direct fit motor (one that doesn’t rely on merely being fixed to the case with screws because the case itself supports the motor from moving), plus the circuit board – supported on one end by being soldered to the endbell of the motor and on the other end supported by the lead exiting the case – is also combined with the potting compound, then these design features team up to work together.

Teamwork

Note, one single feature isn’t more important than another feature, they all work together for the servo to function to the standards – survive in other words. Proof it works? Wayne again;

''Attached an accelerometer to the motor tube by the servo. Seeing 3-axis high frequency acceleration of +/-40-80Gs for about 1/10 second during ignition transient, followed by +/- 5-10 Gs until motor burn out on top of a maximum 14 G of linear acceleration.'

Anyway, once monkey snot sets up, it results in another type of mechanical reinforcement, e.g a reinforcement acting like a buttress supporting a wall, it basically acts like a continuous fillet (but even stronger because it encompasses everything on the circuit board).

Look, these MIL-STDS aren’t participation trophies. They have to be earned! What’s more, our competitors are free to improve their servo designs until they, too, can meet these standards. But for whatever reason, probably to do with costs due to the distribution chain, or possibly because they judge the market for people who are anal about such things (e.g. not sufficiently large, or important enough), they just don’t.

This edge case is where ProModeler thrives. Basically, we service the savvy. Regardless, why they don’t build better servos, well this isn’t for us to say. And what to buy is your call, not ours. And by the way, potting compound isn’t terribly expensive on a per servo basis, the labor (time) adds up, but still, if we can do it, so can everybody else. Anyway, this reminds me of another trite saying; you get what you pay for!

Note; this stuff comes in clear, white, and black (doesn’t matter, which is on a servo, we use them interchangeably). Key thing to remember is this; it’s not decorative like icing on a cake. It serves an important purpose.

Anyway, cooling fins, direct fit motor, potting compound, socket cap screws, o-rings puts me in mind of what football, basketball, baseball, and soccer coaches are famous for saying; there’s no ‘I’ in team. Same thing with these features of a ProModeler servo, no one feature is more important than another, they work together to deliver the performance you depend on.

Gray line delineating sections breaks within a blog entry
Selecting the DS205BLHV

So just like our defense clients, this similarly serious rocketeer specced our servos because he needed ones that would work reliably under incredibly demanding conditions. Maybe you’re not interested in rockets, I get it because we all have our own thing, but I suspect learning about this kind of stuff *may* float your boat even if you like turbine powered jets, scale model helicopters, or extreme performance XA (fixed wing model airplanes).

If quality is important, give us a call. Especially when you deem your servo-application just as deserving of the very best money can buy servo-wise as this rocket-guy!

Gray line delineating sections breaks within a blog entry
Cold, warm, cold, warm, cold

By the way, back to the MIL-STD test methods, 500.6 is one of the tricky ones to earn because the temperature range of our standard grease Super Lube NLGI Grade 000 with Syncolon, which is from -43°C to 232°C. However, while some customers operate at 36km (118,000ft) where this stuff works beautifully, those requiring operation to 70,000ft (-60C) make things more difficult because it’s actually colder at 70,000ft than at 100,000ft (-25C). Surprised?

Don’t be, it’s just because of the fact there’s a higher ozone (O3) concentration in the stratosphere versus the adjacent layers, the troposphere and mesosphere. So here’s the missing bit of information few learn, or forgot if they did . . . temp both decreases, and increases with altitude, and decreases and increases once again on the way to 120km (basically 400,000ft). Next stop, outer space!

So here’s what’s peculiar about the altitude ranging from 10-50km (33,000 to 164,000ft), e.g. the stratosphere. It has more O3 than the troposphere, so when exposed to sunlight, the O3 breaks down into oxygen (O2) and atomic oxygen (O). Then, when the O2 and the O recombine (O2 + O = O3) the resulting reaction gives off heat! And it’s a continuous process, repeating over, and over again, a self-licking ice cream cone, so to speak!

This is why, within the stratosphere, it actually gets warmer the higher you go. Basically down to a chemical reaction between ozone and sunlight! The problem for the grease in our servo is that between 30,000ft and 70,000ft (when it gets colder than a witch’s tit), we’re outside of the temperatures at which it’s rated to perform.

Temperature vs. Altitude

Anyway, we likely first heard all this in middle school physical science, and promptly forgot. Private pilots learn how temperatures decrease with altitude at the rate of 3.56°/1000ft, but because this is true only up to about 10km (roughly 33,000ft), then because they almost never exceed 12-15,000ft, almost as promptly as the kids, forget what amounts essentially is trivia if you get on Jeopardy because it’s outside of practical application.

But commercial flights often operate above 33,000ft, so these pilots know all about this temperature inversion. Ditto pilots of private jets, who sometimes exceed 50,000ft, call it 15km. Still barely into the stratosphere, but solidly within it. And rocketeers, like the guys thinking of soaring them to these heights, know it too.

Note; this graphic may serve as a primer/refresher if you’re curious:

Anyway, since 30-000-70,000ft means quite possibly experiencing temperatures outside of the lubrication range for our servos, this may pose a problem for operating our servo gear trains while using the standard grease. Heads up.

Further to this, whether it’s actually going to pose an issue also depends on how much time the vehicle spends there. E.g. during the period of ascent and descent when temperatures are outside of the grease range. Ascent on a rocket may not be a big deal because it happens very quickly, but a craft that transitions to soaring mode, may spend a considerably period of time experiencing these temperatures, understand? As will never surprise the savvy, the answer to whether it matters is . . . it depends!

Gray line delineating sections breaks within a blog entry
On the rarity of hens teeth

There actually is an old adage, which speaks to something being as ‘scarce as hen’s teeth. meaning really, really rare. It’s this kind of rare I can speak of regarding UAS vehicles like this rocket project, in which we are peripherally involved. Mostly, we sign an NDA (non-disclosure agreement) and mum’s the word from there, forward. So because much of what we are involved with is covered by NDA, I am especially grateful to Wayne for what he brings to the table in allowing this special discourse.

Similarly, recently there was news of a weather balloon transiting US airspace. One, which could potentially release gliding UAS (vehicles which may spend a considerable period of time within these very conditions whilst descending). Meaning, long enough for the servo’s gear train to have a problem due to the grease operating outside specs.

A potential solution, an obvious one, is though the use of thermal strips, e.g. electric heaters (plus the juice to power same). But what if you don’t *want* to use that approach due to the added mass and energy consumption?

So quite obviously, any idiot can find problems, but true genius lays in solutions. As it turns out a practical alternative to strip heating is simply ditching the grease altogether. Or more accurately, replacing it.

So if your servos will see temperature excursions that concern you, first begin by removing the stock grease if your servos will see ultra low temperature (or high altitude) operation. Just disassemble the upper case, remove the gears, and clean them off. Soap and water, or maybe use ordinary spray carb-cleaner from an automotive parts store. Then, substitute automobile synthetic motor oil, which does not freeze. What? Something like ordinary Mobile 1 automotive motor oil? Yup.

We’re not kidding. This works because the pour-temperature for this synthetic motor oil is -58°C, which means it only becomes slightly thicker, sort of grease-like in fact, at -60° (kind of like honey on a cold morning before the heat kicks in). Thusly thickened, it is still *very* protective of metal-to-metal wear and thus, the gear-train is still protected by an excellent lubricant.

Isn’t it nice when the *problem* of operating at 30,000-70,000ft is resolved inexpensively? We thought so and in fact, this is exactly how we met the challenge of 500.6 (and if our customer informs us, we’ll switch lubricants for them during the build). Do you think this is ingenious? Maybe even outside the box thinking? Well, yes, it is! Who’d a thunk it? We did!

Bottom line? Our defense business (as well as certain university and industry clients) require solution to real world problems. If you can’t keep from getting wrapped around the axle when sorting these things, then you’re in second place. And this is also why you pay us the big bucks!

End of gratuitous self-promotion (although in the south, it’s not bragging if you can back it up).

– Precision micro-hobbing – stunning 303 stainless steel bull gears and 304 and pinion gears
Gray line delineating sections breaks within a blog entry
Pitch/Yaw (P/Y) Control Unit

This next image is of the Kraken’s P/Y control assembly complete with servos. This thing slides in through the top of the rocket body and once fixed in place, receives a pair of pitch-canards and another set of yaw-canards.

Take special notice the instrumentation section, the red PCB and how it’s installed. Basically, this is the heart of the vertical seeking inertial navigation unit. If you’re curious, it’s a Sparkfun UDB5 (link). This is a UAV Development Board version 5, which has been obsoleted and is no longer available (but it’s descendant is). It comes with a CPU, and MPU-6000, plus a MEMS 3-axis gyroscope complete with 3-axis accelerometer.

Note the roll control servo lead pigtail (another DS205BLHV) coming out the bottom.

– The red PCB is the vertical seeking Sparkfun UD85 inertial navigation unit w/two DS205BLHV
Gray line delineating sections breaks within a blog entry

So after these photos, Wayne shared he got the servos installed and the control system modified to allow external control of center and span. The servos are working fine. He continued with . . .

'Took the rocket about 150 miles north to Lake Louise, a big lake perfect for a test flight and launched on a M motor to about 16000 feet, circa Mach 1.1. Encountered a control system instability as it approached Mach 1 apparently an issue with excessive authority on the roll axis coupled with rocket body structural flexibility.
'For a period the roll system was dramatically over controlling and the machine was rolling wildly in one direction followed by roll in the other. This caused the P/Y controllers to miss sense tilt and flop back and forth themselves, thus flexing the structure and adding in another factor. All in all, quite a dance. The rocket was laid over quite a bit and at apogee was still moving fast horizontally, so the drogue recovery chute inflation was violent and the structure was damaged. Fortunately, the GPS led us right to it and it is rebuildable, and the servos? They worked great.'

This accompanied by this interesting photo of the rocket body buried in the snow.

– Main rocket body sticking out, with red/yellow drogue and red/green min chutes, Lake Louise
Gray line delineating sections breaks within a blog entry

He added . . .

'Small red/yellow drogue chute deploys at apogee to allow stable but rapid descent. Large red/green Main chute deploys at about 800 feet to reduce landing impact.'

. . . followed by another view of another eventful landing showing the entire recovery systems after deployment. This, after the Kraken came to rest with the aft mounted stabilizing fins buried and the roll-control servo in the water layer between the snow and the ice layer.

– A somewhat sad sight, but the results of a fully successful transonic flight – Lake Louise, AK
Gray line delineating sections breaks within a blog entry

A few weeks later he updated me with progress information and expounding a bit further regarding instrumentation for gathering data and how they got G-data . . .

'We glued an accelerometer onto the motor mount tube, right next to the Roll servo mount, which is also glued to the motor mount tube. So accelerations seen by the sensor are probably pretty close to what the roll servo was seeing.'

The graph below shows the vibration of the tube in G’s for the first second of motor burn (the servo is experiencing 40-80 Gs for about 1/10 second). After that, the levels remained relatively constant till burn out 11 seconds later. (5-10Gs), or as Jerry Lee Lewis once sang it, ‘A Whole Lot Of Shakin’ Going On!’

– Ignition transient acceleration was +/- 40 to 80 G’s in all three axis then +/- 5 to 10 G’s
Gray line delineating sections breaks within a blog entry

He continued with . . .

'I had been having unexplained intermittent failures on the previous servos. Have had zero issues with ProModeler servos. However, the roll servo might become an issue now, though. In the after pic (above) you can see the main body of the rocket sticking upright in the snow. But what it does not show is that there was water under that snow and atop the ice layer. The roll servo got wet. I am drying it out over a heat vent and will give it a try in a few days.'

Later adding . . .

'The roll servo water sealing worked, apparently. The servo checks out fine.'

So this is when we got a better look at the P/Y assembly with this photo.

– Kraken pitch/yaw assembly close up – note the pair of ProModeler DS205BLHV servos
Gray line delineating sections breaks within a blog entry
Then a few weeks later, another update . . .
'Got two more flights on the stabilized rocket this year. Still had a bit of instability on the first one, but that is now resolved with further constants changes in the controls, and the second flight of rocket Kraken was perfect.'

Here are some basic specs;

  • Body diameter 4″
  • Length 104”
  • Launch weight 29.125lbs (466oz)
  • Aerotech M1297 Motor (link)
  • Max velocity 1180 ft/sec (Mach 1.08)
  • Max acceleration 15.3 G’s
  • Max altitude to date 13320 feet

Another conversation he added . . .

'Stabilized to a track on a vertical path by the internal gyro based autopilot driving forward placed pitch/yaw canards and a rear mounted roll control fin through three ProModeler DS205BLHV servos. Three flights to date and zero issues with the servos.'

And to get a sense of scale for the model, this is a photo of the Kraken attached to the launch rail but on its side and laying on the ground. The four forward canard are steering this thing in both pitch and yaw. This, to keep it tracking vertically. And aft between the four stabilizing fins is a third ProModeler DS205BLHV, this one for roll control.

– At 108″ in length, the Kraken has exceeded 13,000 feet of altitude and Mach 1, both
Gray line delineating sections breaks within a blog entry

So a request for further details resulted in this, plus a schematic; The switches on this system are closed in sequence with many seconds between . . .

'S1, UDB (control board) power. UDB boots up and will begin sending out a 1500 uS PWM signal.
S2, Servo power. Servos get power, and will move to center point responding to the 1500 uS signal from the UDB. Note, if I open UDB power while servo power is closed, and the UDB control signal is lost, the servo will move to its internal failsafe point also at 1500 uS. Same if the UDB fails in operation.)
S3, UDB enable. This signals the UDB control board to begin outputting a PWM signal based on current pitch/yaw attitude and roll rate. Signal can vary from 1000 to 2000 uS at 40 hz. This switch carries no power, just pulls two UDB control pins to ground. NOTE: in all of the previous failures this switch S3 was never closed. I always turn the fins off center prior to S1 closure, so that I can see the response of the servos when they get power with S2 closure. In all of the failures the servo failed to move to center point, and I stopped the sequence.
Unit is SD in reverse order.
So: all servos get their power at the same time. But all of the servos are getting a 1500uS PWM signal from the UDB when servo power is applied, so will try to move to that location immediately.
This sequence of S1 and S2 had previously been reversed, back when I was having the initial failures in early 2019. We went this route so that the UDB would be fully stabilized before the servos received power, thinking that perhaps the UDB was sending out erratic commands in the settle out period and if the servo was unpowered it could not respond inappropriately and cause a failure. Failures have occurred with both sequences.'
– Schematic showing the control loop
Gray line delineating sections breaks within a blog entry

Anyway, Wayne also included a closeup of the canard mounting we’d seen in previous photos but now installed using four bolts. These look sharp enough to shave with!

– Close up of the guidance bay exterior and canard mounting – note alignment marks
Gray line delineating sections breaks within a blog entry

. . . and wrote further (briefly) about the last flight . . .

‘We got a good transonic launch last spring and need to get one more before I stick a really big motor in it and a second stage on top.

Your servos continue to give trouble free service in both projects.’

. . . and shared this nifty photo of the other project rocket drifting down to a sand bar as taken by an onboard camera. This rocket reminds me of Yellowbird!

– Note the nose cone hanging just to the left – return to Launch Knik Sand Bar
Gray line delineating sections breaks within a blog entry

He continued with . . .

'The other project (Yellow rocket) is also intended to reduce recovery area, this time by flying home under an RC controlled gliding parachute. This project also uses ProModeler servos, but this time the DS355CLHV.'
– Drifting down, zero breeze, a perfect cloudless day for launch and recovery
Gray line delineating sections breaks within a blog entry

So then I asked about the roll-control servo and he responded with . . .

'This is bare roll-assembly. The roll-servo mount is attached directly to the blue motor mount tube and is thus about 1” from the operating motor during boost. There is an aerocover (fairing), which slips over it from the rear, but the push rod is exposed in the free air stream.'

. . . plus this photo.

– Sitting an inch from the motor, this roll servo has been instrumented and sees 40-80Gs
Gray line delineating sections breaks within a blog entry

And lest you think this a one-man effort, Wayne took special care to share the love with . . .

'Both of these projects are hobby only. I am working both pretty much alone. However, on the vertical stabilization project I have had invaluable help from Bill Premerlani (ex-GE controls and autopilot engineer). Bill did the actual programming of the UDB5 control board. Jim Jarvis is another engineer and rocketeer pursuing a similar project of vertical stabilization in Texas. We keep in touch to share info. Locally, Dave Bruchie, a fellow Alaska Northstars Rocketry Club member and fellow retired oil patch engineer is involved in both projects as a resource, programmer, logistics hand, and idea bouncer.'
Wayne Robinson (constructor) and Dave Bruchie (right) before another Kraken launch
Gray line delineating sections breaks within a blog entry
It’s a wrap

Last thing; key to grokking the whole of this story I’m sharing of Wayne’s effort is that I’d ask a few questions and he’s respond. Honestly, I fumbled at making sense of what he said sometimes so I asked more questions. This article is the work of a couple of years of give and take between us. Him giving, of course.

Like when I asked for a detailed shot of the interior of the guidance assembly and he took it partially apart to accommodate my request. I just wanted to better see the servos and the linkages for the canards figuring this would be an area of interest for readers. Resulted in this close up detail photo

– Guidance bay showing both servos and bellcrank linkage to one cross shaft
Gray line delineating sections breaks within a blog entry

So key to understanding is the above is a couple of years worth of conversation compressed into this brief read. That, and I’ve been mining our conversations basically picking and choosing amongst his words in my feeble attempt to tell the story *and* have it hang together as a cohesive whole. Put another way, any errors or omissions are mine, and mine alone.

What’s next? A second stage. When it happens I’ll create part two, and include a link in this document to make it easier to find. Granted, this may take another year, or two, but no matter because like all modelers, Wayne is a tenacious sort. Anyway, that’s when you’ll know the rest of another story . . . good day!

1 px gray line to each side of ProModeler slogan; Better parts. Better servos. The formula is simple. to help delineate and close an article.
Further research:

Still have questions? Call 407-302-3361 and we’ll help guide you into what’s best for you.

Last thing: for me, Estes was like a gateway drug into the sport of aeromodeling. I’ll be guiding my grandsons into rocketry as I strive to entice them outdoors with hobbies