Hackaday Prize Entry: Very, Very Powerful Servos

A few years ago, [patchartrand] decided to build a robot arm. The specs were simple: he needed a drive system that would be at least as strong as a human arm. After looking at motors, [patch] couldn’t find a solution for under $3,000. This led to the creation of the Ultra Servo, an embiggened version of the standard hobby servo that provides more than ten thousand oz-in of torque.

Your typical hobby servo has three main components. The electronics board reads some sort of signal to control a motor. This motor is strapped into a gear train of some sort, and a potentiometer reads the absolute position of a shaft. This is basically what the Ultra Servo is doing, although everything is much, much bigger.

The motor used in the Ultra Servo is a very large brushed DC motor. This is attached to a 160:1 planetary gearbox and the electronics are built around four reasonably large MOSFETs. The electronics are built around the ATmega168 microcontroller, and the specs for the completed servo include 12 V or 24 V operation, TTL, SPI, and standard RC communication, 60 RPM no load speed, and 60 ft-lbs of torque.

This is not your standard servo. This is a massive chunk of metal to move stuff. If you’ve ever wanted a remote-controlled Cessna, here you go. That said, servos of this size and power will always be pricey, and [patch] is looking at a cost of $750 per unit. Still, that’s much less than the thousands of a comparable unit, and a great entry to the Hackaday Prize.

76 thoughts on “Hackaday Prize Entry: Very, Very Powerful Servos

  1. we’ve used industrial servo’s by B&R to drive huge heavy XY tables, they were cheaper than the price in the article, even with controller. And they would have no problem tearing a human’s arm off.

  2. $750 for 3/4hp seems a lot although the 160:1 gearbox might be quite pricy. The electronics should cost only a little more than for smaller servos and 3/4hp brushed motors are not expensive. I would hope this could be done in the $200 to $300 range. Any good sources for low cost gearboxes?

    1. Ebay for low cost gearboxes, used Bayside or Harmonic Drive units. There are a few other brands to watch out for too. But forget new. A good gearbox can cost as much is not more than the motor.

  3. I dont think water cooling the motor like that will do much. The magnets are not the part that really gets hot with a PM motor, its the armature. If you had a brushless motor with the windings as the stator then you would see some benefit.

    And I have no idea how you are going to get that much power out of that motor, the spec for the RS-775 say 45 watts out at max efficiency and about 150 watts max output. Thats about 1/5th the power you are aiming for.

    1. Judging from the picture (rendering), that motor is of the size of them used in cordless drills. There are some with up to 500W at max. power. Of course that way of water cooling is not efficient – was my first thought when I saw the picture. As you say, that would be useful for a BL motor.

  4. Dear Brian,
    could I please you not to use Martian units or if you have to, please do it consistently.
    Explanation: I had to use Google and a calculator to find out what oz-in and ft-lbs are and how to convert it to something meaningful (70 and 81 Nm with respect to the order of being mentioned in the article) or at least compare those 2 numbers.
    Maybe people in USA and in one country which no longer belongs to EU understand the numbers you wrote here, but for Christ’s sake, how long does it take to compare those 2 numbers in different units even for them? That’ like saying it will be 94° F in the morning and 267 K at night.

        1. Should I mention how the Mars Climate Orbiter was destroyed just because Lockheed Martin made one of its instruments output data in miles imstead of km while NASA expected SI units as stated in its manual? Oh and yes, Lockheed is American company.

        2. You didn’t get there because of USCS, you got there despite of it. US is like the most powerful nation in the world back then. NASA could’ve invented an even more fucked up system of units and they will still reach the moon.

          1. Ah, yes, of course: US space program was direct continuation of Wernher von Braun’s V2 program, so naturally the science was done in continental Europe’s metric units.

        3. Back to my point: If using nonsense units is necessary, don’t use two completely different units for the same thing, especially when they differ only by some 15%. The units you used made it look like the torques are in ratio of 166:1 if ome didn’t carefully read the units.

          1. It’s not going to win anyone over to your point of view to call the units they use “Martian” or “Nonsense” units. Fact is that all units are basically arbitrary and have no meaning to anyone that doesn’t use them on a regular basis. However I do agree that one should be consistent.

        4. Without any help from previous discoveries made in the rest of the world I’m sure. I’m sure no labour and expertise from other parts of the world was used either. Nope, purely a US accomplishment.

        5. The “Lunokhods” (or whatever the correct EN transcription is) were designed, built and landed using metric. Until 2014, Lunkhod 2 held the longest distance traveled on an extraterestrial surface…

          p.s. Imperial units are defined as a multiple of metric standards. There is no standard foot or pound, they are referenced to metric standards. There is a standard kilogram in a US metrology lab. Still care to argue which unit is superior? :P

          1. The Soviet space program actually used kgf as the unit of force because of political reasons. The propaganda bureau was trying their best to pretend that Russian scientists had invented everything in the world, as part of raising a “new socialist consciousness”, and they were particularily proud of mathematics (because it cost virtually nothing to teach) so they preferred to not mention Newton.

            kgf was never part of the metric system. It’s problem is that it requires one to assume standard gravity, and carry that constant through your calculations where it will get mixed up with local gravity, or if you omit it then you’ll have the problem of scales where a kilogram of potatoes in Murmansk is not a kilogram of potatoes in Volgograd.

        6. Plot twist: The USA is part of the metric consortium. It’s been for a while. A quick search shows that the Apollo guidance computer was programmed with SI units, but displayed Imperial ones to the user.

          Which means: Your argument is invalid.

    1. Original definition of a meter:1/10,000,000 of the distance from the north pole to the equator in a line running through Paris, France. Sounds pretty arbitrary to me. Today it’s the distance light travels, in a vacuum, in 1/299,792,458 of a second. Again, sounds pretty arbitrary. This does not even follow the famous metric rule of 10. The definition of a second is another one that is arbitrary, based on the spin of an insignificant planet in a spiral arm of an insignificant galaxy called the Milky Way, of all things! Just get comfortable with inter-converting between units and explicitly state the units.

      1. And yet you use a metric system for the dollar.not so hard eh? Invest some time to look at the old english currency of before 1960 to see the point of all this metric babble and why it is so convenient. Base 10 is it.

        1. Damn. comment-bait again. I hereby declare that i will not, I repeat will not react on a brian benchduff post ever again. my sincere apologies for my last comment here.

      2. “1/10,000,000 of the distance from the north pole to the equator in a line running through Paris, France. Sounds pretty arbitrary to me.”

        Not really, perfectly fine thing to use. That won’t change too much over time. But feet on the other hand…my EU size 44 foot or my GFs EU size 38 foot? And why is the spin of that planet in that galaxy insignificant? It’s a pretty significant planet to us, who defined the second. I mean, we live on this piece of rock, how’s that insignificant?

        You know why the french invented the metric system? Because they had too many different systems/definitions across the whole country back then. Because buying two feet of rope twice for the same price but from different sellers would end up in having two different lengths of rope. It was all a matter of standardization.

      3. >Original definition of a meter:1/10,000,000 of the distance from the north pole to the equator in a line running through Paris, France.

        The original original definition of metre was a pedulum that under standard gravity makes a full swing in two seconds, or one half-swing in one second. (I.e. you count the time it takes to pass a point twice). That would give you a metre that anyone can build and calibrate literally on a shoestring budget, and it still works to a very good accuracy if you know your local gravity, latitude and altitude.

        >Sounds pretty arbitrary to me.

        It’s not. It’s based on dividing the earth into quadrants in a system where a full circle is divided into 400 degrees, so the math of adding up angles for the purposes of land surveying and navigation becomes very simple: a right turn is +-100 degrees. Thereby it became possible to divide each quadrant into factors of ten, which is handy, and from there dividing into smaller and smaller bits until you reach something of practical size, the metre naturally follows.

        The interesting bit is that by coincidence, the earth is just so large, with such a mass, and rotating at such a rate, that the two definitions ended up nearly exactly the same. There’s only couple percentage points difference between them.

        1. Today of course we use radians, because it works better with calculus than degrees. Back then they didn’t have pocket calculators with built-in functions for irrational numbers like pi or Euler’s number, so the use of gradians (gons) made their work simpler.

          For example, on the complex plane (x-i as compared to x-y) multiplying any number by i means rotating it by 90 degrees. If you have a number n along the x axis, and you multiply it by i, it ends up the same distance away on the imaginary axis and it’s now pointing up. If you multiply it by i^2 it ends up being multiplied by -1 and it goes in the negative direction on the x-axis or a half-turn from the starting point. Multiplication by i^3 puts it again on the imaginary axis, but this time pointing down, and i^4 makes it go around back to where it started.

          So you see how useful it is to have a system of angles where the right turns are 0, 100, 200, 300, and 400. Those match your right hand rotations on the complex plane. If we have a point coordinate at 25 + 7i and we want to rotate it around the origin by 147 grad, what do we do? Simple: multiply the complex number by i^1.47

          Back then you’d have a book full of pre-computed results for i^n, and the rest is simple arithmetics.

          So the decision to divide the earth into quadrants and then millionths to derive the metre was anything but arbitrary. It was necessary.

    2. Totally agree. I wish one day americans will discover the advantages of SI and the fact that the rest of the world laugh at their silly units. Not even the English use the imperial mickey mouse units as much as the americans.

      1. My understanding is the English still use imperial for pints of beer (and the obscure unit “stone” for people’s weight), whereas Americans (we still deserve to be capitalized in spit of clinging to our old units) pretty much never use imperial; the US customary gallon and pint differ from their imperial counterparts. Just to add to the confusion, a substantial number of people incorrectly refer to the US customary system as “imperial” or “English”.

      2. Yes. That’s really how it works. it’s Americans the people themselves as a whole, not the few in charge of our industries that decide these things.

        Also we just have to realize the benefits of one system or another and then we get to freely chose to switch to the new system. We have no existing investment in infrastructure and tooling holding us back. No old stuff we need to keep working and no need to make our new stuff compatible with the old.

        Seriously though I suppose that’s the silver lining of having your cities and industries bombed out in a world war. When it’s over you get to start over building your world with the latest ideas rather than being tied down to the old. Then a couple of generations later your grandchildren can make armchair judgements against those who didn’t see much fighting on their soil and therefore are still tied to a legacy that goes back before the industrial revolution.

        Personally I will choose the burden of legacy any day.

        1. Metrication of Europe started well before WW1/2

          France, Italy, Germany, Spain, the Ottoman empire, Portugal, Belgium, Netherlands, Romania, Austria, Hungary… all the greatest powers of Europe were already metric countries by 1875

      3. Those of us (Americans) in the tech fields probably already DO know the advantages of SI. However it seems that most of the rest of us are either afraid of SI because they think it is unknown, or don’t want to admit that a foreign system is better. Nothing will change here until the government requires all consumer weights and measures, road signs, and public schools to change. Not likely in the next 4 years.

        HOWEVER, I (like another commentor here) use both systems every day, and it’s really not hard to do. In fact I’m surprised by the number of tech people that seem to think it is.

        I would guess that getting America to switch to all metric in my lifetime is about as likely as getting the rest of the world to switch to all English.

    3. https://www.whois.com/whois/hackaday.com

      Seriously! Why is it hard for some people to get this? HaD is US owned and the US still uses USCS units which were copied from and are almost identical to British Imperial units. Thankfully we have these immensely powerful computers known as the human brain that can recognize this and even do conversions. Not to mention all the convenient web based tools that can make it easier. Ever heard of this one? http://bfy.tw/AO2C

      It’s a common thing I hear from people around the world that “Americans” (sorry rest of the western hemisphere, I don’t know another word to call us) expect the rest of the world to speak English and don’t bother to learn anyone else’s language. I was born, raised and still live in the US. My family has been here for centuries. I have never been outside of North America (unless you count a couple short trips to Caribbean islands). I haven’t even visited the predominantly French parts of Canada. And yet… I speak four languages and never bitch when I read an article on the internet that uses Metric units!

      Complaining that US owned HaD uses USCS would be like an Anglophone complaining that most of the Chaos Computer Club’s content is in German. How conceited do you have to be to think this way?!?!

      I wish the world all used one system of measurement. I don’t really care which one. There are things I like about each but I am not really that invested either way. The world doesn’t care what I wish for and I’m not asking it to. Who are you to do so?

    4. I’m from the USA, born and raised. Almost everything here that’s made domestically is “Imperial” units, and a lot of the rest (e.g. import cars) is SI. I use both every single day; it’s not that hard.

      1. I think it’s the whole fractions thing where the imperial units become plain silly. They might work fine for e.g. longer distances, but goes completely bonkers when you need precise stuff. Especially when you need to do math with it.

        This is where the advantage of metric lies in for me, everything is base10. It’s easy to do math with things that were made with math in mind.

        1. The US system is often said to be better because it can easily divide into thirds, which is the one thing that metric decimal does not.

          Really? Then what’s 1/3″ expressed in 1/8ths, 1/16ths, 1/32nds or thous?

          None of them can do it, because they aren’t actually following the duodecimal system, but powers of two. The only things that are easily divided by three is a foot a yard and a mile. The rest of the system doesn’t follow the logic. Gallons are divided into quarts, pints, cups, again following the half of half logic rather than 12 parts, which makes it impossible to easily divide by three.

          So the argument about the ease of use because there’s more divisors for 12 than there are for 10 is only valid for long distances, and even there the preference is to stick to decimal representation of one unit, like 25,000 feet for altitude, because mixing the three makes it a total clusterfuck to use.

    5. Guess you made a math error, smart guy. 60 ft lbs is exactly 11,520 oz in, so conversion to Newton Meters should yield the same result for both cases.

      And why, exactly, should Brian re-state the designer’s claims in your preferred units?

      What a childish rant.

    1. Yeah I came here to post the same thing.

      BLDC servos are A Thing. They’re not as cheap as steppers, but they’re readily available and cheaper than designing your own unless you want like 1000 of them.

      1. A servomotor can be any kind of motor you like – commonly brushed DC, but brushless is increasingly common.

        The servomotor is just a motor – nothing special about the motor – combined with mechanical feedback from an encoder and an appropriate electronic driver.

        “This is not your standard servo”

        It really is – it’s closer to the “standard servo” of industrial mechatronics and CNC applications. Those tiny little things you use in RC toys are not “standard servos”.

        1. They are “standard servos” (Motor, sensor and control) but don’t have an _industry_standard interface. Although the interface is a widespread standard in the RC business.

    2. I keep flogging IQinetics .. we have basically eliminated.all of the steppers in our product and replaced with IQnerics closed loop bldcs. The ease of running just power and serial to each motor easily makes.up the difference im hardware cost vs. steppers. And the quiet, fluid motion adds a lot of style.

      Even fist-sized motors are shockingly powerful .. like, fractional horespower powerful (at least in short bursts).

        1. As far as pricing is concerned, I would recommend calling or writing. I am sure the price depends a bbit on how many you are apt to buy, where you are located etc. I think we are paying a tad over $100 per controller, and then add the motor cost. That might sound expensive, but when we started printing our own control boards for steppers and dicking around with all the wiring, we decided it was not bad at all. Now installation involves hooking up 12 volts, a serial line and four screws.

          They do have standard solutions (motor and controller, packaged) for:

          EMAX GB4008 and similar
          http://www.gearbest.com/radio-control/pp_153646.html

          Quanum MT4012 and similar (these motors are ridiculous .. we are replacing with the GB4008 so that we don’t injure anyone)
          https://hobbyking.com/en_us/quanum-mt-series-4012-340kv-brushless-multirotor-motor-built-by-dys.html?___store=en_us

          Exceed Rocket Series 6 and similar (we use these for small centrifuges)
          http://www.hobbypartz.com/86mb82-3010-1700kv.html

          And various IPower Gimbal motors (we use the GBM 2804H .. there are other models, and some are pretty inexpensive)
          http://www.iflight-rc.com/ipower-gbm2804h-100t-gimbal-motor.html

          And probably a few others .. and you can buy controllers “bare” and roll your own design .. you just need to figure out how fasten a suitable magnet to the shaft and hold the AS5047 sensor on the controller in a position where it will properly sense the shaft position.

        2. Just dug up an invoice and it looks like we are paying about $70 per controller .. not bad at all for a closed loop system that is this smooth and powerful, and where the angular position of.everything is known at all times.

  5. Well the French initiative had a totally awesome outcome – used by everyone in the world. Note that the French used (and to some extent still use) a similar inch, foot, yard, pound system ( or systems, it was a bit chaotic).
    The French for inch, foot is pouce, pied (thumb, foot) which explains the origins of the system. The metric system tidied up the muddle – the old systems continue the muddle. What is the volume of a pint? It depends on which country you are in.

  6. No offense to the non-metric people here that feel the need to defend their precious (but inherently illogical) imperial system, I understand your primal response to be protective of the system that you know and grew up with, but the imperial system table of length equivalent units is a complete mess ( https://en.wikipedia.org/wiki/Imperial_units#Units ) especially because it doesn’t even adhere to some kind of multiples of the units within it’s own system.

    The metric system uses multiples of 10. Considering the fact that your number system is base 10, it is the only one that makes sense.

    Please don’t be offended by this simple truth; we welcome you to join us. Your glorious army and brightest scientists already did.

      1. Stop spouting this. It’s absolutely not true. I’ve worked with, and for, NASA for years, and never once have I seen a single official document using the metric system. Literally not a single one.

    1. There is some truth to that. Personally I very much appreciate that when I am working with a project that is in metric and need to do some math.

      On the other hand it is natural to think in terms of dividing something into equal pieces. 1/2 vs 0.5, ok, not too bad. 1/4 vs 0.25, 1/8 vs 0.125. If you have 8 friends to split your cake with does your brain really go straight to 0.125? As you go on it gets less and less intuitive.

      Of course nothing stops one from counting decimal fractions of an inch. I even have a pair of calipers that does exactly that! Conversely I suppose nothing other than custom prevents one from talking about 1/32 of a centimeter.

      That said as an American shopping in american hardware stores who likes to build things I find on the internet I am constantly switching between the two systems. In fact when taking measurements I often finding myself using whichever system happens to be printed on the side of the tape measure that is most conveniently located for me to use at this moment! I really don’t care except for the fact that Metric screws are very expensive to by local and take forever to arrive in the mail. Metric drill bits have no local option, it’s wait for the slow boat or nothing at all. Perhaps if I had a lathe I could build my own?

      This idea that it’s the American people who are so enamored with a measurement system and offended at arguments for another one is laughable. I suppose there are those types whose level of nationalism is greater than their level of education but I suspect this is the wrong forum to reach that crowd.

      At the end of the day it’s the executives in charge of our factories that matter most and they will make the decision based on money. Switching over would be expensive and the monetary benefit would be dubious at best. Unless you poses a magic plan of how to overcome this simple fact there really isn’t much reason to discuss which system is better besides maybe academic challenge and boredom.

      1. >”If you have 8 friends to split your cake with does your brain really go straight to 0.125? ”

        Bad example, because it’s not a case of having to calculate anyhing. You just cut the cake twice in a cross, and not worry about whether someone got a millimeter bigger slice.

        For anything involving actual numbers, yes. It’s such a trivial thing.

    2. Anyone who’d “take offense” over the differing metrology standards should go back on their medications.

      I grew up in Australia where they switched from the British pound/sterling/pence days to base-10 currency back in the late 60’s, and later to metric for all measurements in the mid 70’s sometime. I recall being in school doing drafting (yo, back then, we use paper and pencils and erasers (we called them rubbers) to do our mechanical drawings.)

      So here I am, an engineer living in the US knowing and using both systems.

      I’m a prototyping engineer, so I do a lot of fabrication, build electronic assemblies and often work in code. I routinely convert between the two systems and occasionally work with pure SI. It’s no big deal. Some places I have cutting metal use the old feet/inch system, some places use metric. Since I stopped drawing with paper about 30 years ago, my CAD package today does all my conversions for me. I politely suggest modernizing your tooling so it’ll do either system.

      If you’re getting uptight about which system is “superior” you’re missing the point. Who cares what system you’re using? Use whatever’s comfortable and familiar, just get it right and be consistent. Go make crazy cool stuff and share it with the world. Stop wasting your time and ours nettling over narcissistic nothings.

      1. One thing that strikes me odd is the way newly metricated countries don’t seem to like the deci- and centi- multipliers. The decimeter is a bit of an odd one out everywhere, but otherwise all of them are used regularily.

        So in the UK, Australia, Canada, New Zeland… and in the US as well, they’ll happily go meter, centimeter, millimeter, but then liter … milliliter, and that’s it.

        Why? Especially when they then complain how difficult it is to cook in metric units because everything is hundreds of this and that, or for some reason you need to measure all things in grams. Well, you have perfectly good, useful units and you just pretend they don’t exist.

      2. +1 on the ease of using both systems. There appears to be a lot of whining on both metric and imperial purist camps.

        When I’m making a prototype (here in the US) it sometimes makes sense to use millimeters and sometimes makes more sense to use inches. Some projects I use both. For some reason that REALLY freaks out some people. It makes even more sense in PCB design where some components were designed using mm pitch and some were designed around the 0.1 in pitch or a division of that. In KiCad I switch units and the snap grid quite often – not a big deal.

        For documentation, as long as the units are consistent in the same drawing and are clearly labeled, that isn’t a problem either. It is really not a big deal once you have used both systems for a short time.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s