Thread Tools
Sep 16, 2017, 09:28 AM
Registered User

External Pot "servo" turret drive

Been busy on this stuff lately, I hooked up an external pot servo turret drive, see video. It works okay, I would say, but to be honest I'd much rather deal with the wiring of a stepper instead of the somewhat complex gearing/mechanical connections for this set up. Again it's only roughly thrown together but it demonstrates the external pot servo.

Note that the external pot in this case turns roughly 8 full revolutions, 8 1/2 really, it's a 10k, I assume linear. It's important your pot is linear rather than some non-linear type. It's really a normal pot but the shaft connection to the wiper is like a planetary gear, three ball bearings. It's also only really works at all very well if you use nearly the full range of the pot, I did try the turret to the motor 1:1, but for one rev of the turret it used only a 1/6 or 1/8 of the range of the pot, no good. I had to gear the turret 6:1 to get it to work at all reasonable.

I'm not using any of the servo guts, the wires coming out of the servo case are connected directly to the motor, if you are skilled enough you can snip the wires to the pot in the servo, route them out of the case and connect them to your multi turn pot and then you can use the servo directly with it's three wires like a normal servo, but it's not quite that simple, your pot K's has to match the pot in the servo, and I've not done this but I imagine it might need to be calibrated. If done right you probably get better motion control and again can plug directly to radio reciever channel. This demo isn't really a servo because it doesn't operate on the standard PMW used for a servo, but within the arduino it could fairly easily be made to.

I use an H bridge drive, the little breakout with the red light, very simple, uno pin 9 is pmw for motor speed (which I don't bother varying in this demo), pin 8 is the direction, HIGH is CW, LOW is CCW.

I only used the control pot to make direction input simpler, rather than serial from keyboard or hooking up the compass. So I think the turret motion can be better if controlled more precisely, my control pot is stiff to turn and has no knob so is not easy to operate precisely.

So using the arduino servo example "KNOB" , I just hacked it from there. Important to note, I deleted servo library and servo commands, not used. Mainly I used the "map" function for both the multi turn pot and the control pot. Is all it does is map the pot position to a chosen range, like 0-360. And it goes on from there. If you are interested in the code, no problem I can post it, it's pretty simple. Cap

externalpotservo (2 min 18 sec)
Last edited by capricorn; Sep 16, 2017 at 09:36 AM.
Sign up now
to remove ads between posts
Sep 16, 2017, 04:32 PM
Registered User
Second video shows improvements over first version. It takes quite a lot of monkeying around to get something repaired once it starts off on the wrong foot, almost always a fresh start is most productive. This one starts with a (dual) H-bridge example and hacks in the pots, extermely simplified, therefore, much better overall. Contains only one if, elseif, else control, nothing more.

However well it works I have to say that if one were to use this type of mechanism for just a standard battleship main guns you have to replicate this gearing mechanism no less than three times (I guess you could skip the director and make that two systems ) Again, my obvious preference is steppers and all the wire, eventhough I really like the mechanical aspects of the construction. I concede that it works well though and sort of tracks the exact position of the turret.

In the demo I use a rubber band to connect the motor to the locating pot, I would not recomend that, the band stretches. On the other hand it might not work but for that buffer... who knows?

I sense there are some people out there that view this with some sort of distance, like it's just too complicated or not within their capabilities but that is just not the case. If you have the know how to build scale models, a somewhat logical way of thinking, you too can utilize an arduino, they are cheap, extremely well designed, and have near endless applications. Just buy one, start with blink and "hello world", simple.

Refined external pot servo, all done via specific, logical, yes or no questions/answers:

refined externalpotservo (2 min 37 sec)
Sep 17, 2017, 04:42 AM
Registered User
Separating the functions out opens the door to lots of possibilities, but something like this - - with the right pulley gearing and with its limits set by the arduino might do the job of moving turrets the wanted number of degrees with even simple coding.
Mechanically, to prevent the line on the pulley slipping, it might help to borrow a trick used on old radio dial drives. More than just the one turn round the pulley, at the mid point, pass the line through a couple of holes. Same idea that yotties use on rigging bowsies.
Sep 17, 2017, 09:11 AM
Registered User
!!! mfr02 that's a much better option, I looked all over for servos like that never found one (cheap enough anyway), thanks. I assume when it says 4-4.5T that means it does 4 full revolutions for the range of the servo input. It's very inexpensive too. Beats the external gear box at servo city, $100+. I recall the old dial tuners with the "thread" drive, they work great, the tread on my demo is wrapped twice around the small pulley, it works fine, the rubber band needs to be replaced with a thread like that too and it would work even better. It is just an experiment, sort of test of the multi turn pot I had, I'm still pretty much happier with the steppers in any case. Cap
Sep 17, 2017, 10:57 AM
Registered User
The turns figure is based on an "average" transmitter, some will have a wider sweep than others.
One thing that the winch sellers never state is the drum diameter. What it would cost them to say "1" or 25mm, I don't know, but it does let you know how many degrees of turn you need for a given amount of linear travel. The winches come with a range of output turns, I have seen 1.5, 2, 4 and 6 turn versions advertised in various places. I am fairly sure that the difference is the gearing between the output shaft and the pot spindle.
I have only seen them in "standard servo" size, no miniatures. It might be that for a more compact model, a continuous rotation micro servo (which I have seen mentioned by others) with a multi turn pot under the turret, might be more convenient. Unless appropriately small steppers are around? If the donkey work is being done by the code in the arduino, it doesn't really matter what converts information into motion, does it?
Sep 17, 2017, 07:22 PM
Registered User
mfr02, you are familiar with the 28BYJ48 stepper? Those are relatively small, smaller than a standard servo, if a different shape. I have some tiny bipolar steppers from cameras etc but the steps per revolution are pretty low. That's why the BYJ is so useful the steps per revolution due to gearing is over 2000, and they're practically free.

I find the info provided for servos is always just a tad lacking, not sure why, afraid they'll be one upped by the competition I suppose. They provide plenty of useless info to me, not what I'd like to know. Same vagueness goes for the steppers though, nowhere (I found) is it clarifiied what the actual steps per rev are. For the BYJ I found it's in the range of 4075.8, not the convenient 4096. Someone cut one open and counted the gear teeth, a bunch of 13 and 43 tooth gears etc, no wonder it's not an even spr.
Sep 18, 2017, 04:10 AM
Registered User
I've never had any real experience of using steppers in projects, they only appeared commonly after I had stopped having easy access to bits n pieces. Nowadays I spend more time playing with my toys, but it is fascinating to see what developments are going on, and what is being achieved with the new stuff.
A group that I was involved with in my working days got Christened "The round wheel club" also known as "The boneless sausage club" on account of the number of times that we re-invented things. Hopefully my harping back to the elder times will help others avoid reinventing things, or spot blind alleys before they head too far up them.
Sep 18, 2017, 11:48 AM
Big Boats Rule!
boater_dave's Avatar
Great work! That's what I figured with regards to the 10 turn pot. To make it make work, you need to use almost the entire span of turns, otherwise the resolution isn't there. At my last job, in the test area, we used a 'string pot' to monitor the position of large hydraulic cylinders. It was essentially a 10 turn pot, with a spring loaded spool of fishing line, that behaved like a tape measure. Tied (literally) to your linear cylinder you could use it's output to add safety limits to your test setup. Worked well. So maybe one solution would be to mate a drum of a certain size to use most of the motion?
And yes, the Servo City units swap an external pot for the one inside the servo. The pot inside is electrically abandoned, but remains as the shaft for the last servo gear. I have several of those in my big sailboats. Good price for huge power, like 1700 oz-in. Even did the soldering on one, and it's not an easy swap as the control boards on modern servos are wave soldered during assembly, I'm sure. Not easy to carefully unsolder three contacts at once in close proximity to all the other soldered connections.
I also have several of the 3-4 turn drum type servo winches. They are fast, but the resolution is less than a good servo, but they fit that price vs perf niche for sailboats in the 50" range with almost 200 oz-in of torque. And the included drum in about 1-3/8" with a standard servo spline. I don't know how the pot is hooked up in there, as I've never had one apart. The gold standard for drum sailboat servos is a very expensive unit made in Australia (I think). They also market tapered drums (spiral track) to take advantage of the line force need vs speed that sailboats find useful.

Sep 18, 2017, 04:59 PM
Registered User
Mfr02, You make it sound like you're over the hill, you can still get bits and pieces, on the internet, I like the surplus electronics sites they have pretty cheap stuff and usually they also have arduinos, jumpers, sensors etc so you can get most of what you need at one stop.

"The round wheel club" also known as "The boneless sausage club"
, Ha! that's where I belong. I reinvent stuff constantly, probably for the worse, example... the external pot servo. It was a fun, interesting experiment though, had to know how well it would work.

BDave, sounds like you have much experience with these things. With lasers and all these days I would say the 10 turn pot tape measure is probably a thing of the past but I could see it working pretty well. I was suprised once I got the kinks out how well the multi turn pot works, I figured it was low quality, but they've been mfg pots by the millions for years I guess it's natural that they have a pretty good handle on it now. It's the tiny pots in cheap mini servos that I think are somewhat poor for this use.

I'm not familiar with sail boat winch servos, and don't have any, so don't know how well they'd work for turrets, I can see them not needing very good tracking since they are just winching in sheets. I may buy a few of the 4T's mfr linked me to, just so I have some For the final time I'll say I still prefer the little steppers PS: I personally have operated winches (the hand crank type) on 48' yacht sailing from SF to Santa Barbara then to Hawaii, I was 17, it was great, can't shake the love of boats after that. Cap
Sep 19, 2017, 11:05 AM
Big Boats Rule!
boater_dave's Avatar
Cap, it's the sailboat guys demanding super resolution, power and speed! The common multi turn servo winch is used for some, but the race guys are hard core. Dropping $300 on a winch is no problem, compared to the $2-3K for the entire boat. I get it, if you seriously campaign one or two boats this is how the game is played. I have about 15 sailboats, all for fun, none of them $2-3K, so that kind of cash is set aside for other things, like food and mortgage payments.
I work with some industrial automation, and have had to program, and troubleshoot repairs, on some weird stuff. Like figuring out the air logic of a home grown design that's been working longer than the guy who built it. Or telling the proud owner that his $150K robot servicing a plastic molding machine is capable of more than picking the scrap sprue and carrying it to the waste bucket.
And sounds like you had a great experience sailing! I love it, and have been sailing on Lake Michigan most of my life. My parents had boats all my life. Raced with a friend for a few years. Even worked at Harken Yacht Fittings for 8 years. All of the new tech is great stuff, and there is a generation of kids (granted, not all of them) that are just taking this stuff and applying it to everything. Like us old guys would use a screwdriver. At last years Maker Faire, I had a discussion that revolved around using a GPS device to use as an input in an autonomous boat. We were talking about the problems of how often it calculates position, and how they crop off the course if there was a skipped point. All the little details. He was 15. Didn't have his drivers license yet.

Sep 19, 2017, 04:30 PM
Registered User
A quick calculation says that a 4 turn winch strung to a 4.8 inch pulley under the turret will give 300 degrees of travel. The definition will probably be most affected by whatever is doing the driving, what with the sensor being an analogue pot offering a slope rather than a digital device working in step increments. A 2 turn on a smaller pulley would probably be the same travel, but faster. They have a shrouded double drum making them ideal for push-pull - the shroud is optional. Fairly optional.
From what I have seen (having abused a few and let the smoke out), within the standard servo body, they have all the usual stuff, but there is an extra gearbox between the pot and output shaft. Probably just different gear ratios for different turn winches. This gearbox moves the pot lower, which means that there is less room for the servo control board. In turn, this is shorter, with less space. I do believe that this means smaller components, which, although probably a lot better than what went before, are a lot nearer the limit as regard power handling. I was abusing mine pulling 400 square inches of sail - a model turret should be as nothing.

Dave - about the upcoming generation taking to new generation tools and methods - you do realize that if we had contemplated having this conversation by this means a few hundred years ago, we would have been burned at the stake? I just hope that the next generation has the opportunity to know the pleasure of playing with boats like we have known.
Sep 19, 2017, 04:58 PM
Registered User
Oh, so the winch servos are high quality, I guess the 300$ ones are at least. I know a couple guys that spend more on slip fees and storage for their boats in a year than I spend on all my hobbies, so I don't fret much when I buy stuff, like $260 for 1/200 Bismarck

We have a very talented young guy at work now, was an intern now on full time, very interested in vibration analysis (we are structural engineers). We (he) managed to cobble together a pretty good vibration monitoring device using rasp pi and accelerometer, our main aim is to get the structures natural frequency from the data, which we can. You can buy these for this purpose but they are many thousands compared to our $100 (+time). It's not our companies area of expertise but we may go into it more.

I built a wood model of the boat I sailed on (years ago, probably 1990 or so), close, not exact, no photo's but the inside is all built out too. The real one Phobus, is a great boat, built by two germans in Santa Barbara, has a concrete keel. It's a ketch/motor sailer, has been around the world once. Not sure of its fate, last I heard it's fir deck rotted in the tropics and was replaced with teak.

The round the world trip was by the previous owner (prior to my Aunt and Uncle): Ugo Conti, that prompted me to google him. I assume this is the same fellow (I could be mistaken on the name but am pretty sure):
Last edited by capricorn; Sep 20, 2017 at 04:48 PM.
Sep 20, 2017, 04:15 PM
Submarines, etc.
tsenecal's Avatar
capricorn, your web link is missing a closing parenthesis...
Sep 20, 2017, 04:53 PM
Registered User
Okay tsenecal, I must have missed it, I fixed it. The "spider" seems a bit iffy to me, inflatable hulls and shock absorbers? I guess I'd give a try too, if I could get the financing. Cap
Sep 26, 2017, 10:55 PM
Registered User
Just some helpful info:

The resistance or K of the external pot doesn't have to match. The servo uses it as a voltage divider, so the difference in resistance will simple increase or reduce the angle the servo moves vs the pot.

A few servos on sites like banggood are 120*, 180* or 360* servos from the factory. (ex: JX Servo PDI-HV5932MG 30KG Large Torque 360, $25)

Another option besides sail winch servos is an external multi-turn pot that's driven by a spur-pinion gear arrangement connected under the winch horn.

Quick Reply

Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
Discussion Using an external pot with servo Croga Power Systems 8 Sep 25, 2016 01:26 PM
Mini-HowTo External pot mod flypilot Multirotor Drone Electronics 13 Aug 02, 2012 11:24 AM
Discussion adding external POT dino_squirrel DIY Electronics 1 Jun 16, 2012 03:41 AM
Discussion High-End Gimbals and external Pots DroneBuilder Multirotor Drone Talk 1 Apr 04, 2012 02:19 PM
Question Connecting External Pots Via Trainer Port? Andology FPV Talk 17 Feb 06, 2010 07:00 PM