Robotics Project: Snuggles, The Wolf

[amberwolf]

Thanks to my brother's amazon prime i was able to get six of these

https://www.amazon.com/dp/B0DD4GPYRW

for about a dollar each (they were actually only a penny, but I couldn't find $25 worth of useful stuff to add to the order to get free shipping, so that cost almost six bucks). They appear to go for around $40 each even on aliexpress / ebay (they're almost $60 on the above page now that the sale is over), so I snapped up all of the ones it would let me buy, at that low a price. I'm going to break or burn out some of the servos I have and/or get during experiments and development, so it won't hurt to have spares, especially at this kind of price. Hell, even if they are all broken except this one, it's still a bargain. :lol: But hopefully they'll all work. (at the sale price they were no-returns, so I don't expect them all to work).

For whatever wierd reason each of them is shipping from a different place around the country, none anywhere near here, but the first arrived today. The package label says it's 55kg, but the other info on it matches the model info on the ad for the 35kg, and that's what's actually in the package (or so it is marked on the casing, anyway).

It works as expected; it is a 270 degree brushless servo; the greater rotational freedom means it runs slower than the same thing as a 180 degree version; don't know how it compares to the 180s I already have. The page claims it is 180, but it is definitely 270.

it's strong; didn't try comparing it (by feel, as I have no testing equipment to verify torque) to the others I have that are "25kg" units (vs this one's 35kg), but I could actually stop this one from rotating. It's voltage range is significantly different from those, however, and I'm usiing the same 4-C-cell power source for the servo tester. This one's range is 5v to 8.4v, while the others are 4.8v to 6.0v. I didn't measure the volltage under load or no load for this one, to compare with the previous tests, too tired to fumble around with the meter.


But it is significantly quieter than the brushed versions I already have. The metal gears are still noisy, but that might be partially mitigated by packing the whole gearbox in grease. (haven't opened it yet to see what grease it's already got in there). Plastic gears might also be quieter, but I need the metal ones for strength. If I could get all the servos for this project to be this quiet, it would help a lot with the noise....still more than I want by a lot, but much better.

It is actually a metal casing, and has a metal arm as well as the typical plastic ones with it. Claims IP67 waterproofing, though that doesn't matter for my purposes.

Specs and pics off the sale page below. Didn't take any pics of my own yet.

Waterproof Brushless servo Full Metal Gears High Torque Fast Digital Servo for 1/8 1/10 1/12 RC Car/Shockproof Rc Car Servo (Full Metal Shell 35kg-180°)

• Fast / Strong / Unbelievably Precise maximum torque of 45 kg-cm (624.93 oz-in) (8.4 V) and speeds up to 0.09 sec/60° (8.4 V)
• All-metal Servos /all-aluminum housings/Metal Gears / Full Ball Bearings / Shockproof
• Waterproof IP67 / Digital / Brushless
• High Voltage 4.8V - 8.4V high voltage steering servo, controllable angle range 0-180°
• RC Servo Motor is 40*20*38mm (1.57*0.79*1.49in),

• Product Description​

  
  Electrical Specification
  Operating voltage range: 4.8-8.4 Volts
  Operating speed at no load ( 4.8V ): 0.14 sec/60°
  Operating speed at no load ( 7.4V ): 0.10 sec/60°
  Operating speed at no load ( 8.4V ): 0.09 sec/60°
  Stall torque at locked ( 4.8V ): 27kg.cm(374.96oz-in)
  Stall torque at locked ( 7.4V ): 39kg.cm(541.61oz-in)
  Stall torque at locked ( 8.4V ): 45kg.cm(624.93oz-in)
  Stall current at locked (4.8V ): 3.2A
  Stall current at locked ( 7.4V ): 5.2A
  Stall current at locked ( 8.4V ): 6.2A
  Idle current at stopped (4.8/7.4/8.4V): 5mA
  Control Specification:
  Control System: PWM (Pulse width modification)
  Pulse width range: 500-2500μsec
  Neutral position: 1500μsec
  Running degree: 180°/270° (when 500-2500μsec)
  Dead band width: 2μsec
  Operating frequency: 50-330Hz
  Rotating direction: Counterclockwise (When 1500-2500μsec)
  Product Specification:
  Suitable for: for 1/8 1/10 1/12 Scale RC Cars, Trucks
  Limit angle: 360 degree servo
  Bearing: 2BB
  Horn gear spline: 25T Diameter:5.9mm
  Case : Engineering plastics+Aluminum alloy
  Motor: Coreless brush motor
  Spline:25 Tooth
  Gear ratio: 342
  Bearing：Double Bearing
  Waterproof Performance：IP67
  Connector Wire: 12.6in (320mm)
  Size: 1.57*0.79*1.59inch（40 * 20 * 38.5mm）
  Weight: 2.11oz (60g)
  CE /RoHs: Yes
  Apply Environmental Conditon:
  Storage Temperature Range:-30°C~80°C
  Operating Temperature Range:-25°C~70°C
  Packing List:
  1 x BLS3335kg Servo 180°(270°)
  1 x 25T Adjustable Metal Servo Arm
  1 x Accessories Set (As Shown on The Picture)

 

[amberwolf]

My pics, i couldn't get the top off where the gears are by hand, didn't want to pry at it so left it closed. The bottom came off easily enough, but the seal was just laying inside, crumpled up, so this particulare one wouldn't be waterproof. the bolts do have o rings to seal them and the holes they go thru.

you can see the waffle like thermal pad that sits on the motor driver bridge chips to let them transfer heat to the casing.

the driver board


the package it came sealed in, that indicates it's a 55kg servo (but is actually only a 35, which is what i'd ordered, but the wrong label means those that did order a 55 probably didn't get what they ordered unless the barcode is different on the actual 55's).

 

[and whatnot]

What do you recommend using for the tubing that each individual cable must be separately contained in to pass the many feet from the external box thru the flexible "umbilical" up into the wolfy insides to all the various parts being cable-controlled?

My first choice would be tiny bearings with pressed on grooved disc's. But that's probably hard to realise in this project. Second, tiny metal or plastic tubes. Or a pre 3d Printed cable guide. Preferably I would only use one cable per motion (one direction) and a counter acting spring (other direction). Everything pulling only, I guess pushing ist much harder to accomplish.

 

[amberwolf]

i don't know any way to not use completely enclosed cables to make a safe flexible umbilical for exterior-housed motors running the stuff inside the bot. if you know a way to do that, i'd be interested in seeing it.

simple cable guides, eyelets, guide wheels/bearings, those all work fine if the motorsa re inside the wolfybot...but they wont' work if the motors are outside, not without a mucih more complex umbilical that would rpobably be larger and heavier than a simple bundle of bicycle cables.

(btw, they make teeny tiny (and very large, and all sizes between) guide wheels with bearings, pretty cheaply..., so no need to make them)



just a metal tube would bind or pinch when bent, it wouldnt' flex. just a plastic tube would be cut thru by the cable moving back and forth (or cut the cable itself dpending on the materials chosen), or it would collapse under tension or deform during bending, and crimp.

so whatever is used has to do the job that a bicycle type bowden cable does, with it's coiled-spring outside to keep the bending smooth, and plastic inner liner to prevent binding of cable to that. they don't have to be actual bicycle cables...but they have to do the same job...and that's what i'm looking for, is alternatives that do that complete job.


anything that uses springs requires a place inside for the spring, and each spring has to be made just for that specific part, for that speicfic force, and the motor then has to coutner that force, so it has to be bigger to do taht, cost more, use moe power. eveyr joint would rquire a spring to pull in each direction it could move, to keep it in some default position...any non-simple-hinge joint would need at least four springs....this gets very complex to figure out, and the math and manufacturing needed to make all those springs is beyond me.

that's why i am going for the two-cable-per-motor; it's much easier and actually takes less space.



i've been working this thing out for many years, and have gone thru a lot of potential solutions and figured out why they either wont' work or are less ideal than ohter solutions for whatever reason. im open to any suggestions, but you have to keep in mind all of what this must do and be, and it's limitations, and mine, for wehter something might work. it's pretty complex.... :/


there are very few things about that i'm willing to compromise on functionality / etc for, because if it doesn't do all the reequired things in the right way, it can't do what i need it to, and would have no reason to be built.

 

[and whatnot]

In the modelling/RC world there are really tiny Bowden and even push (hard)wire. Just Google it.

If that is still to big.... ,,

 

[amberwolf]

Thanks--I'd previously looked at some of those, and they didn't look like they could do the work I needed, but I haven't done an extensive look at all versions of them.