Author Topic: water-proofing a servo (Read 62673 times)

Militoy · « **Reply #30 on:** November 17, 2006, 05:39:18 PM »

Just about any "grease" (thixotropic lubricant) will dissolve in a lubricant of a lower viscosity. There are several silicone-based greases, mainly by Dow-Corning, that are somewhat less soluable in organic mineral oil based lubricants - but over a short period of time, they will all succomb to suspension in the lighter oils. The point is, when the mechanism is operated under oil, there will still be adaquate lubrication - with or without the grease.

Admin · « **Reply #31 on:** November 18, 2006, 08:45:24 PM »

Quote

Care to post some pictures or at least a more in-depth explanation of what you're doing?

Unfortunately I cannot, as this is an experimental research project I am working on for the Navy. It isnt confidential, but for patent and intellectual property reasons I have to be careful not to give out details.

But you are in luck, as a paper fully covering part 1 of the project will be published in April at ICRA '07. Ill post it on the site then, and finally end the mystery . . .

The general project involves building a robot fish using a special actuator. Part 1 covers the science, design, and study of the actuator. Part 2 involves building the vehicle with the actuator. We are finishing up part 1 now, just doing final verifications on our data. Part 2 has begun with our first vehicle prototype. The paper on part 2 might not be published till the end of 2007 . . .

JesseWelling · « **Reply #32 on:** November 18, 2006, 09:03:14 PM »

Are you by chance working for NavSea?
I had an interview with them just a while back but they didn't seem to interested in me $:-\$

Admin · « **Reply #33 on:** November 18, 2006, 10:02:14 PM »

Nope . . . Naval Research Laboratory . . .

polar bear6 · « **Reply #34 on:** November 20, 2006, 11:00:21 AM »

Quote from: Admin on November 18, 2006, 10:02:14 PM

Nope . . . Naval Research Laboratory . . .

omg the admin is a navy seal!

JonHylands · « **Reply #35 on:** November 27, 2006, 04:46:20 PM »

Quote from: Cognaut on November 27, 2006, 04:39:05 PM

I like the idea of using bladders and valves to control underwater movement.

I imagine a mechanism without any servos - just electromagnetic or electrostatic "bellows" that pump water through the system.

Most of the AUV gliders do something exactly like this. Some even use the temperature gradient between water layers in the ocean as a power source for the pump.

http://www.webbresearch.com/slocum.htm

- Jon

Admin · « **Reply #36 on:** November 27, 2006, 08:47:05 PM »

the problem with gliders is that they are slow, and have poor agility . . .

you may want to check this post
http://www.societyofrobots.com/robotforum/index.php?topic=278.0

Admin · « **Reply #37 on:** January 12, 2007, 06:52:09 PM »

So I learned just recently that a rubber boot is also often called a rubber bellow. When I google'd it, I found tons of places to buy them . . .

Strangely, hours after I figured that out, someone then emails me with a place to buy rubber boots:
https://buytraxxas.com/product_info.php?products_id=1832

oh well, too late now . . .

JonHylands · « **Reply #38 on:** January 22, 2007, 02:45:02 PM »

Admin,

Don't know if you've seen this site:

http://www.protacraft.com/

These guys are ROV hobbiests, who decided to build inexpensive parts and sell them.

- Jon

sotu · « **Reply #39 on:** April 04, 2007, 06:05:17 AM »

I've been thinking about building a submarine robot, but the waterproofing system is allways a challange. I'm looking forward if ure publishing a water proofing turtoriel..!:D

mhalavo · « **Reply #40 on:** April 04, 2007, 09:50:42 AM »

This might not work for Admin's application but anyone looking into building a seriously waterproof vehicle should concider using magnetic coupling.

This site explains it pretty well:
http://www.dextermag.com/Magnetic-Couplings.aspx?

Most available products are for much larger systems but one could make a simple smaller version on their own for say a propeller.

Here's a crude drawing of a simple setup....

TravelingPencil · « **Reply #41 on:** June 10, 2007, 12:59:18 PM »

What about just coating the PCB inside the servo with plasti-dip or better yet something that will cover the electronics with a very thin but waterproof layer of plasti-dip or similar substance without overheating the electronics.

P.S. srry about the late post

sotu · « **Reply #42 on:** June 10, 2007, 01:43:51 PM »

This doesn't really have anything to do with helping the admin but whatever.
16 servos

woulden't that cost you ALOT OF CASH$$$? I know servos is more expensive then motors (maybe it depends on the motor/servo) but in norway you can get servos in different strenghts up to 6 Voltage for around 8 dollars (for the 6 V) Personally i only use motors, i find it much easier to just take the bus downtown and buy some motors for my own money then ordering it on the internet and pay the servo and then a bit for the transport (They dont sell servos in Norway, where i live)
Any how, you guys seem smart can u please visit my new topic and help me out?:
http://www.societyofrobots.com/robotforum/index.php?topic=1272.0

paulstreats · « **Reply #43 on:** June 10, 2007, 05:38:55 PM »

hi all,

Though this doesnt have much to do with the original problem, i did make a waterproof discovery not long ago...

My electric toothbrush broke a few weeks ago and so i thought id deconstruct it for parts, so when i was doing this i realized that when the toothbrush plugs into the charger section there are no electric contacts from the charger base to the toothbrush itself (obviously for safety around water) and was pleasantly surprised to find parts that are easily available which utilizes the same technology as a usual transformer but where the electromagnetic energy actually passes through plastic from one coil (iron cored from the base) to another coil (in the toothbrush, and also the base coil goes over the core).

I know this isnt related but it demonstrates how you can completely isolate 1 part of circuitry from another, or also use recharging facilities without posing any dangers. Albeit maybe having to do some convertions of ac/dc to use this type of system effectively.

Also mhalavo's idea seems quite a good idea, but the magnets should be powerful enough not to become unstuck from each other, not interfere with other electronics, but the magnetism area as a whole is, if researched probably the safest way to proceed for a watery environment.

Admin · « **Reply #44 on:** June 11, 2007, 06:51:59 AM »

Quote

What about just coating the PCB inside the servo with plasti-dip or better yet something that will cover the electronics with a very thin but waterproof layer of plasti-dip or similar substance without overheating the electronics.

cause I also need to protect the motor

As an update, Ive noticed that my oil filled servos are very slowly leaking oil. Im honestly not sure how much oil is left in the servos, but they still work . . . I guess its still not a perfect solution . . . I recently got access to a de-ionized water (does not short circuits or cause rusting) pipeline in the lab, so I might just cheat and not bother with waterproofing anymore . . .

TravelingPencil · « **Reply #45 on:** June 11, 2007, 12:43:34 PM »

Quote

cause I also need to protect the motor

couldn't you cover the holes in the motor with tape to keep the plasti dip out of the motor then plasti dip the motor case(not the shaft or bearing/bushing) with a thin layer of plasti dip then water cool the servo.

Admin · « **Reply #46 on:** June 11, 2007, 01:20:21 PM »

hmmmmm interesting idea!

it may be a bit hard tho to do because the servos must be reassembled inside the dip . . . ill buy some more servos and try that out . . .

maverick monk · « **Reply #47 on:** June 11, 2007, 01:41:41 PM »

maby you could put silicon sealent around all the leak pouints in the servo case, and replace the bearings with rubber sealed bearings, then it should stay watterproof, or, do what I do, put a rubber balloon around it, zip tie it off at the shaft and put some hot glue around the end.

Nyx · « **Reply #48 on:** June 13, 2007, 08:01:25 PM »

If you put the servo in a box with a rubber "nipple", and have a little pump to actually push air into the box... You could keep (too much) water from getting in.

Admin · « **Reply #49 on:** June 14, 2007, 04:55:52 AM »

This is the mechanical design of what I am doing:
http://www.societyofrobots.com/robottheory/Design_of_a_Biomimetic_Controlled-Curvature_Robotic_Pectoral_Fin.pdf

(on the left of page 5 is a picture of the servo setup)

The problem is that the push rods prevent me from completely sealing off anything. One of my most recent ideas is to cover the entire fin in a glove, but there are issues with that too . . .

rgcustodio · « **Reply #50 on:** June 14, 2007, 11:38:44 AM »

Just trying to feed in ideas.

Why not remove the electronics and use a waterproof motor or stepper instead? The electronics can be put somewhere safe, or if you're using a stepper all you need is a stepper controller or enough pins from your MCU and long (sufficiently water proofed) cables/wires.

I think this site is worth looking at: http://www.empiremagnetics.com/tech_info_indx.htm They have waterproof motors and steppers currently designed for the food industry, etc. Well, they might have something small, as per Admin's size requirements.

(Sorry for not reading the whole thread)

Admin · « **Reply #51 on:** June 14, 2007, 01:45:13 PM »

OK ok ok . . . so a lot of you are responding to this post lately without reading everything already discussed . . .

The entire device is really tiny - 4 motors packed into half the size of a fist. It must have high positioning control. And because I will be doing this to 20 motors (not one or two), I will need something easy to do, affordable, and highly reliable.

I think that summarized everything . . .

Eco19R · « **Reply #52 on:** June 14, 2007, 10:41:59 PM »

why not feed the output shafts through a sealed ball bearing, - those should be water proof as long as there filled with grease right?

bobbylox · « **Reply #53 on:** June 15, 2007, 12:02:21 PM »

I want to try the non-conductive oil version- where did you buy the o-rings from?

Admin · « **Reply #54 on:** October 08, 2007, 10:54:01 AM »

Ive developed a new method to waterproof these darn servos. Its the best method Ive come up with so far . . .

To see pics and a quick description, scroll down 2/3rds of the way:
http://www.societyofrobots.com/actuators_waterproof_servo.shtml

Or just watch the video

[youtube=425,350]UZt7BuiNtc0[/youtube]

aqua_scummm · « **Reply #55 on:** November 27, 2007, 07:36:17 PM »

Hello! First post!

A note on your previous method: Mineral oil is commonly used to degrease things like bike parts, or your hands after working on you bike

. You dissolved the stock lubrication when you added mineral oil t your servo... oops!

I am working on my first robot (freeduino based mini sumo), but used to do quite a bit of RCing, and I have a method that I know works well. It's mildly more complex than your mineral oil, works on the same principles... And one solution was even easier to find than mineral oil

Read on...

I used to fill my servos with various greases. The right grease will:
*Prevent water from leaking in small cracks, such as servo case splits and output shaft holes- even if the shaft is constantly rotating
*Insulate electronic parts (electronically- dunno about thermally)
*Lubricate moving parts

One of the most common ones I used was... Vaseline, or generically, Petroleum Jelly. Found in most homes, and just about any convenience store, drug store, grocery store, etc.

I also used dielectric grease (ask for it at your local auto store), and there's plenty of stuff McMasterCarr has (if you make stuff, you better know McMasterCarr!). While not as instantly gratifying as going to the local shop, MMC does deliver incredibly quickly. You can access their online catalog at http://mcmaster.com . Grease can be searched for on the left, it starts about page 2085 in the catalog.

If you are using non vaseline, look for something hydrophobic. Hydrophilic will attract water. Next look for something water insoluble. Water soluble means once water hits it, it will start dissolving away. Look for something thick. Mineral oil was on the right track, just too viscous. Vaseline lubricates, but it also STICKS. Also you of course want something dielectric, or insulating, so your waterproofing doesn't do what you're trying to prevent water from doing.

You can thin down stuff that is too thick t easily apply, just remember it will re-congeal to the original thickness, so something like parrafin wax (candlewax) would NOT work, or at least well. The best way to do this is scoop some in a glass cup or jar (be willing to dedicate it to this purpose). Then put that in a pot of boiling water. Heating the oil directly (such as on a stove) can cause a fire, so don't do that. Microwaving will probably work, too, but I can't recommend it as I haven't tried it.

You don't need to fill the whole servo. I usually completely cover the circuitry, any vents in the motor, and lather it inside where the output shaft leaves the casing. Don't put it inside the motor, the thickness will slow it down.

Keep in mind the grease will get runnier with heat when looking at options, high use situations will require more re-applying.

I have used this method down to about 2.5 feet of water, and I am willing to bet if properly lathered on, it can go to about 6-8 feet underwater, possibly more (with increasing risk of failure). I usually had to re-apply vaseline to the output shaft section every 2-3 mud races with my RC car (where it would get completely submerged in mud)

It's not the best high use long term option, but it more than worked in my tests.

If you're going to MMC to check out what they have, the dielectric greases will be more than enough resistant for our low voltages. Marine grease will also work. In the about grease section, you can find lots of info about the types/chemicals, thicknesses, etc.

The last tip I have is to get some silica gel (you get it in your shoes when you buy them, some jackets, some food packages). It's the little white packets that say do not eat. They absorb moisture from the enviroment. I'd crack open a packet, take a few beads, and glue them inside the case well away from the gearbox. They'll attract and absorb any moisture that does make it into the case.

let me know if you have any improvements!
matt pace
http://matthewpace.blogspot.com

cooldog · « **Reply #56 on:** November 29, 2007, 09:26:40 PM »

where did you get the o-rings

richard mackie · « **Reply #57 on:** December 02, 2007, 06:52:28 PM »

In the video he said they were from mcmaster but I used to work at Ace Hardware and we sold them individually.

cooldog · « **Reply #58 on:** December 02, 2007, 06:55:11 PM »

thanks what size is best

richard mackie · « **Reply #59 on:** December 02, 2007, 07:23:37 PM »

I don't know, I would just bring it in store with you

News:

Author Topic: water-proofing a servo (Read 62673 times)

paulstreats