reliability of servos

mose

I'd like to use a Futaba servo controlled by a basic stamp to control a·damper on a fan outlet for closed loop temperature controlled cooling·of an electronic device.· I can't control the fan directly through PWM because I don't get a large enough range of air velocity with PWM control on a DC fan for the range of electronic devices I need to cool.· Would the servo be reliable·enough for a 24/7 operation?· There is·very little torque required on the servo to move the baffle (it's only a 60mm x 25mm fan).· This system·will eventually have·'hundreds' of servos, each with closed loop control,·so the servos·need to be pretty reliable.

Anybody seen/done something similar?

thnx.

Ken Gracey

Mose,

Oh yeah, we've done something similar with positive results. Back in 1997/98 we had a project in the works called the "Volume Chip Programmer". This machined-cast system was designed to accept rails (DIP and SSOP) of SX chips in the top. They'd be lowered, programmed and dropped into a tube attached to the bottom of the programmer. This gravity-fed system used servos as an actuator. During the design, the engineer (same one who designed the Boe-Bot) conducted 10,000 cycle tests with the Futaba S-148s. And the servo held precision and torque after these tests. We were surprised. I think the only consideration was a recommendation to add some grease after a while. Therefore, I think you could use a servo for this application. But even if you replaced a servo it's a functional solution considering the benefits of positional control and absence of gearing and additional mechanics which may accompany a stepper motor or linear actuator.

But to find out for sure, rig up a test and cycle the servo to prove it to yourself. It's a cheap experiment.

As for the volume chip programmer we were designing, it was becoming a black hole so we abandoned the project about 60% through the process. We now have some really neat aluminum machined castings left over from the project. Looking back, we should probably have completed the design considering the number of SX chips we're selling today.

Ken Gracey
Parallax, Inc.

mose

Thanks for the info. It sounds like the cycling is what wears out the servo with little 'wear' while the servo is maintaining a position. Although our different electronic devices generate different amounts of heat, each device's heat output is very stable over time. I envision the baffle being adjusted by a PID loop in the first 5 to 10 minutes when the electronic device is first started up, but once the device is at temperature there would be little servo movement to maintain temperature for the next 8 to 10 hours.

coyote

There are many different servos with different torque, speed, size and construction. If you want maximum durability you might consider ball bearing servos. The standard servo has bronze bushings that should last quite a while but tend to wear more than ball bearings and have less repeatability. It does not sound like you will have a large shock load so plastic gears should be fine and speed should not be an issue. You do need to see if the torque of the servo is within your expected load range.

Your damper design should try to minimize the force required to hold it in place. A constand load on the servo causes it to consume more power and to make more adjustments on its own.

mose

I picked up a couple Futaba S3004 servos to try out since they had ball bearings and are readily available. And I do think I can design the damper to require negligible holding torque.

thanks.

John R.

If it makes you feel better in terms of reliability, think about the "normal" use of these things.

They are commonly put in boats, cars and planes, often with a meaningful mass involved and traveling at speeds in excess of 20 MPH, and in some cases significantly higher.

You really don't want your several hundred dollar model going out of control and "crashing" and/or injuring someone because a $20 servo failed.

When applied appropriately, these things last a long time. You might argue that a car, boat or pane isn't used all that often, but many R/Cers will have and use a radio and its servos for their entire life.

If these things didn't hold up (when applied appropriately), they wouldn't last in the market place. Someone would have "built a better mousetrap".

Note that is also possible to ruin one of these things in a few cycles (or less) if applied in an inappropriate manner.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
John R.

8 + 8 = 10

rockin_rick

If they are going to be powered 24/7, will they last indefinately? Just how long (hours)?

Perhaps you could have the micro only power up the servo when an adjustment is needed, and then power down the servo once the damper is in the new position as the servo will sorta hold the last position as long as there the holding torque is low (IOW - as long as you don't exceed the initial 'stiction' to get it moving).

Rick

PJMonty

Mose,

I think you should check out this site:

diy-zoning.sourceforge.net/

It's all about using servos to control heater/air conditioner dampers in your house.

This page...

diy-zoning.sourceforge.net/Advanced/dampers.html

...has information specifically about using RC servos with dampers.

Thanks, PeterM