servo "sqeaky" during centering modification for full rotation

[ozbot]

So I am building my first robot, along the lines of the $50 tutorial.  Have most of the chassis complete, all the electronics complete and installed. 
Successfully programmed the ATMega8 with the .hex file for servo centering during modification.   

(note: I used an MG995.  Id has a metal gearing system, with a metal stop pin.  Had to dremel it off.  Not that easy to modify, FWIW). 

So all seems to work, I turn the pot and the wheels go one way or the other at varying speeds, although it seems to peak out on speed rather early in the pot turn. 

Problem is that at the center point, the servo kind of sqeals, sounding like it cant decide which way to turn, or has just enough power to try but cant.  I am assuming that the servo should be silent on the center point, so, a:) is this abnormal and b:) if so, how to remedy.

[Admin]

Problem is that at the center point, the servo kind of sqeals, sounding like it cant decide which way to turn, or has just enough power to try but cant.  I am assuming that the servo should be silent on the center point, so, a:) is this abnormal and b:) if so, how to remedy.

Well, it shouldn't be moving or making noise at all at the center point . . . are you using the HS-311 or another servo type? Does it make the noise if it's slightly off the center point and moving just a little?

[Gertlex]

Can you stop the noise/vibration by turning the output a slight bit in either direction? (i.e. lining it up to where it's trying to move... and then it ought to fix itself.

I noticed with my metal gear servo that when turning it by hand (with a servo arm on) that it continues to rotate a bit when I let go, indicating that the overall system's friction is very low (my HS311s/422s aren't nearly so smooth in this respect).  Does yours exhibit this behaviour?  If not, it might be that small movements (i.e. that last bit to centered position) have to overcome too much initial friction.

I presume you didn't get rid of the grease on the gears, right?

[ozbot]

admin:

am using the towerpro MG995.  I tried to center it 5 minutes ago, and got one to silent. The slightes bump started it off again. Found this post at www.rcuniverse.com :

I wrote a review of these a while back.

Plus side:
* all metal gears -- pertty tough
* cheap (and I mean "cheap"
* seem to have good torque and are moderately fast

Negative side:
* very poor match between electronics and mechanics produces bad overshoot
* feedback pots simply don't last -- resulting in lots of jittering around neutral
* gear-mesh, centering and stability highly variable from one unit to another

In summary, if you get a good one they're okay and will work (for a while). Chances are you won't get a good one though and you'll notice that they rock back and forth around neutral, overshoot past the commanded position and generally have poor resolution.

I bought 14 of these servos and here are the stats so far:

1 was pretty good and is still in service after about 6 months, although the feedback pot is now so worn that I'll be removing it because it shakes like it has the DTs.

1 failed within 5 minutes of installation -- and almost brought a plane down with it -- the output arm just gyrated wildly for about 30 seconds -- then it came right. Hasn't faulted again but I'm not about to risk it.

4 have not been accurate or stable enough to merit using on any aircraft I value

2 were stripped and I used the servo amps to replace those in a pair of Hitec HS425BB standard servos. This makes a *very* nice little digital servo that is fast, accurate and has great holding power. The difference between these modified HS425s and the original MG995s from which the amps were pilfered shows just how bad the MG995's pot and geartrain really are.

The rest are sitting under my bench. Maybe I'll build an RC doorstop one day and they'll be just perfect for that :-)

Bottom-line: They're okay but they're really cheap and risky.

/quote

So I might just have an edsel........

Gertlex:

mabye friction is a problem, close to "zero", while there is still the noise, it will sometimes increment in ~1/2 second intervals.  I did not intentionally remove the grease, but did have to dremel away the stop pins, probably lost a bit (or lots) of grease, as I haad to totally disassemble the gear assembly.  Ok, I'm off to find some servo lube.........

cheers, ozbot.

[ozbot]

So greased up my gears & while doing so, my pots fell of.......so I put a 2K2 voltage divider in place of, and it works better.  There is still a slight ticking sound in each servo, no movement.  Now I find that my LDR's are not very paired...  Apart from that, My first bot is finished.