Can anyone recommend a good way to reduce noise from small plastic injection molded gears without significantly decreasing efficiency (or better yet, increasing efficiency - since less energy converted to noise)? - without re-designing the gears. The motor is a miniature pager type and the The gears are similar to the ones in the link.
I imagine most of the noise is from imprecise manufacturing tolerances, common in molded vs. machined parts, so unless you pick some that have a better fit, your choices are fairly limited to:
Carefully filing or shaping the gears to provide a better fit.
Adding oil or other lubricant (which can increase resistance and therefore may lower efficiency, not to mention more mess).
Noise is inherent in any gear. Metal gears may be "quieter" because the mass of the metal may tend to dampen the sound, but they may be no more efficient, all other things being equal.
As gears differ as to pressure angle, face width, etc. gears of the same pitch may not necessarily be good matches if you're concerned about maximum efficiency and other factors. For higher end applications pairing gears together is more involved.
My experience in working with hundreds of gearboxes over the years, is that nothing can be done to reduce your existing noise after you have set spaced them correctly. You might try carbon paper interface to see where the high spots are on each gear but that takes a lot of effort. You may be better off just going to belt drives which are really quiet. Take a look inside any cheap Chinese VCR rewinder for examples. You might also just get some surplus rewinders and redo your device with belt drives. They are quiet. And slip instead of break when overloaded.
I thought of belt drive, but it's really difficult at these dimensions. The tiny motor does not have ball-bearings (maybe sintered ones) and I believe any tension on the belt will result in significant friction on the motor shaft.
Well....I did confirm Gordon's suspicion that there were at least 3 components to the noise. One was once-per-rev (of the output gear), another was a buzzing noise at certain speeds and lastly the teeth noise. As advised, I examined the gears under a microscope and used an x-acto knife to remove what seemed like molding defects in the pits of some teeth (more on one side of the gear than the other). This cured the once-per-rev noise. I added some lithium grease to the plastic bearings on the output shaft and this cured the buzzing noise (resonance - shaft/bearing assembly..?). Now, the predominant noise is the high-pitched teeth noise. There is already very little (subjective..) back-lash between the pinion and output gears.
First, if you're not that familar with gear theory, take the Boston Gear "Gearology" online course:
If you're getting excessive tooth noise, there's a good chance that the tooth shapes are wrong or incompatible. This isn't that rare in small hobbyist-grade gears. If the diametral pitch is different or the teeth aren't proper involutes, the teeth won't really mesh properly. The gears may appear to work, but you get noise and excessive wear.
Gears have a quality rating, from 3 to 15. Good injection-moulded gears, (see wmberg.com) have ratings of 5, which is not great. Normally, an aviation actuator would have grade 7 to 11 gears. If you had to trim moulding defects, we're probably looking at an AGMA rating off the bottom of the scale.
Take a look at something like Berg part FFD105-72, which is a
72 tooth 37mm Delrin gear in AGMA quality 5. That may be too thick, though. (Don't go by the picture; those are generic photos.) You could thin it down with a belt sander. Or see Stock Drive Products part A 1M 2-Y48072.