Quite possible - surface finish has a major influence on generated noise - if you have a surface finish measurement device you should be able to pick up a difference between the quiet face and the noisy face.

I don't have the device here, I'm the only mech eng working in an electronic enterprise...  Should I be able to see a difference by eye or by passing a needle on both surface and feel it?

robindbois,
I would suspect that the intenal gear
was not generated but gashed with a milling
type cutter and may be slightly off spec.
Was is cut with a shaper cutter ie. generated?
Sounds like an interesting phenomenom.

One or more of the gears may be off spec in pressure angle and other characteristics. I have detected differences in perceived noise in plastic gears after running under load, where the load is primarily in one direction.

The ABA-PGT tooth form, from the company of the same name, allows for tip relief, which helps reduce the effects of load deflection in plastic gears. It also helps reduce noise after the gears have developed a permanent deformation in one direction. I have measured several degrees of lean of plastic gears after running durability cycles.

I talked to the cutter guy and he told me that the internal gears are cutted with a shaper.  Since it is an internal small gear it cannot be generated by any rotating tooling.

I installed the set with a plastic internal gear and started a run to see if I could have any break-in effect that would reduce noise.  I'll do the same thing with the aluminum internal gear which is really noisier than the plastic one.  Maybe the break-in effect will polish the surface finish and help reducing noise.

Robin
Mech eng.

Is it possible to turn the pinion over and
run it in both directions to see if the
noise reverses itself?  Or is it possible to
turn the gear over and test this condition?
Just trying to get some quick answers.
My thinking is if you do one of the above,
you may be able to isolate which part is out
of spec.

Its impossible to turn any of the gears.  I have two ss pinion, 2 aluminum internal gears and two plastic internal gears.  I have the noise in the same direction with all combinations.  However, the system is quieter when driving it by the pinion with the motor than driving it with the internal gear with my hand on the plastic unit.

I'm not sure the plastic unit is off spec since it sounds quiet for me.  At least as quiet as can be a gear set.  For the aluminum parts, I'm not sure if the sound is still normal but I want to give the systema break-in period to see if noise will go quieter.

Diamondjim, remember the spec I sent you about the helical set, I have the pinion in my hand and I ordered the special tooling to cut the internal gears, I should have them for march I guess.  I really don't know if the helical set will be quieter than the straight set for the plastic material.  It sounds already quite quiet.

Robin

Robin

As a field spook who has to diagnose ill behaving gear sets, I would say your first mistake is having an even tooth count on both pinion and gear.  Cuts way down on the gear life and makes for a noisy set.  Add a tooth to the gear and see if things don't quiet down some.

Vibegeek,

What you said make sense but I did not know that it can make a practical difference in moise and life.  By having an even and uneven number of teeth in gears, you alternate the mesh pair of toth on pinion and gear.  In this way, you wear the teethin a more equal pattern, avoiding repeating wear at specific spot.

I don't have any load so life is not a problem but noise is a big one.

I am waiting an internal helecal gear and by luck the teeth ratio is 12/213:)

Robin

Robin,
The gear set was initially designed
to have the pinion drive and therefore
it was recommended to have a long addendum
on the pinion to make the disign similar
to a recess action gear set.  When the gear
is driving, you have more approach action,
and can expect more noise.  I did not realize
you would be using the gear to drive the
pinions.

Robin,

Any gear which runs at a speed morethan 100 RPM, need to be profile ground.Cut gears profile has the error.When u mesh two gears , theoretically speaking it should mesh at its pich circle dia.normal cut gears will have more variation. So you have to necessarily go for profile grinding operation.Its not just due to surface finish. To have better surface finish u can go for polishing after gera cutting operation.That never reduce noice. So Please go for profile grinding operation and iam sure u will not face this problem again.

ok

Sudev Nair

DiamonJim, what you said make sense, from the fairfield run, I have a slip ratio at sap on the pinion of -2.68 which I think is relativery high (from afairfield guy it should be less than 2 and ideally less than 1) which lead to a more susceptible system to noise mesh.

Probably this fact is explaining why I have less sound when driving the system by the pinion than when driving by the gear.  This does not explain why the sound is not symetrical but the explanation of the non symetrical cutting could be the one for this part.

The system is a lot less noisy with the plastic gear (Ertalyte) than with the alu (6061-T6).

Sudev, since I'm using palstic gear, can I still go to a grinding operation?  Furthermore, the gears are quite small (48dp, 12 and 244 teeth).  The internal gear is running at 60 rpm.

I could profile ground and polish the 6061-T6 gear but it is maybe less expensive to just use the plastic gear.

I'l have to work on this since primary test was done in open loop type control and the closed loop os noisier.

Robin

Robin,

Sorry, I thought you were talking about Metallic gear. Of course I didnt go thru the material u have used.

Can I ask you a fundemental question now? You might have already calculated this. But still I would like to clarify.

Did you calculate the number of teeth required on Pinion under interfernce criteria. When nr of teeth is less as per this criteria, there will be a interference between pinion and gear. I understand that u have 244 teeth in gear and 12 teeth in the pinion.Without calculating itself i feel this will interfere and it will eat material from the root of the pinion too.

So please calculate nr of teeth required based on this criteria. If you want my help pls give me details like nr of teeth in gear(just confirm), module, and like in pinion also. I faced similar situation with metallic gear.Then I calculated for interfernce and decided to go for minimum teeth required for Pinion and solved the problem.One intersting thing is that, its not possible to have a huge redution with spur or helical gear assembly.If you need very high reduction you may have to go for nr of stages of reduction say 3-4 reduction stage.This all due to min nr of teeth required on pinion.One stage you need very high reduction you may have to go for worm and worm wheel or planetary gear train design.As per your gear train you get a reduction of 244/12 ~ 20 reduction one stage itself.I feel its too big..so pls calculate min nr of gear teeth required for pinion.

I dont thing its due to gear shaping problem.Even you can go for plastic injection molded gears(high precision and readily available in market). Also check the center distance and tolerance allowable for center distance variation..okkk

Pls let me know your resonse on my statement

Sudev Nair

I calculated the minimum number of teeth under undercutting and pointing criteria.  This is why I went to a shifting of the profile by 50%  I used the fairfield program and a formula found on an university site teaching gear theory.

I do not have any place to have multistage and I'm under tight backlash control.  I cannot use injected gear since I need really special internal gear (shape, number of teeth)

Just a simple question:  If I do not have undercutting and pointing tips, can I still have interference?

If yes, what are the formulas to calculate it?

Robin

Mech eng.

Robin,
Please be assured that the aluminum gear
will always be noisier than the plastic
gear.  If you can imagine the two part
slapping against each other, the steel
against plastic verses steel against
aluminum, it should be obvious that the
plastic causes less noise.
I am anxious to see the results of the
helical set.  Except for strength, the
plastic will probably be your best bet
as the steel and aluminum mesh will wear
quite quickly from other designs that I
have seen.  I do not know if you can coat
the aluminum with a coating to make it
less porous at the surface to help reduce
the wear.

robindbois, Runout, involute error, lead error, spacing error can all lead to noise. I have found lead error as well as runout to be a cause for noise levels to sound different in opposite directions. My recommendation is to have the gears checked for the above errors first. Helical gear will be quieter that spurs. You can also modify the gear tips to reduce noise.

Gary at Gear Manufacturing, Inc.

I just read quickly this thread and no one has mentioned that the rotation will tend to lift the pinion when spun one way and force it down when spun the other. This then makes for a subtle difference in the location of the axis and "could" account for the difference in noise.
Just my 2 cents
Ralph

Ralph2,

I have the noise with the straight gear set.  I haven't received the helical internal gear yet but should be in a couple of day.  I just found one place in north america that can cut it!  I've asked around 25 places!

Diamondjim, As you said, the plastic set is quieter and I'm happy to have another (other than me!) that think wear will be less with the plastic set.  We will go for the plastic set with ss pinion for sure.  

I was wondering waht is the standard method to deburr a plastic gear that was cutted?

I can see small burrs on the profile where the cutter exited.  The noise is at the pinion frequency and the slip ratio at sap for the pinion is 2.67, which is quite high.

How can I modify the tip to lower noise?

Is it expensive and complicated to measure if the gear if off spec and by how much?  How do they do that with a 48 pitch diameter gear? it is small?

Hum, puzzled by this noise...

Robin

Robin,
The slip ratio is not that high,
anytime the gear od approaches near
the base diameter the ratio get high.
I have seen 14 and the gears seem to
mesh well in slow applications.
Just decrease the od of the gear by
a small amount and see how quickly this
ratio changes.
These are such small gears that you
could probably deburr by hand with
a diamond emery board to see if that is
the source of the noise.  You might try
running it in at a slow rpm for an hour or
so and gradually increase the rpms.

Assuming teeth in internal are shaper (Fellows) cut, it is quite possible to have better quality in one direction than the other.

Best analysis is by single flank testing, or analytical inspection.