GearHeads Corner
September 26, 2020, 06:55:53 PM *
Welcome, Guest. Please login or register.

Login with username, password and session length
 
   Home   Help Search Login Register  
Pages: 1 2 [3] 4
  Print  
Author Topic: Hypocycloid Gear  (Read 34594 times)
0 Members and 1 Guest are viewing this topic.
DesertRunner
Jr. Member
**
Posts: 52


View Profile
« Reply #30 on: December 22, 2011, 06:21:18 PM »

If we want to reduce back lash why not design the pins to have bearings on them this should reduce friction as well.

Got to say love the programs ability to pull this one off, have wanted to use these type of reduction gearboxs in a heap of factory projects so now I can.

Tony
Logged
Dan
Full Member
***
Posts: 228


View Profile WWW
« Reply #31 on: December 23, 2011, 01:28:34 AM »

Tony,

There is no friction (not dynamic anyway) as the pins have a rolling movement against the cams.

Dan
Logged
DesertRunner
Jr. Member
**
Posts: 52


View Profile
« Reply #32 on: December 23, 2011, 02:20:52 PM »

Hi Dan,
On some of the simulatio ns they have bushs on the pins so thats why I figure there nust be friction.  See this video http://www.youtube.com/watch?v=YRATbJs6zac&feature=related
That said there are plenty of units in the real world that done have bushes so I thick you must be right.

Also what happens with formula when both ratios are the same arn't you then dividing by "Zero" which doesn't work.  See the video of the design with the same ratio.
http://www.youtube.com/watch?v=aDifch4iV48

Tony


Logged
Dan
Full Member
***
Posts: 228


View Profile WWW
« Reply #33 on: December 24, 2011, 01:24:12 AM »

Tony,

You know what, I played a bit with the one I made and watched the teeth and I am pretty certain there IS friction. Don't know why the claim there isn't. It did look that way at first, but know I am convinced in the opposite.

Interesti ng point about the ratio. I have no answer for that.

Dan
Logged
DesertRunner
Jr. Member
**
Posts: 52


View Profile
« Reply #34 on: December 24, 2011, 06:06:56 AM »

HI Dan,
I am glad I didn't turn out to be a fool, I though I might have got the formula wrong.   I have been following a tread about 4th axis back lash over here  http://www.cnczone.com/forums/linear_rotary_motion/72261-backlash_free_rotary_table.html
  and there has been some discusion about this type of drives right back at the start because they have miniual backlash.  When I found this thread with this soft ware my brain got excited.

A question if you preload a gear in the same way they do with engines would it increase energy to drive the unit.
Tony
Logged
Dan
Full Member
***
Posts: 228


View Profile WWW
« Reply #35 on: December 24, 2011, 09:16:43 AM »

Hi tony,

Yes. I have seen that thread.

Can you explain your question?

Dan
Logged
DesertRunner
Jr. Member
**
Posts: 52


View Profile
« Reply #36 on: December 24, 2011, 03:35:48 PM »

Sorry Dan my brain tends to work on a different planet to most people and as such I forget to explain myself correctly .

The cam shaft drive on a Subaru engine has a timing belt to drive the inlet cam shaft but there is a gear from that camshaft to the exhaust camshaft.  To get rid of bacl lash the gear on the exhaust is cut in half and spring loaded to stop any backlash.

In that thread they talked about doing this to reduce backlash but they say it will increase load on the drive due to the motor needing to fight against the spring.  I am of the opinion that its a zero or close to increase in load.
My logic is what extra energy required to move in 1 direction is compensat ed by the spring wantingn to push it along.  In simple terms by the nature of the spring load the gears want to push apart so if I drive either direction 1 of the two springloa ded gears will naturally go in the direction I want to go.  This then by fact means the load has come off the other.
Therefor no dead load as everyone has claimed.
The assumptio n here is that the spring loaded gear cannot be the driver gear.

Sorry for the long winded post but I hope I explained myself corrcetly this time.
Tony
Logged
Dan
Full Member
***
Posts: 228


View Profile WWW
« Reply #37 on: December 25, 2011, 10:37:10 AM »

Hi Tony,

You are describin g a sort of a perpetuum mobile ;-)

If you look at it closer you will see that by putting a spring tough enough to maintain zero backlash under load, you're actually increasin g the load on the gears and essential ly the energy required to move them. There will always be friction from BOTH sides of a tooth.

Dan
Logged
danmauch
Full Member
***
Posts: 187


View Profile
« Reply #38 on: December 25, 2011, 11:27:37 AM »

 That technique of splitting a gear and installin g a spring has been around for years. A long time ago an a place far, far away I use to be an  Submarine Ordnance machinist . One of my jobs  was to install the Torpedo Data Computer (TDC) . These were the old WW2 mechanica l computere rs that used selsen motors driving gear trains to solve the Gyro angle for firing the torpedoes . I use to marvel at the the minature antibackl ash gearing that used the internal springs to eleminate the backlash.
Dan

Sorry Dan my brain tends to work on a different planet to most people and as such I forget to explain myself correctly .

The cam shaft drive on a Subaru engine has a timing belt to drive the inlet cam shaft but there is a gear from that camshaft to the exhaust camshaft.  To get rid of bacl lash the gear on the exhaust is cut in half and spring loaded to stop any backlash.

In that thread they talked about doing this to reduce backlash but they say it will increase load on the drive due to the motor needing to fight against the spring.  I am of the opinion that its a zero or close to increase in load.
My logic is what extra energy required to move in 1 direction is compensat ed by the spring wantingn to push it along.  In simple terms by the nature of the spring load the gears want to push apart so if I drive either direction 1 of the two springloa ded gears will naturally go in the direction I want to go.  This then by fact means the load has come off the other.
Therefor no dead load as everyone has claimed.
The assumptio n here is that the spring loaded gear cannot be the driver gear.

Sorry for the long winded post but I hope I explained myself corrcetly this time.
Tony
Logged
DesertRunner
Jr. Member
**
Posts: 52


View Profile
« Reply #39 on: December 25, 2011, 01:01:18 PM »

On thinking about the Hypocyclo id gear,
Some weird part of my brain tells me it should be,
add the ratios for each stage eg 6 + 10 is 16 plus 1 (because the second stage arangemen t doesn't cause the loss of 1 on both stage) and add a half due to the two stage, (being 1 divided by 2)
Net answer is 17.5.  Don't ask me how I know it right it just is (I think).

The reason you can't find much about this design is that I beleive it binds up and I couldn't not find any aviable commerica lly.  Plenty of CAD simulatio ns such as the one that was linked to but if you remember it had high friction according the CAD program.  I searched the net and every comerical ly aviable drive I found had the 2 centre drives moving indipende nt of each other.  For my 2 bobs worth the proper gear ratio is held back by the two centre plates being attached together.

All of what I have said above could be wrong and I am open to being corrected, its just my brain with crazy logic.
http://www.youtube.com/watch?v=_3OwTd0GG78
http://www.youtube.com/watch?v=oEgnOS4Y4zU

Have a great day all.
Tony
Logged
Dan
Full Member
***
Posts: 228


View Profile WWW
« Reply #40 on: December 25, 2011, 01:27:17 PM »

Funny they never mention the main advantage that we ascribe to hypocyclo ids - zero backlash. I am certain that they wouldn't miss such an advantage . But I still believe that the backlash is very minimal since there are always several teeth in contact from both sides and if machined to close tolerance backlash should be close to practical zero.

Dan
« Last Edit: December 25, 2011, 01:31:19 PM by Dan » Logged
DesertRunner
Jr. Member
**
Posts: 52


View Profile
« Reply #41 on: December 27, 2011, 08:54:41 PM »

I did a bit of research and the Cyclo reduction s are around 1 to 3 Arcmin of free play for high reduction s.  The best on plantery at single reduction 10:1 sort of thing and 3 to 5 and the best double reduction at 25:1 double reduction at 5 to 7 arcmin.  My guess is if you wanted to get a plantery out to 100:1 were the cyclo drives are you would end up with a crazy backlast of 15plus arcmin.   When you dig the only way to go is Cyclo drives if you are chasing minual backlash.

If I understan d it correctly there are 21,600 acrminute s in a full 360 I think I can lives with a error of 5.  On the flip side if you could build your own Hypocyclo id drive you wouldn't need it to be perfect and you would still get minual backlash.
Tony
Logged
Dan
Full Member
***
Posts: 228


View Profile WWW
« Reply #42 on: December 28, 2011, 03:46:35 AM »

Tony,

I too have come to an understan ding why they say it's pure rolling movement. If you allow some clearance between the cams and pins then it's true. But if you have too tight a fit (probably not practical) then the (so to say) trailing edge of the cam would rub against the disengagi ng pin.

And just to add to what you said: 5 arcminute s will translate to 0.07mm at the circumfer ence of a 100mm workpiece .

Dan
Logged
DesertRunner
Jr. Member
**
Posts: 52


View Profile
« Reply #43 on: December 29, 2011, 01:10:57 AM »

And just to add to what you said: 5 arcminute s will translate to 0.07mm at the circumfer ence of a 100mm workpiece

Most Z axis would not be that accurate, now I see it in a number I will move on from worrying about backlash on the 4th axis as its not the issue people make out.  How ever I am still intereste d in simple Hypocyclo id drives for some of my packing robots etc at work.

Dan you didn't comment on the fact that I don't think you can have the two reduction s configure the way you and others are doing on the sample you built.  Did you find any  real world commerica l units designed that way.  The reason I ask is that if it is possible it would be easier.

Dan Mauch I keep thinking about your comment, I am guessing but the power the gears took to drive must have been critical so even though they were springed the friction must not have increased that much.
Tony
Logged
Dan
Full Member
***
Posts: 228


View Profile WWW
« Reply #44 on: December 29, 2011, 04:07:54 AM »

Yes, 0.07mm may not be a lot, but it mostly depends on what you're doing. If it's a precision gear then it may be too much.

Dan you didn't comment on the fact that I don't think you can have the two reduction s configure the way you and others are doing on the sample you built.  Did you find any  real world commerica l units designed that way.  The reason I ask is that if it is possible it would be easier.

I can have the reduction s that way merely because I tried it and it worked Wink

I haven't seen any commercia l ones done that way either, but then they don't usually advertise what's in there. I believe that it's mostly due to the complex ratio formula that they aren't popular. It's hard to get round numbers and a logical sequence of ratio steps. Then there is also that guy's (from the other forum I linked to) comment on these drives, which also makes some sense about efficienc y. The head hurts trying to analyze this. A single stage or a true double stage hypocyclo id reduction is far better understan dable.

Dan
Logged
Pages: 1 2 [3] 4
  Print  
 
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.1.11 | SMF © 2006-2009, Simple Machines LLC Valid XHTML 1.0! Valid CSS!