[johngalt47]

I am working on a dual pinion rack and pinion design rack and pinion drive. I have worked our most of the design issues except for the amount of force to push the pinions apart so that teh left sife of one pinion and the right side of the other is in constant contact with the rack.

[jsheerin]

If you want constant contact, I think you'd have to set the spring force to be such that the force applied by the belt to the pinion creates a torque on the each pinion that is equal to or exceeds the torque you apply to turn the drive pinion (the one connected to the motor) via the timing belt coming from the motor. You could either figure this torque as being the max that could be generated by your motors or the max you think you'll encounter given your acceleration settings and anticipated max cutting load.

[PaulRowntree]

This is the first time I have seen a fixed motor axis on R&P. Would this allow you to have a single motor driving a cross shaft to the two exes? Would this be an advantage at all?

Oops, I misunderstood the drawing.

[wildwestpat]

IMO the dual speed belt ratios will also add to the most motion. Why not use one belt to link the pinions and the motor? The dual pinion will drive first with one pinion when the other as the motion reverses. As you have the pinions separated by several times their diameter the variations in rack pitch have to be accommodated as well as the variation in mesh of both pinions to the rack over the length of the axis.

Whilst looking for increased rigidity why not use two plates instead of three? Beef up the spacers and consider using square section tube or solid sides instead of spacers to give a more solid construction.

If the rack has a precision face ground on the opposite side to the teeth why not use this control the mesh of pinion on the rack? Simply fix the rack with through bolts on the free face. The wheels controlling the mesh could double up as the support rail for that side if the dimensions are suitable.

Most machine designs need to work back from the desired machining envelope (length x width x height that can be cut) and the accuracy / repeatability (these are not the same) of the parts to be made. From the material to be cut the feeds and speeds can be factored in to axis motion speed and forces as well as the dynamics based on the total weight needed to maintain the accuracy required. Do not fall into the trap of aiming for accuracy that you do not need unless you have unlimited cash.

Good luck with the rest of the design Regards - Pat

[zool]

Maybe my bifocals are out of adjustment, but it is unclear what you hope to accomplish with this design.

I am a simple fellow, so like mechanical items with few pieces, so just sign me confused.

[johngalt47]

The root cause of backlash is that the pinion tooth fit to the rack tooth fit is
not a good as we would like. The pinion to engage the rack in the direction it is traveling the back lash is caused when you reverse the direction and the pinion will travel from the side it was engaging to the oppisite side. This design will have a seperate pinion engaged on engaged on the right side of the rack tooth and one for the left side of the rack tooth. when the machine reverses there is no back lash as one pinion is fully engaged in each direction.

[zool]

Of course we all admire designs, and I fully admit to WAY over-designing. However, I wonder if your design would cause unintended stresses and strains.

Primarily, how does the changing of direction affect the interaction of the each pinon on the rack. As the direction changes and the load changes from the inside teeth to the outside teeth of the pinon, what is the effect on the entire mechanism and its parts, and what is the effect on back lash.

Ultimately is this just having two pinons really going to diminish the back lash as they are turning in the same direction.

I have looked at your drawing throughout the day [as it is intriguing .. plus it is more fun to think about than write this report...] and I cannot quite totally grasp how your design works; that is, I keep coming up with a different explanation of its operation to myself.

And then again, maybe I just have over thought the entire thing!!


The following are some R&P drive mechanisms that come to mind.


Take a look at this design: NoSleep Studio

I do admit I could be a bit prejudice as I own a Shop Droid with these gear reduction drive on X, Y, and Z-axes.

The whole assemble cantilevers out past the rack and the pinon intersection which load hold the pinon firmly into the rack, plus there is a spring holding the stepper motor and gear drive assemble into the rack, plus there are bolt tensioned springs holding the stepper motor and gear drive assemble down onto the carriage.

Then there is the Nexen RPS system where the pinon is larger that one would expect: Company News

http://www.cnczone.com/forums/linear...ion_drive.html

And the R&P system that Mike Everman came up with: Bell Home

http://www.cnczone.com/forums/linear...e_ever_if.html

[jsheerin]

Zool, when the direction of motion changes, the load shifts from one pinion to the other. Only one side of each pinions teeth would be engaged on the rack. It would be equivalent to a split pinion design, but not as compact. However the pinion teeth would be wider for a given rack width with John's design as compared to a split pinion. The operation is simple - the spring tensions the belt connecting the two pinions which loads the opposing flanks of both pinions' teeth against the rack. This prevents backlash without needing to spring load the pinion(s) against the rack as in the Nosleep design you linked to (and as is commonly done on a lot of the diy r&p builds).

I suppose the one caveat here would be that the belt will have a spring constant, and that would allow some error in the motion produced by the drive. But I imagine it would be pretty easy to find a belt that would make this error insignificant.

[wildwestpat]

Just an idea whilst dreaming! How about using cycle chain as the rack and the smallest rear hub gear as the pinions. Should be cheap and any muck that gets on the chain would be pushed through out of the way. Donor cycles are every where! The chain would have to be tensioned and there would no doubt be some errors in the pitch over a 2 meter length.

The use of toothed belt in place of the rack with the belt taking a loop over both pulleys is another alternative. Belt is available by the meter and is reasonably priced and has been used on some designs. However when I tried it I got into bother with dirt between the gears and the belt. However I did use a HTD tooth form and a coarser one might have been OK but I gave up and converted to ball screw to get the machine out of the workshop.

Regards - Pat

[zool]

jsheerin stated: "[W]hen the direction of motion changes, the load shifts from one pinion to the other. Only one side of each pinions teeth would be engaged on the rack.

The operation is simple - the spring tensions the belt connecting the two pinions which loads the opposing flanks of both pinions' teeth against the rack.

This prevents backlash without needing to spring load the pinion(s) against the rack as in the Nosleep design you linked to (and as is commonly done on a lot of the diy r&p builds).

I suppose the one caveat here would be that the belt will have a spring constant, and that would allow some error in the motion produced by the drive. But I imagine it would be pretty easy to find a belt that would make this error insignificant."

OK .. I think I got it, maybe. It seems there are two loads being placed between the pinions and rack at the same time, one static [so to speak] and one dynamic. The static one is induced by the belt pulling the pinons [because the belt is under tension by the spring] into the rack in opposite directions [if you will]. Then the motion of the pinons moving induced by the motors induces the other load.

Is the belt pulling the pinons against each other and into the rack what is called "loading"? One pulled clockwise and one counter clockwise?

Would there not still need to be a spring to hold the assembly against the rack as one pinion is pulled clockwise and the other pulled counter clockwise by the tension on the belt, which would seem to tend to push the assemble away from/out of the rack.

What about the Nexen RPS system?

Never heard of split pinons, though found this: ATLANTA Ultra-High Precision Rack & Pinion Drive Systems

Still can't see exactly how split pinions work, but then it is after midnight and 50-percent of my cognitive abilities are already asleep.

Remember, if I seem slow it is because I spend my time making things look pretty, so all this engineering stuff overloads my neurons and makes my eyes cross. You get too far ahead of me I will start telling you about root 2 and root 3 proportion, as well as harmonic sequences!!

[louieatienza]

The root cause of backlash is that the pinion tooth fit to the rack tooth fit is
not a good as we would like. The pinion to engage the rack in the direction it is traveling the back lash is caused when you reverse the direction and the pinion will travel from the side it was engaging to the oppisite side. This design will have a seperate pinion engaged on engaged on the right side of the rack tooth and one for the left side of the rack tooth. when the machine reverses there is no back lash as one pinion is fully engaged in each direction.

You should check this out...

http://www.atlantadrives.com/pdf/ads_splitpinion.pdf
ATLANTA Pinion Gears

[johngalt47]

the bottom idler on the pinion tension belt is there for belt maintenance and really not required. one side of the belt will tight to the pulleys and the oppisite side has to have slack. the slack is necessary to allow the pulley tension in oppisite directions. the bottm idler i felt it would be better to take the slack up giving better pulley wrap. the spring for thes idler is very light as opposed to the top idler spring which holding tooth contact.
even though the parts count is higher than most spring tensioned drives i can save machining and welding time by using a cnc plasma cutter to fabrcate it. very reduced cost

[wildwestpat]

Do not think this is so. If the two springs were of equal rate then the driven pinion will tension the lower spring when rotating anti clockwise and the upper one when rotating clockwise. Since your upper spring is much stronger than the lower weak one the weaker spring will do all the work until it runs out of movement. INMO just get rid of the lower spring.

Regards - Pat

[zool]

John:

Would it be possible for you to make the usual three view drawings, with details of the more complex assemblies and functions?

It appears the the diversity of opinion might be based on misconceptions and observers filling in the holes, if you will, with their own ideas.

[johngalt47]

I have added the set screw to cncrouter part rack brackets so the the rach can be adjusted parrallel with linear rails. this means the the pinion can be set to to rack with out a spring. the forces pushing the pinion out of rack will be canceled. allow the wear on the rack to be reduced. the pinion wear is also reduced as the each pinion will be used only half the time doublong the life and thenyou can swap the right to left and double it again.