Help Me Understand Dog Shifting

For some reason, I have the misguided conception I can design a shifting gearbox for my team this season (I blame the mentors.) I understand a lot of the math behind it, for finding optimal ratios and whatnot, and I know Inventor well enough to design it in there, but there is one big problem: I don’t understand the geometry behind the actual shifting.

To the best of my understanding right now, a piece with a couple elevated parts is pushed into a gear with a corresponding milled out parts so that it engages the gear (this is possibly entirely wrong), and when the pneumatic is shifted, the other side of the part does the same thing with another gear; when either gear is not meshed, it is idle. A terrible mspaint of my current understanding is attached.
http://img466.imageshack.us/img466/2745/shifter1nt.png

My problem at the moment is that I don’t understand how the pneumatic goes through one gear (I’m assuming it’s actually through the shaft?), how to keep the gears from moving if they are on the same shaft as the dog shifter (or maybe they are on different shafts…that idea occured to me but seems intiuitvely wrong), and how to keep the gear from moving.

I’ve looked through a number of whitepapers on this, in partiuclar whotek’s, but the cramped shop drawings aren’t really letting me understand what’s going on. If anyone could either explain the geometry of how all of this works or link me to an inventor model rather than drawing of one, it would be greatly appreciated. Thanks!

shifter.bmp (24.1 KB)


shifter.bmp (24.1 KB)

http://www.technokats.org/documents/6Motor.pdf

http://www.technokats.org/documents/shiftonfly.pdf

Check both of these files out. It should help you a little bit.

The TechnoKat-style (e.g. 2002, AndyMark) gearboxes use a rod through centre of the driven shaft to actuate the dog. The Woburn-style one (from 2004) used a movable plate around the outside of the dog, avoiding the need for a mechanism coaxial to the driven shaft. Both types use pistons mounted outside the gearbox to shift.

Take a look at the whitepaper I wrote about the 2004 Woburn design; pages 80-83 might have some useful diagrams of the shift plate method. Also, AndyMark has several diagrams of their gearbox (which uses the through-the-shaft method) on their website; this one is especially relevant.

**Edit: **Bah! Beaten by the sneaky Mr. Gilbert…

so far as i know, there are two ways to shift the dog using a pneumatic cylinder: internally, and externally.

when the internal option is used, a pin connects the dog to a rod that runs through the center of the axle. the rod is connected to the cylinder through an assembly that allows the rod to spin with the axle, without spinning the cylinder.

the external option is simpler; it is simply a stationary ‘claw’ (for lack of a better term) that rides in a groove cut into the circumference of the dog. the pneumatic attaches directly to the ‘claw’ and when it is extended or retracted, the ‘claw’ forces the dog into the appropriate gear.

[edit] ach! beaten* twice*!![/edit]

Here’s a great picture of one that 228 made in 2003 http://www.chiefdelphi.com/forums/attachment.php?attachmentid=1310

On www.andymark.biz you can download a STEP model of their transmission.

You can hit the “Hexamax” link in my sig for the one 696 made in 2005. I’d be glad to send you the Inventor files of it and/or help you with any questions you may have about it or shifting transmissions in general. Send me a PM.

Here’s a great cutaway pic (of 45’s) showing how they work. http://www.chiefdelphi.com/forums/showthread.php?t=20776&highlight=technokat+gearbox+pic

Light blue is bearings, purple and white are bearings, blue and green are the gears, red is the (hex) shaft, orange is the dog, and yellow is the shaft that the pneumatic cylinder pushes and pulls.

Good luck! And once again ask lots of questions and we will do our best to answer them. :slight_smile:

The basic principle behind the dog shifter working is that the gears are free spinning on the shaft. The dog is mechanically engaged to the shaft so that when the dog rotates the shaft rotates. The shaft and dog have internal geometries for this engagement; a keyway for the Who’sCTEKS version, a hex shaft in the TechnoKats version, etc. The shifter is pushes the dog back and forth along the shaft, there is some clearance to allow the dog to spin on the shifter for outside shifters and the internal shifters spin with the shaft.

The mechanism works by the motors driving thru the gears. Both output gears are always spinning while the motors are turning. When the dog engages one of the gears, that gears RPM is transmitted to the shaft while the other gear free rotates. So the shifting is done to select the desired output speed range.

Because of the dynamics involved, there must be a ‘neutral’ between gears. This way you will not lock both speeds on while moving. This can cause the gearbox to break! The simplest way to achieve this is to increase the distance between the gears to the thickness of the dog plus .030+".

There is plenty of help available here at CD, your first gearbox should be a success!

Thanks for your explanations Andy and Sanddrag, that answered pretty much all of my questions, especially the teknocats cutaway. One thing I’m unlcear on is how are the idle gears kept in place? It seems to me that you could only put spacers on the outside of the gears beacuse the dog in between would make retaining them on the inside impossible. Does the motion of the gears keep them in place, or is there a better way of doing this that I’m not catching on to? Surely something is keeping them from moving on the shaft.

On a different tangent , what kind of bearings are used on the teknocats cutaway? If the red is hex shaft, how is it inserted into the bearings in the middle of the shaft? I’m assuming even if those sections are lathed, it would be impossible to move the part of the shaft that is still hexagonal through the bearing. Are they special hex bearings or something? I thought those didn’t exist, but again, I’m probably just missing somethign obvious :stuck_out_tongue:

Thanks again for all the replies, it’s greatly appreciated!

You are on the right track, the bearings are flanged and they ‘PRESS’ in from the side opposite from the @#@#@#@# dog. The drive shaft in this case is hex in the area that the shift dog slides on and round where the gears are. Now, the gears with their bearings can only move in up until the hex area. That is what keeps the gears from collapsing inwards towards the dog. Also the picture in Sanddrag’s was ours from 2003. We have since went to a fork shifting apparatus using servo’s instead of pneumatics. We also changed the dog fingers to be more like Nitro RC vehicles (Larger sweep engagement) you can now literally bang it into gear over a 70-80 degree range instead of waiting for the 3 fingers to line up more precisely.

Look at the cutaway view and notice that the output shaft only has the hexagonal shape where the dog rides on it. The rest of the shaft on either side is a standard cylinder. That answers both your questions. The bearings are put on the normal cylindrical sections, and the thicker hexagonal section keeps the gears from sliding toward the middle.

The Who’sCTEKS use the same setup. The shaft diameter for the dog is 5/8" and for the gears it is 1/2". A bronze bearing is pressed into the gears with a slip fit on the shaft. The gear cannot move inward due to the step.

Yes, I did mean 226. Sorry for the error.

This may help a bit

THis is from our 2005 Nitro RC style dog shifter