11-07-2003 10:53

Andy Baker

Just to let everyone know, I just posted the 2003 TechnoKat gearbox in the white papers section on this site and under the "Resources" section on the TechnoKats site. This is the "2 speed, shift-on-the-fly gearbox" that worked well for us this year.

It is a significant improvement over our 2002 model. Most significant are the changes made to make it easier to fabricate (no EDM needed, etc.). The document includes an extensive writeup, assembly prints, BOM, and detail prints.

Download and enjoy. I will have hard copies of this at IRI (enough for 1 for each team).

If you use or improve upon this design, please let me know.

Andy B.

11-07-2003 10:55

Ricky Q.

Thanks Andy, been waiting for this one

11-07-2003 13:22

Andy Brockway

Thanks Andy!

There does not appear to be any write-up with it. I am interested to know if you have any disassembly/assembly fixtures for removing/installing the press-fit gears on the drill and CIM motors.

I hope to have my simple dual motor white paper posted by the end of the month.

11-07-2003 13:40

sanddrag

Thank you so much.

11-07-2003 13:48

Andy Baker

Initially, I uploaded the wrong version of this paper. My buddy Ricky Q. notified me right away, and then I got the better version uploaded. Thanks Ricky and Brandon for being "on the ball".

Andy B. - the absent minded engineer

11-07-2003 16:53

FotoPlasma

I would definitely not say that, especially considering the quality of this work.

Great job.
Keep up the great work.

11-07-2003 16:58

Cory

Awesome job, Andy. Its always great to see teams releasing stuff like this to share their success. One little discrepancy that I noticed while browsing it is that the stall torque is higher in high gear than low gear. I think the numbers were just reversed. Dont mind my nitpicking Great job!

Cory

11-07-2003 17:46

Tytus Gerrish

what kind of mill do you have?

11-07-2003 18:51

dez250

great release Andy, i cant wait to get enough time to look over everything and see if i can duplicate this in my spare time...
~Mike

11-07-2003 19:33

JVN

Mike,
There are 2 comparable shift on the fly systems sitting in Colonie HS. Check out the Shen 1997, 2000, 2001 and 2002 robots. It's not overly difficult to add a 2-motor combiner to one of these. Believe it or not, 250 has much of this technology in the closet.... literally.



Andy,
Great paper, the reliablility of your "dog shifter" design is certainly making me re-think my "pin shifter" design. We (229 drivetrain guys) are definitely going to have to sit down in the fall and decide what direction we're heading in next. It's amazing for me to think that because of your posts, whitepapers and work, you've had a similar (catalyst) effect on teams across the country. You always help raise the bar!

So what's next for 2004?

Thanks Again,
Your team's openess is an inspiration to us all,

John

11-07-2003 20:35

dez250

john i know we have things like this literally in the closet, i just wanted too have fun at my house/in the shop in my free time and just use some things i have and see if i can do it myself, or with help from some others to make it for a practice bot I'm looking to build... Yeah that prob wont happen though since i don't have much free time...
~Mike

"Learn from the experience of others..."

11-07-2003 20:53

sanddrag

It's not just the time. It's the money. Otherwise my team would be building one right now. We figured it would be about $250 for just gears for one gearbox. Of course that was with some slightly different ratios that we came up with.

11-07-2003 21:10

dez250

yeah the money matters, but the time is first, and having access to what i need now, which for a change i have in my storage...
~Mike

12-07-2003 20:19

Joe Clohessy

Wow ... I came up with valid ratios in about 2 days with spacing considered Casue for your 2 speed you cant have 12 driving a 24
and for your other speed say on that same shaft a 32 driving a 30
All in 2 days work.... Whats taking so long??? he he

13-07-2003 00:59

Andy Baker

Thanks for the replies... here are some answers:

Jim - you would be suprised, I can be a space case sometimes.

Cory - thanks for the nitpicking, it is needed. I will make the change and wait for other stuff that may need updates.... they I will re-post it after a while. Since it is in "public domain" it needs to be right. If you or anyone else finds errors, please let me know.

Tytus - a CNC end mill. Our high school machine shop made the parts that needed CNC'ed on a Haas machine, I believe.

John - thanks for the support. The good thing about the machine dogs is that they are permanent... you don't have to worry about them once they are made. Sure, they take a little more time to machine as opposed to putting in dowels, but it is worth it. As for 2004... it will have more recution and reduce parts. We would've liked to have about 25% less rpm on the output shaft.

Mike - it would be great to see people build these in their free time. Better yet, improve upon the design and post your findings. (one of my favorite quotes: "If I have seen further, it is by standing on the shoulders of giants." - Issac Newton). Take what we have done here and improve it... that is what Newton meant.

Sanddrag - the gears actually don't cost that much. Keep looking around for suppliers. These gears can be purchased, off the shelf, for about $90. Two sources that I used are Martin Gear (20 pitch, purchased through Applied Industrial) and Rino Mechanical (32 pitch, purchased direct).

Joe - good for you. Designing a gearbox in two days is not easy. However, quickness in design can cause errors, just as you made errors while criticising our design. We did not use a 24 or 32 tooth gear. Also, I am not sure why you say "what took so long?" Do you have your design posted on the web? If so, please reply with a link. Also, I trust that you are a good designer, but it would be helpful if you wrote with more clarity. It's somewhat hard to understand what you are trying to say.

Again, thanks for the comments... keep them coming.

Andy B.

13-07-2003 01:31

dez250

Andy thanks for the words of encouragement, i was thinking about changing some of the gearing ratios around and just tinker a little to see what i can do with the output. i am wondering if i make it so the output has higher torque and less rpm's, what it may do on the voltage input, if it may draw more or if it could cause a stall on the motor or what may happen...
~Mike

"Make Failures and LEARN By Them, DO NOT Let Your Own Failures, Get You Discouraged..."

13-07-2003 02:31

sanddrag

I was a little off. I actually came up to $205 for each gearbox using gears from Berg and SPI. My 20 pitch gears were in the following tooth counts 16, 40, 60,20, 80. I plan to use bigger gears than you did for some slightly different ratios. That 80T gear is like 75 bucks alone from Berg. I'll check with your suppliers if and when my team builds this thing.

As for your note about the CNC machines, HAAS are wonderful. They are such a joy to work with and so user friendly. (I am just starting out on one).

13-07-2003 08:18

Matt Reiland

Sanddrag, take my advice and go to your local Motion Industries and get the Martin Gears. We were buying them from Small Parts & Boston and their prices were over twice as expensive as Motion Industries plus they were in stock.

Andy, great job on the prints, to show the level and detail as well as a complete model in Uni-Graphics surely took a great deal of your time, and I for one appreciate it.

I might tinker around with it a bit and see if I can incorporate any of the features we used this year. One thing I really want to test out is the spring loaded shifting linkage with the hobby servo. If I can get it to mount and work on the side of the tranny I will post it to give an alternative to pneumatics. Gear job of eliminating the wire-edm parts, it opens up the design to many more teams.

Along the same lines to teams out there: looking through the design it appears you could make most of the parts on a regular mill with a digital readout. I tried making one of the gears with the pocket dog mates at home on my machine (which is CNC) but I did it by hand with a rotary table and you can do pretty well without the CNC. I can't wait for one of our schools (Troy Athens) to have their CNC machine at school making parts for us this year.

One thing I am not really clear on is the mount of the drill motor. We had to make an aluminum carrier so we could screw into something (Also made a fantastic heatsink too!) , how are yours mounted in the 2003 gearbox? Also love the carrier on the CIM motor, fantastic idea.

Great tranny, ohh so small too.

13-07-2003 08:37

Matt Reiland

Mike,

If you change the ratios to lower the RPM and increase the torque this will only help out the motors even more from stalling and they will draw less current than if you geared it up for higer RPM's. Keep in mind teams out there you can build this tranny exactly to the print and add any output stage to the wheels that you want, up or down. We reduced ours by 3:1 last year with 6" brecoflex sprockets and it was pretty much too hot to handle in high so we will also go a little lower for 2004 which will give you even more torque.

13-07-2003 12:39

dez250

Matt, thanks for that, i want sure if dropping the rpm's would cause a stall or if i had it backward, which i guess i did. This will be a fun project and looks to be quite an interesting one.
~Mike

13-07-2003 14:24

sanddrag

Let's remeber that CNC's do not make parts FOR you; they make parts WITH you. A lot of newcomers to machining (not that Matt is by any means) get into the habit of saying "Oh, let's just do it on the CNC." Well, it's not that simple. Everyone should realize that a great deal of programming goes into making even the simplest part and a lot of thigs could be done more simply and quicker on a good 'ol manual.

13-07-2003 14:59

Matt Reiland

Excellent point sanddrag, even though my machine is CNC in the lab where we build the robot it is fully manual with a Digital Read Out.
However most CNC machines can do things that aren't very easy to do manually like complex curves, speed, and allowing for duplicate parts once the program is created. The best way to learn is to do it manually and have knowledge and respect for the machine, then as a convenience move to CNC if you have it. Up until just recently our schools had no metal machining facilities and they finally went and bought an EMCO with a full shroud. I am going to see it next month.

14-07-2003 09:44

Andy Baker

Matt,

Thanks for the comments.

This points out one thing I forgot to include in the write-up. Mounting the drill motor was not as easy as the prints show. We altered the drill motors a bit in order to make them mountable. Very carefully, we threaded the two holes on the front face of the drill motor for M5x0.8 screws.

In order to thread these holes, we started out with a regular M5x0.8 tap to begin cutting the threads. Then, we switched to a "bottoming tap" (which is the same tap with the end tip grounded off) and finished the threads. We did this in so that the tip of the tap did not hit the motor windings. Also, when we threaded these holes, we used grease on the tap so that any chips would cling to the tap... we did not want chips getting into the windings.

It was kinda tricky, but it worked fine. We did not have any drill motor mounting "issues" once they were installed.

There probably is a better way to do this... 226's way of making a separate piece to hold the drill motor is more of a "for sure" way to go, and I am always for making things better.

Andy B.

14-07-2003 09:58

WernerNYK

Oh so true... We made custom aluminum 72-tooth sprockets on the manual mill; each one probably took less than an hour to manufacture from start-to-finish.

Don't underestimate what can be done on a manual mill with just a little ingenuity

15-07-2003 23:33

dlavery

Andy -

As always, your transmission design and plans are superb.

The approach we took with the drill motors this year was to use two short #10 machine screws in the face plate of the motor mount (i.e. they projected through the small holes in the face of the motor, similar to the plastic pins in the original drill transmission). To retain the motor against the face plate, we used small "J-hooks" made out of 1/4" drill rod that ran from the face plate, along the sides of the motor, to the square notches at the rear of the motor housing. The screws/pins prevented rotation, and the J-hooks prevented lateral motion of the motor. It was very easy to fabricate, and we didn't have to try to tap the holes in the face plate of the motor (I was really concerned about getting chips into the windings), and proved to be pretty robust (we never had a motor torque out of place, and we beat on it pretty hard).

I will post a set of drawings in a few days, as soon as I get them cleaned up. We updated our transmission design for the 2003 larger drill motors, and had good success with it in Annapolis. The design is still bigger and clunkier that the Teknokat transmission, but doesn't require CNC for fabrication (for those of us still limited to manual mills and lathes ).

-dave

16-07-2003 12:56

Joe Johnson

real men hold the back side of their motors...

I like your design, Dave.

2 things. First, I like that it does not require customizing the motor (something that I think is just good FIRST practice if you can manage it). Second, I like that it holds the back side of the motor.

We ended up holding the motor with a plate that fit on the boss on the back side of the motor.

As I said in the subject heading, real men hold the back side of their motors... ;-)

Joe J.

19-09-2003 11:34

Matthew_H

Shift Shaft

The ubergeeks are currently fabricating the parts to try and practice with this gearbox design. The biggest question that our team has at the moment is what do you have the shift shaft connecting to with the thread at M3x0.7
We do not have the this dye nor does the other machine shop that is helping. Is there a specific reason for using this thread.

Great design and my students are very excited about seeing this gearbox work.

Matthew
Team 1028
Ubergeeks

19-09-2003 11:58

Andy Baker

Re: Shift Shaft

This part of the design is shown in the picture on the front of the white paper, but it is not detailed in the prints. The only reason I specified an M3 thread was to keep with our common fastener practice of M3 or M5 metric threads. However, like your team, we did not have a M3 die. So, we used a #6-32 die.

We put a small bearing (1/8" id) and a #6-32 lock nut to couple to that shifter shaft. The lock nut wanted to back off after many cycles, so we drilled through the nut and threaded stainless steel "speed wire" to hold the lock nut in place.

Good luck,
Andy

19-09-2003 18:23

sanddrag

Re: Re: Shift Shaft

Couldn't you have just jammed two nuts up together? Also, I'm not sure if they come in small sizes but there are these really cool nuts called crimp nuts where the top of them is smaller so it fits very tightly on the threads acting like a locknut but it is all metal, no nylon. They are used on aircraft I believe.

19-09-2003 22:44

Ryan Foley

Wow, awesome transmission. Hopefully if we find a good machine shop we can try it out.

::adds this to my collection of Team 45 tranny white papers (this is number 4)::

23-01-2004 06:49

Guilherme

Re: White Paper Discuss: 2 speed, Shift-on-the-fly TechnoKat Gearbox

I was wondering if you could share any kind of CAD format of this file, because we have metric units here in Brazil, and we´ve to see all details of the gears because we don´t have them available to buy here.

Thanks.

23-01-2004 09:37

Matthew_H

Re: White Paper Discuss: 2 speed, Shift-on-the-fly TechnoKat Gearbox

I had one of my drafting students put this desgin into CAD. It should show you the basics of the design. I have not fully checked it out but if you would like it send me an email and I will attach it.

Matthew


Ubergeeks

23-01-2004 14:15

Guilherme

Re: White Paper Discuss: 2 speed, Shift-on-the-fly TechnoKat Gearbox

Thank you all. Andy already sent me the .step file, it should be enough.

Three more questions:

Andy, I was at our high school, but now I´m at home and I don´t have your email adress. May you send me the Unigraphics files? We´re trying to get a sponsorship from a company that has CNC´s, so we could build the parts with these files.

I´ve spoken to this company, and they said that in order to build gears, the should have special tools engaged in their CNC´s. Could ou specify the tools needed to build ALL gears? The models of their CNC´s are: feeler model FV 1100, dyna Myte model 4500

Whats the average time that this gearbox takes to be built?

23-01-2004 15:30

Cory

Re: White Paper Discuss: 2 speed, Shift-on-the-fly TechnoKat Gearbox

When we were asking around at local machine shops, they said it would be around 40 hours of work. That does seem possible, albeit a bit steep. It really depends what's being CNC'ed and what needs to be done manually.

Cory

23-01-2004 20:16

Matt Reiland

Re: White Paper Discuss: 2 speed, Shift-on-the-fly TechnoKat Gearbox

40 Hours seems possible, however do yourself a favor and make extra parts. Lots of the machining time is the setup, once the CNC portion is programmed, making an extra few is much easier and you may need it later.