Hello, This year a two speed transmission was very key to a robot.
I was wondering if teams some how built one or did they buy one.
IF you built one, how, if you bought one, where?
(insert search-before-you-post reference here)
In our case, we bought them from AndyMark. If you’re looking for an off-the-shelf two-speed, you are not going to beat them.
If you search the white papers here for Dr. Joe’s Famous Nothing But DeWalt’s (NBD) paper, after AndyMark, DeWalt 3-Speed/Servo Shifters are the next most popular thing (by a pretty good sized margin too), you build it yourself, but if you do it right, you get pretty good results.
TTLdomination,
Send me another PM. We are in the process of designing and building a 2 speed transmission this summer based off of the andymark shifter but with a different gear ratio. If you can use an andymark product then by all means purchase them since they are bulletproof and relatively inexpensive.
If you search ‘gearbox’ in the white papers you will find several different versions of two speed transmissions including those authored by Andy Baker, Dave Lavery and others.
If you add ‘716’ to your search you can find our 2005 version which I have used a variant of the last few years. My adaptions, not in the original documentation, include: two CIM; one CIM; one CIM with one FP; and repositioning the motors as needed for the robot design.
Should you want to build your own, here is Jim Z’s paper on designing a 4 speed transmission. The 4 speed was a solution when there were not as many powerful motors.
One of the good things about this paper is it talks about making a torque/power gear map. If you are goind to design a transmission this is something to keep in mind.
That being said we did Dewalt’s 2 year (2006, 2007) and then used AndyMark this year. We rebuilt the end plates (steel) into aluminum to save some weight.
For most teams who don’t always have experienced drivers, 2-speeds definitely help in a lot of varying situations. Regardless of what shifting transmission you go with, here are some tips:
- If you modify a COTS transmission, you better TEST, TEST, and TEST it some more. Even a 0.001" shift can burn 40-50% of your power if it’s in the wrong spot. We learned this lesson the hard way.
- Mount the transmissions and the chain tensioners separately – i.e. do not mount the tensioners on the gearbox. This created a wierd off-angle tension for us that further decreased efficiency due to friction.
- If you go with COTS, CAD the mounts, or have slots in order to make minor adjustments similar to the '08 KOP transmission mounts.
- Shifting transmissions have more moving parts, therefore have a higher probability for maintenance. It is imperative you make a frame layout & design that allows for easy transmission removal. Running down the Atlanta tunnel toting a 150lb robot because you’re 90 seconds away from missing your match due to an ill-conceived transmission layout is not as fun as it sounds.
==============================================
Now, it looks like the most relevant information about 2-speeds has been posted, so I’ll go on a tangent for a minute. **Please bear in mind **I say this carefully as I know defensive-style robots gain a large benefit from shifting transmissions whereas offensive style robots see sporadic benefits at best.
2-Speeds weren’t “key” to success in '08 – Case Study:
233 reported throughout various threads in '08 that they were single speed and only went 12fps on an open field. They had a very lightweight frame and ultra-light single-speed transmission. 233 made it to Einstein and were key in the only match in Atlanta that the SimChickenWrangler alliance lost. Not only that, but 233 won every '08 regional they attended. Also of note, the RoboWranglers (148, won SLR and Atlanta), if I remember correctly, were also single-speed.
It still is the overall bot & driver’s performance coupled with strategy that is “key” to success in competition. In some strategies 2-speeds definitely help but should never be thought of as a deciding factor in an alliance’s overall competition performance.