2 Speed Transmission Suggestions

[burkey_turkey]

Hey all, our team just voted today to try out using a 2 speed transmission. While we always have the option of buying one from andymark, we thought it would be neat to try and design one ourselves. I've been looking around chief delphi all night at different desings, and i was hoping for some more input. I understand the basic concepts behind the duel speed transmissions, but i dont understand all the pros and cons between the differet types. This is what ive got so far:

Mash:
easier to build, but the teeth of the gears are more likely to break during transfer - how can you solve this problem?

Dog:
less likely to break, harder to manufacture. we may or may not have access to any advanced machines to build this thing, just a nice drill press and a band saw, but dont worry, we are commited to making this happen.

i know there are more methods, but for only 2 speeds, they didnt seem very good. if im wrong, please enlighten me

thanks for all the input i know the CD community will give.

[Cory]

Originally Posted by burkey_turkey

Hey all, our team just voted today to try out using a 2 speed transmission. While we always have the option of buying one from andymark, we thought it would be neat to try and design one ourselves. I've been looking around chief delphi all night at different desings, and i was hoping for some more input. I understand the basic concepts behind the duel speed transmissions, but i dont understand all the pros and cons between the differet types. This is what ive got so far:

Mash:
easier to build, but the teeth of the gears are more likely to break during transfer - how can you solve this problem?

Dog:
less likely to break, harder to manufacture. we may or may not have access to any advanced machines to build this thing, just a nice drill press and a band saw, but dont worry, we are commited to making this happen.

i know there are more methods, but for only 2 speeds, they didnt seem very good. if im wrong, please enlighten me

thanks for all the input i know the CD community will give.

If all you have is a drill press and bandsaw, you're going to find it very difficult to make a transmission of any sort, let alone shifting. Transmissions generally need to be precisely machined.

There is one transmission I know of (Killer Bees' 4 speed) that can be fabricated with a drill press outfitted with an x,y table. Other than that, I think you're out of luck.

If you have the money, I would strongly suggest purchasing the Andymarks, then using the summer to figure out what makes them tick, and what you could to do improve them for when you make your own.

Good luck,

-Cory

[burkey_turkey]

weve built custom transmissions before, and if absolutely neccesary might be able to get some out of house machining done if its neccesary, but we've got very good at the equipment weve got. My question is more asking what the pros and cons are of the different shifting methods. we have the theory behind shifting speeds down, just not the wisdom that comes with exeriance. for anyone thats built/used/designed a multi speed transmission, i want your stories and advice as we try and make our own. We could always buy or copy a design, but the fun is the journey, and we really want to make this thing. thanks.

[Fab.Master]

take a look in the white pages for nothing but dewalts. very robust, up to 3 speeds and pretty cheap

[burkey_turkey]

Originally Posted by Fab.Master

take a look in the white pages for nothing but dewalts. very robust, up to 3 speeds and pretty cheap

thanks. if what we try making epic fails, thats a good thing to know. I'm still hoping for some experianced CDer to explain the pros and cons of the different switching methods to me though. Ive seen how they work but want experianced input for when i give a presentation of transmissions tomorrow.

[Fe_Will]

General points:
1. Use the same pitch and pressure angle of gears through out (this prevents the chance of to dissimilar gears contacting)
2. The key is proper mesh to transfer motion, too much or little mesh can cause gear damage or inefficient transfer.
3. Avoid pairing gears made of different materials; the harder gear will wear out the softer more rapidly than two gears of the same material.

About "Mash" transmissions:
Yes, they require less machining but the act of moving a cluster of gears (or an entire shaft) linearly is the greatest challenge in my experience. If you were planning to servo shift then “mashing” is not for you… the force required is too great for the servo to do the job. Also, there are several older threads (2003-04 vintage) that discuss the trade offs of beveling the gears vs. a right angled gear.

About Dog shifting:
The dog and its corresponding surfaces are the most difficult parts of this transmission to craft. The dog and mating gears should be made of a hard material (they will take the most abuse). Several designs that I have seen were shifted by servos and just as many that were shifted by pneumatics.

In conclusion:
It’s a trade-off between servo shift and difficult machining vs. less machining and a tricky linear pneumatic shift, IMHO.

Good luck.



On a side note:

Originally Posted by Cory

If you have the money, I would strongly suggest purchasing the Andymarks, then using the summer to figure out what makes them tick, and what you could to do improve them for when you make your own.

Good luck,

-Cory

Seriously, what poor play from a team with experience with shifting transmissions to just go buy an off the shelf component, we in the FIRST community should be supporting each other’s learning…not …. (My grandmother would be ashamed if I finished this sentence in public.)

[artdutra04]

Originally Posted by burkey_turkey

we may or may not have access to any advanced machines to build this thing, just a nice drill press and a band saw, but dont worry, we are commited to making this happen.

Drill presses and band saws don't have the accuracy - to within the thousandth of an inch or so - required to build transmissions.

Unless you can find a machine shop, you won't be able to build your own.

Though if you can find a machine shop that will be able to produce the necessary parts, you should read these White Papers for more information on various kinds of transmissions. The information contained in these will help explain the various options available and the engineering behind a custom gearbox.

http://www.chiefdelphi.com/media/pap...s/transmission

[caraddicted101]

In regards to choosing what type of transmission i stand by dog gear shifters and more specifically the andy mark ones. My team ran with the andy mark 2 speed found here: http://andymark.biz/am-0076.html



We started using them on a prototype chassis before the season and all we had to do to them was put some white grease on them. They are built with 4 mounting holes on both the top and the bottom. We did take them apart a few times to change the spacing on the output gears for the chain, and when we did that, I was very happy with how easy they were to work with. They are a bit expensive, but i feel that they are worth every penny. It's American made too, can't beat that, haha. But really, Andy Baker and Mark Koors know what they are doing and they make great stuff.

If you guys do not choose to use them for this season i would highly suggest that you walk around at comp and look for robots that are running them, and ask them what they think about the transmissions.

- Hope this helps

[Andy Baker]

This is exactly the reason why we sell individual components of our Shifter transmissions. Teams can buy the hard-to-machine parts from us, with a significantly lower total price, after they design their custom transmission.

For instance, the Dog Gear, Large Output Gear, and Small Output Gear are very difficult to machine precisely. If I were to design a custom shifter, I would start with these three parts as COTS items and then design custom from there.

Good luck with your transmission.

Sincerely,
Andy Baker

[ewrado]

Originally Posted by burkey_turkey

Hey all, our team just voted today to try out using a 2 speed transmission. While we always have the option of buying one from andymark, we thought it would be neat to try and design one ourselves. I've been looking around chief delphi all night at different desings, and i was hoping for some more input. I understand the basic concepts behind the duel speed transmissions, but i dont understand all the pros and cons between the differet types. This is what ive got so far:

Mash:
easier to build, but the teeth of the gears are more likely to break during transfer - how can you solve this problem?

Dog:
less likely to break, harder to manufacture. we may or may not have access to any advanced machines to build this thing, just a nice drill press and a band saw, but dont worry, we are commited to making this happen.

i know there are more methods, but for only 2 speeds, they didnt seem very good. if im wrong, please enlighten me

thanks for all the input i know the CD community will give.

its realy good you can go slow on turns and then go fast when you need to

[burkey_turkey]

Originally Posted by Fe_Will

General points:
1. Use the same pitch and pressure angle of gears through out (this prevents the chance of to dissimilar gears contacting)
2. The key is proper mesh to transfer motion, too much or little mesh can cause gear damage or inefficient transfer.
3. Avoid pairing gears made of different materials; the harder gear will wear out the softer more rapidly than two gears of the same material.

About "Mash" transmissions:
Yes, they require less machining but the act of moving a cluster of gears (or an entire shaft) linearly is the greatest challenge in my experience. If you were planning to servo shift then “mashing” is not for you… the force required is too great for the servo to do the job. Also, there are several older threads (2003-04 vintage) that discuss the trade offs of beveling the gears vs. a right angled gear.

About Dog shifting:
The dog and its corresponding surfaces are the most difficult parts of this transmission to craft. The dog and mating gears should be made of a hard material (they will take the most abuse). Several designs that I have seen were shifted by servos and just as many that were shifted by pneumatics.

In conclusion:
It’s a trade-off between servo shift and difficult machining vs. less machining and a tricky linear pneumatic shift, IMHO.

Good luck.



On a side note:


Seriously, what poor play from a team with experience with shifting transmissions to just go buy an off the shelf component, we in the FIRST community should be supporting each other’s learning…not …. (My grandmother would be ashamed if I finished this sentence in public.)

Thanks so much, thats an excellent conglomeration of information, and you;re right we're trying to build our own to learn, not just to have a two speed on the robot.

Originally Posted by Andy Baker

This is exactly the reason why we sell individual components of our Shifter transmissions. Teams can buy the hard-to-machine parts from us, with a significantly lower total price, after they design their custom transmission.

For instance, the Dog Gear, Large Output Gear, and Small Output Gear are very difficult to machine precisely. If I were to design a custom shifter, I would start with these three parts as COTS items and then design custom from there.

Good luck with your transmission.

Sincerely,
Andy Baker

This is something i hadn;t even thought about, but will definitely bring up at the next meeting. I like this option because its gives us flexibility in our own design, but we aren;t constrained by our equipments potential. thanks!

Originally Posted by ewrado

its realy good you can go slow on turns and then go fast when you need to

Precisely. this is why we think this year is the best year to try one out, because that ability we believe will be key.



To everyone who responded, thanks so much for your input, you did make a differnece. expect to see 422 with a shiny 2 speed this year! We'll be at VCU and nat's.

[Cory]

Originally Posted by Fe_Will

On a side note:


Seriously, what poor play from a team with experience with shifting transmissions to just go buy an off the shelf component, we in the FIRST community should be supporting each other’s learning…not …. (My grandmother would be ashamed if I finished this sentence in public.)

I think this is a pretty short sighted ignorant response.

How is it better for me to tell him to make something that I don't believe his team has the tools to effectively fabricate?

In real life, if a COTS item will give 100% of the performance of a custom part, and meet all other required parameters, you're going to use the COTS part. That is a big part of engineering, and a valid strategy for FRC teams to use. This is sometimes just as useful for teaching the kids as it is to make something of your own.

I've been in a situation where my team bit off more than we could chew, and ended up totally sucking. It wasn't inspiring to anyone. If a team does not have the resources to machine a precise shifting drivetrain, they'd be much better served using a great product like what AndyMark offers, and focusing their efforts on making a good scoring mechanism.

[burkey_turkey]

Originally Posted by Cory

In real life, if a COTS item will give 100% of the performance of a custom part, and meet all other required parameters, you're going to use the COTS part. That is a big part of engineering, and a valid strategy for FRC teams to use. This is sometimes just as useful for teaching the kids as it is to make something of your own.

It seems to me like your argument is only correct if you assume that the primary goal of FRC is to create the best robot possible. In 'real life' you would buy the part to make the best robot possible, but in FIRST it is about the knowledge gained from the experiance. In FIRST you can screw up and you can try and bite off more than you can chew because its perfectly OK to learn from your mistakes. IN the real world you need to get it right the first time. Last year we did omni drive for the first time ever, and in the end that was a bad choice, but we all learned alot form the experiance that a two wheel two skid tank drive wouldn't have taught us. You learn so much more by designing and implementing something yourself than you do by entering a credit card number.

[Cory]

Originally Posted by burkey_turkey

It seems to me like your argument is only correct if you assume that the primary goal of FRC is to create the best robot possible. In 'real life' you would buy the part to make the best robot possible, but in FIRST it is about the knowledge gained from the experiance. In FIRST you can screw up and you can try and bite off more than you can chew because its perfectly OK to learn from your mistakes. IN the real world you need to get it right the first time. Last year we did omni drive for the first time ever, and in the end that was a bad choice, but we all learned alot form the experiance that a two wheel two skid tank drive wouldn't have taught us. You learn so much more by designing and implementing something yourself than you do by entering a credit card number.

And you totally ignored my point, which is that if you don't have to design something from scratch (and in your team's case, there is no way to effectively construct a shifting transmission with only a drill press and bandsaw), your effort is better spent elsewhere, like on an arm or lift, or whatever your scoring mechanism will be. You can learn plenty by using off the shelf components that suit your needs. As I said in my original post, if you don't know how to do something, the best time to figure out how is the offseason, not during the 6 weeks.

[kevin.li.rit]

In the past my team has substituted a shifting dog with a disc with 4 3/8" holes drilled into it all along the same radius. And the same holes drilled in the large and small gears. We pressed in pins into the wholes for the shifter. I think this is easier to machine than a regular dog.