Wednesday, October 1, 2008

Chain Drive

With less than a week to go before we need to submit the robot for safety approval, we are still missing 2 vital components.

1) A drive mechanism to deliver power from the motor to the wheels.*
2) A frame to mount all the other components on.
*see update for details

Now, since there is so little time left, I am going to assign everyone on the team to do their best to get what we need. If you are going by a hardware store, or hobbyco, please drop in and see what they have. Everyone needs to search the internet for "drive chain supplier" or "drive chain catalog" etc, and see if they can find anything suitable. Anything that might be useful needs to be posted here, or call me if you have any questions.

Drive Mechanism
After a lot of discussion, I think that a chain drive is superior to a V-belt or gears. They are easier to adjust and more compatible with our design, but if you have reason to think otherwise then let me know. We need two sprockets and a compatible chain:
  • The diameter of the motor shaft is 13mm. Therefore, one sprocket needs to have a 13mm inner arbor.
  • The diameter of the rear axle is 15mm. Therefore, the other sprocket needs to have a 15mm inner arbor.
  • Having said that, BUY SPROCKETS WITH A SMALLER ARBOR! We can always drill it out if it is too small - too large and we have to mess around with bushings and crap. Get 10mm or even smaller if you can.
  • There is no need for anything other than a 1:1 power ratio, but if you must get sprockets with different numbers of teeth, get more teeth on the axle sprocket.
  • Our robot has about a bit more than 50mm clearance from the axle to the ground. Therefore, do not buy a sprocket with an outer diameter larger than 100mm.
  • The chain needs to be at least 200 + 200 + 2*pi*100 = 1028mm. Longer is better. Make sure that the chain is easily adjustable.
http://www.tsubaki.com.au/chain_pdf/G7%20Brochure.pdf has a section on how to choose a chain (pp 29-30). In all honesty though, it is a complex procedure so you are probably better off using your judgment.

Chassis/Frame
We need a chassis that is at least 300mm x 400mm. At this stage, if we have to make it out of wood - so be it, but we need to have something to mount all our parts on! Once again, I'm throwing this task out to everyone - think about what you have in terms of tools and materials... THEN MAKE IT!

Electronics
I am currently working on the remote controller (which I finally picked up), the speed controller, the steering and the power system (AA batteries FTW!). The only thing we don't have is a fuse box, but I'm willing to put that into the non-essentail pile.

Thanks to everyone who has helped with this robot. We are almost at the end!

-----UPDATE-----

I spent friday walking around the city, looking for sutiable stuff. I tried hobbyco, but they don't have anything strong enough for our purposes. I went to a few bike repair stores, and eventually came away with two sprockets and a matching chain.

The catch is that the sprockets will need some work to fit around our axles. I need somebody to help me out by building two bushings out of wood or something, so that we can fit them on the motor and axle.

The sprockets look something like this (I'll take a photo of the actual ones when I have my camera):

The rings that I bought have the dimensions: A = 35mm, B = 57mm.

What I need from one of you is to build something that allows us to connect this to the shaft. Something like this:

Sunday, August 17, 2008

Motor and Competition Info

THERE IS A COMPETITION THIS FRIDAY! The details are here.

On Monday, I will be in the workshop from about 9am till 4pm. I will be bringing all the parts in and putting together what I can. If you have some free time, come down and help out! (If you can't find me, call me). I will leave most of our stuff down there so that we can work on it during the week - let me know when you are available to help.

Here are the specifications for the motor so that we can fit the frame around it. I still can't figure out how to get the flywheel off, but since we need it to use the belt drive that fit's around it, I'll be leaving it on for the moment.



Finally: is "Dreadmill" a good name for a robot?

Monday, July 21, 2008

We have got to use this motor

Thanks to everyone who came on Thursday, and Friday. For those of you who weren't there, here is an update:

On Friday, we found an old treadmill that we took apart to get the bearings. We managed to get 4 heavy duty ball bearings (which I am very excited about, since it saved us a trip to the bearing shop), and some axles that fit nicely into them. There was another surprise in the treadmill as well:

A 230V 2.7hp DC motor.

Now, my first reaction was a negative one. For one thing, the motor weighs about 8kg (although a lot of this weight is from the flywheel). Also, we would need many, many battery cells to run it at 230V. And using our current steering design, we would need a second one to run the other wheel. Overall, it would be too much weight.

Having had some time to think about it, I've changed my stance. We have GOT to use this motor! 2.7hp! That's more than 2kW!! Here are some of our options:

Idea 1: We can use the motor to power our weapon. No messing around with springs or screws - we can run the entire thing electro-mechanically. We can get bursts of high voltage using a bank of capacitors: We can charge them in paralell and then discharge them in series:

This will allow us to get pulses of power at 5x our base voltage. This means we can power the flipper almost directly off the motor, making our build a lot simpler. We will still have to port the heavy motor around, but we don't have to worry about lots of batteries to keep it running constantly.

Idea 2: The above idea is a good way of sticking with our original design, but if we do a bit of a redesign we can make far more efficient use of having such a powerful motor.

First of all, we won't be able to find a matched motor to power the right side of our robot, so we will need to use this big motor to power our rear drive wheels, while using a rack and pinion system to steer. There are a few advantages to this:
  • One motor of this size is more power than we will need to move the robot (pretty much irrespective of weight). A second one to power the right wheel would be total overkill.
  • The remote control that we are getting is designed to operate servo-motors, not continuous motors. Rack and pinion steering is servo-friendly.
  • We already have this motor. Why spend money on two big matched motors, when we could just buy a single small servo to steer?
But what about getting enough power to run this beast? Here comes some maths:
  • First of all, we tested the motor and it will turn with as little as 12V.
  • With 4 x 12V cells in series, we can drum up 48V - about one quarter of the total capacity.
  • 2.7hp = 2.01kW; therefore we would be delivering 0.42 kW just to drive around. This is an enormous amount of energy, especially for such a small robot. We would be able to out push any other robot with this (for reference, our previous plan was 2 x 12V x 2A = 96W).
  • Operating at 0.42kW, for say, 20 minutes means that our batteries will have to have a total capacity of 0.14kWh - 35Wh each if we use 4 cells.
  • Assuming that each cell will operate at 12V, this means we are looking for cells with a mere 2.9 Ah rating.
Of course, we might need to use a different number or type of cell, depending on what we can find. However, these numbers are very reasonable and achievable.

Idea 3: Here is the kicker: What if we combine both of these ideas?

We are running the motor below capacity (about 1/4). If we get 4 capacitors, and charge them up as shown in part 1, we can deliver quick, high voltage bursts of 100% operation.
TWO FREAKING KILOWATTS!!
All we need to do is weld a big metal spike to the front of our bot, and we have a death machine that can stab holes in any robot.

*****************
Update: Found these.

Friday, June 27, 2008

Workshop Robot Building

Hey, since its the holidays we should arrange for some time to build our robot.

What days can people go into the workshop?

********
Update:

Before we can collect the controllers, we ALL need to sign a terms and conditions agreement and give it to Sarah. Since most people don't have any major commitments, I am going to suggest that we meet THIS THURSDAY (17th July) at the Royal for dinner (7pm). I'll print out a copy for everyone to sign, and we can have a progress meeting then too. Let me know if you can make it (facebook or email me if you have to).

********

Wednesday, May 28, 2008

Soldering as an alternative to welding

I was just thinking about robot stuff to avoid doing my System Dynamics and Control assignment.

I've been doing a bit of soldering for an electronics assignment (which I nailed, by the way), and I realised that I am already quite handy with a soldering iron - so I decided to explore the advantages of using soldering as an alternative to welding.

  • Welding is generally stronger than soldering as a method of joining metals, however a large factor in determining the strength is the quality of the join. This means that a well done soldered joint might be stronger than a poorly done weld.
  • Soldering does not require bulky or expensive equipment - allowing us to quickly perform repairs at events and things.
  • Welding can sometimes warp a part or joint because of the extreme temperatures. This can also lead to stress concentration areas.
  • The solder used in electronics is a very soft and low strength type of solder. Other solders such as plumber's solder, aluminum solder and high tin solder are much stronger. This table contains a few types, and their strengths. A particularly attractive choice was a 90% tin solder with a tensile strength of 8000psi = 55MPa.
Anyway, I thought I'd just throw the idea out there. There is no point in using welding in places where extra strength is not required, so soldering might be a good option for certain parts.

Thursday, May 15, 2008

Workshop Initiation Reminder

Hey, for those of you still reading the blog, just a reminder that workshop initiation is next Friday 23 May.

Looks like we're going to be allowed to the student workshops after all.

Monday, May 12, 2008

Metalworking

After many setbacks and trips to the hardware store, I've finally managed to cut the stock steel tubing into pieces that we can use for our frame. One of the main purposes of this blog is to document our progress so that we don't make the same mistakes again. To that end, here is a quick summary of what I did do, didn't do and should have done.

With the exception of stone, metal is probably the most difficult common material to work with, and one of the hardest things that needed to be done was to cut the tube at a reasonable 45 degree angle. The tool I had to use was a small, hand-held 350mm circular saw.

The initial plan was to cut each piece from the next at alternating angles (diagram, top) to minimize the total number of cuts. I needed to figure out a way to clamp down the stock at a fixed angle with respect to the direction of cutting. I ended up using part of a phone book placed on the edge of a table to lift the saw up to the top of the tube. Then I taped the tube to the table at the correct angle with the part to be cut hanging over the edge. I could then run the saw along the edge of the table and cut a straight line through the overhanging part. Long lengths were hard to clamp in this way, so I ended up chopping the stock into sections first (diagram, bottom). A table saw would probably have been the best tool to use for this.



- Use the right type of blade: Initially, the only circular saw blades I had were designed for cutting wood. I made a test cut to see if I could use it, but it just skipped across the surface and shredded the walls - definitely not appropriate. To cut metal, use a fine toothed blade with suitably hard edges. For thick metal, you may even have to use an abrasive blade.

- Measure twice, etc: I used a pencil to mark the places I wanted to cut, but this was hard to see against the steel. I wrapped masking tape along the lines to make it more visible. I also did a quick double check just before I cut to avoid stupid mistakes (ie, cutting on the wrong side of the tape)

- Perform a test cut first: To see how the width of your blade will affect your finished product. I also tried to leave a little extra material on the ends since it would be easy to sand down the ends a bit instead of having to cut a whole new length if it was too short.

- Use a high rotation speed and a low cutting speed: This will minimize the wear on your tool, and result in a better finish. There is really no reason not to do this, unless you are in a huge hurry.