Thursday, February 28, 2008

B9 Electronics: The Inside Story Part 3 - Leg, Hip & Waist

OK, fair warning now... the further we go up, the more complicated the electronics become... sort of...Again, this is what I did. There are several ways to "skin a cat" and this is one of them......
My hip section moves up and down to simulate a power failure from a power pack pull out as well as to bow to pick things up, address royalty, etc. The motor to rotate the torso is also located in the hip section. This is a departure from Mike Joyce's replica robot where the waist rotation motor is actualy inside the torso. The motors used for both the hip and waist rotation is the Dewert 24v motors. Each one is controlled by a Vantec RET 411P single motor controller. These plug in to the RC receiver. I fused each RET 411p to protect it from any overload or spike from the batteries, etc. The manual for the RET 411P is located on the Vantec web site at http://www.vantec.com/retman98d.pdf. It is a very straight forward wiring procedure where the positive and negative leads of the motor are connected to the white and gray wires of the vantec controller. The black and orange wires go to the positive and negative leads of the battery. Note where the switch goes. I will use one of the switches inside of the programming bay to control hip and waist rotation. By switching them off I could save on battery power consumption and also serve as a safety control mechanism if I wanted to limit movement of my robot.

Wednesday, February 27, 2008

B9 Electronics: The Inside Story Part 2 - Soil Sampler

For those using my version 1 drive section (for robots 300 lbs and under) and you have the B9 Creations soil sampler and are using Eric's tread sections, this is the only way to install it. There are some modifications that you will need to make to the face plates of the sampler so it will mate correctly with the drive section. Once this simple and easy modification is done it will fit perfectly. If your robot will be over 300 lbs then you will have to go with a more powerful drive section. Since the version 2 drive section has bigger and more powerful motors the existing soil sampler by Mike Joyce will not fit. I could hack up Mike's existing frame and still use his internals but I decided to design a new soil sampler because Mike no longer offers a soil sampler. I will also add some new capabilities to the soil sampler. Like actually being able to obtain a soil sampler (what a concept). This one will have 2 settings. 1 for indoor "show" use. That setting will allow the sampler tube to come out to the floor, spin and retract like the original. The 2nd setting will actually extend the sampler tube to go below the level of the floor and drill down into the ground. I thought it would be nice for the environmental robot to be capable of actually doing his real/stated job. In order to accomplish this I will be using a 24v power drill motor with clutch. The 2nd setting just extends the sampler an additional 6". There is also a small suction tube/vent in the drill tube assembly to aerosolize the dirt for analysis.

B9 Electronics: The Inside Story Part 1 - Tread Section

The electronics of a B9 can be as simple or as complicated as you want to make it. To assist me I visitied (several times) http://www.instructables.com/id/Circuit-Building-101 to boost my electronic knowledge and skills. There are a lot of sites like this one but I wouldnt let my fear or lack of knowledge hinder me in getting my robot to do what I want him to do. After all, WHO is the master here? Me or him??
Well, anyway, to facilitate this blog discussion I will be going over various systems in am or will be using in my robot. Because this is usually a touchy and potentially complicated subject I will be dividing it up into several parts. As a single point of reference the RC transmitter and receiver I am using is the Spektrum Dx7 with the AR7000 receiver with fail safes set (at binding) to zero forward velocity, forward hip set to "slump" and contact closure set to announce "Warning, Warning, remote control connection lost".
DRIVE SECTION
What makes a B9 go? Money....I mean Motors. BIG Motors! See the heading called "New Drive Section" below. Two (2) NPC Br81 and Br82 24v wheel chair motors with Two (2) 6" 65 durometer rubber wheels for traction and shock absorption. Two (2) Gel Cell 12v 33ah ("Marine" or "Deep Cycle") batteries, plenty of 10 gauge stranded wire and a Vantec motor controller RDFR23 along with a 24v DC power supply (yes, you can plug him in and drive him around with a really long extension cord if you had to).
I started by wiring the two (2) 12v batteries in series to achieve 24v. You do that by taking the positive of 1 battery and connect it to the negative pole of the OTHER battery. Once that was done, you attach wires to the unused poles of the 2 batteries and connect them to a central, resettable fuse block and distribution panel in the left tread section. [I originally had this area in the hip section but then realized how much of a PIA it would be just to reset a blown fuse (pulling the torso off and resetting the fuse would require at least 2 people probably 3, etc)].
The Vantec controller connects to the RC receiver with 2 servo like connections; 1 marked with a "T" for throttle and another with a "S" for steering. Plug those into the corresponding slots on the receiver and you are ready to learn how to drive your "bucket of bolts". [As a side note make sure you do not adjust your power curve for forward or reverse throttle. Try it out at its default setting first. You don't want your "bubble headed booby" popping wheelies or doing somersaults!] The website for the Vantec Manual is this: http://www.vantec.com/RDFR21%20Manual.pdf Take a look at that as it gives you detailed information on how to setup your controller for proper operation. I will also wire in and use one of the switches in the programing bay to turn off the motors for safety reasons and to concerse battery power.
Because the soil sampler will be in the right tread section, you want to take into account that a power feed will need to go to that tread section. Do not forget to put a fuse in line for the soil sampler. Also, the soil sampler is powered by a 12v 3 rpm motor. Since my power system is 24v I have to either draw off of 1 12v battery (which isn't recommended because it will cause an uneven drain on your primary batteries) or have some load balancer in place. The other obvious option would be to have a separate 12v battery that could be used for the bulb lights, neon and sound amplifier. I will option for a separate 12v 12ah battery. This should give me a full days capacity at shows and special events (I am only looking for 8 hrs of use). There was another consideration. I could change the motor of the soil sampler to a 24v geared Tsukasa electronics company motor. I seriously thought of that but then considered the issues with the lights in the torso and a further breakdown from 12v to various other voltages. I found a power distribution board that starts with a 12v source. It breaks down voltages for your RC receiver and other components like the 3.3v for the teeth lights, 6 volts for the belly, finger and brain lights and 5 volts for the RC receiver. Therefore I opted to stay with a separate 12v battery that I will keep in the CSS by one of the vents (so I can R&R quickly or just charge it easily).

CAD Drawings of my B9

The CAD drawings that so many of you have asked me about are rolling in now. This should allow you to have them made at your local machine shop or allow you to make them yourself. There are over 100 drawings and at $25.00 a page you can figure out how much it costs me to have them drawn up.
As soon as I have all of them I will send out an alert to the group. I think this is the first step in getting standardized drawings that fit with established club produced parts. As I have mentioned in previous club posts, these plans are for club use only. If you use them then you are agreeing that you are using them for yourself only and will not take these plans to form any commerical or financial venture without my expressed written permission. I own these designs. They are copyrighted. I am spending a lot of money having these professionally drawn for the benefit of the club. Please do not abuse this gift! I will accept donations for them but I will not demand a charge for them unless you require them to be physically sent out or mailed. There is a PayPal donation button at the top of this blog for these drawings. Let me know what you think about them and if you require something that I may not have.

Sunday, February 24, 2008

The Bubble

Without a doubt the best production bubble I have seen has been Fred Barton's. It is nice and clear but thin and easy to damage. In addition, Paul Felski's bubble from "across the pond" is also very nice! It's good to have quality vendors that allows you a choice for parts. I got to meet Paul last year at ROBOCON 2007 and had the opportunity to look at his bubble from his B9 pictures. I chose Fred's since I am state side. Someday I hope to meet Fred as I hear he is an interesting character!
Anyhow, back to lessons learned....don't bring the bubble out until you are ready to install it. If you leave it out, curious hands, house keepers and other family members will make their way to it where they find out just how light and fragile it is so take it from me ...put it away in a safe, hidden place till you are ready to install it.
When you are ready to start drilling holes in the bubble make sure you "key" the bubble. Mark the front of the bubble with some blue painters tape. Make sure you tape/mark the front of the top and bottom part of the bubble.
Another thing I learned is that it helps to use a special drill bit when making the holes for the lower lid and brain cup holes. The drill bit needs to be made for plastic drilling (yes there is such a drill bit). Drill bit #40 and #50. You can get them at McMaster-Carr.com. When I received mine from Fred, I had to open up the center hole for the neck bracket (just a little bit). Take your time with this. Mike Joyce was kind enough to offer this procedure for fitting the bubble:
Carefully mark the location where you want the 4 holes on the bottom plate (I say four because the original had four.) Drill the 4 holes in the bottom plate slowly using a standard #50 drill bit.Tape the the bottom plate to the bubble using masking tape, etc. in the exact position you want it (centered). Be sure to mark both the bubble and the bottom plate "Front". Put tape on each, etc. Match drill the 4 holes in the bubble using the bottom plate as a guide and again use the #50 bit. Remove the bottom plate and enlarge the 4 holes with a #40 drill bit. Use a counter sink bit on the bottom plate as required for #2 flat head screw to set flush. Attach the bubble using four #2 x 1/4" screws. These are McMaster Carr part number 90006A077. The original used these small screws as well.
Stay tuned for our next episode: "Electronics of the B9. The inside story"

B9 Brain, Lights and Crown

The B9 Brain was from Scott Sanderson. The basic body was copper coated to hold paint better when I received it. I got some white transparency material and sprayer the frost mist from any crafts store to serve as a diffuser for the brain lights. I also sprayed the frost in the brain eyes to give it that same look from the show. Inside I have Tom's brain light kit. His kit has clear LED's. I took some magic markers and colored the LED's for a special, different effect. The beauty of this is if I don't like it I can wipe them off and I am back to the standard B9 white lights.
I also like the LEDs because they only require 6 volts and do not produce any noticeable heat. That is very important so over time my bubble wont glaze. In addition, something not every B9'er seems to know is that the top of the brain is mirrored. So I had my lid polished to a mirror finish. I thought about chroming it but I think I got a good enough reflection to get a good effect when the crown is moving and when the upper lights are blinking. You be the judge.
The crown was from Bill Kendzierski. It is a highly polished stainless steel work of art.

Brain Cup and Finger Lights

The neck support, brain cup, finger light end caps and finger light cam are from Norm Sockwell. They are very well made. The finger light shafts are 1/4" brass tubing (see Craig's finger light drawing on the club website for exact details, length, etc.). I decided to use Tom Wisnionski's brain light kit so I have to have 2 wires coming out of each finger light rod. They can not have a common ground. The sockets I used were from Mouser. They are the best ones! If you get the Radio Shack ones they will be too long and will stick out beyond the end pieces even after you grind them down. What I did was use liquid tape to insulate the hot wire from the ground wire. I found that trying to use electrical tape or heat shrink tubing did not allow me to seat the sockets correctly and added too much weight to the end of the finger lights. I also used a silicon type glue from Home Depot to keep the sockets in place.
I know there is a simpler way to do this but I wanted the option to be able to change the blinking pattern and rate. If that isnt a concern for you you can use the finger light rod as the common ground and just have the hot wire coming out of each socket to a central power source. Make sure you use a conductive, electronic glue so the barrel will have a good, solid, non-insulating contact for the ground. Then the only thing you would need to do is get yourself some auto flashing bulbs and your done.
In the neck support I drilled out additional space for the wires that Tom's unit requires. Remember there is a 7RPM motor that sits in the neck support so I had to make additional room even after shaving the outer sheath off. I am not a fan of drilling through the Hankscraft Motors but even after doing that with my first motor there simply wasnt enough room to pass all of the wires from the brain and the finger lights through the motor. The brain cup drawing is by Craig Reinbrecht while the upper assembly drawing is by Mike Joyce.

Collar & Radar Assembly

Greg Logue makes an outstanding collar for the club. If you have a club torso I highly recommend you get his collar unless you just want to make your own. It is truly a work of art. However, I have a replica torso so my club collar is slightly taller and wider than the replica's collar. The replica torso with the replica collar is to the left. My replica torso with the club collar is to the right. Most people probably wouldn’t notice the difference unless you had the replica collar side by side.
Specifically, the original collar from the series (I would assume the replica collar is the same) had ribs that were only 3.625" tall as compared to the club collar which is 4" tall. It would be nice if the 8 people who got in on the tremendous deal from Mike Joyce months ago for the replica torso & donut could combine forces (I mean resources) to make our own version of the replica collar. If anyone knows of any extra replica collars made then please feel free to let me know. I would most definitely buy one!!
I obtained my radar from Norm Sockwell as well. He did an excellent job of making it. He even assembled it for me since I had surgery on my hand for tendon repair and reconstruction. In preparing to mount the radar for movement I needed to drill mounting holes to mount the 9" rockler bearing. The club collar holes were designed for the 12" rockler bearing. I then cut (4) 1.5" diameter acrylic tubes to the proper height for a smooth rotation of the radar assembly. I would suggest you plastic weld the tube to the radar section and then drill and tap the rods for attachment to the rockler bearing. The collar will also need to be modified to hold the servo motor that actually rotates the radar assembly. The gearing I used was the 9" internal gear with the new small cog put out by Andrew Schwartz. You will need to use this one if you are using my bubble lifter so the neck shaft will fit and function freely. Below is the gear and the servo used (a JR DS8711 Robotic Servo).

Saturday, February 23, 2008

Neon & Back Plate

One of the most sensitive area on the b9 robot is the neon and the backplate it rests upon. Until recently there hasnt been very many choices for a builder to select. There have been some from Fred Barton (See black backplate. This was patterned after the stunt robot). There have been a few that have been of a custom design like Doug Hines (See the grey backplate. He still makes them upon request but is not an approved vendor at this time) and then there were those made by Dewey. Dewey doesnt make them or sell them anymore (see the white back plate). Your only other alternative has been to make your own or modify those identified above.
Not until recently has a new back plate been made available to the group. It is the one made by David Huber. When it first came out it would only fit Dennis Wilbur's neon. So if you had a neon from Craig you were out of luck. Recently I had the oppourtunity to work with David and Richie to remedy that. The result was a neon plate that could hold a 16 or 12 row neon from Craig that was outstanding and had a new twist to it. David's neon has more prominent extrusion at the top plate that just happens to fit the TX2000's neon music synchronizer.

Below is Dave's backplate for Craig's 16 row neon before and after. This plate fits the club standard torso as well as the replica torso. I used a black silicon to adhere the neon to the back plate. One additional note - make sure you trim the backplate as high as possible for the dial lights to pass through. If you dont the excess flashing will prevent the teeth light buttons from being depressed.

And here you can see how it looks with and without the neon installed.

Friday, February 22, 2008

Central Support System

The Central Support System (CSS) of my robot is a fairly straight forward design. It has several circular plates spaced as necessary to accomodate various electronic boards, speakers, etc. This CSS design compresses the torso from bottom to top. A CSS of this design is absolutely essential if you are to have a bubble lifter and are mounting a 4 DOF arm or greater. What you see before you is the prototype for a more sophiscticated CSS that will handle 2 5DOF arms, bubble lifter, 3 BPSSolutions RC Macro boards, multiple brain & chest light boards from Tom Wisnionski. Later on I will get into the electronics of the B9.The black plate of this CSS has the matching waist gear ring on the bottom side of the plate.

Waist Plate, Donut and support rings

Building a fully mobile robot has some challenges that static robot builders do not usually have to deal with. For example, the donut (a fiberglass ring that sits between the torso and the waist plate) has to be able to withstand the impact of bumps and jolts of a greater magnitude and for longer periods of time than a stationary robot or it will split or crack. In addition, my donut is connected to the torso and will disconect at the waist plate. I have designed my robot to disconnect from the lower portion of the donut to the waist plate by a twist lock method. I have seen other builders make the donut secured on the waist plate and the torso seperates from the upper part of the donut via screws from the inside of the torso via the side vents. What I have below are a few pictures of the underside of the donut. Look carefully at the grooved "twist lock slots" on the donut.

To handle this situation, James (my machinist) came up with a way to insert a large delrin ring inside the donut that would mate with the delrin twist lock plates to provide compression support to prevent the donut from being crushed. Standard aluminum spacers would not provide adequate/contiguous support for the donut. I would of made a titanium donut If weight and cost wasn't an issue.

Hip Rotation Drive unit

I have had a few inquiries on how I was able to make my torso "whip" around as Bob May did in the Lost in Space series. What I did was first ask Andy Schwartz (an approved B9 vendor) to make a supersized waist gear so I could get an increase in RPM from the 24v Dewert motor to accomplish this. Thinking of the consequences of such a design I had to ensure safety for the visitors as well as for my robot. So I had James build in a spring loaded retaining mechanism to allow the gear to "skip" or "break away" if the robot were to meet any resistance (presumably from striking any object in the robots rotational path). It is tricky getting the right tension for the "break away" mechanism to function reliably because you also have to overcome the force to actually whip the torso around while "stalling" when hitting a stationary object. Added to that; the tension relief needed to install the torso gear to the leg section can be a challenge. One way I thought of overcoming the install issue is to have a special tool made to pull the gear out of position. I will work more on that later on in the build .




New Drive Section Design

Because of the weight of my robot I had to change my power system from 12v to 24v. In doing so I had to make room for larger motors and an additional car battery. This has caused a redesign of my robot leg section which has some new engineering challenges. The soil sampler will no longer fit as designed by Mike Joyce. The AC power supplies will no longer fit in the upper leg section becasue the additional battery will need to go in the same location and So my machinist and I will have to redesign it. We will probably use a variation of the screw driver approach, but more on that later.
I changed my motors to the NPC BR81 and Br82 motors with the same rubber wheels I had on my original drive unit. The are 24V 182 RPM and can easily move 550 lbs (so long as your robot is less than 300 lbs the original drive section will work just fine).
These motors can be obtained direct from NPC or you can get them from the RobotMarketplace.com. However, modifications to these motors are necessary! Brake removed, length shortened and gear plate replaced with drive section bottom plate. It was dissapointing that I couldnt find any motors strong enough that could be installed in a plug and play fashion. Below is a picture of the batteries I use.

Don't let the tape worry you. It was just a temporary modification to ensure fitment.
Here is the final version of the new drive section & battery compartment:



This is a VERY POWERFUL drive section! High torque AND Speed!!!

Thursday, February 21, 2008

Have Bubble Lifter will travel

One of the hardest things for my mind to wrap around was the bubble lifter. I can remember trying for hours how to figure out how I could get it to move the way I wanted it to. But no matter what what I tried it just didnt seem to work right. Using the eliptical wheel with the center mounted machined post that Mike made interferred with the internal workings of the finger lights, crown and brain light mechanism and wires.
Then one day I saw an A frame upper control arm hinge in a hobby shop for a RC car and the idea popped into my head on how to control the upper and lower movement without allowing it to sway or rotate the bubble while also allowing me to use the existing motor and electronics for the upper assembly.
This is the result:
Also note the robot servo. It is a strong one! It is a GWS 777 capable of lifting 20 lbs easily. This will allow me to more accurately control the speed, height and control of the bubble and will also allow me to do pre programmed combination actions involving the bubble, radar turning and voice sync to movement for a better presentation.