I was going to use the same receiver that I had in the past, but I couldn’t get it to fit in the design (it is an 8-channel Futaba FP-R148DP). Instead, their is now a smaller 3 channel PCM receiver that I was able to fit. I’ll be able to use my existing crystals. My primary channel is 73, with 74 as a backup.

This is a R113iP 3-Channel 75MHz Micro PCM1024 Receiver “Without Crystal”.

FEATURES: Narrow-band spacing and superior signal filtering combined to provide exceptionally clear, interference-free reception and amazingly smooth, accurate control. Designed for modelers who want to use a single radio transmitter to run several different models.

• Narrow Band: 20kHz Intermediate
• Frequency: 455kHz
• Power Requirements: 4.8 to 6.0 volts (shared with servos)
• Current Drain: 18mA
• Length: 1.69″ (42.7mm)
• Width: 1.13″ (28.7mm)
• Height: 0.63″ (16.0mm)
• Weight: 0.74oz (21g)

Description: Futaba FP-R113iP
Vendor: Tower Hobbies
Part Number: LXAKL2

In RS1, I used four Dewalt 24v hammer-drill motors and ran them at 24v. They offered good power, but were very difficult to mount. The gearbox does not securely attach to the motor, they instead rely on the case to hold them together. Though I thought I had them securely mounted (a number of times), they still came loose, which tends to crack the plastic gearbox casing and allows them to spin, reducing the output.

In RS2, I will be using Dewalt 18v hammer-drill motors. I will power them at 21.6v (they have been used up to 24v without problem). I was considering using the older model motor and gearbox available at Team Delta. I then noticed that my top speed would only be 16mph. I need more speed. Dewalt makes a new motor combo, that should give me approximately 20mph. They don’t secure together as well as the older model, so I will need to devise a foolproof mounting method.

One problem with using drill motors is that they are timed to run faster ‘forward’ than they are ‘backward’. When arranged in my design, one side of the drive will be going forward, the other side backward, which means they are running at different speeds. I will ‘retime’ all four motors by rotating the magnets inside the case so they are perpendicular to the brushes, making the forward and backward speeds the same.

I will be running the gearbox in ‘MAX high’ to get the speed I want at the wheels. At 18v, the output at the gearbox is 2000 RPM.

I am using Victor 883 speed controllers. With the exception of fan mounts cracking, they have worked flawlessly so far. I will use two controllers to power my 4 18v Dewalt motors at 21.6 volts. This configuration has been used before, namely by Wedge of Doom. I will reinforce the fan mounting to prevent failure.

Here is info from the IFI Robotics website:

The 24V Victor 883 is a speed controller specifically engineered for robotic applications. The high current capacity, low voltage drop, and peak surge capacity make the 24V Victor 883 ideal for drive systems while its braking options and precise control meet the demanding needs of arms and lift systems. This controller safely handles the high continuous current draws and extreme current surges produced by robots. Innovative FET switching architecture and an integral cooling fan ensures cool FET junction temperatures. The low voltage drop and high switching speed ensures the motor receives maximum power, providing significant improvements in acceleration, direction changes, and lifting torque.

24V Victor 883 PWM Signal Information

The Victor 883 (for use with standard R/C receivers) comes with a PWM Signal Driver. This PWM Signal Driver ensures the signal from your receiver is compatible with the Victor. The PWM Signal Driver may not be necessary on some receivers or with some external mixers. The Victor 883 (for use with the IFI Control System) does not come with a PWM Signal Driver.

24V Victor 883 Fan Information

The Victor 883 comes with a cooling fan attached to the top of the unit. The 24V Victor can operate over a 6V–30V range, however, fans cannot operate over this wide a voltage range. When you purchase a 24V Victor 883, you must select either a 12V or 24V Fan.

Wire

I’ll be using W.S. Dean’s Wet Noodle wire. It is 12 gauge copper with silicone insulation. It is highly flexible.

• 1603 strands in 7 bundles
• Composition: High purity OFC (oxygen free copper)
• Insulation: Silicone – Ultra flexible, thus the name

Description: W.S. Dean’s Wet Noodle
Vendor: Team Delta
Part Number: RCE42

Connectors

Wire to Wire

Where two wires need to be connected, such as connecting the battery packs, I’ll be using PowerPole Connectors. They are compact, and secure well. The connectors will be both crimped, and soldered to the wire to ensure a secure connection.

• Average Contact Resistance (mated pair): 600 microhms (5/8″ #12awg)
• Current Rating (2 pair): 30amp #12awg
• Voltage Rating: 600V Wire Size: #12-16awg

Description: Powerpole Connector (Red)
Vendor: McMaster Carr
Part Number: 8026K11

Description: Powerpole Connector (Black)
Vendor: McMaster Carr
Part Number: 8026K12

Wire to Screw

I will use ring type wire connectors where a wire attaches to a screw, such as the terminals on the speed controllers. Ring type connectors are less likely to come loose than a spade type connector. The ones I’ve chosen include shrink wrap insulation to help secure the terminal to the wire, and to protect the metal, and guard against shorting.

Description: Harsh Environment Insulated Ring Terminal 12-10AWG/10 Stud
Vendor: TerminalTown.com
Part Number: HS1210-10

Wire to Blade

The connectors on the Dewalt motors are 3/16″ Quick-Disconnect Terminals, which generally are meant for 14 gauge or smaller wire. I’m using 12 gauge wire, which generally requires a .25″ wide connector, which would contact the frame of the motor. I will use the 3/16″ connector, and shave a few wires to allow it to fit into the smaller connector. Because of the tightness in my design, I will use ‘flag’ style connectors.

• tin-plated copper
• Rated to 600v
• Rated to 221° F

Description: Flag Quick-Disconnect Terminal w/Fully Insulated Double-Crimp Barrel
Vendor: McMaster Carr
Part Number: 7820K33

Switch

I’ve struggled with switches on both of the previous versions of RS. I wanted to handle all of the functions of the switch with one interface. I’ve decided to depart from that in an effort to keep the switch as simple as possible, which should lead to extra reliability.

The functions that need to be accomplished are:

• Connect and disconnect the two battery packs to the electrical system
• Separate the two battery packs for individual charging

I’ve decided to use two removable links, one near each battery pack. Removable links are required in some robot competitions, and could eventually be required in BattleBots as well. Two wires from the battery, and two from the electrical system are grouped together into a quad arrangement with the PowerPole Connectors. The link is made of 4 PowerPole connectors with small sections of 12ga Wet Noodle connecting each set of connectors, to create a pass through. When the link is removed, power cannot pass from the battery pack to the electrical system. While it is disconnected, a battery charger can be connected to the battery packs connectors for charging. The link fits snuggly, with no chance of it coming loose. This ‘switch’ is about as simple as can be, with almost no chance of failure.

Interference Suppression Capacitors

To help prevent the motors from generating RF noise that would interfere with radio reception, capacitors are used across the positive and negative leads of each motor. The caps are rated at .1uf@275v. The leads will be crimped and soldered with the wire to the terminals.

Description: Interference Suppression Caps
Vendor: Team Delta
Part Number: RCE71

As the design has evolved, a modular system has developed that will make construction and maintenance of the bot easier. The modules are:

Power Module

The Power Module (total of two) includes 18 1.2v cells (21.6v), and a removable link connector, as well as connectors to the Control Modules. The Power Modules sit in a foam lined compartment, and lift directly out of the bot without fasteners.

Control Module

Their are two control modules; Master and slave. The Master Control Module contains two Victor 883 speed controllers, a power converter, the receiver, and an inverter. The Slave Control Module contains two Victor speed controllers and connects to the Master Control Module for it’s signal information. Both modules connect to the Power Modules for their electrical source. Each module feeds two motors in the Motor Module. The modules will sit between the wheels on each side, and have a flange that prevents them from moving forward/backward, or from side to side. After disconnecting the wires from the Motor Module, and Power Modules, they simply lift directly out of the bot without fasteners.

Motor Module

The motor module contains 4 Dewalt 18v motors and gearboxes. It is constructed to hold these components in the correct relative positions, and restrain the parts from any movement. With the spindles removed, the module will lift directly out of the bot without fasteners.

Drive Train

The drive train includes four spindles with belt sprockets attached. These spindles release from the body with four retaining rings, which allow them to slide from their connection to the Motor Module (Control Module must be removed first). The wheels can be removed by loosining two setscrews on each sprocket, and by removing the 5 screws holding the wheel to the sprocket. The axle will then slide out allowing removal of the wheel, sprocket, and belt.

Body

This bot uses a unibody comprised of 7 pieces welded together, and a lid that screws on. Because of the modular nature of the components, everything is easily removed from the bot for servicing, negating the problems of working in such a tightly packed layout.

My general design philosophy for RS2 is similar to RS1:

• small: less of a target for the opponent
• tight: the smaller the package, the less weight used on structure and armor
• tough: components must not fail in combat, period

Layout

RS1 was designed as a two wheel bot with a rudder (the rear two wheels). The wheels were offset, moving the pivot point closer to the front of the bot. This proved to be difficult to steer. RS2 will utilize a standard 4 wheel based skid steering system, with the wheels as close to square as the design can allow.

The size of RS2 will be reduced, namely, the height. Providing a smaller target will thwart some of the opponents weapons, and reduce weight in structure, possibly allowing the use of tougher materials.

All of the wheels will be moved inside the structure. This will reduce the chance of wheels being attacked from the side. RS1 intended to protect the front wheels, but this had to be eliminated due to excessive weight.

Components

One of my first decisions was to ditch the 24v Dewalt motors in favor of the 18v version. This was done for a number of reasons. First, reliability. As hard as I tried to mount these in my design, I couldn’t get them secured reliably, resulting in the gearboxes spinning, and self destructing. Second, I found the torque that I originally sought wasn’t gaining the results I desired. Third, the 18v are smaller, but when run at 24v (which I’m limited to by using the Victor 883 speed controllers), they are as powerful as the 24v version.

Since making this decision, a few things have changed. First, I realized that with the motor and wheel combination that I would be using, my top speed was approximately 16mph. This is slower than I was hoping for. I found that Dewalt had come out with a new version of the 18v combo that provides 2000rpm in it’s highest gear. This would take me near 20mph when powered at 21.6v. The downside is that the new combo is as hard to mount as the original 24v combo. I’ll be devoting extra attention to this mounting, as it is now apparent how important it can be.

Another decision was to ditch the home made battery packs based on Dewalt XR+ battery packs (2.4ah cells). Because the design will change substantially, the old packs will not fit, and their is newer technology available, so it’s time for a change. I will instead use BattlePacks, based on the new 3.6ah cells. This will provide a battle tested package, with far more efficient cells.

I will continue to use the Victor 883 speed controllers. They are smaller than other alternatives, and have proved reliable, except for the durability of the fan mounts. The new design will reinforce the fans so they do not break off due to shear.

Structurally, I will stick with my choice of steel angle iron as the attack weapon, but will change it from Cold Roll Steel to Stainless Steel. At 3/16″ thickness, I don’t believe any lightweights can significantly damage it (Ziggo and YU812 would be the ultimate tests of this). I will switch from 3/8″ aluminum to 3/8″ Stainless Steel for the main structural members. They are thick enough to be stiff, and allow tapping for mounting of the armor. The armor will be .06 Stainless Steel. I will have all of it cut with a waterjet to save my tools, and make them much more precise than I was able to achieve in the past..

Construction

The spindles, and their connection to the sprockets, have haunted RS1. RS2 will hopefully solve this. I’ll be using spindles machined by Team Delta. They have done strength tests that show them to have a 120% safety margin before failure. Mounting the sprockets will be done through a keyway and set screws.

The connection between the structural members and the angle iron must be improved. I will be welding the structural members to the angle, which will provide superior strength. I will also weld the bottom armor to the structural members and the angle, providing a very stiff frame. The top armor will then be screwed on with alloy screws, tapped into the structural members.

The design of RS2 (that’s the short name for Ramming Speed 2 for those of you just getting out of bed) is evolving quickly. I’ve got it close enough to final that I can start ordering some of the parts. Since I’m still paying off the first two, I don’t want to extend the plastic any further. I made a deal with myself that any money I can make selling stuff on e-bay can go towards RS2. I sold 109 batteries from the first two for $250, about half what I originally spent on batteries. I’m using that, and some money from a motorcycle helmet to buy the new batteries. As the design gets closer, I’ll begin posting more images of the model.

The robot I brought to San Francisco in November 2001 was a slight redesign on the original which fought in May of 2001. It was smaller, sleeker, and slightly better designed. After fighting in November, the short comings of the design were apparent. An observer commenting on fights noted that “Ramming Speed is surprisingly slow, and more a pusher than a rammer”. Though both of my losses could probably have been overcome by better driving skills, their are still underlying physical changes that must happen to improve the design.

Time for Ramming Speed 2. The goals for RS2 are:

1. Increase the drivability: The original design was based on a two wheel bot, and used the narrower rear wheels mainly for stabilization. I found steering to be less predictable than a more squared off 4 wheel design. RS2 will be a true 4 wheel design, with the wheels oriented very close to a square, to allow for predictable skid style steering.
2. Increase the reliability: The original design utilized Dewalt 24v Hammerdrill motors for power. They were chosen based on their high torque, which I hoped to use to win pushing fights. These motors are hard to mount, and failures in this area resulted in both SF trips. I also found that my ability to out push others was not significant. I am switching to Dewalt 18v Hammerdrill motors, which have been proven in combat by many. They are smaller, lighter, and most importantly, the components secure to each other well, making them easier to mount securely.
3. Increase the potential speed: The original design used the drill motors in low-gear (for a focus on torque), with sprockets gearing them up slightly for a potential high speed of 11mph (17 feet per second). This is fine for a pusher, but not a rammer. The new design will switch to high-gear, but with smaller wheels (both for stealthness, and to reach a more reasonable final speed). The potential final speed will be 16 mph, better enabling me to duck and RAM.

I’ve been working on a new version of Ramming Speed, which I’m referring to as Ramming Speed 2. It will be faster, smaller, and tougher.

Here are descriptions of my two fights at the November event, as reported by spectator Chris McVey:
Ramming Speed (won) vs. 401k

Ramming speed was not real fast, either. In the beginning, 401k could only circle, and Ramming Speed came over and pushed him. That must have shaken something loose, he regained control afterwards. It was a basic pushing match then; Ramming Speed had an advantage but not a clear one. He was, strangely enough, not very fast. Maybe it was just an illusion because he could not hit very hard on the soft tire-body of 401k, but he seemed not to be a rammer at all, more of a pusher. He looked slow.

Ramming Speed vs. Firestorm (won)

Ramming speed took a run at Firestorm, and went right over him. You can’t ram what you slip right over like a road dot! This wasn’t like Buddy Lee and TRK; in that fight Buddy Lee Was able to hit TRK before he went over him. Ramming Speed actually couldn’t hit at all. He was easily wedged under and lifted up, and slammed into walls. At one point Ramming Speed got under Firestorm by hitting him from the side, and pushed him around in circles, but it didn’t last long. As soon as Firestorm got free he slammed Ramming Speed into a wall again. The bots circled and pushed each other as the match ended.

It’s been a while. I’ve been working on a redesign of Ramming Speed. Why? Well, I think it’s late enough in the TV showings to disclose that I won my first fight, but lost the second. In both fights, I was disappointed in the drivability, the speed, and the power. I’m going to simplify the design some, going with proven solutions. Note that I won’t start building till I pay off the first two, so I may not make the announced May date, but will try to attend as a spectator.