last modified: Monday 8 February 2016 (UTC)
author: Hales
markup: textile

Rebuilding Tankbot

'Tankbot' has floated around my student group for many years. In the last couple I've been repairing and modifying it for events.

In its natural habitat:

Electronics atop the tank in the photo above:

Beneath the tank are a pair of geared motors, one per track.

Traditionally tankbot has been controlled from people's phones via the bluetooth module. This makes it great for introductory workshops and other shenanigans where we want to convert people to the dark side.

Throughout the year it gets used as a testbed (read: cannibalised) for other projects.

Wheel-motor adaptors


With the tracks off:

Small aluminium adaptors connect the motors to the plastic wheels:

Adapting wheels to motors is a massive hurdle for people starting out making robots. Everyone can enjoy wiring a motor to a battery. Gluing a motor to a body is not that hard. Attaching wheels to motors stumps everybody, because pre-made solutions are often not (easily or cheaply) available and the toolset required for making your own adapters has a cliff in its learning curve.

I'm not sure who made these adaptors, but I can take a guess at how they were made. If you look at the transition between the large and thin sections you will notice a circle cut into the aluminium:

This, along with the circular scratches around the thick end, suggests the part was turned on a lathe.

The flat sides of the thin sections have a slight 'pyramid' to their shape (especially visible on the left one). To me it looks like they were filed flat by hand.

One day at an event in the powerhouse museum one of the motors on the tank seized up -- a cog in its gearbox got into a fight and lost all but a couple of teeth. My student group didn't have any more of the old type of motor, so I retrofitted on some of our newer ones:

The new motor shafts were shorter than the old ones and did not work properly in the original adaptors. The worm screws were too far in to get a bight onto the shafts.

I butchered the adaptors to be shorter, but in so doing made the holes slightly larger. They held onto the motor shafts, but only barely, and would eventually wiggle themselves off. You can see a little of the misalignment in this photo:

Since then the wheels have been falling off every chance they can. Coupled with the fact I had cable-tied the motors onto the chassis, tankbot was left with a very unreliable drivetrain.

Making replacements

I started by cutting a 1×1x5cm block of aluminium using my band-saw:

The blue permanent marker makes scratches more visible. To mark dimensions onto the metal I scratch them on with the tips of my vernier calipers:

After a bit more careful cutting:

My band saw is a cheapie with its original blade, which I suspect is losing some of its edge. Cutting these parts heated them up very quickly, so I had to keep a bowl of water on standby.

You can see a little bit of irregularity in the thickness of the thin bit in the photo above. It's difficult to tell scale from the photo, but this section is supposed to only be 3mm thick. The unevenness is more visible from the opposite side:

I tidied this up with a little more haphazard cutting on the bandsaw.

The next few cuts were different:

This section of aluminium is much thinner and easier to bend. Importantly I can't sit it directly on the bed of my saw as it will want to fall over or get pulled down by the blade. I didn't want to risk my fingers so I decided to make a small sacrificial jig to hold it whilst it gets cut.

I started the jig by marking up a bit of spare pine. Chiselling it to shape was difficult:

I didn't have any 'wall' or edge of my bench to press it against whilst chiselling. This meant it would travel away from me every time I whacked the chisel, with very little material actually being cut away.

Time to make another tool!

Chiselling stop

Some more scrap wood, a bit of cutting & drilling, some screws and a little more time later:

I used roofing screws. These are normally used to hold down corrugated iron to wooden frames. They're nice and thick, but most of all they were on-hand.

I pre-drilled their holes into the wood and countersunk the screw heads so they would not stick up:

It's important to make sure that the screws don't bite into both pieces of wood. The top piece -- what's in the photo directly above -- should allow the screws to freely travel through, so that when you tighten the screws they clamp this piece of wood against the other. Originally I had the screws biting into both pieces. This left a gap between the two pieces of wood no-matter how tightly I screwed them together.

Special limited-time-only offer from our sponsor: Use only the best clamps.

Get whaling. Get Whale.

Back to the wheel adaptors

With the chiselling stop done I was able to quickly finish the jig. After strapping the partly-complete aluminium adaptor onto it the last few cuts were tremendously easy:

Before I could stop myself I had the adaptors hammered into the wheels.


Getting them out again would be fun.

In a similar spirit to 'I should have made this removable' I hammered the ends of the shafts over to permanently lock the wheels in place. Sucks to be the next person that has to fix these. Bring a dremel.

The second adaptor was as easy to make using the same jig and process. The part got even hotter when cutting this time around, so I'm thinking my blade did not enjoy the aluminium.

Mounting the motors

Previously a couple of cable ties held the motors in place. They stretched. They bent. The motors would move, rotate and dislocate the tracks. Nasty.

Nothing some scrap pine and a band-saw can't fix:

They are held in by roofing-screws from the top-side of the tank. I angled the screws to provide a little thrust against the motor as they are tightened. Loosening the screws allows the motors to be rotated.

Track tension is an important part of the performance of this robot. Too tight and the motors run against a lot of resistance, whilst too loose and the tracks constantly fall off. 'Too tight' is remarkable easy to achieve, partly because the plastic tracks are slightly elastic.

Adjustment to track tension is done by removing/adding extra track links and rotating the motors. The output shafts from the motor gear boxes are off-centre, so rotating a whole motor moves the driving wheel away/toward the vehicle. Once the right tension is achieved the wooden motor chocks simply need to be screwed on tight.

Ending notes

I rewired and re-mounted all of the electronics shown in the first picture on this page. I've done this a few times before and it's started to become habit.

Programming and testing the tank will have to wait a couple of weeks until I find someone with a compatible phone. A have a few variants of the code lying about, one of them ought to work :P

I need to mention the hard work put in by a particular dynamic duo of power tools:

Without these fellows making small metal parts would be a nightmare. With them it's a dream. They're not new nor expensive, and I've had to perform extensive repairs and maintenance to each, so I think it's a two-way relationship.

I tend to skip over lots of areas of what I did and how I thought about things in my write-ups. If you have any questions about anything you see here (or otherwise), from the simplest to the most crazy, please feel free to ask below. I'll happily answer.

No one has commented on this page yet.

Add your own comment:

Email (optional):
URL (optional):
Enter the word 'irrlicht' (antispam):
Do not enter anything into this box (antispam)

Comment (plaintext only):

If you provide an email address: it won't be made public. I'll only use it to contact you regarding your comment. No spam, mailing lists, etc.

If you provide a URL then your username will hyperlink to it.