ok.
I have the means to mold metal, plastics, fiberglass, carbonfiber composites and poly. But molding a continuous length of track accurately can’t be done with home equipment.
There are ways to make your own tracks though using individual links like you mentioned, or using chain to drive it.
If you’re gonna go big, why stick with rubber? It’s much easier to cast and work metal parts, especially at home. With a simple open faced mold and a mill (to remove the top layer of crystalization and drill the hole for the rivet) one could make small to medium tracks in little time.
Also I found this site backyardarmor.com/Elite%20Armor%20Tamiya%20Tank%20Accessories.htm which sells metal track upgrades for model tanks if you just want to buy. Not sure if these are big enough.
I actually always wanted to get started with aluminum casting mainly since my school throws so much out I’d never run out of material. However, this project never really got off the ground.
I’ve thought about those before. If you want the 1/16 scale metal track upgrades and metal spurs, Mato has them right here in Canada for about half the price.
matotoys.com/canada/product_ … ucts_id=88
They are only about 2" wide and 24" long. So they would work for smaller robots. I suppose you could buy them then use one of the links to make a mold. Then make a bunch more links so you have a longer track.
Or just buy 2 sets. 
They have about 3 or 4 different styles of track and sprocket for different tanks as well.
They also have the metal gear upgrade which makes the gearboxes in these tanks look pretty impressive:
matotoys.com/canada/product_ … ucts_id=86
Man somebody already thought of this too ,well i,m still going too make mine for my tracked go cart
Me thinks Jims Tracks have the advantage of ajustible size where as with the others ur basically going to have to build around ur tracks and ur significant others probly wont like the bike chain chewing up the living room floor 
Hmmm… this looks familliar. 
For the idler axle shafts I used Du-bro axles #248 which come in different lengths and diameters of course. Look on page 28 of this dubro.com/DUBRO2005CATALOG/index.htm catalog.
Alternatively you can buy some 6mm shafts from mcmaster or other metal supplier, cut them to length, and then use a die set to put threads on one end. Now use a pair of locking nuts to capture the shaft on your side panel. like NUT - PANEL - NUT ---- unthreaded portion of rod. then you can use wheel collars to hold your idler tires in position on the rod.
The difficult part about the lynx tracks is the sprockets are wide enough to support no more than about 30% of the 2" tracks and about 20% of the 3" tracks. Unless you can find a way to put them towards the middle of the track width you either need two sprockets per axle (this is what I did and makes the sprockets about as expensive as the track they support) or you need an idler arrangement to compensate for the lack of track support. The latter appears to be the way Lynx is thinking about it by using large plastic sections for the track to ride upon and is how their pictures so far seem to be illustrating assembly, including the new jonny base. custom assemblies are complicated by the requirement to make these idler sections since they need to be pretty accurately sized, smooth on the edges to reduce friction, and made of a durable material. my point is that while the lynx tracks are a cool design and their inclusion into the product line helps simplify finding mating parts to work with them, they are not without their own particular design hurdles to overcome.
The Sprockets for the Lynx tracks can be doubled up as many times as you need. If you had a 24" wide track from them, you can use a long shaft and mount 12 sprockets every 2 inches if you want. And idlers can be made out of anything, wood, aluminum, model airplane wheels, R/c Car/truck rims or slicks, etc.
If you bought 4 2" tracks and 8 sprockets, it’s still cheaper than a comparably sized track assembly… Except for there isn’t any other commonly available tracks that big.
Yes I am quite aware of how they can be stacked ad infinitum and the cost of doing so. However if you had never held track and sprockets in hand it might not occur to you how little support a single sprocket provides for the track. You can also certainly make idlers out of anything you want so long as your motors have the torque and you use big enough batteries but a low friction material and good fit will help your performance. The Lynx tracks are a good size and the price isn’t bad, it just is not necessarily as inexpensive as it might initially appear so you need to look at the whole design picture and figure out where your budget goes, if you have to work within a budget. 
Tracks systems generally are expensive compared to a 4WD equivelent system.
Not exactly… Here is a resubmit of an image of the way the tracks are utilized. The sprockets (driven or passive) are placed back to back, so they cover much more area than you wrote. I know I need to make some assembly guides for the sprockets… We are doing idlers made from nylon tubes that fit over the 1.5" standoffs. This works great for indoor use on flat level surfaces, and great for outdoor use with grass, dirt, etc. But there is no suspension so it can be pretty bumpy on some surfaces.
I don’t know that I’ve ever found that picture of the track opened up such that the details of how the motor are mounted are revealed before. You appear to have literally created an inner idler and motor support section. then it looks like you bolt this onto the main chassis where the side has another hole the diameter of the motor gearbox to allow it to pass through. this illustrates how the sprockets are not really designed to be used individually on an axle. as a pair they support 60%, 2x what I said individually, of a 2" track which when suspended in the middle as you have arranged is probably just fine. My observation from designing and building my own tracked rov using your parts was that how all this fits together has not previously been clearly explained and figuring out a way to mount the tracks without the benefit of an inner mounting assembly was not simple using simple and readily available materials and tools. I saw Joe was heading off to make something similar to what I had already worked through and thought it would be helpful to shed some light on some of the design issues I encountered. Costs… each axle is either a $12 or $14 deal putting the cost to actually implement one track (i.e. two double sprocket+hub assemblies) at close to what the track itself costs. Given the track itself is 21 pcs of a reasonably complicated double shot molded part I personally think the sprockets themselves are a bit steep pricewise, but that is just one customers opinion.
Ah, ok… The story is, the tracks are mass produced! The sprockets on the other hand are custom made by an injection molder friend of mine. Such is the life of niche market injection molding. The quantity involved really doesn’t justify the investment, but I’m the sort of guy who will do what it takes to make something happen without regard for the bottom line. Eventually I will pay for the mold! 
lol
Here is an image of the Triangular Track assembly with the tracks removed. We will be launching this version first. It shows how simple it can be setup. The motors attach to one side of the assembly, the passive sprockets and the idlers float between the two panels. 
It makes sense about the cost of the sprockets, and there really isn’t any way to make a double sprocket that would release from a mold so until your quantities ramp up (releasing the tank rov kits and the johnny kit should help draw attention) you live with the lack of the economy of scale.
The track arangement in the last johnny pic makes a lot of sense. Are you going to put a couple different mounting hole patterns for the motors or just put the hseng gearbox pattern? I am thinking specifically of the planetary drives here. It does look like there will be room for shaft encoders on the motors though so that’ll rock for odometry.
you can always make the tracks bigger across too (not sure if this is relavent to the topic 
) u can use a 1.5" tube and then put 2 of those hub things on the end and presto, it works u then just screw in the first sprocket to the motor hub, then u have the tube assembly, put the other hub on th tube, then put the other sprocket on, then u then screw it on
COUGH THREAD HACKERS!!! HMM, EXCUSE ME! lol,
OMG! I completely forgot you did the EXACT SAME DESIGN!!! 
Though it isnt a copy. The design was actually inspired by the FX ROBOTICS’ nagotiator. Dont worry, there are four major features on this robot that I didnt include because it would take too long to draw at 10:00 PM. I am worried though, I dont like the idea of copying someone, even by accident. I hope these features will make it much more original as they are very good and original Ideas I have cooked up! 
Big time sorry on the thread hack Joe… 
Heh we did kinda get OT a bit there but hopefully some good info and discussion came out of it.
Don’t worry about copying stuff as undoubtedly the construction and control details will be different. Just don’t put a mini-ITX and an SES arm mounted pan and tilt camera base on it too.
Lol. Well Im putting “spikes” on the back of the bot. BAssically SES brackets with servos that pull up and down. When the Paintball gun is fired, they come down and dig into the ground so the bot dosent slide back from the force “recoil” after the shot.
Are anchor blades really going to be necessary once you figure in the mass of the batteries, electronics, motors, driveline, and structure versus the recoil of a paintball gun? Admittedly, I’m not a paintball player, but from the few times I’ve fired a marker, I don’t seem to recall all that much ‘kick’ - certainly not enough to warrant anchoring a fully-loaded bot chassis against movement due to recoil. If it tends to rock around upon firing, maybe redesign the turret to mount the marker a bit lower, and give the recoil a shorter effective moment upon which to act.