Would connecting the 2 inner rails of a drive base with standoff triangles be much stronger and lighter than connecting them with aluminum c-channel? Off the top of my head, it seems like it would be very expensive (if buying all new 6" standoffs), but it would allow a lot of modularity with spacing. The robot wouldn’t be restricted to the standard 1/2" spacing increments, so it could have the maximum space between rails. However the possible increases in strength and decreases in weight could be mitigated by the increases in weight and decreases in strength from the collets that would be necessary to get the proper angles. Is a 6" standoff stronger (per weight) than (2) 3" standoffs coupled together, or is the difference very small? How about using all 1" standoffs coupled together? Does the 1/2" coupler provide the same strength as the 1" coupler? Would using bar to replace some standoffs be as good? Is there a reason that no one has done this kind of base before?
3" or 4" standoffs should be just as strong coupled together. To get the proper distance you need for an angle you can put a spacer between the 1" couplers. Like shown here [ATTACH]7481[/ATTACH]
This picture also shows you can double up shaft collars to increase the number of possible ways to mount it.
The 1" couplers are the ones you should use except for connecting to shaft collars where you need the .5" ones. You would also need them for standoffs less than 1" in length. The reason the 1" ones work well is that standoffs longer than 1" are not threaded all the way through and only have threads .5" deep then it is solid.
Now for standoffs you can purchase them from McMaster Carr (up to 4" ones) ( they are about half price compared to vex) you just have to be sure that they are the proper size in both length hex size and thread and also be aluminum . The thread is #8 and the hex size if I remember correctly is 1/4".
The reason I used standoffs in a lot of areas on my robot was just due to the lack of aluminum available. And the need for custom lengths that metal would just not work in.
Also I would not be so worried about making your chassis that robust. ( yes this is just a high school robotics competition and at most if it lasts 30 matches its good)Yes triangles are nice, and very structural but most of the time a rectangle works just as well and is simpler to build and attach other things to like your cortex and such. I’m not saying you should not worry about how strong and reliable your robot is, but time could be spent elsewhere.
Says the guy with cantilevered wheels.
Reasoning for cantilever wheels direct drive:
-
I find they are way more responsive than when chained or geared. ( probably due to less loss from friction.)
-
Its lighter
-
Its easier to fix.
-
Keep it simple stupid
-
It looks cool
So don’t be hating on my wheels 
to be fair, you can direct drive wheels without cantilevering:D
Easier to fix, but probably also easier to break.
I didn’t know that mcmaster had cheap standoffs, but grainger has some clearance sales on standoffs 1/4" to 2", and they’re really cheap, the quarter inch standoffs are .30 for 10, and they have some other cheap standoffs. It would be much cheaper for me to use all 1" standoffs because they are .90 for (a pack of) 10.
I’ll try out some base configurations with standoffs and see how it goes.
There’s nothing wrong with a cantilevered wheel. They’re easier to do wrong than dual supported wheels (too small bearing ratio can spell disaster), but when done right, as these appear to be, they work just fine, and add the advantage of pushing your wheels to the maximum possible dimensions of the robot footprint, increasing stability. And of course, changing wheels is a painless process with cantilevered wheels.
Yes they have pretty cheap standoffs 4" ones being $0.66 each if you purchase 10 if you couple a lot of 1" ones together not only will it be pain to assemble but the couplers will also make it heavier than using longer standoffs.
Also I would recommend grabbing some of the weird sizes like about 4 of each of these sizes .75" 1.5" and 2.5"
Also if you order from them about 500 .5" screws as this is the size I assume most people use a lot of. Because that is what you use on bearings and with nylocks.
Here is the link to the section that has standoffs ( all but the 2" ones are a good deal compared to vex) http://www.mcmaster.com/#91780A059
And last but not least since I am throwing products all around thin nylocks. These not only reduce weight but can basically replace your keps nuts. [McMaster-Carr
I have looked for other stuff like thin Teflon washers and they have a similar size that is within tolerance of vex ones but just never got around to ordering any but they are really cheap compared to vex ones.
But just remember when trying to save money buy screws elsewhere.](McMaster-Carr )
The teflon washers that they have aren’t close enough to the vex washers, however, they do have nylon spacers that are in spec McMaster-Carr . I’ve found the screws longer than 1/2" (and all set screws) cheaper on ebay, but the shorter screws are cheaper on McMaster. McMaster also has some good shoulder screws (5/32" dia shoulder up to 1" long). Grainger has a clearance sale on standoffs below 2", making them much cheaper. I think I’ll just get some .5" and .25" standoffs to go with the 1" and 3" standoffs to get those odd lengths.
Their thin lock nuts look like they would be good for lower strength applications, (while their regular lock nuts are the same thickness as the ones sold by vex). Another application for the thin lock nuts is when connecting (2) c-channels together face to face (with their flanges pointing away)(or when putting a standoff between the flanges), you can use the thin lock nuts and a 1/4" standoff to get the correct spacing (the lock nut would replace the 4.6mm spacer that is commonly used for that purpose). However, there is no 13/64 (or 7/32) nut, so the flange side to plain side that is used for outer and inner rails on a drivetrain won’t fit perfectly without some slight bending or hole widening.
I just built a trussed base (in inventor) and it is about the same weight as a 4 c-channel base. It will probably be more rigid, and it allows dimensions in increments smaller than .5, allowing you to use the full 18" width, and doesn’t interfere with things mounted to the top of the c-channel. It is only practical if you’re using the 1x5x1x* c-channels for the drivetrain. Triangulated standoffs are also useful for other support, as androidforlife mentioned. I’m using them to reinforce the towers.
The thin lock nuts are actually just as strong as the regular lock nuts, because they both meet the minimum length of engagement for an 8-32 screw, which is .1" (the thin nuts have ~.14" length of engagement).