What makes something good build quality?

no, kepler build quality is trash


Reduce friction, maximize strength.

Use this mindset in your builds. When putting a bot together, think to yourself, “is there a better way to do this that will result in

  1. Less friction?
  2. More strength?”

And as always, use the resources around you. Look at pictures of others’ robots, chat in discord servers, browse the vex forum, watch YouTube vex reveals. These sites have plenty of good inspiration for quality build quality.


Just make sure it does not fall apart.


nylocks, bearing flats, and having at least two screws secure metal together, using high strength shafts when needed, and screw joints when applicable

1 Like

this is a fairly basic concept but still something I had to learn in my early years, everything you build needs to be carefully planned before it’s built. even better, make a prototype to test before you make it officially part of your robot

also just don’t use hexhead screws, it’s extremely hard to get the right amount of torque to be able to unscrew it and they strip out very easy


I really love the spacing

First, to specifically answer your question. Good build quality is when the bot operates at the limits of the materials & design NOT the quality of assembly, the motor power is put to use moving stuff not creating friction/heat, and you don’t have to keep fixing stuff that breaks.

To achieve that, here are some guidelines.

  1. Bearings should be checked to make sure the shaft rotates without friction or feeling ‘notchy’. If you feel notches, the bearing is damaged (usually from overtightening). Check them BEFORE installation and AFTER.

  2. Shafts should be straight and free from burrs. Sandpaper or a flat file will help with this. Do NOT assume shafts are all the same dimensions. I have gotten entire PACKS of shafts that were over spec .5/100"

  3. Braces are important. However, a drivetrain that does not flex slightly is a liability as the competition floor is never 100% flat.

Note the braces for the lift. They are set at a 45 angle in BOTH axes. That’s all it took to hold a 4 stage lift in place that could reach over 3 feet high.

  1. Wire routing is important. Wires should be as short as possible. They should not rub against anything sharp. They should be tied to structural pieces to keep them tidy.

  2. There should be NO sharp spots on the robot. Period. File all your metal when cut. File the flats of the cuts. File the edges. File the corners. File/sand the ends of cut shafts.

  3. Robot take impacts. A 1x2 or 1x3 c-channel chassis should be made with the ‘flat’ facing outward. This provides more protection against bending when hit.

  4. MEASURE the spacing on all arms/etc to make sure they are parallel. Measure spacing on your chassis /etc also.

  5. Tighten all fasteners to almost tight. Then come back and do a final tightening. This keep the ‘slack’ from being pushed all to one side of your bot as if you tighten completely at first before all fasteners are in.

  6. Put code in your bot to check & display motor amps & temps. Temps between similar motors (drivetrain/lifts/etc) should be VERY similar. This also allows you to check gearing/load/etc to get the most from your motors WITHOUT overloading/overheating them.

  7. ALL shafts get at least 2 bearings supporting the shaft. No exceptions.

  8. ALL shafts get DOUBLE shaft lockers where needed. No exceptions.

  9. Shaft spacers/lockers should still have room for a fingernail between them and the metal they are against. Otherwise your friction will be too high.

  10. Rubber wheels should be cleaned periodically. This will help grip. Also, do NOT run your bot on the dirty floor.

  11. Stand offs should be tighened with a nut driver if possible. It’s very hard to get them tight using wrenches, even 2 wrenches double stacked together.

  12. TOOLS: Proper tools will 1) speed assembly 2) increase quality of assembly 3) protect/prolong life of stand offs, screw, nuts, etc.
    a) nut drivers (you need 2 sizes, get them)
    b) elec screwdriver with torx tip AND socket adapter for nuts with a torque limiting clutch
    c) medical hemostats (even better than needlenose pliers, can be bought in fishing section of wal-mart or gotten from amazon).
    d) cheap calipers can be a real time saver for checking dimension. this is also how we found why our out-of-spec axles were not fitting into anything.

A pair of hemostats will speed up assembly more than any single item you will ever buy. They can hold nut in tight corners. They can ‘fish’ screws through tight spaces into holes with a speed and precision you won’t believe.


Could you explain this one; I do build in the way described but I never thought about the direction of the channel really having an impact. Many Worlds robots that I saw had the flanges facing outward.

While redundancy is great, I have not found the need for double locking shaft collars. If you put spacers on the shaft with one collar you don’t need more.

I like your metric for measuring how tight the spacers are. The way I test this is just to see if the spaces can freely rotate on the shaft/screw.


With the edges of a ‘c’ channel facing outwards, an impact can easily bend the edge or even nick it causing a sharp spot. With the flat side out… this is not going to happen. It makes a smoother safer chassis. Also keeps your metal in great shape (no bends, nicks, etc) for future re-use.

Locking metal collars are held in place w/ a set screw. Those are famous for 1) stripping 2) fracturing 3) falling out 4) being easily overlooked when checking the bot. The consequences for a loose collar are high. Doubling up removes that problem. Have seen plenty complex bots… have not seen where double protection cannot be done.

Furthermore, when placing back to back locking collars, but sure to orient the set screws the same way so it’s easier to check them both at the same time.


This is the only one I disagree with. One of the most important aspects of a robot being good build quality is serviceability. In my many years of vex, I’ve always done it like this ] ] - [ [ , because of how serviceable the drive will be WHEN you have to make a repair. Worried about bent metal? Pair of pliers. If your metal is bending that much though, I would recommend better bracing on your drive channels.


Pickup a pair of hemostats and you have a robot’s that tougher AND easily/quickly serviceable. If you protect impact items… the need for service drops off a cliff.


] [ – ] [ is perfectly fine, there really isn’t much of a difference if you compare strength of ] [ — ] [ against [ [ — ] ]
Doing or ] ] — [ [ is absolutely cursed

only one if it is connected to a motor, otherwise only two


it’s great and all until the moment your port dies, and your perfect length cable can’t reach any other port. give them some leeway

you can alternatively use loctite, or simply by adding a washer under the screw head. never had any issues with loosening standoffs even though all i use is a hex wrench

i will be sure to try this out myself

you can literally buy good quality set screws that won’t strip from mcmaster for like 15$ per 100. the stripping is not really a big problem if you have good quality tools too. Have seen plenty of complex bots… have not seen where double protection is needed.

then maybe this is the year for you to stop doing this

while torx are indeed great, hex is perfectly fine if you have the right tools. i recommend the wiha precision screw drivers along with the stainless steel screws from vex (black allow steel have been bad from my experience). i’ve used hex for the past 2 seasons without stripping a screw

building with all nylocks is perfectly fine. this exists for a reason


Currently crying over the bracing and halfcuts :skull_and_crossbones:


While I agree that having the outside pieces flipped can have a positive impact. I personally have never had any issues with the outer rails bending even with heavy defense.

NEVER flip the inside piece. Having the inside peice flipped make it so that the motors take up alot of unnecessary space and the screw holes/mounts do not have the protection of the c-channel to prevent damage.
Ps. Probably not the best picture to show good build quality.




I can post one word comments, without underlying reasons, that do nothing to move the conversion forward also.

The pic was to illustrate the bracing on the lift, nothing more. If you’re reading stuff into it based on 1) a partial picture of 2) an incomplete bot, you are basing/extrapolating conclusions on bad data.

If you are NOT using 2 bearings on shafts into your motors you should be aware the gear cartridge is carrying the load of the bot… something it’s not designed to do. Yes it’s made of steel gears, no they are not as smooth as a nylon bearing.

You take a bot with a 100% rigid frame, hooked up to a bot with a very slight amount of chassis flex, and I’ll pull you backwards every single time.

As far as the ‘single shaft lockers are good enough’. Do you also drive around with only 1 lug nut on your car?

Yes I should have mentioned that loctite has its place… and stand offs are often one of those.

You missed the point of the way chassis ‘c’ channels face. It’s not about chassis strength. It’s about the open edge of the channel being much easier to deform in an impact than the flat side. Additionally, with the flat outside, you can put a screw in the very corner between side/front or side/back for reinforcement.

To the original discussion, I would add that where possible, each subsystem of the bot should be fastened back to the chassis as opposed to being mounted to another subsystem. Usually makes for a stronger bot. Also, it allows one system to be removed/changed without impacting others parts. Sometimes this works, sometimes space/design won’t allow it. One place it usually always works is the radio. Can usually mount the brain, transmitter, & battery all on a flat piece of metal, mounted to the chassis w/ 4 standoffs.


jokes on you, I’m in VEXU…

1 Like

You can go ahead and read the edited posts. It is fairly hard to edit posts on a phone

A person’s build quality comes from the small things that add up. If your incomplete robot have bad build quality, the final robot will be equally as bad.

Even if it is not a final build, if I see anyone with only one cross brace across the robot, base plates being used as bracing and mounting support for a 4 stage lift, drive with c channels positioned like this: , and bracing with bent, flanges cut L channel, I have strong reasons to seriously doubt the person’s ability in building. Anything that you build, even if it is incomplete, represents how good you are at building. I can easily pull up photos of my incomplete robots and all of them will still have good build quality. Here is an example:

The complete robot looks like this if anyone is curious

Maybe post a picture of your complete robot too and try to redeem yourself?

The braces you recommended are not great either. The cut flanges on the L channel means that the channel loses all its structural strength. The better option here would be to have two separate triangle braces. It will be a lot more rigid compared to having braces that takes on force in two directions.

One thing you should notice is that the shaft’s free spin is extremely sensitive to extra contacts. Having three constraints on the shaft risks one of them being slightly misaligned. Even tiny, 0.01” misalignments will completely kill the free spin on the shaft. The motors in this case acts as a constraint as well, meaning having two bearings and one motor can risk one of the component being misaligned. This is an even bigger issue because the motors tend to not perfectly center. Overall, using only one bearing will always be smoother than using two. You lose more than what you gain with two bearings.

Source: I have lots of experience with friction tuning. This comes from my experience and extensive testing.

Creating a hypothetical scenario, “without underlying reasons, that do nothing to move the conversion forwards”

I can easily say that you take a bot with a very slight amount of chassis flex, hooked up to a bot with a 100% rigid frame, and I’ll pull you backwards every single time.

Do you put 5 high strength gears to drive your lift in fear of one of them cracking? No. Do you add 5 cross braces on your drive in fear that one of them will end? No.

What you are suggesting is called overbuilding - adding extra precautions too scenario that won’t happen, resulting in unnecessary weight.

If you have the right tools and knowledge to deal with collars, it will not loosen. Any team with good build quality will not have loose collars in the first place.

You wouldn’t add extra gears to drive your lift because it won’t happen. Then why add extra collars when loose collars won’t happen? Just take the right precautions, tighten it once, and save some weight.

The difference is absolutely negligible. I have built lots of drives with open face on the outside. None of them bent. I can easily pull lots of photos of past successful robots that have open faces on the outside.

Well, don’t get me wrong here. Doing ] and [ on the outside of the chassis are both perfectly viable options. However, strength is certainly not the distinguishing factor between them.

Having the C face outside means that you can make repairs easier since the nut is right on the outside. On the other hand, having the flat side on the outside gives more space on the inside to deal with shaft spacing. They are both viable options and you can choose to your liking. The inner channel should always have the flanges pointing toward the motor however to make use of the negative space. Or in other words, the opposite of what you did.

Lastly, to add onto this thread, here are my two tips for improving your build quality

  1. The more shafts you have on your robot, the worse the robot’s overall build quality is.
  2. By not taking any advices given on forum

Good build quality is a hoax


You’re forgetting your target audience. The OP asked for good build quality. If they are asking that… experience will show that they are likely having problems with shaft collars coming off, wheels falling off, chain skipping/breaking, frame misalignment leading to massive friction, too tight shafts, wire tangling, gears skipping, and the list goes on.

And again, you’re extrapolating from a partial pic on an incomplete bot. The plates are not there for structural reasons. They protect the bottom of the 4 stage lift from impacts (good build quality) and the field from accidental gouges (responsible robot construction & good build quality).

If the L channel lost all its strength (as you say) the lift would have failed/bent/moved. Never happened all season long. No adjustments were required either. It loses some strength, yes. That’s why there are 2 braces on each side.

And the bot in question must have been pretty good as it dominated matches all year long, won skills & excellence at state, and would have gone worlds were it not for covid. It also had a primary lift chain break and didn’t suffer performance issues due to running dual chains in tandem (good build quality).

The chances of a high strength gear cracking and a metal shaft collar coming loose are about 100,000x different so equating them as similar examples is not the best comparison

I’d suggest tabling this unless we have specific items to add to the OP’s question.


this is a great application of redundancy, but couldn’t one make the argument that not using chain at all and using gears instead would’ve likely prevented the need for a redundancy in the first place?

there are good building principles that i see you’re using on that bot, but with pretty mediocre execution. for example in the picture with the lift braces, they’re attached at different positions using custom bent halfcuts. how are you supposed to guarantee that each brace is providing the same amount of structural integrity? its just inconsistent.

OP, all you need to understand is that good build quality is achieved in tandem with consistency. Do you know why teams gear together tank drivetrains even though it adds more friction? Because it guarantees even distribution of power between your powertrain, ensures that all wheels are rotating at the same speed regardless of differences in traction, and (as a consequence of the other things) makes autonomous programming much more consistent. There’s a large net benefit from doing so despite adding potential sources for more friction to your drivetrain.

The goal is to build to ensure the most consistent mechanisms while doing the bare minimum needed to achieve that. Ensuring channels are aligned properly, making sure spacing is correct and doesn’t create unnecessary friction, making sure screws are tightened properly, et cetera.

we built what I would argue is a pretty successful robot this year using a 3/32 screwdriver, a 3/32 with a ball tip, a 5/64, alan keys for each of these sizes, a wrench, a hacksaw, and a dremel and drill for polycarbonate. You don’t need super special tools to have good build quality, you just need to have a good understanding of mechanical principle. Having a basic understanding of topical physics comes a long way here.

And also, out of obligation, use screw joints when possible. hope this helps