I wanted to share an update about our robot’s drivetrain. Recently, I decided to experiment with screw joints instead of the usual axles, based on positive feedback from the forum. However, the results haven’t quite lived up to my expectations in terms of reducing friction in our chassis.
After some investigation and closely examining our robot, I’ve identified a few potential issues:
The screws I used aren’t the official Vex ones and are slightly thicker.
There might be some irregularities in the drilling of the green inserts.
The main concern, in my opinion, is the misalignment of the bearing flats. We’re using a gear ratio of 36T to 72T, which requires spreading the bearing flats across different rows of holes. To accommodate this, we opted to lower the wheels by one hole. Without low-profile bearing flats, I suspect they might be causing some drag on the screws.
I’d like to get your input on whether this theory holds water or if I might be overthinking it. Any thoughts or advice would be greatly appreciated.
“1. The screws I used aren’t the official Vex ones and are slightly thicker.”
You could test the thickness of the screws by setting up a wheel assembly as you would on the drive train in a separate c-channel sandwich, and spin test the wheel. (edit: only one wheel assemly for a spin test)
Also if you did it how my team did it, we capped the end of the screw joint with a nylock, if the nylock is tightend to much it will create friction. you want a loose connection against the c-channel, but tight enough to keep it from backing out.
That’s you problem. “Slightly thicker” means that your screws won’t fit through the free-spinning inserts, bearing flats, and shaft collars as well because those VEX parts were designed with tolerances around the thinner VEX screws.
Also, for bearing flats, in a screw joint the screws aren’t supposed to spin as much. If you’re talking about shafts, bearing flats cause some friction, but are still better than the merrry jingle caused by a metal square spinning in a metal square.
For 36:72 ratio, we shave down the bearing flats so they sit flush against the top of the c channel. Without shaving it down, the bearings will be pushed out slightly, adding friction.
I’m not sure how easily you can see it in that picture (from our CAD). We just used a belt sander, but I’m sure you could do it with something else too.
Today, my team driver and I dug into more possible issues with our drivetrain, especially since we’re dealing with overheating problems, particularly on one side.
As we looked closer, we noticed that some wheels turn more easily than others. Our first thought was that certain screws might be tighter, but it seems the screws aren’t causing the issue.
Our coach pointed out that the spacers on the bolt should spin freely, but ours don’t. Initially, we thought it might be because the bolts were too tight. However, even after loosening them, the spacers didn’t change much.
In my opinion, the spacer problem might not be with the bolts, but with the spacing being larger than the distance between the bars (which are held at the ends with L-channels), causing the spacers to tighten.
Our trainer recommended removing the 36T gears and testing until all the wheels were smooth and uniform.
I think this is a better option to optimize our time, and not have to disassemble a large part of the drivetrain.
According to what I have heard in other posts, in general, teams drill the inserts even if they are using the official screws. So with these screws, we have done the same thing.
I totally agree, but when the flat bearing does not fit because it is pressed by the upper part of the c-channel, the screw does not enter perfectly but is pushed slightly upwards, rubbing against the plastic of the flat bearing.
Oh, I hadn’t thought of that. It makes a lot of sense really, when I was assembling it I realized that it didn’t fit perfectly, but I thought that with the low-profile bearing flats, this problem would be solved. But until today I had not had the opportunity to ask my trainer to place an order for parts. The low-profile bearing flats perform the same function, don’t they?
I attach some pictures of one of the wheels of the drivetrain, to see if you can visually see any errors.
One additional thing you can do is add standoffs (with spacers/washers to fill in the extra gap) on the front and back, which works a lot better than L-channels. (Sorry for my previous post, I didn’t know about the bearing placement).
One other note, don’t fill in the space between the two c-channels compeletly (noticing those washers in pic 4). I recommend removing 1 washer from each axle/screw and see if that helps. A little bit of back and forth wobbling is negligible, and it drastically reduces friction.
Based off 2nd picture, I think I see a bearing flat there on the right and a screw going through. However, in my experience, 2.5in screws aren’t that long when combined with a bearing flat, with it instead being just barely long enough to get to a bearing flat on the other side (when the drivetrain is that wide). How long are your screws? If it’s longer than 2.5in than it may be illegal. The last thing you want is an inspector calling out something from the most imporant part of the bot, the drivetrain.
In the previous drivetrain, we used standoff, but we changed to L-channels because we saw some teams using them and thought that they were better.
Don’t worry
Thank you a lot. I was a little suspicious about the spacing of the drivetrain, and today my team designer and I were looking at our CAD and realized that the spacing of the wheel is bigger than the space of the inside of the drivetrain (the space inside of the sandwich)
This problem was bending 4 mm our c-channel, squeezing all the spacers and washers. The solution that we figured out is to change the large spacer for a smaller one (from a 0.375" to a 0.250") and remove one washer, that was your advice for us. So thank you very much for that .
Wow, I forgot about that rule. We were planning to cut them, but at level with the c-channel, but now we will have to disassemble the drivetrain to cut them to that size. The ones we are using are 8 cm, which would be about 3.1496063 inches. Thank you very much to you too, I really mean it .
Knowing all this I am going to try to solve my drivetrain problems.
Something I forgot to mention… SQUARE THE CHASSIS. If it isn’t squared there will be a lot of friction and misalignment. That could be part of your issue.
Based on the wording of R20 I would argue that just because you cut down your screw to below 2.5 in it is still illegal because that screw that you have fabricated is not commercially available. I checked for a Q&A on this and couldn’t find one, I don’t know how this has been ruled in the past either.
But in my opinion, this question doesn’t make sense, you can cut the screw to any size and I’m sure some nut is marketing that size. Besides, it would simply be enough to cut it to a more standardized size otherwise.
The truth is that it makes me very angry and frustrated to see this kind of stupid rules.
Based on my interpretation of the rule, it doesn’t prohibit cutting the screws to legal size.
The rule for part a (the one about screws) does not say, “commercially available”, rather it just specifies the size and “up to 2.5 in”
Since it does not say the screws have to be originally >=2.5 in; and seeing as there is no Q&A or other game rule prohibiting cutting screws to legal length, I would deem it legal.
Honestly, I think that r15 helps my argument, it states
“Most modifications to non-electrical components are allowed. Physical modifications, such as bending or cutting, of legal metal structure or plastic components are permitted.”
This is another reason why cutting down the screws would be illegal by the current wording, as physical modifications cannot be made on an illegal metal structure, in this case, the screws. (on that topic, there is a typo in the manual; it should be of a legal metal structure Does anyone know how to flag that for change?)
You should use bearings even with the motor. If not, the load rests on the motor’s gearbox instead of a bearing, increasing friction. The unsupported length of shaft is greater also, increasing shaft deflection/bending.
.
.