Barróg - Featherweight Lifter Build Diary

[RogueTwoRobots]

On point number 2, the easiest way would just be to use a servo to turn on a couple of microswitches, one for each direction. It allows you to control forward and back from one radio channel and the stock actuator motor will have a low enough current draw to not require any additional relays etc. Here's a setup in its simplest form; I usually tend to just hot glue the switches onto the sides of the servo and wrap a cable tie around them, but you may prefer a more robust/maintainable/aesthetically pleasing alternative

DSC_2215.jpg

[Shooty]

Thanks for all of the advice, especially the diagram. I've ordered some m5 barrel nuts, and will see if I can source the steel locally.

[Shooty]

Right, so quick question. I'm planning on getting the Banebots T81 3 7/8 wheels, and I see that you need a custom hub for them as opposed to a simple hexnut. The only thing that has me worried is that the shafts are not threaded, whereas the drill gearbox is. Would I be able to use a 3/8 T81 hub on a 3/8 threaded drill shaft, or should I look for another solution, such as different wheels?

[Maxamuslead]

you could use those hubs and tap them to use the tread, but is there any reason you for 3 3/8 wheels, as you can get the 2-7/8 wheels in the t40 format which you can use with standard 3/8 unf JOINER nut, also you can get them in the uk http://www.robotshop.com/uk/rb-ban-94-orange-wheel.html.

[Shooty]

Cheers for the link to the UK store. The main reason I was going for the 3 3/8th is because the mounts I'll be using combined with the 10mm HDPE base puts the centre of the shaft 37mm off the ground, which leaves the 2 7/8ths just too small to make contact with the ground (radius is approx 36.5mm). I would prefer a hex mount, but it seems that Banebots don't do any hex wheels above 2 7/8. Are there any other options for someone who doesn't have the equipment to make some?

[Shooty]

Huzzah, visible progress! So over the weekend I was finally able to start cutting some of the HDPE strips, and have made progress on the sides, back plate, base, and lifter face. I still have to round out the back of the side plates, and see if I can shrink the design further, but it's something.









I have also finally been able to order speed controllers for the damn thing, and will hopefully get the wheels soon, which should be the final part to get. I'm still a little iffy on the actuator controller though. Microswitches seem compact, but it doesn't look like I could stop a linear actuator halfway with them, which may be a worry If I'm say, clamping down on another robot, and am a little worried that I might do damage to something inside the machine. a speed controller seems like it would give me that functionality, and if I were to go down the ESC route, would the Botbitz 30A controllers be a bit overkill? I'm asking because they are what I'm using for the drive, and since I'm only just ordering them today, it would be convenient to send everything in the one package. If they are overkill, could I find another speed controller that could take the 18.8V outputted by the battery?

[R9000]

You're using 30A for the drive? Is that four individual controllers? As for the actuator, I don't know what kind of current it's pulling, but it shouldn't be a huge amount. I should think a 30A would be ok, but there might be a better option. If you have a ammeter with a decent capacity, you could try putting the actuator under load and seeing how much current it draws straight from the battery. Otherwise Jamie's point here: http://www.fightingrobots.co.uk/thre...l=1#post468452 is an excellent solution, and would allow you to stop the actuator halfway.

Also you weren't kidding. This machine is immense.

[Shooty]

Yeah, it certainly is big. It's a side effect of the mounting of the actuator, if I ever make a Barróg 2, then it will almost certainly use a high torque motor system instead. I'm hopeful I can reduce the weight enough though while also keeping it at least somewhat sturdy. ALso I'm planning to run each sides motors in parallel, so an ESC will be shared on each side.

I'm not completely certain that my readings were accurate, but I measured the current draw of the LA with a multimeter, and I found it drew roughly 1.7-1.8A when in use, bursting to roughly 2.5-3A at the start of the connection. That sounds right to me considering that the no load current of the LA is .85A, and it's being overvolted. It certainly seems 30A is a little excessive: the issue for me though is if a brushed speed controller exists that A) is more cost efficient then the Botbitz ESC B) can take the 18.8V provided by the LA and C) can still take the current drawn by the LA.

If Jamie's solution can do the job though, then that'd be excellent, I have a few servos left over from my antweights that would do the job.

[Shooty]

So quick update, brackets for LA have been mounted, and motors are (slowly) being fitted. Progress has been slow due to guesswork of how exactly to go about certain things, but many of the construction methods have been figured out by now. One of the problems I faced was how to drill a completely vertical hole through the sides of the HDPE so as to mount the barrel nuts. An issue made harder by the fact that I don't have a drill press. Fortunately, we've been able to MacGyver up a solution. In terms of Getting the hole drilled vertically, I got two angle brackets, fastened them together to a piece of wood, and am able to use the resulting shared hole to drill vertically. and using some spare wood and poly, I've gotten a spacer that puts the center of the hole at 5mm away from the spacer: just perfect for drilling into 10mm HDPE.





As for locating the position for the barrel nut, my old man came up with something pretty clever. By sticking a drill bit into the hole, and shining a torch underneath the plastic, you're able to see the drill bit inside, and thus locate the center of the hole (sorry for poor quality of the pic, my phone is rubbish with contrast, I promise I'm not building the robot in a cave).



Chances of getting the thing ready for Insomnia are very slim, but I am finding the construction to be moving at a far faster rate than I was expecting, and I should have accommodation should I get the machine done and get selected. If not late August, then my next goal after that is to get it in a battle before the end of 2017.

[Shooty]

Nearly a month after my last update, and I can safely say that I have been sleeping just fine recently, because Insomnia is not in my near future. Progress on Barróg was slowed down by my family going on a 2 week holiday to a small island off the west coast in early August, but since then construction has been going forward at a steady pace.



The sides and back have all been fastened together with barrel nuts, the motor mounts have all been placed, the LA can be securely mounted to the base, and the lifter brackets are in position. I'm also testing out how the lifter will rotate along the brackets, and while I've been getting away with low tech solutions so far, I'm seriously considering getting a drill press to deal with these bits of 20mm HDPE. I've even been able to shorten the machine by a few cm, and that should help with the weight loss.



I also got the curves at the back done, which I was wondering for a while just how I would get a clean across. As it happened, we had a frying pan that, when offset a little as shown in the pic below, gave us the perfect arc!



I also got the servo/microswitch setup that Jamie suggested working, and am going to see if I can 3D print a mount to keep everything in place. I'm still learning how to channel mix the ESC's, and will hopefully have that down in the near future (side note: first time using drill motors as drive, damn these things pack a serious punch when brought to full power!)

[daveimi]

Your robot is awesome. Nice to see another builder having a go with basic gear like me lol.

Get the wattage of your actuator motor and divide it by the rated voltage. That will give you the Amps it will pull at full chat as:

wattage = volts x amps

[Shooty]

Update time! Progress has slowed down since I moved across the country for college at the start of the month, but we were still able to get some good work done on the machine. We were finally able to place the hinges for many of the important parts, including the base/lifter hinge, and the linac/lifter hinge, so now we know exactly how much Barróg can lift (over 20kg at the tip, which is enough), as well as how high it can go. It seems a little overboard, but it still has as much power as it needs, will be able to self-right, and just looks over kill, but in a fun way.



the other main goal that was accomplished was the shape of the lifter supports/spines, which I am thrilled with how they've come out. Obviously a lot will be trimmed from the bottom of them, and holes may possibly be drilled into the spines in the future to lose weight, but overall I'm thrilled to see one of the trickiest parts of the build just come alive. Shown here in a picture alongside the plywood test piece.



The trickiest part of making the supports was to get a straight cut along the lifter side of them, so as to bolt the front plate onto them for support. That would usually be simple, but unfortunately our skill saw gave up the ghost a while back, so with no other easy alternative, we finally shelled out on a CNC machine to do the job properly!



...just kidding, instead we bolted a straight sheet of plywood to a T square (now just an F square), and used that as a guide for our jig saw. The cut ended up straight, if a little at an angle.



My target for the completion of this bot is before the end of October, mainly so that I don't have to wait until April to see it in action. I would love to be able to use it at Mechatrons in January, especially as that seems to be the only Irish robot combat event at all, but given that they've only had heavies for the past few events, that seems unlikely. Hopefully, there will be enough feathers eager to go that they'll end up hosting them in the end, but that seems like a long shot.

[Shooty]

Well, I'm finally back, and with some actual progress to boot! Since the last time I posted (very optimistically, 2 and a half months ago), a lot has been achieved in the weekends I've actually been in the same county as Barróg. Possibly the most important thing achieved has been the electronics, which have now been properly wired up, and allow the machine to drive, as showcased down below.

https://www.youtube.com/watch?v=o2JbVjo3vMA

Here is a closer look at all of the wiring. Obviously, everything is still a little loose, but will be tied down in due time. In theory in its current state it can drive and operate the linear actuator, but unfortunately we don't actually have an ESC to spare to do such a job. Hoping to order one soon though.



At long last, the wheel situation has been solved, sorta. We're using 4'' "Stealth Wheels" which have been tapped with a newly acquired drill press. I'm obviously still skeptical how well the wheels will hold up in actual combat, but for quick drive tests, they seem to hold up.

Finally, we got the front face onto the lifter brackets, although it seems to be a little misaligned still. The actuator brackets are now also made of HDPE instead of wood, and they've came out quite nicely. We were hoping to fix the lifter brackets , but unfortunately I had to leave before anything else could be done. Still though, bar the "grabbers" and a bar to prevent the machine from getting stuck on its side, this is what the machine should look like, and that's pretty cool to see (at least for me).



One final shot of it driving around with the top on, just for fun:

https://gfycat.com/SneakySaltyAztecant

[Maxamuslead]

looks good. Are you going to swap out the rs550 motor on the Linac for a brushless motor (like what Rory did) to get more power and speed to it? As at the moment its looks rather slow or was that due to running a low charged battery

[Shooty]

Rory's solution would be ideal certainly, but given what would be required to set it up, it's just too low on my priority list for the near future. The test above is at full power, and while it is slow, I've tried to design the grabber mechanism around it, so that it will have a far shorter travel time to be effective. The only times I would need to extend it completely is either to self right, or to showboat. It would be slow, amd would certainly leave me vulnerable, but at least it means that I wouldn't be in a situation where I couldn't self right at all. A linac like Rory's is an upgrade I would love to implement, but at least for now I'm more concerned with getting the rest of the machine in a working condition at all.