Team Darklight - Broadax 2 (Featherweight Axe/Hammer)

Here we go again!

As a lot of you that attended Extreme Robots Manchester in 2018 will know, that whilst the previous iteration of Broadax potentially could have made a great brushless axe, it failed to deliver, having a number of design issues and being rushed to make the event.

As a result, I decided to redesign again, with the aim of ironing out a number of the issues.

On the hit list:

• The axe mechanism - no longer Chain/Sprocket, this time around, it's Spur gears all the way - no chains to break of go slack or out of alignment. Also dead shafts all round, to remove the need for bearings pushed into the bulkheads which makes for a much more sturdy and robust overall configuration.
• The drive mechanism - raised up substantially so that it doesnt rub the ground and beach itself. Again spur gears, removing the great but too wide planetary gearboxes used previously. Ratio change to 6:1 to give it greater top speed.
• The armour - not symmetrical any more (boo!) but designed such that it doesn't cause clearance problems and to lean the front into the floor more reliably. Also, not welded but one piece bent, and attached with anti vibration mounts.
• The electronics - switched up a gear to VESC 4.12 ESCs for drive

Since April, work has been ongoing on the CAD design for this bot. Here's a more-or-less final render:

Attachment 8220

Towards the end of May and early June, this was finalised and the whole design has now been water cut, and the front armour bent from a single piece. Assembly has started but hasn't yet finished.

Attachment 8221 Attachment 8222

Still on the do list is some brackets to attach the armour to the body, wiring, VESC configuration, and a paint job!

Gee man, Broadax V1 was already a pretty impressive upgrade over your previous bot, but this looks like a really substantial upgrade. Feather axes are always great to see, and it looks like this machine could easily rank among the top ones in the UK right now if the improvements prove to act as they should. Can't wait to see more!

Looking great and not far off finished either!

Thanks, that's very complimentary! Will have to actually finish it first!

To that end, got the VESCs configured and the drive propped up for a little test. After trying VESCs I don't want to go back to SimonK!

https://youtu.be/JhNsUQtb_p8

Great upgrades! Did you get the 1 piece armour cut and bent from Kcut?

Looks awesome! I like the brushless hammer

Here's the first axe test on a VESC. Still more tweaking to do, the robot wasn't fully charged here yet it still bounces.

https://youtu.be/GIKlLC_9QkA

Quote:

---

Originally Posted by Sam_Gad

Great upgrades! Did you get the 1 piece armour cut and bent from Kcut?

---

Yes, I used KCut for this. There was a bit of a delay on the bending but the result is pretty good

More progress made today, finished the armour mounting brackets and the drive plate attachments.

Did a quick second drive test:

https://youtu.be/9xAXIvsqEto

Looks very much like the Original version of Kashei.

Purely by coincidence! There are a number of 'inspired by' design elements of a number of different Axe bots

With KOB at Insomnia 63 fast approaching, I've been concentrating more on the actual build than keeping progress here. I have taken a lot of build photos which I'll upload here at some point, but in the meantime, here's a current photo of work to date. It's painted up and wired, and more or less ready bar a few minor bits to do.

Attachment 8350

So now back from KOB UK, here's a writeup on the event from us!

The good
- The drive
- The polyurethane cast wheels
- The axe mechanism itself

The bad
- The drive
- Hobbyking VESCs
- Losing a battery
- The link
- The current wiring for serial batteries

The ugly
- The lids

Preparing for KOB UK
Once I knew Broadax 2 was in the event, I started to make extra preparations to give it a better chance.

First of all, we cast new wheels out of 60 shore Polyurethane onto HDPE hubs. This made wheels that we extremely grippy, with a cost that they wear out very quickly. I used liquid casting rubber from MB Fibreglass (http://www.mbfg.co.uk/polyurethane-r...60-rubber.html)

Attachment 8382

I replaced the axe motor with a Racerstar 5060 BRD5060 280KV from Banggood (https://www.banggood.com/Racerstar-5...r_warehouse=CN). This was a gamble, but I think that this is a better motor than the Propdrive 5060 it was originally designed around. It also doesn't have a tapered 6mm shaft but a full 8mm shaft. I made a custom mounting plate for this out of 6082 aluminium.

Attachment 8383

We painted up the axe(s) to make it pretty. Note we did the one from Broadax 1 at the same time in preparation for when that's put back together at some point.

Attachment 8385

After we heard that there would be a total of 5 Axe bots going to KOB UK, we added extra plate to the lid, laminated with a 5mm HDPE insert, to mitigate overhead blows.

Attachment 8386

We also added a front wedgelet ground into a fine blade, which rides on the floor with the aim to make the front much closer to the ground and thus makes it much harder for things to get under us. I was a little concerned about a side effect of this, that the side clearance got raised when it was added, but we ran with it anyway.

Attachment 8387Attachment 8384


The event itself

Attachment 8377 Attachment 8378

Tech check
Broadax 2 was only passed to run at 4s, rather than the 6s/8s it was originally designed and demoed for. This was due to an circuit oversight on my part which was the lack of a second link on the Axe which meant the VESC for the Axe remained live with the link out - still safe as the receiver would not be connected or powered up so the VESC 'does nothing', but was not satisfactory to the stewards. There are diodes in the circuit to stop this, but this should also have a diode in the ESC receiver servo line, but this causes VESCs to behave erratically.

So unfortunately I had to loop out one of the batteries which boosts the voltage up from 4s to 6s/8s before feeding the Axe, and ran at 4s the whole event. As a result, the axe was still hitting far harder than Battleaxe (a previous scooter motor based axe Featherweight of ours) from last years event, but nowhere near what it is actually capable of.

System check
We managed a system check prior to kickoff, and it worked brilliantly, albeit at 4s only. The drive is FAST but controllable, and the amount of grip we had was incredible. However, the axe twitched if you went too hard on the drive throttle. I did some transmitter changes to mitigate it, but there wasn't enough time to resolve it correctly before the first fight.

Fight 1 - versus Rocket and Saber-X
Going in we weren't confident about this fight, as we thought we would just get ejected from the arena. As it turns out, Rocket stopped working, and although Saber-X tossed us about and over around a dozen or so times, it kept running.

The axe was still twitchy, but we compensated as best we could with it. A larger problem was however exposed - the lids both came off. This was to be a problem all weekend.

2 Points!

Fight 2 - versus Binky and Microswitch
Blink and you miss it - we took one hit off Microswitch and the link came out. That was it, we were immobile in the corner for the fight after that.

However, this also did more - as we found out when we got it back to the pits, both drive ESCs had been damaged, both with the #1 issue with VESC 4.12s - the DRV chip had failed on both. I attribute this down to a less than great build quality on the Hobbyking VESC 4.12

Luckily, Kevin from the Microswitch team delivered me the spare Maytech VESCs that I had bulk ordered with him, so they would have to go in Sunday morning. Comparing to the Hobbyking VESCs, which are identical in design given the open-source nature of these controllers, the Maytech ones are put together MUCH better, with better component selection and soldering work overall.

1 Point

Fight 3 - versus BFT Remix and Forge Master
Sunday morning came, and I had to resolder and rewire the drive ESCs, reprogram them, reinstall all the rest of the internals, reattach the lids, and make sure it was running again. Unfortunately, this took longer than the 3 hours I had to fix it, so had to forfeit this fight.

0 Points

Here's a pic after we had the Maytech VESCs installed, just a little too late to enter the fight
Attachment 8379

At this stage, we were out of the competition, with only 3 points scored.

Whiteboard
We had it working again for a Whiteboard with the drive fixed, the axe twitch was resolved, and so this time, it performed magnificently, winning a 5 out of 5 Judges score!

One side of the lid came off again though!

Tag Team with Apex FW
Finally, we went in for a tag team with Apex FW as our tag team partner. This is what Broadax 2 looked like pre fight. I nylon thread tape and gaffer taped the lid on - no way it was coming off this time!

Attachment 8380

I was more worried about hitting a working Apex FW by accident than I was the opponents, with only Take Cover being the only other cause for concern. However, I found I could push it around with ease, it giving me very little damage beyond scratches.

This was another win for Broadax 2 and Apex FW!

The drive fail

At the end of the whiteboard, I noted it wasn't turning correctly, but post match cease, I did a little spin for the crowd anyway. The drive was clearly not working, so once I got it out of the arena, I turned it over to see what was wrong, and noticed that the motor bolts had sheared off entirely. The drive certainly works, but perhaps a bit too well!

Attachment 8381

The motor also clobbered the back of the LiPo battery, so that had to be thrown out as it had worn all the wrap on the base of the battery away and was perilously close to being pierced through, so would be unsafe to ever charge again.

Conclusion
Overall I'm pleased with how Broadax 2 performed. We will have to resolve the issue with the lid, and remount the drive motors with larger bolts, but otherwise there's nothing fundamentally wrong. A few tweaks here and there and it's ready to run again.

The axe mechanism itself has held on for the whole event without an issue - no jamming or pinion slip problems which were initially a concern. It does need the electronics sorting for the 8s it is designed for to show it off at full capability.

It's still up for debate if Broadax 2 is entered into next years competition, or we build something new that is not-an-axe for it; given there were 5 Axe entries this year, it will have to be on top form to be accepted next time. It will still go to live shows though once repaired!

May I ask how you sure up the axe? Does it detect an end state or is the fully manually controlled?
I'm working on an axe got and how to control the weapon is a big question for me. Do I simply let it stall the motor?

Your not looks great! Very professional looking build

There's a few ways to do this, for me, the clutch is literally on the transmitter control in that I pull it to stall either direction, and it has rubber stops either end of the travel. If I were to go up in power again I would probably look at a mechanical clutch

It would be an interesting project to turn off the motor when it hits the end stop. Not sure how useful it would be though.
Electric actuators use switches and diodes to limit travel, but that will only work with DC motors.
It should be possible to have some micro switches (or similar) to detect when the axe is at its end-stop, then use a microprocessor to modify the signal from the transmitter to the axe motor to stop it moving.

This might just be another way for the robot to fail so might not be worth it, but it would be interesting.

I thought about this when I built previous robots. I think simplicity is the key, the less there is to go wrong the better generally!

Those HDPE hubs for your wheels, did you make those yourself or did you get them done else where? I'm asking as I'm thinking of making some wheels like that for my FW.

These were CAD designed by me but I didn't print or make them myself, I got them water jet cut as part of a larger parts order.

I will be making a V2 hub for these that will be 3d printed, that will be lighter and more well designed to mould the rubber around

Quote:

---

Originally Posted by qbnut

These were CAD designed by me but I didn't print or make them myself, I got them water jet cut as part of a larger parts order.

I will be making a V2 hub for these that will be 3d printed, that will be lighter and more well designed to mould the rubber around

---

Ah I see, totally forgot water jet cutting was a thing :crazy:. Thanks

Preparing for Extreme Robots Cheltenham

Attachment 8529

Many, many tweaks, changes and updates have been made to make it (hopefully) perform better than it did at KOB UK.

The lids would hit the front brackets if the axe hit a target hard enough, given the flex in the armour. Cropped the corners to provide much more flexing space

Attachment 8522

This was the motor that came loose at KOB UK. The motor is fine, the leads not so much. The red lead had frayed, so soldered to repair it rather than replacing, then resleeved all the leads

Attachment 8523


These side brackets hold the armour on, and would flex into the lid on a hard axe swing when it missed the target, causing the lids to be pushed off, the #1 issue Broadax 2 has so far had. Cut these corners off to allow for much more flexing space before they hit the lids

Attachment 8524

The first bracket iteration just wasn't up to the job. New mounts made from 10mm bar threaded as needed and bolted into the mid section should see the lids stay on properly

Attachment 8525

New rubber cast onto the wheel hubs. 60 shore has great grip but poor durability. We have a fallback to the 75 shore blue rubber castors if these wear out prematurely

Attachment 8526

3d printed parts! We've added motor guards for the drive outrunners and a new mount for the link

Attachment 8527

I have also rewired Broadax 2 from scratch. This eliminates the dual voltage setup, so Broadax 2 now runs at a fixed voltage for both drive and weapon, simplifying things. For Cheltenham it will run at 4s only, as I have run out of time to source new batteries that fit the space now available post new link holder and guards being added. Two of the original 4s 2200 MaH packs it uses, that allowed it to run at 8s for the axe, will no longer fit the new space unfortunately.

I have however now split the link so that the weapon has a separate one, as advised by EOs, as it means they can test all but the weapon in the pits, saving time.

Left is the new loom, right the one used at KOB UK

Attachment 8528

Looks rock solid.
When does the next uk kob event taking place?

Quote:

---

Originally Posted by Eidg

Looks rock solid.
When does the next uk kob event taking place?

---

It's not been announced yet, but probably August next year

Thank you for your answer.

Post Extreme Robots Cheltenham

The Cheltenham event was really good fun, and was a high for Broadax 2, followed by a low.

So in melee 1 of the Saturday session, Broadax 2 was declared the winner! This was a fantastic result, the bot ran exactly like it was supposed to, the drive was very on-point and it racked up the points! I have setup dual rates for the drive speed, at 60% and 100% setting - however, I ran this at 100% all the time given how large the arena at XR is.

https://www.youtube.com/watch?v=PY1fnsGcAco

Coming out of it though, as suspected, the cast rubber on the wheels had perished. It turns out it ate its own tread, such was the drive torque:

Attachment 8568

Back to the blue castors it started life with for the following fights

Attachment 8569

Here's where things started to go south. During melee 2 on Saturday, it was clear that the Axe was not firing every time, eventually it packed up altogether. But worse was to come, as with a hit to the rear of the bot, the rear lid bent inwards, and with it, one drive ESC bent in the same way, writing it off.

Attachment 8571

This happened in a corner of the arena that I couldn't see, and it happens off-screen on all the current videos I can find, but from what most other roboteers have told me, this was the work of Dave Westons Satanix - it might be an ageing bot, but it's still powerful!

https://www.youtube.com/watch?v=G7HEioMsVGE

Once I got the bot out of the arena to inspect the damage, it was clear that I wouldn't be going back in this weekend :( To change the Axe motor requires a 70% percent strip down, something that I simply didn't have the time to do. Whilst I currently live a short distance from Cheltenham, I really didn't want to pull a late nighter to fix it, and to do so in the pits on Sunday would mean I would miss the first melee anyway, leaving only one remaining fight for the whole event.

So at that point, I called it quits, and just became a pseudo team mate to any teams that needed a spare pair of hands.

---

This weekend, I did a teardown, so the bot is a box of bits again. This was to get the axe motor out for a closer look.

Attachment 8570

The axe motor can is basically the wrong shape and has seized solid; when the axe was hit, the can couldn't spin which meant the only way to go was to shear the shaft away from the mounting to the can.

Attachment 8572

No magnets were harmed, but the can is no longer round, so it does not spin around the stator. You can't really fix the warping, so this motor is essentially destined for the bin.

Attachment 8573

With that, Broadax 2 is now shelved until sometime next year, whilst I work on some new bots for the 2019 season. It can definitely be fixed, and may well see a live show or two again in the future.