Team Darklight - Broadax 2 (Featherweight Axe/Hammer)

[qbnut]

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:

Broadax-2.0 v94.jpg

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.

IMG_20180625_200043.jpg IMG_20180618_191105.jpg

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

[Shooty]

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!

[Al_]

Looking great and not far off finished either!

[qbnut]

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!

[Sam_Gad]

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

[Andy]

Looks awesome! I like the brushless hammer

[qbnut]

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.

[qbnut]

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

[qbnut]

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

Did a quick second drive test:

[maddox10]

Looks very much like the Original version of Kashei.

[qbnut]

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

[qbnut]

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.

IMG_20180809_193814.jpg

[qbnut]

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)

IMG_20180811_172157.jpg

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.

IMG_20180727_220641.jpg

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.

IMG_20180804_202416.jpg

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.

IMG_20180819_171349 (1).jpg

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.

IMG_20180819_190343.jpgIMG_20180821_205828.jpg


The event itself

FB_IMG_1535401820376.jpg IMG_20180825_142218.jpg

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
IMG_20180826_153755.jpg

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!

IMG_20180827_140426.jpg

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!

IMG_20180827_145225.jpg

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!

[Deathly Hallows]

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

[qbnut]

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