(FW)Nibbler- Horizontal Spinner

So with parts starting to arrive and things starting to happen, I thought I would start a thread for the build of hopefully what will be a horizontal bar spinner. The reason for documenting the build on here is so I can ask for advise from you guys and for you to hopefully catch any beginner mistakes I’m about to make before they happen. Especially considering I’m kind of running before I can walk jumping to a spinner instead going for normal wedge/rambot first. To make sure I don't end up with a half finished robot I plan to first build it as a wedge then add a weapon, so if things do go belly up I should at least have a working wedge robot. Any way here the parts i’m planning on using.

Mechanical
Armour- 20mm HDPE for sides with 10mm top and bottom
Wheels- 125mm Nylon wheels with bike tread and hub inserted with Ellis’ melting in a nut method.
Motor pulley- HTD5 15mm 12 tooth
Weapon pulley- HTD5 15mm 40 toothed (Giving a ratio of 3.3:1)
Shaft- 20mm silver steel
Spinning bar- 500mm*50*15 s7 or hardox. Should weigh about 3kg.
rectangular metal(attaching bar to body)- 1.5mm 20mmx10mm rectangle aluminum as shown in the cad.
Bearings- 20mm standard deep groove ball bearings

Electronic
Drive motors- 18volt drill motors (already brought and had the clutch bypassed with grub screws).
Weapon Motor-NTM Prop Drive 50-60 Series 380KV / 2665W planning to run it on 7s (Should spin the bar at a theoretical 3000RPM after reduction)
Drive ESCs- 2 x 30 amp ones from Rangle Box
Weapon ESC- ESC- 125 amp red brick
Weapon Batteries- http://www.hobbyking.com/hobbyking/s..._5S1P_40C.html plus http://www.hobbyking.co.uk/hobbyking..._2S1P_40C.html in series to give about 26volts and 3000mAh
Drive battery- http://www.hobbyking.com/hobbyking/s..._5S1P_20C.html
18.5volts + 3000mAh
Connectors-XT60s

Maths
These are done with the assumption the bar will be at 2500RPM (15% loss from motor)
Energy-2.2 kj
Bite(5 mph impact)- 5.5cm (quite good right?)
Tip speed- 145mph (scary!)
Questions (sorry there’s a lot)
Am I completely mad to try and build a spinner with no previous robotics experience?
Is hardox a good choice for the bar and does the grade matter?
Could i get away from only having spares of the cheap things e.g bearings/drive motors? What parts are prone to breaking?
Do brushless motors have mounting holes at the bottom? Is it a good idea to mount only to the bottom, or is supporting both sides a better idea?
Is the stall current of a drill motor over 30amp? Is the ESC i’ve mentioned be able to handle the stall current or is the Tz85a a better choice?
Where do people normally buy hardox from? Struggling to find any online sellers that sell something other than digger wear plates.


Cad (minus wheels)


http://www.fightingrobots.co.uk/atta...tid=5448&stc=1http://www.fightingrobots.co.uk/atta...tid=5449&stc=1

You're mad to ,start with Robot Wars whatsoever. Build something that takes a lot of time and efford, and then hope it won't get smashed in 5 seconds.
Join the assylum.

Hardox is a fine material for the spinner.
The grade, as long it's higher than the opponents armor, it's good. so aim for 500+. Valkiries disk is hardox 550, 12mm thick.

Expect everything to break sooner or later.

Brushless motors have mounting holes on 1 side.
The NTM outrunner has threaded holes on both sides. Mounting to the frame is done with the 4 M4 holes on the side with the wires coming out. The other side, on the spinning body, has 3 M3 holes to mount propellers to.
Oh, and ruggedize the NTM. The weak point are the bearings and the magnets. bearings can be replaced with decent brand bearings, and the magnets really love a lot of glue.

Hannibalito4 and Valkiri3 used a 25A per channel esc, and they moved fairly well; but with higher graded esc's it went better. the 30amps could work. But take the advice, go for the TZ85's.

Hardox, small bits of high grade Hardox I buy at a place that makes excavator buckets and simular.
Expensive, compared to buying a big sheet and cut from that, yes, but a full sheet goes a few £1000, and a 350mm diameter disk goes about £100 in raw material.(and around £150 in lasercutting cost)
The proven and renowed Hardox 450 3.2mm thick we buy in bulk with the DRG, and then we get a pasture sized steel plate of 2.5*6 meter (last time it was even a 2.5*8m), And then the cost is "less". If I recall correctly, divided in sheets of 1.5*1m it was about £160 a sheet.

Carry out your bite calc at 1m/s. The vast majority of collisions are at low forward speed.

That NTM motor may not have enough low speed torque to get the bar spinning up in a competitive time and will likely get dangerously hot. The NTM motors are good quality but I would try to use something with a wider diameter and closer to 4KW peak power (or at least leave room for an upgrade).

The main concern to me are the aluminium arms supporting the weapon - With the energy on tap they will very likely bend on the first full speed impact. Try to increase the wall thickness and diameter as much as you can.

Thanks for the advice so far.
So 30 amp ESCs out and Tz85a in. Thanks for the information on mounting brushless motors, understand that a lot better now. For hardox I might make it a problem for future me and sort it out when I need it.

At 1mps collision speed there would be about 2.4cm of bite. Much more realistic and thanks for the spot.

As for the weapon motor, shame the NTM wont do as the next step up gets a bit pricey. The outrunners seem to have a huge jump in price once in the 4k power level, so I might end up having to go down the the inrunner path. I originally didn't want to do this as the increase Kv makes single stage reduction harder, but it looks like I have little choice. Would something like http://www.hobbyking.co.uk/hobbyking...Motor_4kw.html be more suited for the job? Or I could be irresponsible and use the NTM motor and have it hate me and look to upgrade it as you mentioned. The aluminium is a bit of a shame as this was one of the few parts I've already brought. Never mind I could also just keep it for when the bot is in wedge configuration.

Valkiri uses an NTM5060 with a gearratio of 1.3. Spin up is slower than wished for but it has a very high top speed on 6s.

a bigger gearratio will solve that. On getting hot. Not really. We measured it and it stays inside any reasonable limit.

That is a damn good start! You have your list, you have a design and you have a plan!

"Am I completely mad to try and build a spinner with no previous robotics experience?"
- Yes, but that didn't stop me from building Conker 1 ;-)

"Is hardox a good choice for the bar and does the grade matter?"
- The higher the better, 450 minimum - 600 at Best. After tossing a basic bar design through Solidworks, you can get your 500mm bar in at 12mm thick, 60mm wide, with a 90mm dia hub in for around 2.6-2.7Kg.

"Could I get away from only having spares of the cheap things e.g bearings/drive motors? What parts are prone to breaking? Do brushless motors have mounting holes at the bottom? Is it a good idea to mount only to the bottom, or is supporting both sides a better idea?"
- In an ideal world you would swap out the stock bearings in the brushless, glue in the magnets, put in a Titanium shaft and then have a bearing to support the back of the motor as well. That said many machines run stock motors and just have lots of spares. The NTM's are so cheap that its more effort/cost to reinforce them than replace them.

"Is the stall current of a drill motor over 30amp? Is the ESC i’ve mentioned be able to handle the stall current or is the Tz85a a better choice?"
- TZ85A every time. If you look after them they will last you several years and several robot iterations. Just check the insides every so often for swarf and loose components

"Where do people normally buy hardox from? Struggling to find any online sellers that sell something other than digger wear plates."
- That is a tricky one. I got the Hardox for my first weapon from JL Steel in Scotland. They were very kind and provided me with a plate for my weapon (20mm Hardox 500) and my armour (4mm Hardox 450) and asked I donate £60 to MacMillan Nurses rather than paying them. However - http://3.bp.blogspot.com/-65GE6TQ2E_.../s1600/041.JPG - That is what I did to a 12mm thick bar spinner last time I met it in combat so you will want a spare.

Hiya, Nice start. You are crazy indeed, but get stuck in and make it, then find what fails and improve it till the whole lot eventually works! That's half the fun.

For a bar spinner like that, 500mm is in my opinion too big, the robot will be huge and in the robochallenge arena at least, will barely be able to get up to speed + will throw itself out immediately if it ever does. Decimators was about 350mm effective diameter for reference :)

The aluminium section on the front is way too weak sorry, that will crumple the first time it hits something 25x25x1.6 or 2mm thick steel would do though, very cheap too.

The NTM is fine to get you going with the aforementioned mods for reliability as long as the weapon sits on efficient bearings. They have decent enough torque especially the super low Kv ones like that. Check out the SK3 motors too, they are substantially better made and tend to have more power IMHO.

Probably have way too much battery as well. Lots of weight can be saved by using 1x for the entire lot, or at least a way smaller battery on the drive. 4s 2000-2200mah is plenty, the drive will be fast especially on 125mm wheels (they might be too big btw, probably will break the gearboxes) and you shouldn't burn out motors. Or go for a bigger 6s pack and run the weapon and drive off it.

Absolutely go for the TZ85s over the 30As on drills, trust me on that one.

Good luck :) Keep us posted

For the record, that bent bar wasn't hardox, it was a different kind of steel which turned out to basically be only just better than mild. It was also 8mm not 12mm. So, don't build your bar from 8mm mild steel and you'll be fine. Also wasn't that the hit that slightly bent your 20mm hardox beater which caused you to shit even more bearings up through the rest of the comp, Alex?

Echo Glens comments, esp the ones about the tz85. I was an idiot and designed myself into a corner meant I had to use 30a. Don't do that lol.

Quote:

---

Originally Posted by harry hills

For the record, that bent bar wasn't hardox, it was a different kind of steel which turned out to basically be only just better than mild. It was also 8mm not 12mm. So, don't build your bar from 8mm mild steel and you'll be fine. Also wasn't that the hit that slightly bent your 20mm hardox beater which caused you to shit even more bearings up through the rest of the comp, Alex?

---

I hit you so hard I bent my own beater, which was under designed because I had estimated at top speed of 8000rpm with losses, not 11'000 :P And I was using tapered thrust bearings which had to be in compression (something I didn't know at the time) so that's why they kept exploding. If I ran that again I would use regular deep groove and I doubt it would break/explode. As for the thickness, I just looked at the Pic and guessed :D And I failed :P

I'd be careful reducing battery capacity. The C discharge rating is a multiple of the capacity of the battery. Smaller battery = lower available current.

I would also recommend running the brushless at as high a voltage as it will allow. Going on the black box data from boner when I ran it on 6s and then on 10s later, the current draw difference was enormous. On 6s the motor drew easily around 80a if not more. On 10s the motor was only drawing 10a peak at any time. The 10s was run on a larger motor so that may have had something to do with it but always run brushless at their top rated voltage.

I would also echo the comments regarding the A frame. Aluminium box will just crumple and send the bar into your own machine. A 500mm bar is doable but you would want a much larger motor behind it to get up to speed.

The spinning bar of massacre was Creusabro 8000, and should have been the equivalent of hardox 500. Unfortunatly, the ones selling it to me didn't know what we do with it. Now they do.
It's perfectly possible that the Creusabro in real world wear resistance applications will perferm like advertised.

And perfect bearings. Every spinner I have build uses sinterbronze bushings. Seems to do the job.
Advantage. Smaller, lighter and cheaper than the equivalent in bearings.
Disadvantage, it gets messy with the extra grease after each fight.
But for every bearing/bushing, installation by the book, or else you get rubbish.

As said before huge thanks for all the info it's a big help.

I think you're right Glen about the bar size, I was being a bit optimistic with the length. Reducing it should also help the spin up time, power needed and weight. With the only downside being the reduced energy which, as you said, might not be a bad thing. The dimension Alex talked about also seem a good alteration too (reducing the thickness to 12mm and increasing width to 60mm). This thickness appears to be easier to find and the increase in width should make it stronger (hopefully not having the same fate as Massacre's bar- nice to know that was 8mm and a weak metal :) ). Out of interest what sort of RPM does the bar spin at on Decimator and how much does the bar weigh (sorry if these are answered elsewhere)? I ask this because it has a great balance between delivering good hits and not pinging itself out the arena.

You all are unanimous in thinking the tz85a is a worth while upgraded and Aluminium not being up to the task, so these have been changed. Probably go for the 25x25x2mm steel arms as talked about by Glen.

With a few more of yous opinions on the motor choice, I think I'll stick with the NTM50-60. If it sufficient for the disk on Valkiri, it should be okay for me even if the spin up is a little slow. The reduced bar size should also help with this. I'll also try to design it so it can be upgraded in the future. After checking out the SK3 range they definitely look like a good upgrade for the future.

Might spec down the battery capacity a bit too (thanks again Glen). Also thanks for the reminder about increasing the discharge rate Gary. Something like this
http://www.hobbyking.com/hobbyking/s..._5S1P_40C.html I think would do with the increase to 50c and decrease to 2200Mah (drive battery). [Edit: ignore the following bit(miss read the post)- like the idea of having one battery, but don't think it's easily possible due to the different voltage between drive and weapon motor and the huge peak current that would be drawn from it].- Thanks for the information Gary about increasing the voltage, but I'll like to keep it 7s or lower. This is solely down to the limit of cheap brushless ESCs seems to be 7s.

As for the bearings vs bushings debate, I think I'll stick to bearing as I vaguely remember reading somewhere they don't have to line up as well (could be wrong about this).

Also out of interest why do larger wheels break gearboxes? Is it because of the increase in stresses from faster acceleration? Finally is they any way of reducing the likenesses of them breaking e.g metal gears?

The information so far as been brilliant and gives me a much better chance of making it an arena with a working robot. So once again thanks

Larger wheels can break gearboxes as larger radius provides more leverage on the pins and gear teeth. Think of it this way: if you have a stuck bolt, its hard to shift it with a short spanner. Get a long enough spanner and the bolt will shift or the head will break off.

Bearings VS bushings is a complicated comparison. they will both wear faster if they are misaligned but bearings will usually take longer and fail more gracefully (unless you are Alex). The big problem with bushings (which many people don't know about) is their maximum speed rating called the PV value - here is a quick explanation: http://www.mcmaster.com/#about-plain-bushings/=y529dd

The other problem with bushings is that they can't be sealed and pick up grit, leading to wear & jamming. For the speed you are planning, you can use sealed bearings that will be dirt-proof.

The bushings I use are on the edge of their spec in Valkiri3.
But that's a monster in weaponpower; Diameter 356, 4 kg going 5.5Krpm. (theoretical 6300).
Start up amps, around 120, running amps, 20.

The grit issue is solved like any good bushing in a backhoe, bulldozer or any other heavy equimpent dealing with massive amounts of dirt and sand. Grease it so the grease pushes out the crud after each use.

Quote:

---

Originally Posted by Cameron_S

As said before huge thanks for all the info it's a big help.

Out of interest what sort of RPM does the bar spin at on Decimator and how much does the bar weigh (sorry if these are answered elsewhere)? I ask this because it has a great balance between delivering good hits and not pinging itself out the arena.

---

The entire rotating mass is about 4.5kg, the bar alone is ~3.5. The original motor i had rewound for it was about 7500rpm on 6s, the one it ended up running through the event measured at 5500rpm which was a 12s motor running only on 5s. Turned out well though, will probably keep it around that speed next time and just adjust the gearing / motor wind to suit :)

Quote:

---

Originally Posted by GlenR

The NTM is fine to get you going with the aforementioned mods for reliability as long as the weapon sits on efficient bearings. They have decent enough torque especially the super low Kv ones like that. Check out the SK3 motors too, they are substantially better made and tend to have more power IMHO.

---

Basically that, there's a 50mm one on Hobbyking that's 4000w on 6s which is very nice indeed. You'll want to get the shaft upgraded from 6mm to 8mm and to a material that isn't cheap steel/plastic-with-some-kind-of-metal-on-it so it's stronger - get some 8mm ti and it should be a pretty powerful thing. And they're not really that much more than an NTM either... I ran one in Overdrive this year and was happy with how it ran in the fights where I actually had a drum haha

Good to know why the large wheels break gear boxes (analogy helped). Unfortunately there seems to be little way of reducing the size of the wheels while keeping it invertible (expect from having two small wheels on top of each other which seems difficult). Might just be a case of have lots of spares and hope it doesn't happen too often.

Thanks for the specs on decimator's bar, faster than I thought.

The design has also been changed to fit the smaller bar and the arms upgraded to 20x20x2mm steel bars. The weapon motor has also changed to this- http://www.hobbyking.co.uk/hobbyking...ize_heli_.html on 7s which I believe is the one Matt is talking about. This seems a worth while upgrade with the increase in peak power. However the increase in kv, puts having a small pulley on the motor going to large one on the bar impracticable. So would it be possible to have a small pulley on the motor going to a large pulley on a shaft near by which also has a small pulley that then goes to the bar? I think an image will be clearer. (The smaller pulleys have 12teeth-the large on on the shaft has 40-the one on the bar will have 36) Should reduce the RPM to about 3500.

Edit:Any way of making images smaller :) http://www.fightingrobots.co.uk/atta...tid=5455&stc=1

Valkiri uses a Z40 around the motor can, and a smooth ali 112mm diameter pulley on the disk.

The technical term for that is a jackshaft. A double reduction will work but has some drawbacks: It will add weight and complexity and takes up valuable space. There is also a potential problem when you have a small pulley close to a larger pulley. The small pulley will have very few teeth fully engaged with the belt (maybe only 4) and those teeth have to transmit all the motor power. this graphic explains it better:

http://www.nswrfc.org/Nick/belt_wrap.jpg

When the blade hits something, there will be a huge amount of force on those few teeth and they may strip off the belt, or they might slip and start to wear down. Either way, the belt will wear out much faster than it should. The way to prevent that is with an idler or 'back-breaker' wheel, which increases the belt wrap around the motor pulley. Its also a good way to adjust the belt tension.

The alternatives to using an idler are to use a wider belt and/or larger pulleys. You can also use a smooth pulley on the weapon shaft to act as a clutch and limit the amount of force on the pulley teeth. I don't really trust this option as its hard to know what the right amount of tension is.

All the above solutions will work but the best or at least most elegant solution is to find a motor with a lower KV and have a single reduction to get a weapon speed of around 3.5K rpm, with a 4:1 reduction ratio and a 7S battery, you will need a motor with around 600 to 650KV.

Thanks Nick for the info about jackshafts (new word you add to my engineering know how). The diagram illustrates perfectly what you mean by having a idler wheel to increase the amount of teeth in contact when the pulleys are close together. With all the draw backs you mentioned, I'll take your advice and try to use the more elegant solution.

So it's back to the pesky weapon motor dilemma. I have quite literally looked at every motor on the HobbyKing website and I can't find something that's high power and low Kv (except from a very expensive Scorpion motor that's a bit out of my price range) of course i've probably missed a perfectly suitable motor. So something going have to be compromise and if the 2 stage reduction is too complex then it's the power. Because of this I've gone back to the original plan to use the NTM50-60 380kv motor. This will most likely to be one of the first parts to be upgraded in the future.

These posts wouldn't be the same without some questions:
Is the anything glaringly obvious that would make this battery unsuitable for the weapon and this be unsuitable for the drive?

Is under volting the drive motors, suggested earlier by Glen, worth the reduced speed to hopefully stop the motors burning out?

As always thanks in advance

Both batteries are up to the task.
But a 2200 is on the low capacity side. I would go for the 2700mAh compact.

You could use a lower KV motor and a higher voltage pack to get the RPM you want. For instance this 530KV motor: http://www.hobbyking.co.uk/hobbyking...ter_Motor.htmlan 8S battery and this ESC: http://www.hobbyking.co.uk/hobbyking...PM_Sensor.html will get the job done. I prefer to use two 4S packs in series as they are easier to find and easier to charge on two smaller chargers.

Just when i thought I've finally figured out what motor to use. Must of missed that one, but it ticks all the boxes. So thanks for the spot and ESC you've mentioned would certainly do the job.

I see where you're coming from with preferring two smaller packs over one larger one. This would also mean only one type of pack is needed for the whole machine. So the current plan is to use 3 of these (1 for drive and 2 in series for the weapon).

The build log might actual become just that! My HobbyKing, RangleBox and DirectPlastic orders have just arrived. No problems with any of them as they all arrived quickly and well packaged.
http://www.fightingrobots.co.uk/atta...tid=5472&stc=1http://www.fightingrobots.co.uk/atta...tid=5473&stc=1
Looking quite cramped should I be worried?

In case the image isn't too clear there are:
12Guage Red Wire (2m)
12 Guage orange wire (black was out of stock :( )
3-zippy 2650Mah 4s 40c lipo battery (scaring me before they have even been used!)
2-Tz85A ESC
1-spektrum dx6i
1-OrangeRx GR300 3ch (The 620 was on back order so had too look for a replacement. I would only be using 3 channels so think this would work- Correct me if i'm wrong)
15-Male and female XT60 (definitely have spares)
1-Imax b6 charger (Instead of buying a 'proper' power supply for it would a small 12v car/motorbike charger work?)
50-m6 50mm socket head bolts
50-20mm m6 barrel nuts
2-cordless drill motor and gearbox
1-100x25cm 20mm HDPE (sides)
1-100x25cm 10mm HDPE (top/bottom)

Before starting to cut the chassis/Armour i'm going to wait for the wheels to arrive as I fear they won't fit. Probably being a bit paranoid as they fit in CAD design.

EDIT: Forgot to mention I didn't order the brushless motor or ESC. This is because I've yet to fully decide on a motor and want to get a working rambot before adding a weapon.

Some progress at last. So after converting a drill press adapter thing into a ghetto press, I went about heating the hubs. The first attempt was with a hot air gun, however this struggled to get the hub much above 100c.
http://www.fightingrobots.co.uk/atta...tid=5502&stc=1
So it was blow torch time which worked far better. One thing worth mentioning is that having a bolt with some washers the same size as the hole in the wheel makes lining it up much easier.The photo is a bit clearer.
http://www.fightingrobots.co.uk/atta...tid=5504&stc=1 http://www.fightingrobots.co.uk/atta...tid=5506&stc=1
A bike tire from a road bike was used to increase the grip. It was attached with screws through the side into the wheel. The end result isn't too pretty but should work- I hope.

So far only the base and top have been cut out of the HDPE sheet. The mess it makes is incredible and the piece was only 10mm. Hate to think how bad the 20mm sheet will be like.

Progress seems good Cameron, one thing I did notice is that GR reciever which is a Ground based one. I know they have odd failsafing issues so I'd run a few checks with servos to make sure all 3 channels go to neutral before you put it in something dangerous.

Also, bike tyre usually gets screwed in through the top, the heads sit within the tread on most peoples.

With the champs approaching, I've had a bit of extra motivation to try and finish this in time. So the progress so far:

Firstly, You were right about the receiver, it didn't even want to bind yet alone failsafe, so I bought a lemon rx instead. The reason for the the screws going into the side was because the tire was from a road bike which meant the tread wasn't deep enough to hide the screws' head.

The 20mm parts then got cut out making even more mess. I then went about cutting the arms (not very well). Then holes for the bolts and barrels nuts were drilled, most of the barrel nuts lined up well, but a couple didn't so the holes had to be widened.
Attachment 5742Attachment 5743

So with the body mostly done, I went about soldering the wiring. However, in my haste I completely forgot about the led being needed so that needs to be added. The motor mounts were then cut and drilled as were the corresponding holes in to lid and base. After the motor mounts were cut and fitted, it was time for a quick test drive which was interesting considering it was my first time driving it and the patio was very bumpy.

https://www.youtube.com/watch?v=dcLR...ature=youtu.be
Attachment 5740Attachment 5741
With it now moving, it's time to buy the spinner parts. Luckily since the last time I searched for Hardox, there's an online ebay seller. So the current plan is to use a 400*50*15mm hardox 500 bar (2.35kg) with this at a 1:4 reduction to have a bar RPM of about 3500 (2Kj). The only concern I have with this motor choice, is that is it going to be run at 8s which is quite a bit under the 12s max. Should this affect the outputted power too much?

Without delving too far into Ohm's laws about power, when you double a motor's voltage the power increases by 4 and when you halve the voltage, power decreases to a quarter. Since your motor is running at .6667 of it's rated voltage, the power will be .444 of 5,500 watts or 2,444 watts. That is a large reduction but still a very respectable amount of power.

The NTM Rotor drives apparently run very hot anyway, so a bit of an under volt is probably not a bad thing. A better bet for feather spinners is probably the NTM 5060, but if you've already got the Rotor drive then it should still be fine.