The New 120 - a few questions

This year the RoboJackets are designing a brand-new 120 lber from the ground up. I'm the head of the weapons design team, and I had a few questions:

1. McMaster-Carr has very limited supplies right now as far as drum materials go - what would you suggest as the material for the drum? Schedule 80 pipe? Also, what are some other good materials sellers?

2. What are the advantages/disadvantages of 2 vs 4 wheel drive? We're having a bit of a debate about which actually performs better. [I personally favor 4 wheel drive for redundancy reasons](and yes, I did see the topic on 2 vs 4 wheel drive here)

3. Electronics - what's the best way to calculate a robot's power requirements? (I imagine looking at what the draw would be from all the motors stalled out for 5 minutes would do the trick)

speedymetals has been good for my team whenever mcmaster hasn't

And as for your other 2 questions you may want to check this out from OLATHE i found the site to be very helpful during freshman year and it does discuss the advantages of 4wd
http://www.ravenengineering.net/alumni/#9

online metals is just like speedy too..

hmm... sorry i missed the meeting today, napping and catching up on history reading for the test tomorrow.

1. Material-wise, I never order metals from mcmaster because it's all overpriced. The vendors I choose from are Speedy Metals, Metal Express, Online Metals, Yarde, and Southern Tool Steel. The first 3 for smaller quantities of aluminum and common grades of steel while I buy bulk Al from yarde and all my tool steels from Southern. From experience, they seem to offer the best prices on almost any grade of steel ranging from S7 to A2 to 4140

2. 2wd has the simplicity aspect going for it, but 4wd as you said has the redundancy. From experience, 4wd always seems to be better because 2wd systems always find some way of not having proper traction at all times especially since arena floors are never flat.

3. Battery-wise, I havn't gone wrong with Steve Judd's calculator for drive. Plus, we should be running separate battery systems for drive and weapon anyways so that can all be figured out. Also, I can collect "experimental data" from others who have already built 120's using similar set-ups.

I'm sure Jeff is thankful for your metals supplier lists, and I just want to say I am too. I recently bought some Al and tool steel stock from McMaster and was really surprised at the expense.

Jeff--I think that all motors stalled for 5 minutes is pretty extreme. At worst, the match runs like what, 3 minutes? (Maybe RoboGames has different rules, or if you are designing specifically for a Rumble, then maybe there is some logic for 5 minutes, I dunno). Also, if all your drive is stalled for more than, say, 20 seconds, you are knocked out anyway, right? And suppose your weapon stops running altogether, only a terrible driver would leave it a full throttle for the rest of the match, in which case they deserve to lose.

I'd probably go for 1 minute full stall as a total capacity, but I'm sure there are better ways to estimate this.

And this analysis is only good for total capacity--don't forget to figure 'surge' requirements. 'Back in the day,' Lithium Ion batteries had great energy density, but couldn't handle a current draw above 5C, so (prior to LiFePo/A123 technology) they were useless for battlebots. Make sure the batteries can handle full stall at all without dropping voltage so low that controllers drop out.

If we've got 24V 3600 mAh batteries with a 22C rating, do you think it would be possible to support the current draw from two short mags and one large brushless outrunner if there were two batteries in parallel? (24V, 7200 mAh, 22C)

Well, this link suggests that one S28-150 MagMotor can draw 285 amps in a stall.

7200mA x 22C = appx 150 amps max current.
So, you are more than a bit short for stalling two of those puppies plus an outrunner.

While previously I was arguing for substantially less than 5 minutes at full stall, this is perhaps going too much in the other direction.

Lets say 1 minute at all stall, thats 2 x 285 amps plus, I dunno, another 200 amps for a large outrunner?, times 1/60 of an hour.
We get in the neighborhood of 13000mAh, or about twice what you suggested. Granted, at 22C, you are still falling short of the peak surge of having everything stalled.

A pair of these in parallel would get you just about halfway between my suggestion and your original proposal, and they have a surge of 40C actually, tho I'll bet that applies for substantially less than 1 minute.

But I sound like I might know what I'm talking about here, and I most certainly do not. I suggest a good internet search for the total on-board powerplant for some winning 120's. And keep asking questions!

Um, I was rereading my post and realized there might be confusion--

The two 22.2V thunderpower 5000mAh in parallel (10,000mAh total) would put you about halfway between the 13,000mAh I had estimated and the 7200mAh you had proposed in the previous comment (NOT the staggering 65,.000-plus mAh--or 5 minutes at full stall--you had originally proposed).

clear as mud, right?

Ah. Well, it just so happens that we have four of those batteries, so it's a 24.4V 14400 mAh 22C system. So that would get us 316.8 Amps max draw instead of the 150. The thing is, these are nicad packs that we've got left over from the wedge they had previously (the packs are still in excellent condition). The reason I was looking at outrunners was to minimize current draw & price - the other weapon motor candidate is a s28-400 [long mag] (which draws 390 Amps in stall!!!).

Now comes the devilish part of it--if you have four, what do you after the first match?

NiCads don't like to get too hot, so you can't really charge them too fast, or too many times in a single day, but without spares you got no choice.

Also, what ESC's are you looking at?

Sorry, I should have clarified: we can run four batteries at once - that was the configuration for the old wedge (and have more than four total [99% sure we do at least...]).

If it helps any... On Copperhead we had 3 3000 mAh 24V NiMH battlepacks, and ran a long mag and two short mags.... and nothing died until...



...until the fan on an IFI caught the fan guard, stalled, overheated, melting the plastic, solder, and circuitboard, the voltage regulators touched, shorting, sending a burst of current out of the battery into the motor, busting a winding out of the short mag's armature, and melting the battlepack. It was a fairly catastrophic consequence of such a minor failure...

OK, finally the voice of experience (rather than my wild guesses)--one long mag and two short mags can successfully run in a 120lb battlebot on 9000mAh of battery capacity at 24V.

My "1 minute at all stall' would have argued for significantly more for Copperhead: If I am assuming correctly that a "long" magmoter is the S28-400, and a 'short' magmotor is the S28-150, and using Robotmarketplace's guide, then the total stall current would have come out to 960 amps. For 1 minute (1/60 hour), this gives total capacity of 16000mAh, or over 175% of what is battle tested.

So, what do we say a 'rule of thumb' is 'everything stalled for 40 seconds' is a tolerable, probably conservative, starting point for selecting battery capacity?



Surge currents these days are in the 25-40C range, so if all the motors went from stopped to full stick, one might expect voltage dropout ('cuz 960 amps is nearly triple the 360 amps that 40C would yield for 9000mAh batteries), but again, practical experience shows that to be an unrealistic assessment.

Bottom line again--Looks like about '35% of all stall for surge current rating for batteries' would do as a starting point estimate for keeping things running.



Dan, would you be willing to divulge the surge rating (or chemistry, or even mfg info) on those batteries? What did you do for receiver power? (separate BEC, BEC integrated with the ESC, or a receiver battery pack?)

they were NiMHs, Saft VH Cs 3200 XL cells, (http://saftbatteries.com/Produit_VH_VHT___Super_high_power_series_329_52/Language/en-US/Default.aspx)

The website doesn't give specs on discharge rating, but NiMHs can take quite a beating...

We used a separate BEC for receiver power, RobotMarketplace's ParkBEC.

Thanks, Dan. I took a look at the datasheet on the Saft site, and for those batteries it lists continuous discharge of 40 amps per cell, and peak (for 0.3 seconds) of 150 amps.

Three packs in parallel make for 120 Amps continuous and 450 amps for 0.3 seconds. Umm, converting to C rating, that's 13C continuous and 50C for burst/surge (0.3 seconds)--pretty tough, sure enough.

More importantly for the topic at hand--450 surge rating divided by 960 amps all stallled gives us 47% of stall is the surge rating proven to work (at least in Copperhead). This is greater than the 35% estimated in my prior post (because the NiMH's actually used could take a greater surge-to-capacity than the LiPoly's I was using to do the estimate),

So, based on this information, as a design starting point, the surge rating of your batteries should be about half of the full stall condition of the battlebot.

DISCLAIMER:If you do this and your robot blows up, don't blame me, 'cuz I already said I don't really know what I'm talkin about. Also, anyone familiar with fire codes would say that "half the max" is crazy talk--nothing should exceed continuous capacity at any time, but we build battlebots, so I guess that kind of logic might not apply. Probably because it only has to run for 3 minutes at a time, and only has to last for a couple of hours of total lifetime running. And we generally expect it to get destroyed sooner or later anyway. Either way, do this at your own risk.

Oh, and a design value for continuous current delivery of only one-eighth of all stall.

Just nuts, absolutely scary.

But it did work.

Which actually brings up a good question--Jeff, was 22C the continuous rating, or the peak/surge/burst for those batteries?

And one more thing--use a separate BEC for your receiver, (or even better a separate receiver battery). If the main/drive/weapon battery's supplied voltage drops momentarily due to load, then the BECs in the speed controllers might stop working (if the controller's required voltage is higher than what the batteries can deliver under big loads), but if the loaded battery still has a higher voltage than the receiver, then the dedicated BEC should keep you connected.

I'm not sure, the rating wasn't listed. I calculated it from the peak current delivery, so I'm assuming peak. Personally, I think the electrical team needs new batteries ASAP, but I don't have much control over that.

Quick question - what type of bolt would anyone recommend to hold drum teeth in? The teeth are set into the drum, and held on the opposite side by a 3/8" 'nut-bar.'

Grade 8 flat head socket cap screws. Largest diameter the design allows.

Grade 8 steel--pretty hard, but with some toughness too.

Flat head--i.e. requires a countersink hole, but doesn't stick out

Hex socket--drive in with a hex/allen wrench, even if it takes some damage you will likely be able to remove it.

How many are you thinking?

I think Fluffy had/has four 1/4 inch screws holding each tooth. It was designed to have 7, but we ran into some unexpected countersink problems. That means one for every two inches.

Of course there are some important differences to remember: Foremost is that the heaviest thing Fluffy's tooth would have to move is Fluffy itself (supposing it hit the arena wall). Also those screws are with no nut bar, just inset and with holes countersunk from the inside of the drum (had to have space for internal motor) And the drum was Al.

According to what I've been informed elsewhere, you can have the same diameter of material around your fastener as the fastener (so a 1/4" bolt should have 1/8" of material on all sides, unless I'm mistaken). That would mean I can use a 1/2" diameter fastener every inch or so (6 bolts for the large tooth, 3 for each small tooth). I'll have to see if that number of bolts will be able to take the impact though.

You're not giving us enough info.....size of blade, etc. joules of energy, single, double, drum material ???

Fluffy is a 2 blade design and I believe that with a single blade or staggered blade drum you need twice the holding power.

Riobotz has been at or near the front of the pack with their true single blade drums and haven't been able to keep a blade on any of their bots as of a few months ago.

I believe that their 120 was using M12 or M13 flat shcs and they didn't hold....threaded into solid stainless!!

Here's a pic of what happened to their feather




Between flat, button and reguar shcs's, the flat head has the weakest crossection (in the head area), but it is the most used for this application.

I like dowel pins rated well beyond the forces, forget the grooves and just use the screws to keep the blade on the dowels..at least for a single blade design....however, you need to beef up material around where the dowels will be installed....it can get tricky

jmo

Well, we've got options for teeth. We can either:

1 - stick with the single-tooth setup and have one bolt per inch (3 in the 3" teeth, 6 in the 6" tooth), or
2 - switch to two 12" teeth with one bolt per inch (12 bolts per tooth)

The drum itself will be low-speed this year. We'll max out at 4900 RPM, which oddly enough still gets us to four times the energy capacity of Touro (120 lb, specifically the version being run in 2008). Not bad, considering the technical issues we've had, and still well above standard drum energy storage.

The drum itself has an outer diameter (tooth to tooth) of 6.25", and the teeth themselves had to be reduced to 5/8" above the surface of the drum to meet space requirements. At the current speeds we're using (and heck even double that speed), bite will not be an issue. I'm expecting drum performance not unlike a giant Fluffy (based on drum size), but perhaps less dramatic due to the increased opponent mass.

From dubard's suggestions and our specs, I'm currently looking at these:
http://www.mcmaster.com/#91253a377/=5mru47
http://www.mcmaster.com/#91253a716/=5ms0f9

Hey, keep in mind that I'm just a high school teacher tryin to run a program, while richard is an actual, experienced machinist. I would follow his advice over my limited experience any day.

rdubard wrote:

Hey, keep in mind that I'm just a high school teacher tryin to run a program, while richard is an actual, experienced machinist. I would follow his advice over my limited experience any day.

Don't sell yourself short Bob......common sense goesa long way and you have plenty...

Jeff..24k joules ???? what's that drum weigh? I might be more concerned with wheel placement and balance....that bot might have a hard time moving with the wheels set so far back??

It's around 32 lbs (call it 15.7 kg), with a 5" drum OD and half-inch walls (4" ID) at 12" long. It gives us a pretty decent moment of inertia. As for driving, the center of gravity is almost dead-on for the wheels; we were pleasantly surprised on that account. I wouldn't be surprised if it actually improves a little more, because the last time I saw the center of gravity was before we added the gears into the model.