September 20, 2017, 10:32:49 am
Welcome, Guest. Please login or register.

Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.


Messages - zacodonnell

Pages: [1] 2 3 ... 30
1
Progress Reports / Re: Team Brain Damage
« on: August 26, 2017, 10:02:34 am »
Got my tickets in the mail. Looking forward to the event.

I don't think I'll have time to do anything especially interesting this year, unfortunately. Spare time has just not been widely available lately. It's still going to be fun though!

-Zac

2
Progress Reports / Re: Team Brain Damage
« on: August 18, 2017, 06:36:50 pm »
I'm in now. Thanks for the reminder!

Zac

3
Progress Reports / Re: Team Brain Damage
« on: August 13, 2017, 12:37:00 pm »
 The board is definitely damaged. That is almost definitely the cause of the symptoms that I experienced. My trouble is figuring out how it happened and how to make that particular thing not happen again so I can find whatever the next failure in the system is.  Either way it should be fun, and I have a feeling Jerry Clark and it's gonna show me why pneumatic flippers driven by someone with a clue dominate silly fly wheel powered flippers driven by an old-school spinner builder.

4
Progress Reports / Re: Team Brain Damage
« on: August 12, 2017, 11:09:56 am »
I don't think I'll have any major mechanical upgrades to Magneato for Franklin. In the two fights where the control system was working properly it performed great. It self-righted reliably, was very mobile, and the trigger worked well.  The only thing I'd like to investigate more in the trigger is the arm position sensor. I would like to polish up the light-sensor mount (instead of the ghetto ziptie setup I did the day before the show) or maybe add an encoder or magnetic switch to the arm so I know when to disengage. I'm thinking something like this might work:
http://www.kr4.us/Magnetic-Door-Switch-Set.html?gclid=CjwKCAjwzrrMBRByEiwArXcw28uN1OJMyRSMGHK-B3yOiD8W8U8roYkwd5VdBZ269X5qd9d7WHPIyBoCNIIQAvD_BwE

As far as the problems getting the flywheel to work reliably, I'm 99% sure that is an issue with the Arduino nano control board. When I was trying to get it to work for Shuttle Launch I spent a few hours trying to debug a complicated failure behavior where it would work if only the throttle was plugged in, but as soon as I plugged the trigger channel into the receiver the throttle would go to zero.  I think something happened to the circuits inside the controller that's causing a bad connection somewhere.  I can just swap the nano out, but I need to figure out what happened. I doubt it was shock loading but it is possible.

My best guess is that it has something to do with electrical feedback from the giant magnet running the clutch. I think it might have eventually toasted my flyback diode and then proceeded to damage the rest of the control hardware.  I don't know enough about electrical engineering to even know how to test something like that. Anybody have any suggestions?  The magnet draws ~5A @ 18v when it is on, and is about a pound of copper windings. I'm guessing there is a very large voltage spike when I turn it off. It damaged one of the little toggle switches I was using during my tests before I put a little diode on the leads. Can somebody recommend a specific diode for this?

-Zac

5
Progress Reports / Re: Team Brain Damage
« on: July 18, 2017, 07:38:55 am »
Fight 2: Plan F vs Driving Practice - https://www.youtube.com/watch?v=iYXpeNEpCdo

Fights like this are the reason I tried the wedgelet. The blade I'm using doesn't get much bite unless it can get under somebody, and a smooth wedge is a problem. Right off the bat they teamed up and rolled me over, but since I fixed the blade binding thing I had no problem powering the weapon up to self right. It took some slamming but the motor is very powerful.  Naturally they immediately "jammed me up" again and then in all the pounding I lost a drive motor.  At this point I wasn't feeling very hopeful about the fight.  At the end I decided to spin down so that if they pushed me into the wall I'd bunny hop out, and the gambit worked. I flipped over and took off a tire. Then the other bot came after me and I popped a tire off there too. Still, I wasn't in control at all during this fight and the judges knew it - loss by decision.

At this point I was pretty down on the Kitbots 1000 RPM drive motors I was using. I lost one in each fight (left first, then right). The motors themselves are very electrically noisy, such that even with the caps on them driving around was causing the blade to kick on and off, and when I drive forward it messes with the throttle and I can hear the blade speeding up and slowing down. Mix in the busted gearbox from this second fight and I'm gonig to be looking elsewhere in the future. Kyle pointed me toward some of the ServoCity planetary motors (14:1 and 19:1 ratios) and the boxes themselves look much nicer. I'll probably give them a shot.

Fight #3 vs Land Shark - https://www.youtube.com/watch?v=b4m7H5ZDAdY
I figured I was in big trouble here. I just got beat up pretty bad by a pair of plastic-wedge beetles, so going against the low slung metal wedge of Land Shark was not looking good. I had to leave the wedgelet on, which made the bot drive poorly, but if I didn't get a bite on the wedge I was toast.

Miraculously I managed to get under the lip early on in the fight and then it wasn't nearly as effective as a wedge. That let me go on the offensive until I skiped over the dent and got hung up on the damage. I think he must have been having some battery or drive troubles because he never really pushed me much even when I was stuck on top. We had to be separated twice, but in this fight my drive motors held up the whole time and I was able to gnaw at the sides and back of the bot until time ran out. Win by judges' decision.

-Zac

6
Progress Reports / Re: Team Brain Damage
« on: July 17, 2017, 08:14:08 pm »
Ok, we're going to do these one or two at a time as I have spare time:

Fight #1 - Plan F vs Super Kamikaze - https://www.youtube.com/watch?v=BN2h_o1lyNc
This was a tough first round draw for me. Super Kamikaze bashed me up pretty badly last year and I expected nothing less this year.  His wedge also made it tough for my blade to get a bite and my little wedgelet made it tough to drive.  About 45 seconds in I ended up upside down in the corner and I thought I was in big trouble. Then he charged in and accidentally rolled over his own wedge. This left both of us with pretty limited mobility. The damage to the weapon axle caused the bearings to jam up a little bit which made self righting using my blade very difficult. The motor and controller are so over-spec that they managed it, but there's a little burn mark on the ti shaft where the housing of the bearing was rubbing on it.   I managed to get myself upright but lost a drive motor in the process. SK was bouncing around trying to self right too and I happened to get a lucky strike on his battery pack (despite being unable to really attack on my own) that knocked him out.

-Zac

7
Progress Reports / Re: Team Brain Damage
« on: July 16, 2017, 06:24:00 pm »
I also had some other videos from the event added to my channel for other bots:

https://www.youtube.com/user/robotdesigner/videos

-Zac

8
Progress Reports / Re: Team Brain Damage
« on: July 16, 2017, 06:23:24 pm »
I don't have time for a full write-up for now, but I'll give a quick summary. Jeremy and the Campbells are awesome. The event was a blast (heh) as always and I think pretty much everybody had fun. I was pleasantly surprised to get first place with my bot, which I did end up naming "Plan F".  There was a wedge multibot that ended up doing unexpectedly well by making it to the final in the winners bracket. They schooled me pretty well in my first fight but I managed to beat them in the final. I got a lucky hit on the battery pack against Super Kamikaze the first time I fought him but fared better on the second try. Seven fights in a day with one bot is a lot, but other than a couple of drive motors and a dinged up weapon shaft I took very little damage.

Here are the videos:
https://www.youtube.com/watch?v=BN2h_o1lyNc&list=PLATw5d_vDtZkTdW2VN-Zw1S2WzlQMNzRp

-Zac

9
Progress Reports / Re: Revelation Robotics
« on: July 13, 2017, 04:00:09 pm »
If this problem is intermittent it could be a loose wire or bad solder joint between the motor and controller. I know you said you got it fixed but if it comes back that's what I would check.

Zac

10
Progress Reports / Re: Team Brain Damage
« on: July 13, 2017, 07:19:33 am »
Yeah, they're only 1/8". It's what I had weight and material on hand for. Because of the way I packed all the guts in foam I'll probably feed my back end to a horizontal spinner instead of the blade, at least until I can get it slowed down a bit. 

The wedgelet is pretty great, but it bites into my rough OSB floor pretty badly and makes it hard to drive the bot forward. On Jeremy's floor (at least at the beginning of the competition) it should be ok. I can always take it off if it is too annoying.

-Zac

11
Progress Reports / Re: Team Brain Damage
« on: July 12, 2017, 10:50:42 pm »
I'm already using the coarse-est screw I could find in the 3/8" range without getting into the multiple-thread-start or ultra-preci$ion version. The larger diameter screws offer additional options but I'm not sure if coarse is better for that first start or not.  I may toy with the idea some more in the future, but for Bot Blast I'm definitely not looking to do anything else that might not work. I've got to have a bot that can at least enter, and a barely-adequate drive system with no weapon is too lame for me to stomach unless I have to.

Sooo... In the last two nights I added a weapon to the platform and it actually came out super well.  There's a ton of empty space in the frame so I moved all the guts to the back and added a bunch of padding. The weapon mount is pretty weak but otherwise everything is actually well beyond me expectations for a two-night retrofit. Here is the catch-up build report
https://plus.google.com/photos/107279152885550274671/album/6433529956027270705/6442082293372338194

And some videos of the final version of the bot. Tomorrow is a frisbee game then packing on Friday. I'm getting pumped!
spin / drive
https://youtu.be/cKtuQWrxn50

self-right (that's a lot of motor!)
https://youtu.be/vDUzSqNpyi8
-Zac

12
Progress Reports / Re: Team Brain Damage
« on: July 11, 2017, 10:14:40 pm »
Thanks, I agree that the concept has a chance to be very cool. I spent most of this project trying to get it to spin faster because of the small size of the linkage I was using. In a bigger robot, with significantly more travel on the arm, the RPM required might be much lower to get the target tip speed out of the flipper. I think a lower RPM combined with a more reliable trigger mechanism than the silly servo-crank setup I have here could be a winning combination.  I have some even crazier ideas with it long-term but for now I'm a little worn out on the whole thing so they'll have to wait.

I spent tonight retrofitting the frame with an alternative weapon. I toyed with several different flavors of what I could do in 8 hours of build time with the parts I have on hand. Fortunately for me "the parts I have on hand" is a fairly extensive list, including twenty or so timing belt lengths, a half dozen pulleys, three or so spinning weapons in the right size range, and lots of motors and controllers to choose from.  With the lathe, mill, and random material stock I have on hand it was a question of what I could attach to the current frame.  I really wanted to do a horizontal midcutter with Devour's blade but I'm pretty sure the 1000 RPM drive gearboxes would explode from the recoil forces.  I opted to set devour's blade up in a vertical configuration because I already had it mounted to a hub/pulley assembly and had the motor, belts, and pulleys in the right lengths to fit it.  All I had to do was make a pair of weapon rails to support it and steal a hinge pin from Magneato for the blade axle. 

I'm going to remake the top plate tomorrow and move some of the guts around to be as protected as possible then I'm done.  Hopefully it has enough oomph to self right the bot (I bet it will with the simonk software).  If not, then I'll just try to keep it running and hope for the best.  I'm reasonably happy with how it fits in for this late in the game and it looks like the weight will be right on. I might have to take a little of the frame material out but it's going to be very close. I forgot how much simpler regular spinners are when you just ignore all the hard details :)

-Zac

13
Progress Reports / Re: Team Brain Damage
« on: July 10, 2017, 10:52:40 pm »
Well... I spent the last two days, in their entirety, on the robot.  I have decided to officially call the experiment "unsuccessful within timeline". I prefer to avoid the word "failure".  I have two build days left before one night to pack and the show on Satureday, and it just doesn't work.  I didn't leave enough time in the schedule to deal with "stuff" that comes up.

I'm still planning to have a bot, but it'll be something thrown together at the last minute out of spare parts.  Not sure WHICH spare parts yet...  I'll bring the weapon assembly to show around.

See you all Saturday!

-Zac

14
Progress Reports / Re: Team Brain Damage
« on: July 08, 2017, 08:41:04 pm »
Ok, now for the bad news.  I managed to rig up a test at 28v with the new controllers, but like somebody suggested here I'm having troubles with the screw stripping / getting misaligned when I initially engage the nut at very high speeds.  It was fine at 16v, but at 28v everything is moving so much faster that I've done 3 tests and two of them have had the problem.  I took some video of it flipping at 4S and firing with no target at 7S. 

The 4S flips are reliable, but pretty lame as I expected. The 7S flip wasn't as fast as I hoped, but the reliability is a much bigger deal. I think what is happening is the nut is just getting part way onto the first thread and then the screw is coming around and instead of grabbing the first thread it's getting crooked and then the whole thing cross threads and jams.  I tried to fix the problem by getting rid of any partial-height threads I made with the lathe and grinding a counter-ramp into the first thread to encourage engaging or retrying without getting "half way".  That seemed to help but still failed on my second try. I am going to remake the screw tomorrow and start with a nice sharp counter ramp first thread to see if that helps. My description stinks but I tried to show it in the pictures.

I'm open to suggestions on ways to improve the engagement. The most obvious is to make it happen faster (higher speed servo, solenoid, etc) but I'm not sure I can fit that in the design in time. I will think on it some more.

4S lackluster flip @480 FPS
https://youtu.be/4Z68djhGX5I

7S only working flip (no target)
https://youtu.be/DTqQIJ5_GlY

7S Jammed flip (still uploading)
https://youtu.be/0-aiZcbHWTY

Pictures:
https://plus.google.com/photos/107279152885550274671/album/6433529956027270705/6440560546237723762

-Zac

15
Progress Reports / Re: Team Brain Damage
« on: July 08, 2017, 08:32:29 pm »
Well I have good news and bad news. The good news is that thanks to Kyle I have been able to successfully flash my red  brick 70A controllers with SimonK and set them up to be reversible.  It turns out these are third, newer version of the board and they swapped CpFET and BpFET in this version.  The entire mapping file for these new controllers (from Kyle Singer) is below. My boards had the 500403-1040-02 markings on them just like the comment says. The red bricks can take 8S A123 which is super-duper close to 7S Lipo. Because ThunderPower took so long to get my batteries out I won't have them until Monday so I had to improvise and use an 8S A123 pack to do my testing today. I'll post the rest of my progress in another entry


;********************************************
;* Red Brick 70A first seen November 2016  *
;* Blue circuit board labeled 500403-1040-02  *
;********************************************
.equ    F_CPU           = 16000000
.equ    USE_INT0        = 1
.equ    USE_I2C         = 0
.equ    USE_UART        = 0
.equ    USE_ICP         = 0
;*********************
; PORT D definitions *
;*********************
.equ    c_comp          = 6     ;10 i common comparator input (AIN0)
.equ    AnFET           = 5     ;30 o
.equ    ApFET           = 4     ;31 o
.equ    rcp_in          = 2     ;32 i r/c pulse input
.equ    INIT_PD         = 0
.equ    DIR_PD          = (1<<ApFET)+(1<<AnFET)
.equ    ApFET_port      = PORTD
.equ    AnFET_port      = PORTD
;*********************
; PORT C definitions *
;*********************
.equ    mux_c           = 7     ; ADC7 voltage input (47k from Vbat, 2.0k to gnd, 10.10V in -> .402V at ADC7)
.equ    mux_a           = 6     ; ADC6
.equ    CpFET           = 5     ; ADC5
.equ    CnFET           = 4     ; ADC4
.equ    BpFET           = 3     ; ADC3
.equ    mux_voltage     = 2     ; ADC2
;.equ                   = 1     ; ADC1
.equ    mux_b           = 0     ; ADC0
.equ    BpFET_port      = PORTC
.equ    CpFET_port      = PORTC
.equ    CnFET_port      = PORTC
.equ    O_POWER         = 47
.equ    O_GROUND        = 2
.equ    INIT_PC         = 0
.equ    DIR_PC          = (1<<BpFET)+(1<<CpFET)+(1<<CnFET)
;*********************
; PORT B definitions *
;*********************
;.equ                   = 7
;.equ                   = 6
;.equ                   = 5     (sck stk200 interface)
;.equ                   = 4     (miso stk200 interface)
;.equ                   = 3     (mosi stk200 interface)
;.equ                   = 2
;.equ                   = 1
.equ   BnFET            = 0
.equ    BnFET_port      = PORTB
.equ    INIT_PB         = 0
.equ    DIR_PB          = (1<<BnFET)

Pages: [1] 2 3 ... 30