Well...it happened. I was 99% sure it would happen and it did: Electronics problems. Here's what happened:
About three months ago, all 4 motors and controllers were working perfectly. Two weeks ago, all 4 motors and 3.667 of the motor controllers were working perfectly. We (Shane and I) discovered that one of the phases was messed up for no apparent reason. Anyways, I put the board under the work bench and got it out 2 days later. Literally nothing was different and suddenly NONE of the controllers were working properly. This is why I hate electronics. After many hours of trouble shooting (mostly shane being a genius), he came up with some stuff to try. We think the noise from the sensor wires being so close to the power wires is messing everything up. So I bought some shielded 5 conductor wire (to be installed next week) for the sensors, and Shane added some capacitors to the sensors. This cleared up the problem for 3 out of the 4 MC's. However, the one that was having phase issues still didn't work...and when one of the gate drivers started smoking...well, yeah. You get the idea. My shitty soldering job on a single resistor caused one of the gate drivers to smoke after awhile. The controller requires replacing 3 IC's, after which it should* (cross fingers) be good to go.
So as usual, it's not the electrons' faults that it didn't work...
Anyways, other stuff that's been accomplished: The Wii Nunchuck has been hacked apart so that a Xbee + board can fit. The trucks and risers have finally been properly bolted on. I used every single riser pad I bought to get enough ground clearnence for turning. Besides that, nothing has happened really.
Left to do:
1. Finish Wii-nunchuck controller (8 hours?)
2. Test and program motor controllers (8 hours)
3. Finish fixing electronics problems (10 hours)
4. Finish mounting motor controllers (45 minutes)
5. Test ride around a little inside (10 min)
6. Cut/Glue polyurethane tires on (24 hours)
7. Test drive
8. Install fans (5 hours)
9. Install lights (20 hours) (optional)
10. Waterproofing (4 hours + drying time)
Estimated work time left on project: ~80 hours (including lights).
One month down, two months to go. O.o
Saturday, June 25, 2011
Sunday, June 12, 2011
Weekly update
The motor controller mounts have been installed. I made them from 1/16" polycarbonate strips: measured the height the controller needs to sit, heated up the plastic with a heat gun until it was soft, then used some square blocks to form it into a little bracket. The fore and aft mounting brackets have a little hole drilled in them that allows a 4-40 screw to go through and into a mounting hole in the MC. Small pieces of super vibration dampening rubber will be placed between the mounting brackets and the MCs.
The little white wires running from one end of the board to the other in the picture below are wires from a magnetic deadman's switch that controls the contactor coil. A magnet on my foot closes the switch, thus activating the contactor...so when I fall off (not if), the contactor will open, and the board won't keep going.
Below are pics of the partially assembled board.
I tried standing on it and turning, but discovered that there's not enough ground clearance (I get edge bite). Luckily, I bought a ton of riser pads and installed all of them...it looks kinda funny, but that got the ground clearance up to a more reasonable 1 3/8 inches. It'll also give me more suspension, which is much needed with such a stiff deck.
While all of this was going on, I was balancing/charging my giant lipos. It turns out that one of the cells in one of the packs was much lower than the others. At first I thought it might be a bad cell, but the charger has been able to get it up from 3.4v to 3.94 at the time of writing this. I also learned that putting the charger on top of the power supplies is a bad idea (I know...duh...). I started smelling something awful about 15 minutes into charging and found the duct-tape that I used to hold the power supplies together in a block melting/burning...they got really freakin hot. Unfortunately, despite the separate power supplies, each individual charger maxes out at 50W (~2A for these packs), which implies that it is software limited and not power supply limited. I was hoping to get 5A out of them so I could charge the packs in under an hour...oh well.
I also did a bunch of wire routing. I'm about 3/4 the way done with getting all the motor wires routed. All of the essential inner wiring is complete (in other words, everything but the fans and lights). The nice sleeving stuff I bought works like a charm. Something I thought of is that I could have used ribbon cable or something like ethernet cable for the sensor wires, instead of running 5 separate strands...doing that probably would have made my life easier (yet another lesson learned).
I really really REALLY hope that none of the electronics have decided to stop working. Debugging this system takes forever, and taking apart a motor is a whole day (and destructive) ordeal. We'll see soon enough (queue ominous background music).
Left to do:
1. Finish wiring (3 hours)
2. Design/Make Wii-nunchuck controller (15 hours?)
3. Test and program motor controllers (8 hours)
3.5. Fix electronics problems (I'm 99% sure something will go wrong...it always does)
4. Finish mounting motor controllers (45 minutes)
5. Test ride around a little inside (10 min)
6. Cut/Glue polyurethane tires on (24 hours)
7. Test drive
8. Install fans (5 hours)
9. Install lights (20 hours) (optional)
10. Waterproofing (4 hours + drying time)
Estimated work time left on project: ~80 hours (including lights)
The little white wires running from one end of the board to the other in the picture below are wires from a magnetic deadman's switch that controls the contactor coil. A magnet on my foot closes the switch, thus activating the contactor...so when I fall off (not if), the contactor will open, and the board won't keep going.
Below are pics of the partially assembled board.
 |
| It's starting to look like a longboard. |
 |
| Hell yeah. Note: BWD is in the background |
While all of this was going on, I was balancing/charging my giant lipos. It turns out that one of the cells in one of the packs was much lower than the others. At first I thought it might be a bad cell, but the charger has been able to get it up from 3.4v to 3.94 at the time of writing this. I also learned that putting the charger on top of the power supplies is a bad idea (I know...duh...). I started smelling something awful about 15 minutes into charging and found the duct-tape that I used to hold the power supplies together in a block melting/burning...they got really freakin hot. Unfortunately, despite the separate power supplies, each individual charger maxes out at 50W (~2A for these packs), which implies that it is software limited and not power supply limited. I was hoping to get 5A out of them so I could charge the packs in under an hour...oh well.
 |
| Compare the 2 motors' wiring. Also note the massive stack of risers. |
I really really REALLY hope that none of the electronics have decided to stop working. Debugging this system takes forever, and taking apart a motor is a whole day (and destructive) ordeal. We'll see soon enough (queue ominous background music).
Left to do:
1. Finish wiring (3 hours)
2. Design/Make Wii-nunchuck controller (15 hours?)
3. Test and program motor controllers (8 hours)
3.5. Fix electronics problems (I'm 99% sure something will go wrong...it always does)
4. Finish mounting motor controllers (45 minutes)
5. Test ride around a little inside (10 min)
6. Cut/Glue polyurethane tires on (24 hours)
7. Test drive
8. Install fans (5 hours)
9. Install lights (20 hours) (optional)
10. Waterproofing (4 hours + drying time)
Estimated work time left on project: ~80 hours (including lights)
Sunday, June 5, 2011
June work
Well, bad news...I've decided to suspend work on the ELB indefinitely.
.
.
.
.
.
.
.
.
.
.
.
Just kidding.
I've actually just restarted work on it. I spent about 9 hours figuring out how I'm going to do all the wiring. The following picture shows about half of the result.
You can see the 4 blue lipos in the middle of the deck. Instead of linking the batteries together and running only a few plugs out of the box, I decided to go the "run-every-single-plug-out-of-the-box" route. While this meant more plugs, more drilling, and more soldering, I really wanted the ability to separate all of the packs (safety, maintenance, etc). Having each pack's terminal and taps also has a charging advantage: Since my modified HobbyKing 4 port charger (see previous post) can charge 4 packs at a time at 200W each (I think...) , I can charge faster than if I was charging the packs in parallel.
Anyways. It'll be apparent what the mess of wires is doing when it's done. The black thing on the right is a MAXI blade fuse holder. The black in the box is an Albright SU-60 contactor.
Left to do:
1. Design/make/instal mounts for motor controllers (6 hours)
2. Finish wiring (5 hours)
3. Install fans (5 hours)
4. Install dead man's switch (3 hours)
5. Blue-RTV holes (2 hours)
6. Mount trucks with risers (2 hours)
7. Design/Make Wii-nunchuck controller (15 hours?)
8. Mount ESC's and motors. Test and program motor controllers (8 hours)
8.5. Fix electronics problems (I'm 99% sure something will go wrong...it always does)
9. Test ride around a little inside (10 min)
10. Cut/Glue polyurethane tires on (24 hours)
11. Test drive
12. Finish waterproofing (2 hours)
Estimated time left on project: ~72 hours
I'd say ELB is about 80-85% complete at this point.
So, in accordance with the laws of project-dom, I've started designing my next projects before finishing this one. 3 more electric hubmotor powered longboards: a "light" version, a "heavy" version, and a secret version. "oooo, I wonder what it is..." :-P . I should have some preliminary CAD drawings by the end of the summer for "heavy". So, by extrapolating my current project work schedule, I should be done with the next three longboards by...oh...say...2020? 2025? haha.
In other news:
MIT Rocket Team update: We got 2nd place at USLI!
DBF: Top 10!
I won the Unified Engineering Flight Competition! Our plane carried a record breaking 941g of payload.
I'm really happy with the way the semester turned out...even though I was stressed beyond belief (and couldn't find time to work on ELB :( ) . I'm interning at Aurora Flight Sciences this summer designing/building UAV's...it's really awesome. I'll be doing ELB work some nights and on the weekends, so look for weekly updates.
.
.
.
.
.
.
.
.
.
.
.
Just kidding.
I've actually just restarted work on it. I spent about 9 hours figuring out how I'm going to do all the wiring. The following picture shows about half of the result.
You can see the 4 blue lipos in the middle of the deck. Instead of linking the batteries together and running only a few plugs out of the box, I decided to go the "run-every-single-plug-out-of-the-box" route. While this meant more plugs, more drilling, and more soldering, I really wanted the ability to separate all of the packs (safety, maintenance, etc). Having each pack's terminal and taps also has a charging advantage: Since my modified HobbyKing 4 port charger (see previous post) can charge 4 packs at a time at 200W each (I think...) , I can charge faster than if I was charging the packs in parallel.
Anyways. It'll be apparent what the mess of wires is doing when it's done. The black thing on the right is a MAXI blade fuse holder. The black in the box is an Albright SU-60 contactor.
Left to do:
1. Design/make/instal mounts for motor controllers (6 hours)
2. Finish wiring (5 hours)
3. Install fans (5 hours)
4. Install dead man's switch (3 hours)
5. Blue-RTV holes (2 hours)
6. Mount trucks with risers (2 hours)
7. Design/Make Wii-nunchuck controller (15 hours?)
8. Mount ESC's and motors. Test and program motor controllers (8 hours)
8.5. Fix electronics problems (I'm 99% sure something will go wrong...it always does)
9. Test ride around a little inside (10 min)
10. Cut/Glue polyurethane tires on (24 hours)
11. Test drive
12. Finish waterproofing (2 hours)
Estimated time left on project: ~72 hours
I'd say ELB is about 80-85% complete at this point.
So, in accordance with the laws of project-dom, I've started designing my next projects before finishing this one. 3 more electric hubmotor powered longboards: a "light" version, a "heavy" version, and a secret version. "oooo, I wonder what it is..." :-P . I should have some preliminary CAD drawings by the end of the summer for "heavy". So, by extrapolating my current project work schedule, I should be done with the next three longboards by...oh...say...2020? 2025? haha.
In other news:
MIT Rocket Team update: We got 2nd place at USLI!
DBF: Top 10!
I won the Unified Engineering Flight Competition! Our plane carried a record breaking 941g of payload.
I'm really happy with the way the semester turned out...even though I was stressed beyond belief (and couldn't find time to work on ELB :( ) . I'm interning at Aurora Flight Sciences this summer designing/building UAV's...it's really awesome. I'll be doing ELB work some nights and on the weekends, so look for weekly updates.
Subscribe to:
Posts (Atom)