http://www.techranch.com/blog/index.php Copyright 2008, Jim Greenhaw Jim Greenhaw en-US SPHPBLOG 0.5.1 http://www.techranch.com/blog/index.php?entry=entry080904-220453 blog , so all new post will be made there...

I hope you stick with me. The new site should allow more interaction...

Thanks,
Jim
]]> Other http://www.techranch.com/blog/index.php?entry=entry080904-220453 Jim Greenhaw Fri, 05 Sep 2008 02:04:53 GMT http://www.techranch.com/blog/comments.php?y=08&m=09&entry=entry080904-220453 http://www.techranch.com/blog/index.php?entry=entry080804-173355
Summary:

Nine robots actually competed
No excavator put regolith in the collection box
Only a few even moved out of the corner
Only one dug any material at all

Fortunately for me, the Tornado was in the last two categories, and was awarded the first place trophy. This is an honor, and better recognition than last year, but the $500k would have been nice.


Details:

I had the excavator pretty much finished last weekend and spent most of the week testing in my front yard, the damaged grass is proof. The machine made the trip up and down the ramp at least 100 times during the week, but I didn't realize that the cute BaneBots planetary gear boxes were getting worn out with the testing.

Friday night, around midnight, after cleaning up the machine, I did one more dry-run with the excavator up on blocks. During the run I could see that the right-hand track was not running, and in fact, the gears were stripped. It took about two hours to disassemble and repair the gear box. I thought a lock-up that had occurred a couple of weeks before is what caused the failure, plus I wanted to sleep a couple of hours before the competition, therefore I didn't inspect the left-hand gear motor... The robot ran for about 10 minutes Saturday morning before the left-hand drive completely failed.

What happened was the teeth of the final ring-gear stage simply stripped out, the strangest thing being that both failures happened essentially within minutes of each other. I will still use gear motors sold by BaneBots for projects, but regardless of whatever product I use, I will perform inspections before engaging a critical task.

This is very frustrating, and one starts to ponder all of the things that could have been done differently to effect a different outcome. I know that other teams were having the same thoughts.

But there is more. After setting up for the excavation attempt in the sand box, the judges "noted" that the marker panels that I attached to the ramp extended past the allowed area for markers. I had built the panels for a 31 inch ramp, and the actual size of 36 inches wasn't even announced until the week before. It never occurred to me that they would be too wide. Later I asked a judge if I would have been disqualified because of the violation, and he suspected that I would have been. That would have been worse than the failed gear box...

It was a lot of work, but I'm glad I did it. Every machine had at least one interesting design feature, and I met some nice people, too...

I put some pictures on the web here: Tornado Photos

I didn't get the huge cash sponsors that I had hoped for, but I did get some support:

Hapman provided the custom-made auger at no charge, and openly offered their support to the effort. Thanks!

Wire Mesh Belt Company did a quick turn-around on the custom wire mesh tank treads, and generously discounted the price.

Robot Power supplied the motor controllers at a discounted price.

Bob CAD/CAM made designing and milling the parts a lot easier

Other companies that made this possible: McMaster-Carr , Mouser Electronics , BaneBots , Kirshaw Designs , and The Perfect Stitch.

I guess I'll take a little vacation before working on next year's machine... ;-)
]]> Regolith Excavation Challenge http://www.techranch.com/blog/index.php?entry=entry080804-173355 Jim Greenhaw Mon, 04 Aug 2008 21:33:55 GMT http://www.techranch.com/blog/comments.php?y=08&m=08&entry=entry080804-173355 http://www.techranch.com/blog/index.php?entry=entry080707-152247
I am happy with the progress, but still lots to do. Finalizing the drive train, and still looking for the "just right" motors. I am using a conveyor to lift material into a hopper, and stumbled upon a key piece of information that finally led me to an excellent design. Preliminary testing yesterday gives the green light...

I received 500 lbs of JSC-1A in March, and built a 4x8 sand box. It doesn't allow driving around much, but makes a big difference for testing mechanisms. I suspect that others bought some, but I don't know. A five month lead time didn't help matters.

It sounds like everything will start Saturday morning, I'm not hearing about a Friday night mixer like last year. With the quarantine at the start of inspections, it doesn't look like there will be much time for on-site, last minute tweaks.

I was reflecting about the absoluteness of the whole thing while thinking about actual NASA missions. We get one shot; put the machine in the box, hook up the power, say "Go", and hope for the best. It was the same with the early Apollo rockets, some went up, some blew up. I'm sure the engineers were gritting their teeth before launch just like we will be… ;-)
]]> Regolith Excavation Challenge http://www.techranch.com/blog/index.php?entry=entry080707-152247 Jim Greenhaw Mon, 07 Jul 2008 19:22:47 GMT http://www.techranch.com/blog/comments.php?y=08&m=07&entry=entry080707-152247 http://www.techranch.com/blog/index.php?entry=entry080623-190643
The video went well, as most sub-systems are operational. I'm still working on the navigation, and it's going fairly well, however, I still need to examine plan B.

After the registration close date I expected to hear about the remaining teams, but was surprised to hear that 5 more signed up. As with last year, I suspect that some won't make it to the competition, perhaps 50%, but we'll see. I don't know what the refund policy is for those who opt out. Regardless, it will be a busy weekend.

With so many teams involved, it will be fun to see all of the designs. Things have been quiet on the web, as expected, so it will also be interesting to what gets published after the competition.

Be sure to check the FAQ often, or set up an RSS feed to it. I was surprised to see the talk about passive components. Since I hadn't really included any in my design, it made me wonder what I was missing. I have been whining about the ramp being a side effect of an un-filled sand box, but Matt assures me that the design challenge offered by the ramp is something NASA wants specifically addressed. I told him I would stop whining about it now.

I also expressed some concern about voltage drop at the robot power connector, and that due to the long power umbilical this year, it could be a real issue. I suggested using the third contact of the power connector as a sense line for a regulated power supply (perhaps better than last year's) to make sure we actually have 24 VDC delivered to the bot. If others share this concern, please send a message to CSEWI…

I guess that's all for now, the countdown continues…
]]> Regolith Excavation Challenge http://www.techranch.com/blog/index.php?entry=entry080623-190643 Jim Greenhaw Mon, 23 Jun 2008 23:06:43 GMT http://www.techranch.com/blog/comments.php?y=08&m=06&entry=entry080623-190643 http://www.techranch.com/blog/index.php?entry=entry080530-120352
The next beam power challenge is coming up this Fall, and it's not too far away. I once thought about entering, but I just don't have time for both contests. So, I will give some design secrets away to those who are in the contest, and happen to read this. They might not be new to some, but they are design approaches that I intended to use.

PV panels exhibit some interesting properties regarding the relationship between their voltage and current outputs, and some side-effects because of the way they are built; individual cells wired in series. This means that unless the entire panel is illuminated evenly, you may get only as much as the least illuminated portion, which might be dark, hence zero output.

Here is what I wanted to do.

First, make your own panels from individual cells. Wire them in series, and put a reverse-biased germanium diode across each cell. This way, if a cell is dark, current generated by other cells can pass through instead of being blocked.

Next, wire the cells in long strings to produce 70 - 100 volts, and connect them to a power point tracking (PPT) regulator, which is really a DC-to-DC converter with a constant output voltage. However, there is an important addition. PV cells output their peak power at a certain voltage and current, which is controlled by the load impedance. The true PPT regulator adjusts its input impedance to operate the cells at this point.

The other part of the above design is that even when illuminated poorly, you might still get 20 or 30 volts, for example, and while this won't translate into a lot of POWER, at least you will get something. Whereas, with a 12 volt panel, running at 18 volts fully illuminated, you might get 10 volts when poorly illuminated, and that does nothing for a 12 volt system.

So, in my opinion, these three features can add greater chances of success to beam power challenge teams. It isn't rocket science, just squeezing every watt of power possible out of a PV system.

In summary:

1. Construct panels from strings of cells with reverse diodes across each one.

2. Make the strings long for higher available power.

3. Use a PPT regulator to step-down the high voltage, and operate at maximum power.

Good luck,
Jim
]]> Other http://www.techranch.com/blog/index.php?entry=entry080530-120352 Jim Greenhaw Fri, 30 May 2008 16:03:52 GMT http://www.techranch.com/blog/comments.php?y=08&m=05&entry=entry080530-120352 http://www.techranch.com/blog/index.php?entry=entry070928-115318
The 2007 challenge was a little strange. The first prize was $125,000, and that should have been enough to cause a lot of interest. Initially, I heard that over 100 teams expressed the intent to enter, but by the time I entered, only 9 were actually planning to attend, and only 4 showed up at the competition. Cal Poly San Luis Obispo is 30 miles from the contest site, why weren't they there? I can only imagine that they all thought it would be too difficult.

I didn't enter until I determined that I could use a stationary machine, but I expected other teams to have fancy rovers picking up material and dumping it into the collector. I was very surprised when the other three teams all had stationary machines.



After the 2007 contests I talked with one of the people running the event and he said they too were surprised that no one made a roving machine, hence the 2008 rules that necessitate one.

The 2008 challenge is definitely a new ballgame, but not outside of what can be done. Power and weight shouldn't be an issue, as they have bumped both of those limits up considerably. However, the navigation problem makes up for that.

Avoiding the rocks should be pretty easy, and if a machine is small enough, it won't have problems going around them. The requirement for a vision system to find the ramp is the big one. I suggested using different colored lights instead of flat targets, which would simplify this a great deal, but I will be surprised if they go that way.

Another big issue is driving up a 20 degree ramp. This is difficult even under human control, autonomously will be a challenge. Last year the JSC-1A settled more than they expected, and that is why there was such a high sidewall margin above the surface. They could cut the sidewalls down, or add some sand to the bottom first, as a way to decrease the ramp slope.

I sent about 10 emails with comments about unclear or potentially problematic points in the draft rules. Last year when I pointed out a particular ambiguity, or difficulty, the answer was sometimes, "That's part of the challenge", so I expect to hear that some more.

With the first prize being 1/2 million dollars, I suspect that this will get the attention of more people. I do think there will be more teams this year, but perhaps not more than a dozen. The 2008 rules filter out a lot of amateurs that might not be up to the task, which implies that there might be some exciting high-tech competition going on.

That prospect scares me a little, but I am committed to enter. I have most of the initial design worked out, and am starting to prototype some components. My intent is to get something built soon enough to have time to run it for a couple hundred hours. I would like to tell everyone about it, but it is difficult in a competitive environment...

I will post answers to more questions, and expose a few tidbits about my machine as time goes on…

Jim
]]> Regolith Excavation Challenge http://www.techranch.com/blog/index.php?entry=entry070928-115318 Jim Greenhaw Fri, 28 Sep 2007 15:53:18 GMT http://www.techranch.com/blog/comments.php?y=07&m=09&entry=entry070928-115318 http://www.techranch.com/blog/index.php?entry=entry070821-114555
No other member of Congress has the integrity, or the consistent voting record of this guy, I am totally impressed. Most of his ideas to repair the United States are a little on the fringe, but this country has been taken to the fringe by a long history of greedy politicians.

Ron Paul can sell himself, he doesn't need my help...

Watch his videos on YouTube, and see if you can find a more qualified candidate for President of the United States.

Part with a few dollars to help his campaign.

Jim
]]> Other http://www.techranch.com/blog/index.php?entry=entry070821-114555 Jim Greenhaw Tue, 21 Aug 2007 15:45:55 GMT http://www.techranch.com/blog/comments.php?y=07&m=08&entry=entry070821-114555 http://www.techranch.com/blog/index.php?entry=entry070628-180929
A photographer from the magazine spent a day at my house a couple of weeks ago, so I am eagerly waiting to see how it looks, what they said.

Check it out... NY Times Magazine online

Jim]]> Regolith Excavation Challenge http://www.techranch.com/blog/index.php?entry=entry070628-180929 Jim Greenhaw Thu, 28 Jun 2007 22:09:29 GMT http://www.techranch.com/blog/comments.php?y=07&m=06&entry=entry070628-180929 http://www.techranch.com/blog/index.php?entry=entry070515-012037
There were 4 teams at the competition on May 12th, and my machine was the only one to run the entire 30 minutes. It also moved about 75 kg of JSC-1A, 65 of it going into the collector, 10 on the ground. It also ran exactly at the specified limit of 30 watts average. Rob Mueller of NASA made the point that is the power used by a refrigerator light bulb, and it is amazing that any real work can be done with it. Even though it didn't move the minimum 150 kg to win a cash award this time, I think that it can be done.

This is one of the cases where simple was better. The machine can be described as an inverted bucket, continuous chain driven drag. The buckets are moved to the end of a 4 foot horizontal boom where they flip over and contact the soil, and drag it back towards the delivery ramp at the root of the pivot. The machine uses 3 motors driving the bucket chain mechanism, and one to pan the boom from side to side. The boom pivots vertically as the hole it digs becomes deeper. I'll have some pictures posted on a web site soon, but I'm still recovering from the last couple of weeks...

In the meantime, here are links to a couple of small avi files that show an earlier version in action during a test on some clay soil:
Front View
Back View



All of the machines relied on a stationary base that secured a mechanism that moved across the simulated regolith and moved material to the collector. I think this was a surprise to the contest hosts, and they expected more autonomous vehicles driving around the sand box gathering up the JSC-1A and dumping it into the collector. I suspect that rules for the 2008 contest will try to include an element of transportation in addition to excavation, we'll see.

I don't want to give away too many secrets about the JSC-1A yet, since I will be competing again, but I will say that testing an excavator in hard clay soils won't work; it's too hard. The simulant behaves more like fine sand, but it has some properties that make it different from sand; it is some strange stuff.

The stated goal of the Centenial Challenge program is to get inventors involved that might not have the resources to deal with NASA and the grant process. The entry fees are very resonable, and the excavation challenge in particular is accessable to a wide audience, unlike some of the other challenges, so I encourage everyone to give it a shot.

I will post some more info as soon as I can...

Jim
]]> Regolith Excavation Challenge http://www.techranch.com/blog/index.php?entry=entry070515-012037 Jim Greenhaw Tue, 15 May 2007 05:20:37 GMT http://www.techranch.com/blog/comments.php?y=07&m=05&entry=entry070515-012037 http://www.techranch.com/blog/index.php?entry=entry070415-174714 link to the information site.

This is a NASA sponsored contest to create a machine that can excavate lunar soil to be processed, one process being removing breathable oxygen from the soil.

One key reason that I decided to enter is that the competition takes place about 15 miles from my house; I couldn't pass up the opportunity! It doesn't hurt to have a total purse of $250,000 available, but in reality, the most value will be from all of the people that I will meet at that time.

This particular challenge involves moving simulated lunar soil from a 4 meter square sandbox into a 1 meter square collection box. The machine that moves the most material, over 150kg, wins first place. Oh yea, did I mention that you get 30 watts of power for 30 minutes to do this?

I don't want to say much yet about my design approach, but I will say that I have a prototype machine that already digs pretty well, and there is plenty of time to work out the details.

More later...]]> Regolith Excavation Challenge http://www.techranch.com/blog/index.php?entry=entry070415-174714 Jim Greenhaw Sun, 15 Apr 2007 21:47:14 GMT http://www.techranch.com/blog/comments.php?y=07&m=04&entry=entry070415-174714