Sawdust not ready for prime time

Designing a Mini2440 Faceplate for Chameleon 1

landon@360vl.com — 2009-11-09T08:23:40-07:00

Synopsis: ESawdust has had several requests to create a Chameleon 1 faceplate for the FriendlyARM Mini2440 ARM9 development board that’s suitable for running embedded Linux, Windows CE and other embedded OS’s. This article describes the processes we go through to layout a faceplate and create cutouts for shapes we need to allow connectors to protrude through the face. By the end of the article you’ll see a completed prototype faceplate for the Mini2440 and how we got there.

The article is broken down into sections including measurement, layout, prototyping and finally, there’s a video tutorial on how to create a complex cutout shape, a DB9 cutout, using a consumer grade C.A.D. package, ViaCAD (Windows and Mac versions.)

ESawdust follower, @BusError (Michel), maintainer of the embedded Linux distro for the Mini2440, put me in touch with Charlie Springer of Industrial ARMWorks who was kind enough to supply us with a Mini2440 and FriendlyARM displays to start sizing up.

The Mini2440

In case you’re not familiar with the board, this is a top-down shot of the Mini2440. I put a physical ruler nearby so I could calibrate photos. I have a virtual caliper in the image also which is showing the board to be around 3.94” square.

Measuring

The first order of business is to determine the orientation of the board and identify the side that makes the most sense to expose via the faceplate. In the case of the Mini2440, the side with power, audio, RJ45, USB, RS232 was the obvious choice.

I measured the board with physical calipers and got a 3.94” square board. Also, using the screen caliper software, I calibrated the virtual calipers to the photos I took of the connector edge. I then began measuring all the typical dimensions such as:

1) Offset of the left side of each of the major connectors (except DB9) to determine the relative position of the connecter from the left side of the board. Here’s an example of measuring (virtual calipers) from the board left edge to the left edge of the mag jack.

While measuring the key dimensions, I take down all the numbers in a spreadsheet (shown later) for eventual faceplate placement calculations.

Layout

The overall Chameleon 1 faceplate is a pretty simple shape on bounds. The baseplate on which the board is mounted is 4.9” wide and the enclosure lid and faceplate are example 5” wide so the lid fits snugly around the baseplate. You can see an example of the Mini2440 resting on 1/4” standoffs on the baseplate of the Chameleon 1.

Based on the width of the board and the width of the baseplate, assuming the board is centered on the base plate I can easily compute the “easements” or margin between the board sides and the side of the baseplate. (4.9” - 3.94”) / 2.

The baseplate is 0.4” tall. The standoffs add another 1/4” and the PCB is about 0.063” thick, so that is the base offset for the height of all the connectors on the faceplate.

So, the two main layout tasks are:

1) template of the board hole patterns to drill on the baseplate so the board perfectly lines up with the faceplate cutouts we’ll do in the next step.
2) faceplate connector cutouts so the cutouts are exactly the right height and right offset when the board is mounted according to the spec.

It’s convenient, when possible, to have the board connectors protrude slightly from the faceplate rather than be very recessed. It helps make the faceplate fit feel solid and positive. Fortunately, with the Mini2440, they did a good job of extending the connectors past the edge of the board so we can line the board’s mounting holes up pretty close to the edge of the baseplate and get a nice fit for the faceplate.

Using a CAD package called ViaCAD, I designed for mounting the Mini2440 on the baseplate of the Chameleon 1. Here’s a scaled down image of the template for the base plate mounting for the Mini2440:

The template is made to cut out and overlay over the Chameleon 1 base and has the center-punch markings in the middle of each mounting hole which you need to easily center-punch the mounting holes for a perfect fit.

The faceplate template is a lot more challenging than the baseplate.

I start with a blank faceplate design, again done in ViaCAD, save it as a new faceplate and remove all the dimensions from the drawing (planning later to prototype with it, so using a laser cutter, for example, I don’t want anything except the cutouts as lines on the design.

Here’s the blank faceplate design I start with:

Using all the connector height, width, and offset measurements I mentioned earlier, I create a spreadsheet that helps to easily identify the X / Y offsets of the various connectors as they map to the 5” x 2” faceplate. The spreadsheet accounts for the width of the board, the margins on the side of the baseplate, the baseplate height, PCB thickness, standoff height, and full width of the faceplate.

Here’s a quick example of a working spreadsheet I use to start laying out the connector cutouts on the faceplate:

This serves as a pretty accurate but not final guide to initially placing connector cutouts on the faceplate.

Most of the cutouts are pretty simple - holes and squares. Love those. The harder cutouts are for odd shaped connectors like DB9. So, one of the things I had to create was a DB9 cutout for the Mini2440. Later on in this article I developed a video tutorial on how to create a DB9 cutout from scratch using a CAD package. Once I have the DB9 cutout, I can add it to the faceplate design.

Using the spreadsheet computations and DB9 cutout, I can complete the Mini2440 faceplate design which looks like this (image is not 100% scaled):

Prototyping

Before submitting a design like this to have manufactured at a sheet metal fabricator or laser cutter for a lexan version of it, we like to prototype it by using cardstock or some type of heavy paper. This lets us see exactly how everything’s lining up and tweak any positions or sizes if needed.

So, the beautiful thing about using CAD programs is you can print a design at scale and use that scale pattern to test and prototype.

Using an 8 1/2” x 11” piece of black card stock, I cut out an oversize boundary around the printed faceplate design and tape it to the card stock.

Using an Exacto, I cut the holes out of the design along with all the connector cutouts. The last thing is to cut the faceplate boundary out of the card stock from the template and remove the white plate off the cardstock.

Next the Mini2440 can be mounted on the baseplate. Using the baseplate template, I line it up with the physical baseplate of the Chameleon 1 and tape it in place directly on the baseplate. This will guide the next step, creating the center-punched holes for mounting the board.

The baseplate template in place, I center punch the holes for the board. It’s best to cut a short, thin piece of wood (0.35”-0.4”) over which the baseplate can straddle so the center-punching doesn’t bend the baseplate. It supports it so you can easily drill it later, too.

Next, flip it over and using a sharp awl or nail, tap a small center-punch dent into the base plate precisely where the template center holes are (there’s a small dot in the center of the mounting hole.) The center punch will keep the drill bit from skating when you drill out the holes. This is really important to make sure the faceplate lines up perfectly in a later step.

With the 4 mounting holes center-punched, you can see the small indents in the baseplate where you drill:

Next, simply drill out the center punched holes with a drill press or hand drill.

I used a hand drill. 3/32” bit for 4-40 holes works well. Keep the wood block underneath the baseplate during drilling to catch the drill when it breaks through.

With the mounting holes drilled, I can add 1/4” standoffs and the 4-40 screws to attach the Mini2440 to the baseplate:

Using the black faceplate prototype I cutout a few steps back, I can literally screw it onto a Chameleon 1 and start checking it out. From a distance it looks remarkably like a real faceplate. On my first prototype, the DB9 cutout came in too high and too small for the hex studs, so I learned right away what I had to do, trimmed the cardstock and rechecked the fit.

Here’s the prototype faceplate on a Chameleon 1 taken in a light studio to help see what the production version might look like:

We’ll be creating a large cutout in the top of the lid for the Friendly ARM LCD display, but for those who use the board standalone without the display, a Chameleon 1 basic + a production version of this faceplate is all that’s needed to have a sweet lid for the Mini2440.

For those of you interested in how to create complex shapes for faceplate cutouts, you can watch the video below where I do a tutorial on creating a DB9 cutout with a CAD tool.

Tutorial for Creating a DB9 Cutout

One of the key skills you need to create your own faceplates is to be comfortable with a CAD tool. In our case, for the production units, our designs go into SolidWorks because our Chameleon 1 manufacturer, Advantage Manufacturing (Colorado Springs, CO), has the technology to take the SolidWorks CAD models directly into sheet metal CAM files needed to punch it all out to spec in industrial volumes.

However, if we’re making a few faceplates and can’t justify punching several hundred that are required to make going through sheet metal manufacturing, we can have the faceplate laser cut on Lexan in a number of colors like dark smoked or even sold black. This is a very cost effective useful way to make custom, professional looking faceplates for Chameleon 1 enclosures.

Fortunately, the kinds of functions and features needed to create complex cutouts on a faceplate can be easily achieved with a consumer level CAD package. We use ViaCAD for these.

The video tutorial below takes you through making a complex cutout shape like a DB9 using ViaCAD and this is the design I’ve used for making the Mini2440 faceplate’s DB9 cutout:

Make a DB-9 Cutout with ViaCAD from Landon Cox on Vimeo.

Wrap up

We have a few more things to wrap up this week on the design including getting a lid cutout for the FriendlyARM touch display. Our next steps are to get a Lexan version of the faceplate made and do our final adjustments. We’ll probably first offer the faceplate in a smoked Lexan and see how sales go before committing the faceplate to sheet metal. Will update the blog and site when the Mini2440 faceplate is available for sale.

If you have a popular development board for which you’d like to see a new Chameleon 1 faceplate, contact us and we’ll see what we can do to help you out.

Landon Cox
www.ESawdust.com

Sparkfun VS1000 Ogg Vorbis Player Review

landon@360vl.com — 2009-09-19T10:56:36-06:00

Summary - This is a review of the Sparkfun VS1000 Ogg Vorbis audio player in the context of using it for a microcontroller sound extension. The review and use of the Sparkfun Ogg Vorbis board is primarily contained in the video piece accompanying this article. Sparkfun released this board the week of September 14th, 2009 and this is an initial review of the device. [Full disclosure: we do business with Sparkun - we’re a reseller and they resell a product of ours, the Chameleon 1.]

Speed Climbing Timing Lessons Learned

landon@360vl.com — 2009-07-14T07:29:32-06:00

Synopsis - the speed climbing timing system got its biggest test yet when we deployed two systems for the USA Climbing nationals in Sandy, UT near Salt Lake City this past weekend. This article is about some of the technical lessons learned. Previous articles in the speed climbing timing series discussed many other lessons learned as we developed, but this one is specific to the Nationals event. It includes a video of various speed races showing how the system works in a real event.

Overall I had some very positive feedback on the system, particularly the laser based hand sensors. The climbers, coaches told me, really liked them. They’re very sensitive and positive and easier to hit than a small button switch they’re used to. They’re easy to coach to, also. You have to touch the yellow line or it will not stop, simple as that.

The System in Action

Video after the breakl...

ESawdust Tour (Electronic Sausage Making)

landon@360vl.com — 2009-07-06T08:46:06-06:00

Synopsis - the whole point of this article is: “projects are messy”. When I’m in the middle of building, I am always thinking - “A ‘professional’ machinist, woodworker, or EE would know where everything is and would have the decks clear except for what he’s working on at the time”. I envision this ideal engineer out there with a pristine machine shop or clean-room quality lab amiably knocking out cool stuff right and left. I labor thinking everyone who’s doing things like this must be orderly. Me? Ha. I’m always looking for my wire strippers, phillips, or multimeter buried under the archeological layers of schematics and spools of wire. Maybe you’re the pro who has it orderly, but I throw this out for public consumption because for me, this is my reality when I’m in the middle of prototyping or building projects. Look at this as a small tour of ESawdust labs - such as they are.

[Afterthought: maybe this article is more about what defines an ideal “office” for a “maker” or “builder” which involves embedded hardware, software, metal, wood, power tools and paint. Not saying I have the ideal setup (far from it), but if you have thoughts on productive prototyping environments - tweet me at @esawdust and let me know]

Many more pictures after the break

Speed Climbing Laser-based hand sensor

landon@360vl.com — 2009-06-25T21:10:10-06:00

Synopsis: Based on more test results from the USAC Northeast divisional championships, the IR-based sensors were fairly effective (195 out 200 climbs on the qualifier day worked fine, 5 obviously didn’t). There were some issues getting it running during installation. I needed to rethink the hand sensor. There were 2 problems. One, as I’ve demonstrated in other articles, there are weak or dead IR zones a climber could hit if he didn’t hit it with a flat hand - the sensor would not stop the clock in that case. Two, there were glitchy problems when the long (50ft) cable we made was used from the display to the hand sensors. The behavior was similar to EMI problems though as I was not at the New England competition, I had to surmise based on descriptions I heard. This is the redesign of the hand sensors to fix these two problems before the USAC national climbing competition in Sandy, UT July 10-12, 2009.

Larger images later in the article.

Flip UltraHD - It's Not Just Me Who Has Problems With It

landon@360vl.com — 2009-06-24T16:51:42-06:00

Synopsis: My last article, Flip UltraHD - It Doesn’t Just Work, must have struck a chord because it’s been the most popular article on the site since I posted it. Judging by the search strings I’m seeing that land people on the sawdust website, it’s clear others are having some of the same issues. Here’s a sampling of FlipHD related searches I’m seeing hit sawdust:

"flip ultra hd An error occurred while interacting with your camcorder. Please make sure your camcorder is connected to your computer."
"flip ultra hd charging paused"
"flip firmware ultrahd"
"Flip ultraHD check battery pack error"
"flip ultra hd, battery doesn't charge"
"flip ultrahd upgrade"
"why does the flip ultrahd say check battery pack?"
"check battery pack charging paused Flip UltraHD"
"Error code: 27, 2 An error occurred while interacting with your camcorder. Please make sure your camcorder is connected to your computer. Error removing file '/Volumes/FLIPVIDEO/System/VIEWER/FlipShare.ico'. Thread info: Upgrad",1,1.1,""
"flip camcorder error code 27,2"
"flip ultrahd firmware version"
"why doesn't my flipshare camera always come out HD
"flipshare 4.5",
"how much time left on flip ultra hd battery"
"Flip ultraHD battery doesn't charge"
"flip video problems"
"flip cam error 27,2"
"flip video error 27,2"
"Flip Error code: 27, 2"

Flip UltraHD - It Doesn't "Just Work"

landon@360vl.com — 2009-06-18T19:57:02-06:00

Synopsis: I recently bought a Flip UltraHD after reading many good reviews for it. However, my experience with this device was completely different. The bottom line is that I could never upgrade the firmware in the device on either Windows or OS X. I wasted a lot of time with Flip's tech support. After having installed the latest FlipShare desktop software, the tech support had me download it all again (115MB), uninstall everything, reinstall it. There were some real surprises, too, like no reset switch...if you want to hard reset the UltraHD, you have to pull the battery and wait 2 days (not joking.)

Improving Screen Caliper Accuracy

landon@360vl.com — 2009-06-17T20:11:06-06:00

Synopsis: In the prior article, Real-world Measurements in Virtual Space, I introduced the concepts of using virtual calipers to measure real-world objects on your computer screen. In this article I show how to improve the accuracy of the measurements you take on the screen by carrying forward the same example of measuring various dimensions of a small printed circuit board. It's possible to greatly improve the accuracy of Screen Calipers by wisely choosing your calibration points.

In the first video of the previous article, I showed the calibration process by calibrating using the known spacing of the breakout board pinouts. Because this board has standard 0.1" pin spacing, I calibrated based on the outside edge-to-outside-edge of a pin pair that represented what I knew to be 0.1". This is a good start but can be improved. Here is a demonstration of the percentage errors and improvements in accuracy based on better calibration.

First, here's the actual caliper measurement of the physical board in order to show what the goal is. Obviously, if you have the physical board and calipers, you wouldn't necessarily need Screen Calipers, but I'm using real calipers to compare to Screen Calipers in order to illustrate how accurate Screen Calipers can be. The only way to do that is to compare actual measurements or an engineering drawing.

These measurements are inches because I calibrated to 0.1" and wanted to keep the units the same throughout the illustration. So, the actual width is 1.35 inches according to my real-world calipers. The height is measured:

and is 1.105 inches.

Compare these actual values to the initial Screen Caliper measurements when the calibration was done on a single 0.1" hole spacing:

This shows 1.424 vs the real 1.35 inches - off by 7/100ths of an inch. Doesn't sound like much, but it's quite a bit and we'd like to do better. The height was measured by Screen Calipers to be:

1.156 inches versus the actual 1.105 - off by a little over 5/100ths of an inch.

Instead of calibrating to a single 0.1" hole spacing, I calibrated to 10 holes or exactly 1" on the board. After setting the calibration to a larger length of known dimension, it's possible to get very accurate readings. After the new calibration, here are the results:

Screen Calipers shows 1.352 vs the actual 1.350 - we reduced the error from 7/100ths to 2/1000ths of an inch just by picking a better calibration point pair. The height compares like this:

Screen Calipers reads 1.098 vs the actual of 1.105. This reduced the error from 5/100ths to 7/1000ths...again a great improvement.

The reason calibrating on a smaller scale, even if the distance in the photo is well known, is that there are both accuracy and precision errors in setting the length and that error is a larger percentage of a small distance than it is a large distance. So, by calibrating to a small distance and then measuring a larger distance, you multiply that error over and over across that larger distance. By calibrating to a larger known distance, that error is a smaller portion of the distance and therefore any error is much less when measuring smaller or similar distances.

Conclusion

The moral of the story is that you can get highly accurate Screen Caliper measurements by calibrating to the largest known distance within the subject photograph.

Real-world Measurements in Virtual Space

landon@360vl.com — 2009-06-10T13:44:37-06:00

Synopsis: This article shows how to use software calipers to measure the size of objects of any scale in images on your computer screen. I was looking for this exact solution for a long time and didn't even know what to call it in order to search for it on Google. The software is called "Screen Calipers" and it's exactly what its name says.

I searched for "Virtual Calipers", "Software Calipers", "Virtual Measurement" and other variations before finding this solution. I also tried some other software packages which were no where near as good as this, so I hope this review is helpful in identifying a genre of software that's a bit difficult to find but extremely useful.

In this example, I demonstrate how to measure the distance between two mounting holes on a circuit board without knowing the dimensions of the board or having an engineering drawing for reference.

Screen Calipers is cross-platform software for both Mac and Windows - Screen Calipers Macintosh and Screen Calipers Windows version are only $29. It's really a great tool if you spend much time sizing up physical objects and images.

The Problem: I'll use a real-world example: Lets say we wanted to know the dimensions of the mounting hole pattern of a Sparkfun XBee board, the XBee Explorer Regulated. We start with photos of the board which look like this:

Accurate dimensions are not published for this board. The goal is to find the distance between mounting holes on a circuit board using just the top-down image of the PCB.

Discussion

While Sparkfun helpfully superimposes a ruled gauge with top-down pictures, those aren't sufficient to get real dimensions for the hole pattern. Also, they don't supply the actual physical (CAD) dimensions on their site. Referring to just these pictures, we can still figure out fairly accurate dimensions of this board even if the superimposed gauge didn't exist.

The gray rulers in the image are added in post-production with Adobe Illustrator, so while I think they are good for general size considerations, they're mainly there as a scale "hint" rather than a full engineering dimension. We'll be able to get a lot more accurate measurement with Screen Calipers and measure anything we want on the board while we're at it including the hole spacing so we can design something that mounts this board.

Step 1) Start with the largest image you can get to display on your screen - this will help the accuracy of this technique. Also, the less oblique you can get the image, the better. Top-down is great. Lots of images are top down including maps, so this same technique would work with many maps and miles or kilometers (assuming a Plate-Carree equirectangle geographical, ie: square projection is used) as well as it works with PCBs and millimeters.

In our case, the image itself is larger than actual size which is good:

Step 2) Calibrate the Screen Calipers to some known dimension on the board. It's tempting to calibrate to the gray rulers since they're right there, however, since I know those are added to the image in post production, I don't trust their accuracy and take them only as a hint. It's better to calibrate to something directly on the board itself. In this case, we know the hole patterns on the breakout board are on the standard 0.1" spacing which makes this board easy to plugin to protoboards and breadboards. I know this board was produced with a schematic capture tool, Eagle, so I know the hole spacing is very accurate.

To calibrate the calipers, launch the Screen Calipers and overlay the device on the 0.1" spaced holes. Theoretically it's the hole centers that are 0.1" apart, but in practice you can measure from the outside left of one hole to the outside left of the adjacent hole and that should be 0.1" as well and it's easier to align a caliper guide to a hard boundary like that.

This movie demonstrates this process. Once the calipers have been positioned, you can set the calibration which is essentially telling Screen Caliper software that X many pixels equal Y many "units". In this case it's about 25 pixels to 0.1" inch.

Screen Calipers Calibration Demonstration from Landon Cox on Vimeo.

Step 3) Now that we've calibrated the Screen Calipers to a known dimension on the board, we can start measuring stuff. In our case, we want to figure out the dimensions of the center-points of the mounting holes on the board. To do that we could estimate by placing the calipers into the middle of the open holes by sight, but it's a little more accurate to measure the outside dimension of one hole to the outside of the other hole, then measure the inside to inside of the same holes. You can also get the diameter of the hole itself with the calipers.

Once we have the outside to outside and inside to inside measurements, the center points can be derived.

This movie shows the process of measuring the outside-to-outside and inside-to-inside dimensions of both the horizontal and vertical mounting holes on the board.

Real-world Measurements in Virtual Space from Landon Cox on Vimeo.

If the outside-to-outside horizontal measurement is X and the inside-to-inside horizontal is X', then X-X' is equal to 2X the diameter of the holes. The center-point is 1/4 of that number offset from the outside of the hole towards the middle. Same with the vertical - Y and Y'.

Using the other measurements such as the physical width and height of the board, the center-point distance of the mounting holes can be computed based on the edge of the board also.

All this is much easier to do off a high-quality, enlarged photograph than it is the physical object itself. Sometimes you don't have access to the physical object or its engineering drawings but you can determine a great amount of physical data using this technique. If you can find an object in the photograph that has a known size - say a coin, a Bic pen, hole spacing, soda can - then you can calibrate the calipers to give you some interesting distance information.

I still consider these measurements estimates, but they are quite accurate in lieu of not having engineering specs.

Conclusion - the Screen Calipers provide all the information needed to determine the dimensions of the board and various offsets on the board to determine things like center points for mounting. You can click the links below which take you to an ESellerate commerce page if you'd like to buy the software.

Buy Screen Calipers for Mac

Buy Screen Calipers for Windows

Speed Climbing Timing - Installation

landon@360vl.com — 2009-06-10T10:29:24-06:00

Synopsis: This article provides installation instructions and hints for how best to install the speed climbing timing system.

Both timing systems arrived at their destinations for the southwest and northeast divisional championships. Yeah!

Here is the wiring diagram for the system. All the cable to connect these modules is regular CAT5 straight through patch cable.

Tools you'll need for the install:

1) Diagonal cutters (used to trim tie-wrap excess)
1a) pliers - helpful for tightening down the tie-wraps.

2) 3/8" button head bolts: 2 short button head bolts work the best with the 3/8" washers in the package to secure the hand sensors from rotation. Other bolts will work, but the button head bolts have the widest, flatest surface area that's best for holding the hand sensor brackets to the wall.    The flanged, martini bolts are a no-go - no surface area to hold the bracket.

3) an extension cord that has at least 2 outlets.   I usually bring one of those short 1-2 foot extensions that expands from a single 3-prong to 3 outlets.   Heavy duty, flexible...they cost about $10 at home depot. Then I run a regular extension to the expansion extension.   An extension to a power strip will also work.

There's enough power cable leading to the display to eliminate the need (in most cases) to run any extension cord up the wall.

4) It would be ideal to plug the system into a filtered power source - something to catch line spikes.   Typical UPS or lightening suppressed power strips would be best.

It's quite simple to hook up, but here are a few tips for install.

1) Always attach the left lane foot and hand sensors' BASE jack to the base unit.   This makes the LEDs on the starter console line up left and right (unless you flip it over/upside down.) It works either way (you could go from the base unit to the right lane instead), but one way is better. The diagram shows it going from base to left lane.

2) Make sure you bolt the hand sensors offset from the rope - far enough that the loop and slack of the rope that is slapping around as they climb will not break the light beam of the hand sensor.   This is a very important installation/ positioning consideration.   A fast rope flick in the target area is enough to trigger the sensor, so configure the routes and lanes so this possibility is eliminated.
2a) it's also good to make sure the rope cannot get under the wires leading into the sensors.   One of the improvements we made after the regionals was we added another bracket to the back so the sensor can be flipped over. This lets you route the cables to the sensor from the left or the right depending upon what the best way is to keep the cables out of the way of the rope.   The hand sensors work just as well in any orientation, so use the one that makes the install the safest to keep cables away from the rope.

3) For the long run of cable from the base up to the left hand sensor, if you are unable to drop the cable behind the wall and come out the front at the bottom, it works good to screw in some bolts along the path on the face wall up to the top. Then you can use the plastic tie-wraps included in the package to secure the cable to the bolts. It's an effective cable chase that keeps them tied back. Same works with running the foot cable along the wall until it needs to come out to the foot sensors.

4) Make sure to leave a little slack in the wire to the foot sensors (don't tie it down to close to the wall), since the footbeds need to move around some where it's comfortable for the climber.

5) The sensors are interchangeable - they all have the same jack configuration. Not all jacks will be used on a single sensor as shown in the diagram.

6) It would be good to tie-wrap the cables feeding into the display base unit to something on the wall close to where they connect to the base. This will provide strain relief and especially, if someone trips over a cable it won't immediate yank the display off the wall or rip the cables out of the display.
6a.) Also makes sense to tie wrap the cables leading into the hand sensor jacks to a bolt in the
wall or something secure so those cables have some strain relief as well.  This will make
incidental hits to the cable less likely to cause an issue.

7) If the cables included aren't a length you need, you can use whatever length you want - the cables are basic CAT5 straight-through patch cables which you can buy at CompUSA or home improvement stores like Lowes or Home Depot.

8) Finally, make sure you have at least one backup stopwatch on each lane - I'd hate to have some malfunction in the system hose up an event.   We should have some spare sensors built by nationals.

The demonstration video is still valid for how the system works (in operation) from a starter official's point of view as well as the climbers.

The main things that have changed since that video was done:
1) the hand sensors now have LEDs on the end-cap that light when the sensor is activated - giving the climber some clue that he hit it OK.
2) the starter console LEDs now not only light when the foot is in the footbox, but also will light when the climber has activated the hand sensor.

Speed Climbing Timing Part 1 - Sensors
Speed Climbing Timing Part 2 - Controller
Speed Climbing Timing Part 3 - Integration
Speed Climbing Timing Part 4 - Touch Pad Construction
Speed Climbing Timing Part 5 - Schematics
Speed Climbing Timing Part 6 - Perf Board
Speed Climbing Timing Part 7 - Display
Speed Climbing Timing Part 8 - Hand and Foot Sensors
Speed Climbing Timing Part 9 - Demonstration
Speed Climbing Timing Beta Test Boulder Rock Club
Speed Climbing Timing - Sensor Improvements
Speed Climbing Timing SHIPPED!
Speed Climbing Timing Schematics (shipped v1)
Speed Climbing Timing Installation
Speed Climbing Timing - Laser-based Hand Sensor Design
Speed Climbing Timing Lessons Learned

Speed Climbing Timing Schematics

landon@360vl.com — 2009-06-03T19:38:02-06:00

Synopsis - These are the latest schematics for the Speed Climbing Timing system. In the process of improving and refining the system, the schematics changed substantially from the original schematics I posted. One of the biggest changes was eliminating the LM339 comparator circuitry for all 4 sensors (2Xhands 2Xfeet). It was just not reliable and too twitchy. Added an LED as feedback to the climber on the hand sensor and added the starter console schematics and base station schematic changes.

There are 6 pages of schematics that cover the system (currently). Click on each image to get an enlargement.

Base station display controller:

Hand sensor:

Foot sensor

Starter’s console:

Add capacitance to the power supply on the ET-AVR stamp dev board:

Example wiring diagram that shows two hand sensors attached to the Base Station through RJ45, CAT5 cabling:

I may scan and post the schematics of the large LED display board later. That board is a pre-assembled module I purchased based on the Maxim 7219 LED display chipset.

AVR Code for the firmware for the base station and sensor microcontrollers will be posted next week. All code will be open sourced as well as these designs.
Speed Climbing Timing Part 1 - Sensors
Speed Climbing Timing Part 2 - Controller
Speed Climbing Timing Part 3 - Integration
Speed Climbing Timing Part 4 - Touch Pad Construction
Speed Climbing Timing Part 5 - Schematics
Speed Climbing Timing Part 6 - Perf Board
Speed Climbing Timing Part 7 - Display
Speed Climbing Timing Part 8 - Hand and Foot Sensors
Speed Climbing Timing Part 9 - Demonstration
Speed Climbing Timing Beta Test Boulder Rock Club
Speed Climbing Timing - Sensor Improvements
Speed Climbing Timing SHIPPED!
Speed Climbing Timing Schematics (shipped v1)
Speed Climbing Timing Installation
Speed Climbing Timing - Laser-based Hand Sensor Design
Speed Climbing Timing Lessons Learned

Speed Climbing Timing Systems - SHIPPED!

landon@360vl.com — 2009-06-03T18:05:37-06:00

Today I shipped a speed timing system to USAC representatives in Texas and New England - one for use at the Southwest Divisional Championship and another system for Northeast Divisional Championships. One should be in Texas in 2-3 days and the other in Mass in 3-4 days.

Here's what was in the box:

Large display with power cable (includes 2 wall warts - one for the digit displays and one for the electronics - integrated into a single cable).

White CAT5 cables - various lengths from 50', 30', several 25' and 20' and 10'. Since I don't know the gym and installation, a variety of lengths were enclosed. You will not use all the cables sent.

2 Hand sensors (those with the yellow dots).
2 Foot sensors for false start detection (white triangles).
1 Starter console - using one of the CAT5 cables, connects to the main display
1 Bag of installation helpers (white cable ties to tie CAT5 cable to bolts in the wall to keep it out of the way of the route, 2 3/8" washers.)

This is a momentous occasion as it represents hundreds of man-hours of design, testing, improvements, and build-outs.

We hope that others may improve and build off this design for the benefit of speed climbers everywhere.

Landon

For more on the design and development of the Speed Climbing Timing System see these links:
Speed Climbing Timing Part 1 - Sensors
Speed Climbing Timing Part 2 - Controller
Speed Climbing Timing Part 3 - Integration
Speed Climbing Timing Part 4 - Touch Pad Construction
Speed Climbing Timing Part 5 - Schematics
Speed Climbing Timing Part 6 - Perf Board
Speed Climbing Timing Part 7 - Display
Speed Climbing Timing Part 8 - Hand and Foot Sensors
Speed Climbing Timing Part 9 - Demonstration
Speed Climbing Timing Beta Test Boulder Rock Club
Speed Climbing Timing - Sensor Improvements
Speed Climbing Timing SHIPPED!
Speed Climbing Timing Schematics (shipped v1)
Speed Climbing Timing Installation
Speed Climbing Timing - Laser-based Hand Sensor Design
Speed Climbing Timing Lessons Learned

Speed Climbing Timing - Sensor Improvements

landon@360vl.com — 2009-05-22T08:39:42-06:00

Synopsis: We’re into the 20% of the project that takes 80% of the time. In this case, after the regional championship and another day’s session at the Boulder Rock Club, I decided I needed to make further improvements to the detection at the hand-sensors. In this article I have two videos - one showing a typical detection failure of the first generation hand-sensor and the second video showing some standalone tests of the 2nd generation sensor after an ATMega8 micro was installed in the end-cap of the hand-sensor in order to do the voltage comparison. The second video also shows the 2nd generation sensor which includes a bright LED on the end cap to help show the climber that the detection occurred.

The first video below shows an analysis in slow motion of a typical 1st generation sensor detection failure.

Videos below the break.....

Speed Climbing Timing - Boulder Rock Club Beta Test

landon@360vl.com — 2009-05-13T21:38:33-06:00

Synopsis: The speed climbing timing system was installed at the Boulder Rock Club (BRC) May 12-13. There were three objectives in doing this. One, Sparkfun wanted to do a photo shoot of the system with my daughter climbing and me belaying. The results of this shoot will end up in an ad for Sparkfun in four geek magazines this fall and some banners on their website. Second objective, the BRC ABC and youth climbing teams would have a chance to practice with it and get in a good speed workout. Third, I would get a lot more testing done on the system, in a different gym installation, and in a setting where we could more easily replicate issues if they came up.

Prior to this install at the BRC, I made several of the improvements I listed in the trip report from the beta test at Albuquerque. Namely, I added the second T bracket to the hand sensors so that the sensor was completely symmetrical and could be installed with the jacks on the right or left side. Secondly, I made a power cable extension that took power from the base display station out about 15 feet to the two wall warts (9VDC and 18VAC). This made a cleaner power install on the wall by not having to run an extension cord and multiple outlets up the wall to the base station.

Videos and photos below the break....

Speed Climbing Timing - Regional Champ Test

landon@360vl.com — 2009-05-13T21:34:20-06:00

Synopsis: This article is a redux of the lessons learned from the first trial of the speed climbing timing system at the regional championships in Albquerque last Saturday, May 10.

Overall, we were very pleased with the performance. I didn't do a final count, but out of 10 categories each category probably averaged 5 kids - some older categories had just a couple but the younger ones were busting at the seams.

Given that estimate, there were approximately 100 climbs (10 X 5 X 2 climbs each) we did with the timing system. Once the results are posted, we'll know exactly, I could be off 10 or 20....it's always chaotic.

Speed Climbing Timing - Part 9 Demonstration

landon@360vl.com — 2009-05-04T15:01:03-06:00

Synopsis: In part 9 I demonstrate the speed climbing timing system in a “Compact” form. In this configuration, all the sensors are within an arm’s reach so it’s possible to demonstrate the operation of the system from both the climber’s and the starting official’s point of view. The video below demonstrates a typical starting sequence, race finish, clock reset, and false start detection. Finally, the video has a very brief overview of how the system is wired up when it’s installed.

The speed climbing timing system is going to receive its first real-world test this weekend at the USAC Youth Regional Championship in Albuquerque. Once that event is completed, we’ll assess the performance, make any adjustments or fixes, then build a second system. These first two systems will then be sent to the USAC Southwest Divisionals and the NE Divisional championships. Assuming all goes well, both systems will ultimately perform the timing duties at the USAC Nationals 2009.

Video below the break....

Speed Climbing Timing - Part 8 Hand and Foot Sensors

landon@360vl.com — 2009-04-06T06:49:28-06:00

Synopsis - In Part 8, I cover the finished hand and foot sensors to the speed climbing timing system. Way back in Part 1 and Part 4 of this series, we covered the IR sensors and touch pad design and construction....the “bones” of what you see here. This article includes a system block diagram and photos and explanations of the final design of the hand and foot sensors.

These sensors are “finished” in that they are more done than some of the other parts of the system, and finished enough that this is what we’ll install in the rock climbing gym for our first major tests.

Diagrams and images below the break...read on...

Speed Climbing Timing - Part 7 Display

landon@360vl.com — 2009-04-03T18:00:37-06:00

Synopsis - This article covers the speed climbing timing display. In the architecture of the timing system, the display is the base station and holds the microcontroller (ATMega128) and sensor board presented in Part 6. The microcontroller drives the LED display, manages the real-time clock, and responds to the touch sensors described in Part 4.

All sensor cables and the starter’s console will connect to the base station through CAT5 cable.

Dale Herbert and I got together one weekend and finished off the two hand sensors, two foot sensors and did an initial fit of the LED display in the enclosure that Dale built. Here are some photos that show these components.

This is Dale working on the hand and foot sensors in the living room of my house. You can see the upside down display enclosure on the far right of the picture. More on that in a minute.

More below the break....

Speed Climbing Timing - Part 6 Perf Board

landon@360vl.com — 2009-03-10T14:34:02-06:00

Synopsis: Over the past weekend, I spent the time to take the sensor analog and digital interrupt logic off the breadboard and onto a more semi-permanent perf board. In the process, instead of a single sensor circuit which I’ve been using on the breadboard, this perf board incorporates all the circuitry for 4 sensors ( hand and foot X 2 lanes.) This perf board implements the schematics in Part 5 of this series with one exception...I added 4 LED indicators through another driver to show on-board when a sensor “fired” due to a touch.

Also, in this article are a couple of movies that show a single-sensor alpha test on a climbing wall.

2 Lane, 4-sensor Perf-Board

Here’s a picture of the perf-board that contains the sensor detection and microcontroller interrupt logic for a two-lane speed climbing system. The microcontroller is on a separate board. The schematics for this board can be found in Part 5 of this series. The 4 orange potentiometers on the upper left of the board are there to adjust the sensor sensitivity.

More below the break....

Speed Climbing Timing - Part 5 Schematics

landon@360vl.com — 2009-03-06T12:44:08-07:00

I took a short break from the prototyping workbench to catch up the project documentation. This one in the form of schematics. Using Eagle, I created a short set of schematics focused on various subsystems of the speed climbing timing system.

They are:

IR Sensor and Interrupt Logic
Power Regulation
Microcontroller Interfacing
RJ45 pinouts for the Modular Connections

I won’t spend much time explaining the schematics. They could well change as I get further into this.

More below the break....

Speed Climbing Timing - Part 4 Touch Pad Construction

landon@360vl.com — 2009-02-25T19:20:19-07:00

Synopsis: In Part 4, get a sneak peak of the mechanical design and construction of the sensor for stopping the clock at the top of a speed climbing route, the hand sensor. A video outlines the thinking behind the design and how it can work well for the climber and the routesetter to mount the sensor on the climbing wall.

Some of the critical requirements for this sensor are:

Robustness - it has to be able to take a direct, forceful hit by a climber’s hand. Often a climber will launch for the finish and smack the finish switch. It’s this behavior that has destroyed more speed climbing timing systems than any thing else. So, this touch pad has to be robust even though there’s no mechanical switch. A corollary to the robustness is safety. It has to be robust, but it needs to be safe for a climber to hit, even if he or she hits it very hard.

Assuming this IR light based sensor works well, climbers will quickly figure out they don’t need to practically destroy the sensor in order to stop the clock. The finish technique and behavior for finishing a speed climbing route will change to more finesse as a result.

Inexpensive - Another major requirement is the touch pad should be inexpensive to build. This mechanical and mounting solution uses easily attainable and inexpensive components from your local home improvement store. The sensors themselves are about $10 each (GP2Y0A21YK from Sparkfun), so there are $20 of sensors and about $10 of other materials for a total per-pad cost of about $30. For a 2-lane system, there are 2 touch pads for hands (one for each lane’s finish) and two base sensors for feet to detect false starts, so the system cost can get quite high if the per-pad cost isn’t kept low.

Easy to Mount - Finally, it had to be lightweight and easy to mount. This sensor, because it is installed at the top of the speed climbing route, needs to be ported up a ladder by a routesetter and mounted to the climbing wall. The solution presented is both lightweight and extremely robust and can be mounted with the standard climbing hold hardware - a 3/8” bolt. The bolt can be tightened as much as desired to keep the pad from moving on the wall without any damage to the sensor mounting hardware whatsoever.

Video below the break....

Speed Climbing Timing - Part 3 Integration

landon@360vl.com — 2009-02-21T19:39:03-07:00

Synopsis: This is a video to show the initial integration of the IR sensors with the stop switch interrupt of the microcontroller. There’s more work to do as you will see. Later in the article are a couple videos which show the system working after an EMI issue caused spurious interrupts.

The basic integration of the IR sensors with the AVR is:

IR Sensors have a high frequency component and the overall wave goes from about .8v to 2.5-3V.
I take the IR signal into a low-pass filter to clean up the signal and remove the high frequency components
That signal goes through a non-inverting comparator and amps the signal to the rail when the IR beam is reflected with a hand
The output of the comparator goes to a hex inverter (74LS14) schmitt trigger
The output of the schmitt trigger goes to the interrupt of the ATMega128.

Video below the break....

Speed Climbing Timing - Part 2 Controller

landon@360vl.com — 2009-02-19T17:04:34-07:00

In Speed Climbing Timing Part 1, I introduced the touchless sensor concepts of a speed climbing timing system. In this part, I’ll demonstrate the basic controller and the timing functions as well as the display driver.

The display is based on a Maxim 7219 LED multiplexor. The controller is an Atmel AVR ATMega128. I’ve developed the Max7219 driver code and the basic timer to deal with a two-lane speed climber competition.

Speed Climbing Timing - Part 1 Sensors

landon@360vl.com — 2009-02-19T13:34:34-07:00

Most of my kids climb competitively and do an event called speed climbing. The name pretty much says what it’s about, but in general, 2 climbers go head-to-head on 2 25-50’ routes. After a climb is finished, they switch routes and climb again. The addition of the times of the two routes is the total time for the climber.

A problem that has plagued the event is the speed timing systems used are very unreliable. Most of these systems rely on some type of mechanical switch for the foot pedal and the top hand sensor to stop the clock at the finish, the top of the climber’s “lane”. Because the systems malfunction so frequently and can ruin the flow of the event and drastically extend the time it takes to execute the event due to re-climbing or fixing the system, I set out to design a system that did not use mechanical components. The goal is higher reliability, ease of installation, and also safety for the climber because sometimes mechanical switches have injured speed climbers hands.

Below is a video of a sensor concept I’m working on which will apply to the hand sensors at the top of the lane and the foot sensor at the bottom that’s used to detect a false start. These use an IR technology that’s relatively cheap (about $10 per sensor, 2 sensors per “pad.”) and immune to most IR emitted in ambient lighting situations.

Build an IDC cable from an IDC PCB socket

landon@360vl.com — 2009-02-17T13:35:09-07:00

I recently needed to create a cable that had an IDC socket but only had PCB mounted IDC sockets available to work with. It would have been very tedious to try to directly solder wire to the short pins of the PCB IDC socket. It’s easy to find IDC ribbon cables with the plug end, but not the socket end.

What to do? Hack a cable using a 2x5 molex type connector, attached that to the PCB socket and super-glued the two together. The result was effective and what I needed.

This is also a good way to build your own JTAG squid cable if you need one or lost yours that came with the STK500. The cable below is not for JTAG, but a custom cable I made for a SPI interface.

I used the Hansen Hobbies molex connector kit to build the 2x5 molex cable termination. Their crimp and cable making kit is awesome - their site has lots of videos to show you how to build cables. The 2x5 works great to plug it directly into the PCB side of the socke, but the PCB pins are not very long, so it won’t stay by itself. Superglue the bodies together or if you really want it to be sturdy, you could epoxy the two bodies together. Slightly rough up the plastic to help the bond.

Here’s how it looks when it’s done: