Discussion in 'Armory - Q&A' started by reawl, Aug 29, 2006.
I am curious why you have changed your mind about this project?
Perhaps it has something to do with requiring 18 banana jacks in order to distinguish between each of the three weapons as opposed to one switch (2 fencers who could be fencing any one of 3 weapons, each of which will require 3 banana jacks - plus a 19th if you allow for the possibility of a grounded strip).
I can't even imagine the confusion when one person plugs in to the foil socket, the other plugs into the saber socket, and they are both holding epees.
As Mr E. said, it might make a nice science project but it would be difficult to market.
Properly marked 2 sets of 3 banana jacks (for each plug) for each of 3 weapons could be very easily done with simple graphics on the face of the box. Not a big deal -- especially for a version 1 of a project of something like this.
L <--> R
--- -- ---
xxx -- xxx (Foil)
xxx -- xxx (Epee)
xxx -- xxx (Sabre)
I agree that it might work as a prototype for some future design. However problems such as switching between weapons would still have to be addressed before the product could be considered marketable.
Actually the switching is very easy because the software only needs to see where signals are actually present... If could also instruct the user in setting up the system correctly for a given weapon and actually instruct the fencers to test to ensure that the setup is correct. Software such as this is not in the least complicated and can be made to be very user friendly--going so far as to walk the user thru steps of setup.
Remember this is software so much more is possible that with a hardwired timing system. That is the beauty of using a computer/laptop.
But in order for the software to see the signals they must first pass through the hardware.
Looking back through earlier posts it appears that hardware limitations associated with the parallel port may be the source of the problem.
Actually not: The software can experience a LACK of a signal in the port and instruct the person setting it up as to how it can be corrected--e.g. plugging the reel/cord into the correct plug for the weapon selected. This could be done with a diagram/jpg of the box with highlights.
rac was spot on with the design (except I've only drilled the hole for the piste connection, it's not wired in this particular adapter). For the parallel port adapter I want to keep the assembly as simple as possible so that any fencer with access to Radio Shack, a soldering iron, and an afternoon can make one of these. Could I rig up a switch to do the same thing? Maybe, but it wouldn't be as straight forward.
Cost of parts from my local electronic supply store: $25.
Time to assemble by a guy who barely knows how to use a soldering iron: 3.5 hrs.
Yes it's a little funny looking. Yes, if you're switching a lot between weapons, it would be tedius.
It works well for Foil and Epee. I only have pseudo-code for sabre in place since I ran into these hardware issues. Now that I've got an adapter to work with, that should fall in to place soon (this weekend?).
rac is also corect in that this solution opens up a whole new door for feedback to the user. In addition to the previously brought up points of tracking actual tip depression times, his point is vaild in the feedback you could provide in standard use (thanks!).
rac: I like the way you think, any chance you have any experience designing hardware to interface with a PC that I could pull on? Shoot me a PM.
Nice looking prototype. How about a wiring diagram?
Also, if you have separate colors for each weapon (as opposed to the rainbow your prototype has) it'd be a bit easier. Just a suggestion, not a critisism.
How about a 6 or 7 line 3 position switch to set the weapon and change circuitry. You may also be able to make the program recognize the switch position so that it can set itself accordingly.
Got a part number & dealer for one? I ddin't realize such a beast existed in an off-the-shelf form.
Try these I think.
*sigh* those switches run $14 on up.
I like Mergs's suggestion about the colors, these just happened to come in 6 packs of 6 different colors. I kind of like the colors of the jacks matching the colors of the lights as an alternative color scheme.
Sabre "works" after some effort tonight (woo hoo! all 3 weapons!). I need to program in the whipover, but basic functionality is there.
Keep the ideas and critique coming in the mean time!
I think its a great idea, and if the software can be ported to a mere Win95 machine, any old laptop for under $75 on ebay can be used. Let me know on more details about this.
How much are the jacks, though? You save the money of 12 bannana jacks, and it's more compact and nice-looking.
The jacks were $15 off the shelf. If I had the patience to order them from a catalog/internet it'd be somewhere closer to $10. The larger box was about $3 more than the smaller boxes I was using before. So by using the $14+ switch you might save $10 to $13 in parts on my current design.
That's assuming the $14 switch would work; I don't think it would. I have 6 sockets and 3 variations of those 6 for the 3 weapons. I think that means I need something like a 24-line (6 in, 18 out), 3-position switch, or six 4-line, 3-position switches. Cost wise it doesn't really make sense. It would be far cheaper to just use a PIC and a push button, but now we're adding complexity if we do that.
The goal of this project is to keep cost and complexity to create the adapter to an absolute minimum for those wishing to build their own. If I decide to manufacture and sell an adapter, then some of these things become an option. But on that note it has been brought to my attention that I have exposed liability if anyone uses my software/design regardless if I charge or not. So I'll be having some beta tests with people I trust while I pursue the appropriate insurance coverage.
I think you are missing something.
The problem you are having is enabling the outputs, and selecting the inputs.
So, what you want to do is use the outputs to control what your box does.
There are very cheap (less than a dollar) chips that can give you control of outputs and selection of inputs. So, for example, you can have 3 bits that control WHICH PIN you put on an input and WHICH PIN you drive for output, That means you use 3 output pins to control an input multiplexer (1 of 8 selector), 3 bits to control output drive (1 to 8 demultiplexor), and one output pin for value out, and one input pin to read value in.
This takes one 8:1 mux, one 8:1 demux, and two quad tristate drivers, $3-4 of chips.
You wire three output pins to the selector of the mux, 3 output pins to the selector of the demux. You wire the mux inputs to your 6 box pins. You wire the demux outputs to the tristate controls of the drivers. You connect all of the inputs of the tristate drivers to your 7th output pin, and connect the outputs of the tristate drivers to the 6 box pins. You wire the output of the mux to your parallel port input pin.
Then you cycle through the pins, setting the output select to drive a signal, and the input pin to look at a signal. If you have a grounded strip, it can be a 7th input.
That solves the materials cost issue. It doesn't much work with his goal of making this easy to assemble by people of only moderate skill.
brtech, any chance you'd be willing to draw me a diagram? I'm willing to give it a shot, but I'm not quite wrapping my head around it. If you'd rather email it to me just shoot me a PM and I'll give you my address.
Howdy all. I've been working along similar lines and have written a Java version of the software and added some voice control (e.g., starts timer when you say Fence!, stops timer when you say Halt!, etc.). Instead of using serial/parallel ports, I cut up an old keyboard to act as the input instead since there is usually something like a 9x14 matrix for over a hundred inputs. Unfortunately, I ran into masking/ghosting problems with foil that I couldn't seem to solve by stacking resistors. The software was pretty slick, but I haven't figured out how to do the hardware cheaply and easily enough to make it economical.
Anyway, please post details on the hardware you've got and I'll do the same later this week when I assemble it. I can post the software to the web as well.
Separate names with a comma.