My home built metal DDR pad (Part IV)

[geckoinc99]

Actually, the chips that you use for the touch activation could easily be implemented later and you would just secure the arrows down. So there's no reason to go ahead and build your pad, then add this later if it works out well.

The complicated designs were from people trying to take the project further. For those that would like to have a flashy pad, these designs will fill in greatly. So definitely keep those designs coming.

PSXZombie, what chip are you trying to use? I would like to work with it also. With two of us working on it, we might both come up with some ideas.

Happy building everyone!
_________________

Life is very different 3000 feet above the earth with only a wing to hold you up....
Can't wait to be there again. Wanna join me? Visit http://www.hanglide.com

[HitokiriX]

Could you explain how to create a touch sensitive pad? It doesn't have to be exact, but just tell me what I would have to change in the design of a riptide pad to implement them. When I think of touch sensitive the 3G iPod touch buttons come to mind. I don't know how they work, but I assume that if you use touch sensitive buttons there's no need for internal moving space. So in a riptide pad where the arrows are held down by corner brackets, you'd need to use extra wood as filler and in the stoli pad you would pin down the arrows to the base with the screws. Is this correct? Oh and just as a cheapo question, how much would this cost approximately? Say 4-8 touch sensitive contacts under each arrow, and 6 arrows (trying to see how high the price of this system could go)?

[fid]

[-Scorpion-]

According to this page http://www.qprox.com/products/qtouch.php
a QT113 sensor has a response time of 30ms, which is imo way too much for a time-sensitive game such as ddr. It would create a feeling of stepping on the arrows before they actually hit the receptors on top of the screen. Adjusting the globaloffset would help correcting the offset of the song itself but wouldn't get rid of the remaining visual delay.

[psisarah]

Wouldn't a touch sensitive pad be extremely over-sensitive? It's a neat idea, but wouldn't anything from a speck of dirt to a dangling shoelace trigger the button?
_________________

[geckoinc99]

Hmm... I don't know if you could notice 30 ms or not. That's something that we would have to see when a prototype is made. I'm thinking about making a prototype arrow for one of my pads and seeing if I can notice a difference. Might be awhile since I'll have to play around with settings. As far as the sensitivity, you can adjust how sensitive the chip will be. A speck of dirt isn't enough to trigger a responce. If properly adjusted, it will be looking for a large change (ie: a shoe) so that small things like shoe laces would not trigger a false reading. There would be some adjusting and fine tuning in the beginning, but I think there's still great potential here. I nice thing to notice on this chips is that it has a pin for turning LEDS on to give visual feed back. For anyone wanting to add light, this solves two things.

The concerns are valid, but we need to see what it does in real life. Remember that a chip that responds at 30 ms can respond around 33 times a second, and I don't think anyone's foot responds that quickly. The delay may or may not be noticable. If it is, then maybe another faster solution can be found. We'll see.

David
_________________

Life is very different 3000 feet above the earth with only a wing to hold you up....
Can't wait to be there again. Wanna join me? Visit http://www.hanglide.com

[Matrlx]

30ms is actually a lot to a game like DDR. It's not a matter of hitting the arrow faster than 33 times a second(never doubt though, by the way), but it is possible to only touch the arrow for that fraction of a second. Additionally, 30ms is HUGE in terms of the accuracy needed to score better. For the DDR arcade (5th mix and onward), the approximate window for getting a Perfect from a Great is 33ms. Getting a Marvelous requires you to be inside about 17 ms of the mark. In the Groove has smaller timing windows to achieve Excellent's and Fantastic's respectively.

While it sounds like a good idea, I don't think it would be suited to players who want to have a highly accurate response time. It sounds like a 30ms variation would enough to significantly affect gameplay for those play for accuracy.

[fid]

[stoli]

I think this has potential. I agree that 30ms can be an eternity in certain apps, but I'm going with the Gecko on this one - we need to check it out and see how it works. I've always hated the mechanical switches, although they do work well, I much prefer a solid state alternative. Up to now, the alternatives that I have looked at are rather costly but it looks like this could be done for less that $20 (based on a 6 arrow pad). It also looks like the 113s are the way to go. More costly in the long run since you need a chip for each arrow but they list a much faster response time than the chips that accomodate multiple sensors. The 113s would actually be cheaper for the fat fingered solderers out there who like to fry electronics - each egg in different basket so to speak.

Now, I haven't really analyzed the documentation in depth, but the response times that they list are nominal. It looks like you could get the 113's response down to 10ms with the right capacitance. This would decrease sensitivity but I think that might be a good thing. It might also fill the empty half of Matrlx's glass and get him closer to the Fantastic that he is talking about.

So, fid or anyone else, can you recommend a supplier for the QT113s?

-Stoli
_________________

My Chrome Pad Website

[fid]

[fid]

[fid]

I got qt113h's from digikey ( http://www.digikey.com ). Note that they charge you an extra 5 bucks if you don't spend over 25.

Anyone else know any places you can buy them in quantities of less than a few thousand?

[marcan]

I already thought of the touch sensitive thing too

You can do it straight with my circuit too, with a couple simple mods. Essentially the way touch sensors work is they send a very weak electrical pulse to the metal. If you touch it you add capacitance and delay the pulse. It then measures back, and if the pulse is not there, say, within a couple of microseconds, you touched. The microcontroller in my mod can do it alright (I've gotten it to work before). This simple arrangement only works for direct touch sensors (touching metal barefoot), but with a more developed system you can sense through thin surfaces (or using a dedicated chip like psxzombie used). Using the simple method (playing barefoot required) would also fix the latency problem. You can get latency down to less than 1ms if you do it like I did, but I'm not sure what the technical requirements for sensing across larger distances are, and if latency can be reduced in that case.


Would make for one extremely cheap pad too. I don't know how much those chips cost, but the simple version with a generic microcontroller could be less than $10 for the circuit, plus a wooden board and metal contacts.

[Matrlx]

You'll have to excuse the skepticism, I'm just one of those players who can get mad when I'm not getting AAA's and just more prone to blaming the equipment.

The time delay in a mechanical switch would seem to me to be much less than 30ms with consistency, and the delay for debouncing the contact doesn't actually affect gameplay.

The impression I'm getting from how these touch sensors work though, is that they 'check and respond' in certain timed intervals. So every 30ms the sensor would would say 'on' or 'off'. Is such the case? Or is it a consistent delay from when you touch the panel and when the sensor realizes it?

A variable sort of delay and the need for myself to compensate for the such delays when I'm playing would be problematic for myself, I'm not saying everyone else will have a problem with it, I'm just talking about a very tiny window for inconsistency, something that's probably negligible for most people.

The idea sounds great and I really do hope everything works out, I'd love to be able to use something like this.

(Btw, that comment on having problems with shoes really worries me.)

[stoli]

[raytracer]

I finally finished my pad. It looks VERY unassuming, but it feels like a touch sensitive pad and works like a charm (not bad for something that only cost me $30 total. I got the sheet metal free from my uni.)

http://www.maj.com/gallery/epsilon/dance-mat/dsc03234.jpg

got my first AA shortly after finishing this pad - duality on normal (stepmix song) (heh, I'm just starting!)

so what do you people think?

:edit: for spelling

[psisarah]

I just wanted to show off my metal pad;



(Sorry 'bout the quality, my phone cam sucks)

Most of the pad was built following ddrhomepad's design, but I omitted the corner braces and added X,O, Start, Back/Select. I used the PCB from ym Ignition so it works with Playstation, XBox (and shows up as a pad on XBL) and USB. I wanna give a shoutout to those who contributed to this thread, thanks. I don't know if I could have finished without searching this thread for help.

..Not bad for a girl in 3 days, I'd say?
_________________

[geckoinc99]

psisar,
Nice job!! Good for you doing custom arrows. Also, way to show that girls can build too ;-)

Matrix,
I didn't realize that the timing was that short on the gameplay, so I retract my original statement. Thanks for the info there. Yeah, 30ms would be too long, and since it couldn't respond again until after that, then it really wouldn't work. However, if it's possible to adjust it to respond faster, down to 10ms, then it might work just fine. That's where some R&D would come in handy. I could do some to a point, but since I'm an A to AA (haven't gotten that AAA yet) I can't judge it completely like some of you can. Also, if shoes are going to cause problems with the reaction, that too would be a major problem.

Again, the concerns are definitely valid, but there may still be a way. I may talk to the people at QProx and see what they say. What would need to be the minimum time for acurate gameplay? Would 10ms do it? When you say that Marvelous has to be inside 17 ms, does that mean that you you can strike 8.5ms before or after the point, or is it 17 ms before or after the point? I'm just making sure I understand what the worst case scenario is here.

Here's a list of where the chip can be bought:
http://www.qprox.com/about/can_usa.php
Don't know if the other places charge you if you don't order a certain amount though.

EDIT: I just read up on the QT113. It is possible to get its response down to 10ms based on the size of capacitors you use with it. Now, if it will work with shoes is another question that I'm still trying to get answered.

There are definitely some limitations to overcome, but I don't think it's time to give up yet. We'll see.

David
_________________

Life is very different 3000 feet above the earth with only a wing to hold you up....
Can't wait to be there again. Wanna join me? Visit http://www.hanglide.com

[slvrshdw]

OMG A GIRL BUILT A PAD

good job

i liked the arrows too.

anyway, you know i really like my homepad now since my friend got a TX-?000 pad and the left arrow was messed up in about 9 days, NINE!

he paid 200 including shipping and i paid about 150 so ya, homepads rock
\m/_ heh, ive always wanted to do that
_________________

[-Scorpion-]

About worst cases consider this:
USB game devices are polled at 125HZ, or every 8ms.
So even if the game device produces virtually no lag, the computer still produces latency, which adds up.
For example, let's say your get your sensor down to 10ms, connected to a USB device that makes the signal arrive at the computer between 0-18ms, moving the timewindow a whole marvelous setting. And that's without the computer latency, which is probably about another 10ms under Windows!


Fortunately you can make a couple of optimizations:
There's a USB Patch that can set the polling rate of USB devices up to 250,500, or 1000HZ, moving it up to every 1ms.
Linux is a much better System for a time sensitive game such as Windows. There are low latency and USB polling patches for Linux available, too.
I have a Linux System with low latency and USB patches and the difference is amazing.