UltraSound 2.0 After upgrade..

[Eandori]

Well that is too bad about the switches. It really can be fixed with a simple software-based switch debouncing program. For what it's worth, I tried and tried to contact the actual developer that UltraSabers is working with and he never replied to me, even once.

In case anybody is curious, you CAN work around this by adding a few simple circuits that electrically debounce switches. Perhaps we should post a few solutions so those using UltraSabers boards can implement them.

[Darth_DevilGuy]

I'd like to know how to do that

[Darth Leximus]

as would i

[TD-2272]

Sounds good, If anyone has insight on this please share

[Eandori]

This page talks all about doing it via hardware, and even some software methods. The one I used to use in College is using the 2 NAND gates (listed on page 11). We also used to use the D-Flip Flop method. (not shown in this document.)

http://www.ganssle.com/debouncing.pdf

Here's a link to a NAND gate sold on Digi-Key:
http://search.digikey.com/scripts/Dk...e=TC74AC00P-ND

The resistor capacitor method works, but you need to wire up an inverter to get rid of the slow transition. If you are going to be using a chip anyways, perhaps the NAND gate would be best because most of the time if you buy a discrete NAND chip, you get 4 NAND gates per chip which would let you debounce 2 lines separately.

When I get my 4x Ultraboards back in the mail finally I'll test out some methods and reply with which one worked for me. There are many ways to skin this cat though, too bad it was not done in software. I 100% guarentee that CF is doing the fix to bouncy switches in software.

[Roy's Blues]

From Ultra at FX-sabers.com

[quote author=Ultra link=topic=17547.msg271578#msg271578 date=1241563341]
Yall know I would never just leave you hanging, right? After all these years yall know what kind of operation we run around here, right? Of course yall know that. If it ain't right, we (Strydur/Yoda/Ultra) will make it right.

So we're working with Nova right now to clear this up and have already taken the steps needed:

I have been talking to Tim about the continued problems with the switch debouncing. He has offered to send in some
various types to ensure that the new boards will work with any switch. This is, in my opinion, very easy. I'm not
at all worried about it, except for the possibility of having to reprogram all of these boards again! In which case,
if that is necessary, I was thinking of constructing some stand alone circuits to clean up these noisy switches.

The above quote is from the programmer.

Switch debounce circuits will likely be the easiest solution since it doesn't require reprogramming the boards. We were already working on that idea before, but scrapped it because it didn't solve improve pulse-mode/clash-lock. Guess we'll go back to that idea.

I'll keep everyone up to date as get more info. Remember, this isn't Wal-Mart, yadda, yadda, everyone should know the drill by now.

[/QUOTE]

[xwingband]

NAND's are pretty easy... I'd just go with that.

[Darth Xusia]

I'm all for whatever fixes the problem, as long as someone tells me what to get and how to do it. If someone came up with a kit, I'm sure they would be flooded with people wanting it. If it's simple to mass produce, maybe Tim would sell it in the store. That would be cool.

[Strydur]

If someone can post a diagram with how this would all work for us I would be glad to run with it. We might get the manuf to make some of these though as he has already suggested it.

[Kant Lavar]

I admit, I consider myself at least somewhat intelligent, but I looked at the diagram for the double NAND gate solution and I just scratched my head and went "huh?" I mean, I kinda-sorta get how it works, in a vague sense, but I'd need more explicit instructions as to what gets connected where before I tried it myself.