That video shows what I'm expecting to happen. On mine, both LEDs blink right when I turn it another tick. By default they're both off.

I'm looking at articles on rotary encoders, and I think mine may be a bit different. Instead of the ouputs staying connected between ticks like in , it just blinks very rapidly right at every tick.

While salvaging a broken stereo I came across what appeared to be a potentiometer behind the volume knob. The only difference appeared to be that it could turn infinitely in both directions. I hooked it up like a pot, with hot on one side, ground on the other, and an LED on the wiper. I was very surprised to find that the LED would blink every time I moved the knob a tick.

The problem is that the behavior appears to be the same when I move the knob clockwise or counterclockwise, so I don't know how I would actually use it with the arduino. Does anyone have any experience with these and/or know how they work?

Picture of the dial:

The blink program works for the frequency, but the pins on the arduino can only supply 40mv of current. Unless you want to put 12 of the IO pins in parallel, you'll have to use some transistors, capacitors, and the PWM system.

I would recommend 4 1.5 volt batteries. It should provide enough current, and in my opinion 9v batteries are a ripoff anyway.

This probably doesn't apply to many people, but when I used my nxt in conjunction with my arduino, I found that my rechargeable lego battery could power both at once. Also, 6 batteries worked, though it may not be recommended to use 12 volts.

Guys,

before we go all crazy I think you deserve an explanation on how the SSO works and why we haven't implemented the feature you request (yet). This doesn't mean we are looking into feasible methods, but for now it is not possible because:

- the SSO (single sign on) system resides in its own separate server protected with a certificate to ensure there is a secure connection and that your usernames and passwords don't fly over the internet unprotected

- the different websites are having their own way to identify users and we make them request a secret token from the SSO server to identify the user and make sure he/she is logged in

- the scenarios you guys suggest imply that, at this point, we should let the passwords fly freely from e.g. the forum into the SSO server and that is a solution we are not willing to implement because of its high risk

In other words, we understand he usability issues mentioned in this topic, but it is nothing we are going to implement now since it would mean putting the whole Arduino website under https, which according to some reports it could slow the server up to 5 times depending on the case. We need to make a lot of performance tests before we can do that.

His is coming from my severely limited knowledge base, but would it be feasible to submit the login information via a javascript that can make a separate connection to the secure server?

Also, many logins don't have https. They use post. Considering the fact that we're not storing out SS #s on here, I think that post would be fine. =)

Again, this is out of my limited knowledge. I hereby release this post into the public domain. 

Oh. I see. Arduino is cool like that, making the millis perfect. 

Cramming millis() into an unsigned integer means it will go screwy in little over a minute (65.536 seconds to be precise). All the variables relating to millis() must be defined as the unsigned long type.       

This is probably beside the point, but I'm pretty sure that the timer actually counts 1024 milliseconds per real second, so the integer actually overflows after exactly 64 seconds. Again, beside the point, but true.

About the code, pluggy is right. You'll have to define all variables relating to time with the long identifier. If you're crammed for space (Which I doubt you are) you could divide the number of milliseconds by 1000 (Or 1024 to be precise) if you can work with just knowing how many seconds have elapsed.

Hello, forum! I have been working on a project that lets me debug analog and digital sensors for the NXT robotics system. During the development, I had the need to convert digital to analog signals that could be read by an ADC unit. I found it very difficult to find the information on the subject, and now that I've found it, I thought I might share some informational links for smoothing out PWM signals.

Note, I didn't write these.

Websites:
http://www.datadog.com/pwm_tutorial.pdf
http://www.ehow.com/how_7867398_design-circuit-smooth-pwm-signal.html
http://hyperphysics.phy-astr.gsu.edu/hbase/electric/filcap2.html
http://ww1.microchip.com/downloads/en/AppNotes/00538c.pdf

Here's hoping that others can find the information on this topic more easily than I did.

Please could people vote for their preference on the poll? If we get enough votes the team might take notice.

Your text says "New Forum:" and then right under that it says "Forum 2005-2010 (read-only)" which would appear to completely negate your theory.
If you check the REAL new forum home page you will see a tiny fraction of those message counts.

Actually, going down into those categories shows that all of the posts were in fact migrated, although they were not sorted. This means that theoretically, if only the new forum were searched it would still get all of the old posts, just on the new forum. Please actually look into what you're talking about before you say it.

This would actually be VERY easy to implement.  Just replace the help, login, register thing with a tiny little login form.This doesn't even require a change to the css. (idea 2) Maybe if enough people like it it will be implemented.

idea2 uses no more space than the current site. It just replaced the help and login links with credential fields.

That should theoretically work. If you just want the raw values sent over, hook aud to an analog pin, and then send the analogRead value of that over xbee constantly. Then have the target device use PWM to send this to a speaker. Of course, I don't think that this mic picks up sound. It just does levels of sound. (I'll need confirmation on that.)

Here's a really easy improvement that could be made to the template. Please add a username and password field to the userbar section when a user isn't logged in. Just to show people what I mean, I used firebug to insert my idea (A simple copy of the form on the login page) into my version of the site. Here's how it looks. Of course, improvements could be made to the system, but that's a start. Just a mod of the template and a bit of css change to make it look okay.

The second idea is to have the username and password input be on the top bar. This would make it available on every page including the home page. The image idea2 shows a rough idea of what that would look like.