We are developing a embedded system (a electronic door lock) that uses different LED's and a Buzzer for signaling different events to the user. There's a very limited opportunity to communicate to the user. A signal is just a sequence of LED's and Buzzer. Always only one at a time.
For example: Access OK: Green LED (100 ms) Buzzer (150ms) Access OK: Buzzer (200ms), Pause (100ms), Buzzer (200ms), Red LED (150ms)
Now the problem is that the timing is not very accurate. So we try to define the acceptable timing differences in the signal. We want to figure out how we could define some tolerance on the given times so the user doesn't see any difference.
The main questions are:
- What is the minimum perceptible difference in the duration that an LED or a Buzzer is on?
- What about pauses between them?
- How long is it acceptable (for a responsive device) to wait to signal after a user action?
I know this is a very fuzzy topic. It's not realy measurable and has also to do with psychology and other factors.
But maybe someone has already done some research on the topic or there is some known related work.