| Kai Klaas 02/07/10 08:10 Read: 146 times Germany Msg Score: +1 +1 Informative |
#172956 - Never discharge a cap directly by a switch! Responding to: Joel Ward's previous message |
Joel said:
@ everyone else - thankyou very much, I have tried another implementation of the switches on a separate piece of veroboard and changed the (what i thought to be brand new but cheap) switches for the ones I had ordered to be used on the final version. These work far more reliably! The mistake is, that your switch directly discharges a cap without any current limiting! That burns down the contacts and results in a little, non-conducting oxid spot. Always limit the current rushing through such a tiny switch to less than a few mA. So, don't discharge the cap directly, but via a resistor. Just put a resistor in series to the contacts, like shown in this example: 
I think the evalboard from SILABS has 100k pullups? Then a current limiting resistor of 2k2 is adequate. The current through such a switch should be more than 100ľA, but less than 5mA, for reliable switching. Kai Klaas |
| Topic | Author | Date |
| Buttons - Hardware | Joel Ward | 02/05/10 11:51 |
| do double check .... | Erik Malund | 02/05/10 11:55 |
| Debouncing | Andy Neil | 02/05/10 12:55 |
| Have you considered that it could simply be the "breadboard" | Richard Erlacher | 02/05/10 13:56 |
| I (dis)agree | Erik Malund | 02/05/10 14:48 |
| Relevance to pushbuttons | Andy Neil | 02/06/10 03:01 |
| Consider the objective | Richard Erlacher | 02/06/10 08:03 |
| Agreed | Andy Neil | 02/06/10 08:39 |
| that's why there's so fluid a definition for "working" | Richard Erlacher | 02/06/10 13:08 |
| PCB's to match the contact arrangement on a "breadboard" | Andy Neil | 02/06/10 13:25 |
| Not exactly ... | Richard Erlacher | 02/06/10 18:05 |
| Solved: | Joel Ward | 02/07/10 06:18 |
Never discharge a cap directly by a switch!   | Kai Klaas | 02/07/10 08:10 |