| Per Westermark 06/19/09 15:05 Read: 199 times Sweden |
#166265 - Interference Responding to: Erik Malund's previous message |
Was all hardware designed long-time ago, or is your banking pin a recent hardware change?
As already noted, you get a number of problems if you let the same processor pin switch both data and code banks. It cuts off a lot of your sw design options. You either must access the banked variables from a common code bank, or use helper functions for the switching. And you do not want any ISR to touch anything banked. This is a big reason why you always have to look ahead when designing the hardware. And even if your initial software don't need all features of a hardware platform, you have to write test applications that validates all parts of the design - this includes possible interference between code bank switching and data bank switching, trying to maximize current consumption, trying maximum signal frequencies on all relevant signal lines, submitting the board to noise levels way above the expected installation environment etc. |
| Topic | Author | Date |
| Code and XDatabanking | Sandeep Gupta | 06/16/09 02:55 |
| Manually handle a few large variables? | Per Westermark | 06/16/09 03:39 |
| Code and XDatabanking | Sandeep Gupta | 06/19/09 01:54 |
| You missed the point! | Andy Neil | 06/19/09 06:16 |
| Do you *have* to use an 8051? | Andy Neil | 06/19/09 02:07 |
| Code and XDatabanking | Sandeep Gupta | 06/19/09 05:35 |
| Fair enough, then! | Andy Neil | 06/19/09 06:10 |
| no banking | Erik Malund | 06/19/09 08:33 |
| Cross reference | Andy Neil | 06/19/09 08:49 |
| Verbiage | Erik Malund | 06/19/09 09:48 |
| P3_3 commonly for ROM and RAM to switch banks? | Frieder Ferlemann | 06/19/09 14:09 |
| that will not work | Erik Malund | 06/19/09 14:44 |
Interference | Per Westermark | 06/19/09 15:05 |



