| ??? 02/13/07 20:51 Read: times |
#132830 - Keil? Responding to: ???'s previous message |
Erik Malund said:
They told me that they couldn't accurately predict cycle counts in Maxim/Dallas MCU's because some of them were not predictable.
1) to expect ANY toolmaker to fully support all 4711 derivatives of the '51 is ridiculous. That they don't do it is one thing. That they come up with a nonsense argument is another. Erik Malund said:
2) the 'cycle stretcher' in the Dallas chips DO make "cycle counts ... not predictable". The only thing which can make cycle counts unpredictible is an external cycle-stretching input (aka WAIT) which AFAIK is in no commonly available '51. Even if there would be such, a "best case" cycle counting is always possible. Erik Malund said: How did you come to that figure? Why not 5 or 7 or 9?
If you run your thingy as a 'standard' '51 and divide by 3 you should be close enough JW PS I have to repeat my shameless self-advertising: for cycle counts why don't have a look at the "fast '51s" overview on my pages. It's available as excel sheet, too, and also an example is given ... |
| Topic | Author | Date |
| Cool assembler/compiler IDE feature.... | 01/01/70 00:00 | |
| No, but... | 01/01/70 00:00 | |
| the other way round | 01/01/70 00:00 | |
| Keil has something | 01/01/70 00:00 | |
| interesting.... | 01/01/70 00:00 | |
| there is something in the simulator | 01/01/70 00:00 | |
| I asked KEIL about that ... | 01/01/70 00:00 | |
| derivatives, derivatives | 01/01/70 00:00 | |
| Keil? | 01/01/70 00:00 | |
| 12/4 = 3 | 01/01/70 00:00 | |
on 4-clockers | 01/01/70 00:00 | |
| eh, on the simulator | 01/01/70 00:00 | |
| Pinnacle 52 ? | 01/01/70 00:00 |