| ??? 11/29/07 15:08 Modified: 11/29/07 15:13 Read: times |
#147568 - ok, let me try Responding to: ???'s previous message |
please elucidate how you suggest to measure the %error
why bother, there will be none. The only precision component needed is Rfix (see below). I would make Rfix out of two resistors in series with opposite tempcos, that way you can get virtually 0%. make a setup like this (example illustrates reading from a voltage divider, you, probably use a bridge, but same method applies) Vsomething *--------------------- ADC1 | Rfix | *--------------------- ADC0 | Rvar | gnd with this setup adc1 will read x and adc0 wull read y so whatever Vsomething is (cheap reference) x (Rfix + Rvar) --- = ------------- y Rvar Thus, the only thing affection your precision will be Rfix which you will need anuhow, the error from Vref has disappeared. Erik |
| Topic | Author | Date |
| Cygnal/Silabs MCU Adc reference | 01/01/70 00:00 | |
| WHICH? | 01/01/70 00:00 | |
| WHICH? | 01/01/70 00:00 | |
| Research? | 01/01/70 00:00 | |
| Research? , yes you are right | 01/01/70 00:00 | |
| Maxim, AD, LT | 01/01/70 00:00 | |
| thanks | 01/01/70 00:00 | |
| Tall order | 01/01/70 00:00 | |
| REF3225 == The price of F206 :( | 01/01/70 00:00 | |
| What kind of sensor? | 01/01/70 00:00 | |
| go relative | 01/01/70 00:00 | |
| go relative | 01/01/70 00:00 | |
| ok, let me try | 01/01/70 00:00 | |
| ok, let me ... | 01/01/70 00:00 | |
| explaining | 01/01/70 00:00 | |
vsomething | 01/01/70 00:00 |



