View Issue Details
| ID | Project | Category | View Status | Date Submitted | Last Update |
|---|---|---|---|---|---|
| 0000311 | SpeedFan | Other | public | 2005-07-12 16:48 | 2005-07-21 00:41 |
| Reporter | klebom | Assigned To | alfredo | ||
| Priority | normal | Severity | feature | Reproducibility | N/A |
| Status | acknowledged | Resolution | open | ||
| Summary | 0000311: Feature wanted: Auto ranging FANx divisors | ||||
| Description | As the computer is heating up after a cold start, fan speeds can vary of a broad range. Since the best FANx divisor to use (the one giving optimum reolution) will vary with fan rpm, I suggest implementing auto ranging FANx divisors. | ||||
| Tags | No tags attached. | ||||
| Motherboard Model | |||||
| Video Card Model | |||||
|
|
This is something I coded long time ago, but the net result was quite poor. When you get improper readings, they wouldn't be fixed by changing the divisor. Those odd readings are due to PWM control (there is an article about PWM on SpeedFan's web site). The most difficult part would be to detect whether SpeedFan is reading 0 RPM because of a wrong divider or because of stopped fan. This is something I'm investigating. |
|
|
I suggest an additional FANx divisor value of AUTO. That way the current functionality is preserved. On my ASUS K8N-E deluxe (IT8712F) there are very stable readings and very good PWM control. I guess this mobo isn't alone. Autoranging of course must be based on several readings. Also some sort of hysteres must be included to avoid range toggling :) A reading with count < 123 => select next smaller divisor (d/2). A reading with count > 250 => select next bigger divisor (d*2). Assuming divisors d/2 and d*2 exists of course. Will a stopped fan need special handling? At least with IT8712F even a divisor of 128 will give readings up to 10500 RPM, although at very rough resolution I admit. |
