Fix formula for automatic fan control and adjust minimum fan speed #59
Conversation
… c, and associated min fan speed to 33 (up for 25) for lower temps.
There was a problem hiding this comment.
Concept ACK
The original code does have an error with values at the minimum temp and your code fixes it and appears to do what the original code intended. Have some minor comments. The main thing is to at least fix the comment since that can be very confusing if you have a comment that doesnt match the logic.
If you replace all hard coded values with descriptive local variables you can likely delete the comments above the function as well since it should be self evident at that point what the logic is doing
|
Please change the min temp back to 50 as it is still very cool for the chip and reduces fan noise. |
* reworded comment to be clearer * added min_fan_speed local variable * added fan_range local variable and rennamed existing range local variable to temp_range
|
Great, i'll test it tonight |
|
I have tested this successfully with a Noctua NF-A4x10 5V PWM:
@benjamin-wilson my Bitaxe ultra feels warmer than the reported chip temperature: at 44 degrees c reported, an infrared thermometer shows 55 degree c for the heating surface temp, and 59 degree c at the back of the board under the BM1366. |
|
@ozbibi Which version of bitaxe ultra do you have? Sounds like V1? V1 won't work with automatic control as the temperature sensor is not in the correct location. |
|
@benjamin-wilson I indeed have a V1. I'll use manual fan control for it then. |
|
Checked this out tonight. @ozbibi I think 33% starting is a little aggressive, I tested with 20% and it seemed to work well settling the chip temp at 55C. Could you change it to 20? Then I think it's good to go. |
|
@benjamin-wilson I have lowered the min fan speed to 20 as requested. |
…kot#59) * Fix formula for the automatic fan control
Fixed the formula for the automatic fan control #58.
Adjusted the minimum fan speed to 33 (vs 25 previously), and min temp to 40 degrees c (vs 50 previously), as tests with a Noctua NF-A4x10 5V PWM at 24 degree c ambient, show that this keeps the chip temperature cooler (~42 degrees c) than the previsous minimums (~51 degrees c).