We have discontinued this forum and moved support tickets to the GitHub issue tracker. More info

Feature request: improvement of sensor-based algorithm

Control fans on Apple hardware: iMacs, Macbooks, etc

Moderator: Steven

Post Reply
theomoto
Posts: 33
Joined: 22 Jun 2013, 00:52

Feature request: improvement of sensor-based algorithm

Post by theomoto »

Hi guys! Love Macs Fan control on my Mac mini 2012 i5.
A good idea for a feuture request is the ability to choose different algorithms for automatic fan speed regulation. What I mean by that is the ability to choose the sesitivity of the fan changing his speed.
For example you could have like three options: "extreme" where the sesitivity of the fan is high, "normal" where the sensitivity is normal, and "low" where the sesitivity is low.
That will be usefull when choosing it from the drop down menu of the status bar.
User avatar
Steven
CrystalIDEA Developer
Posts: 2294
Joined: 21 Dec 2009, 11:48
Location: CrystalIDEA headquarters
Contact:

Re: DIfferent algorithms

Post by Steven »

Hi,

Do you think that the current algorithm is sometimes slow or fast? Slow is what I'd think
theomoto
Posts: 33
Joined: 22 Jun 2013, 00:52

Re: DIfferent algorithms

Post by theomoto »

Steven wrote:Hi,

Do you think that the current algorithm is sometimes slow or fast? Slow is what I'd think
I'd think slow to normal, more to the slow side.
What about the idea and the implimitation that I suggested?
User avatar
Steven
CrystalIDEA Developer
Posts: 2294
Joined: 21 Dec 2009, 11:48
Location: CrystalIDEA headquarters
Contact:

Re: DIfferent algorithms

Post by Steven »

Yes, I'm thinking of improving the algorithm. And it's in the todo list, though not sure of exact implementation yet.
theomoto
Posts: 33
Joined: 22 Jun 2013, 00:52

Re: DIfferent algorithms

Post by theomoto »

I think the improvement of the auto algorithm should be number 1, then profiles.
Also should you go with the extreme, normal, low settings, an auto setting would be great, where the app itself decides what alogorithm to use depending on overall coditions such as temperatures, ram being used etc.
User avatar
Steven
CrystalIDEA Developer
Posts: 2294
Joined: 21 Dec 2009, 11:48
Location: CrystalIDEA headquarters
Contact:

Re: DIfferent algorithms

Post by Steven »

theomoto
You're right that improvement of fan control algorithm is very important. It's going to be re-written in a new great way.
theomoto
Posts: 33
Joined: 22 Jun 2013, 00:52

Re: Feature request: improvement of sensor-based algorithm

Post by theomoto »

Another idea!

You could implement a graphing section where the user can map his own personal curve of how the temps affect fan speed! That can be really usefull when looking to fine tune your cooling according to what you do. The ability to save different curves as modes is a plus!

I think that what holds macs fan control back is the algorithm and then the linear response of the fan speed according to the temps.

ps. i know crude drawing but you get the idea.
Attachments
idea.png
idea.png (53.74 KiB) Viewed 19440 times
User avatar
Steven
CrystalIDEA Developer
Posts: 2294
Joined: 21 Dec 2009, 11:48
Location: CrystalIDEA headquarters
Contact:

Re: Feature request: improvement of sensor-based algorithm

Post by Steven »

theomoto
Thanks but I really want some better math for that.
theomoto
Posts: 33
Joined: 22 Jun 2013, 00:52

Re: Feature request: improvement of sensor-based algorithm

Post by theomoto »

Yeah I'm at math university myself so I know what you mean..!

You can actually make use of this idea by breaking down the curve to small linear parts! How small is all up to you! Pic like 5-or more- points on the curve, make linear equations using two consecutive points, and you will have like an estimate of the curve. The math could be simpler that way!
I'm not sure but in that way you could make speed changes in fan speed more dramatic, depending on the tilt of the linear equation.


ps: I write a little bit of code myself and I could helf you out with the math part, but now I'm like in a bit of a struggle with some final exams so it is out of the picture for me. Maybe in the end of 2014- start of 2015.
Post Reply