Counterforce: Difference between revisions
No edit summary |
|||
| (2 intermediate revisions by the same user not shown) | |||
| Line 10: | Line 10: | ||
This is less a single software project than a movement, perhaps a manifest. Changes will be necessary in a variety of tools. A single database of coolant hardware is likely to emerge, though that is not an explicit goal. | This is less a single software project than a movement, perhaps a manifest. Changes will be necessary in a variety of tools. A single database of coolant hardware is likely to emerge, though that is not an explicit goal. | ||
==Stage I: Linux sensor/control support== | github: [https://github.com/dankamongmen/counterforce dankamongmen/counterforce] | ||
==Goals== | |||
===Stage I: Linux sensor/control support=== | |||
Various coolant elements require closed source to properly use them. Perhaps foremost among these are the excellent products of Germany's [https://shop.aquacomputer.de/index.php?language=en Aquacomputer] and California's [https://www.corsair.com/us/en/ Corsair]. Almost all overclocking management requires closed-source Windows-only tools. This is ridiculous: the space of controls and data exported by these devices is fairly small. A fan exposes a single scalar control (PWM level) and generates a single scalar datafeed (revolutions per unit time). A pump is the same. A thermistor generates a single scalar datafeed (measured temperature). | Various coolant elements require closed source to properly use them. Perhaps foremost among these are the excellent products of Germany's [https://shop.aquacomputer.de/index.php?language=en Aquacomputer] and California's [https://www.corsair.com/us/en/ Corsair]. Almost all overclocking management requires closed-source Windows-only tools. This is ridiculous: the space of controls and data exported by these devices is fairly small. A fan exposes a single scalar control (PWM level) and generates a single scalar datafeed (revolutions per unit time). A pump is the same. A thermistor generates a single scalar datafeed (measured temperature). | ||
| Line 21: | Line 25: | ||
* Development of a cross-platform UI for control and data import | * Development of a cross-platform UI for control and data import | ||
==Stage II: Data discovery and visualization== | ===Stage II: Data discovery and visualization=== | ||
Logging and visualization of time series data is thankfully pretty much a solved problem, and I expect to leverage various open source tools such as [https://oss.oetiker.ch/rrdtool/ RRDtool], [https://prometheus.io/ Prometheus], [https://grafana.com/ Grafana], and others. Anything that can work with a time series ought suffice for our needs; I do not expect to personally spend time on development of such tools. | Logging and visualization of time series data is thankfully pretty much a solved problem, and I expect to leverage various open source tools such as [https://oss.oetiker.ch/rrdtool/ RRDtool], [https://prometheus.io/ Prometheus], [https://grafana.com/ Grafana], and others. Anything that can work with a time series ought suffice for our needs; I do not expect to personally spend time on development of such tools. | ||
What <i>does</i> come under the purview of this project is discovery of applicable sensors, automatic extraction of that data, and storing it on disk. Furthermore, it is expected that custom microcontroller setups will be in use; defining a lightweight protocol by which those micros can publish structured data, and making it easily integrated into e.g. [[Arduino]] projects, will be critical. | What <i>does</i> come under the purview of this project is discovery of applicable sensors, automatic extraction of that data, and storing it on disk. Furthermore, it is expected that custom microcontroller setups will be in use; defining a lightweight protocol by which those micros can publish structured data, and making it easily integrated into e.g. [[Arduino]] projects, will be critical. | ||
==Stage III: Volume-based coolant control== | ===Stage III: Volume-based coolant control=== | ||
Competent watercooling has three advantages over a pure air solution: | Competent watercooling has three advantages over a pure air solution: | ||
* more total heat can be removed from the system | * more total heat can be removed from the system | ||
| Line 33: | Line 37: | ||
For many people, the second advantage (a quieter system) is the primary goal. Yet I know not a single coolant manager that takes noise into account. <i>Noise ought be a fundamental concern of coolant management!</i> It ought be possible to measure ambient noise, possibly taking that into account for a "noise budget". It ought furthermore be possible to determine the noise being generated by cooling, and actively use that as feedback. Once this is accomplished, the user can specify a "cost of noise" function. | For many people, the second advantage (a quieter system) is the primary goal. Yet I know not a single coolant manager that takes noise into account. <i>Noise ought be a fundamental concern of coolant management!</i> It ought be possible to measure ambient noise, possibly taking that into account for a "noise budget". It ought furthermore be possible to determine the noise being generated by cooling, and actively use that as feedback. Once this is accomplished, the user can specify a "cost of noise" function. | ||
==Stage IV: Intelligent coolant control== | ===Stage IV: Intelligent coolant control=== | ||
Over time, the system will be able to build a <i>predictive</i> noise model, and optimize cooling based off of that. Doing so requires discovery of noise-to-perf relationships of various components by themselves and in combination. | Over time, the system will be able to build a <i>predictive</i> noise model, and optimize cooling based off of that. Doing so requires discovery of noise-to-perf relationships of various components by themselves and in combination. | ||
==Stage V: System discovery and visualization== | ===Stage V: System discovery and visualization=== | ||
The most speculative part of the project requires establishing the relationship of various cooling and heating components in space. Discovery of absolute geometry is probably not necessary, but effective topology is. In my current design, this will be accomplished through internal microphones and thermal imaging controlled by custom microcontroller firmware. | The most speculative part of the project requires establishing the relationship of various cooling and heating components in space. Discovery of absolute geometry is probably not necessary, but effective topology is. In my current design, this will be accomplished through internal microphones and thermal imaging controlled by custom microcontroller firmware. | ||
| Line 42: | Line 46: | ||
Noble work, that, [https://www.poetryfoundation.org/poems/45392/ulysses work not unbecoming men who strove with Gods]. | Noble work, that, [https://www.poetryfoundation.org/poems/45392/ulysses work not unbecoming men who strove with Gods]. | ||
==Implementation== | |||
My current implementation is set up on [[Schwarzgerät_III|Schwarzgerät]]. Control and visualization takes place on a Libre Computing [[Libre_Le_Potato_AML-S905X-CC|Potato]] using a [[Waveshare_AMOLED|Waveshare AMOLED]]. | |||
==See also== | ==See also== | ||
* [[ | * [[inaMORAta]], my custom controller for the [[MO-RA3]] external radiator. | ||
* [[Waveshare AMOLED|AMOLED]], my custom frontbay 1920x1080 display, and software for it. | * [[Waveshare AMOLED|AMOLED]], my custom frontbay 1920x1080 display, and software for it. | ||
* [[Schwarzgerät_III|Schwarzgerät III]], the machine which inspired this effort, and on which it is being developed. | * [[Schwarzgerät_III|Schwarzgerät III]], the machine which inspired this effort, and on which it is being developed. | ||