Originally Posted by IT Diva
I've written a lot regarding the Aquaeros, the A6 in particular, regarding their PWM implementation.
The PWM version of the D5 pump, and the PWM Corsair fans have been a bit problematic.
If you do an advanced search with IT Diva as the username and PWM, Corsair as the keywords, (tick the" posts only" box, not "entire site") you'll get 3 pages or so of relevant posts.
If you do an advanced search with my name and PWM, "sink current" as the keywords, you'll get pages about PWM. (tick the "posts only" box)
A quick copy paste from one post explaining briefly how to look at PWM in a PC:
In actuality, a PC PWM controller based on the Intel PWM standard, "pushes" no power at all.
It sinks current to ground.
Each PWM controlled device is responsible for supplying its own 5V level.
Therein lies the problem as you add greater numbers of controlled devices . . . . . the amount of current that the controller has to sink to ground rises, and manipulating Ohm's law, voltage dropped equals I squared R, the drop across the protective resistors on the controller's PWM input raises the logic low voltage level that the controlled devices see above the threshold to where it isn't seen as off-time any more, and you loose progressively more low speed control as you add devices.
It's a much better analogy, and much more accurate to think of it as, "the signal current level of multiple devices overwhelms the controller," than that "the controller signal becomes too weak as more devices are added".