Joined
·
5 Posts
Long story short, here's a log of my lataest useless endeavour.
I wanted to see what would be required to build an ACD pump from "scratch"-without using any of the parts from an OEM pump.
It is possible, for less than 300EUR, with a few downsides, which could be fixed if given more effort.
Volvo EPS pumps have been used in DIY projects and had been already figured out, so I started from there. Idea was based around making pressure in a continuous-flow loop, by adding an arbitrary restriction somewhere, then by controlling flow, it controls pressure before the restrictor.
Here's the plan:
I built the controller from an Olimex ESP32-EVB devboard, as I had a case and connectors designed for other purposes. (It's not pretty on the inside)
The ACD ECU checks for a specific behavior of the solenoid, so my initial plan of replacing it with an equivalent 3.7Ohm resistance failed.
I had to compromise and keep my broken-neck solenoid in line, as I didn't want to figure this one out.
At idle the solenoid receives an average of 50mA. Specifications say 50-1000mA can be commanded, so I mapped that range to a 0-100% request.
ACD ECU sends a 200hz control signal, with an initial duty cycle of 50%, but the "high" part of the duty cycle is composed of another 18kHz PWM signal, with varying amplitudes apparently.
The simple way to measure something in this case was to add a ~400Hz 1st order RC LPF, with an included 1/6 voltage divider to keep my ADC happy.
Simulation of what I'm measuring after this is applied:
I wanted to see what would be required to build an ACD pump from "scratch"-without using any of the parts from an OEM pump.
It is possible, for less than 300EUR, with a few downsides, which could be fixed if given more effort.
Volvo EPS pumps have been used in DIY projects and had been already figured out, so I started from there. Idea was based around making pressure in a continuous-flow loop, by adding an arbitrary restriction somewhere, then by controlling flow, it controls pressure before the restrictor.
Here's the plan:
I built the controller from an Olimex ESP32-EVB devboard, as I had a case and connectors designed for other purposes. (It's not pretty on the inside)
The ACD ECU checks for a specific behavior of the solenoid, so my initial plan of replacing it with an equivalent 3.7Ohm resistance failed.
I had to compromise and keep my broken-neck solenoid in line, as I didn't want to figure this one out.
At idle the solenoid receives an average of 50mA. Specifications say 50-1000mA can be commanded, so I mapped that range to a 0-100% request.
ACD ECU sends a 200hz control signal, with an initial duty cycle of 50%, but the "high" part of the duty cycle is composed of another 18kHz PWM signal, with varying amplitudes apparently.
The simple way to measure something in this case was to add a ~400Hz 1st order RC LPF, with an included 1/6 voltage divider to keep my ADC happy.
Simulation of what I'm measuring after this is applied:
Attachments
-
3 MB Views: 6