Parasitic Powering of a Link Plus G3

[bootstrap]

G'day. Been helping install a Link Plus G3 into a Rover Tomcat over the weekend. Most of it is straightforward wire-to-wire stuff, but the relays and switches were where most of my effort was spent. The problem we were having was that the Link ECU would power up for no reason, albeit at only 8 volts, causing it to do 'strange' things. Pondering the problem I figured the ECU must be getting parasitic power (more on this at the bottom if you don't know what I mean). Since the fan relays we perminantly powered, (relay coils grounded to run the fans), the digital output pins of the ECU would have 12 volts on them all of the time. Problem Description: With the ignition off, the ECU would start up (its power rail sitting at 8 volts), close the main relay (powering it up fully) and run the fans briefly. After 10 seconds the ECU would open the main relay and the system would power off, except the power line goes back to 8 volts (instead of open-circuit) and the cycle repeats! Pull the fuse for the fan relays and the problem is gone. Problem solution: Move the relay power so that is comes from the ignition/turbotimer output rather than permenant to the battery, though this did involve splicing into the loom, fusebox and relay module and running a short wire etc. bit of pain. Has anyone else encountered this problem? The ACTUAL problem is probably due to the fact we hadn't programmed the ECU properly (damn you vista64), but 12 volts at any input could cause this to happen right? A cautionary tale? Notes on parasitic power: Parasitic power can be done deliberately (such as in iButton/OneWire systems) or can be unintended. One simple way of protecting digital inputs is to use a pair of diodes, creating a 'diode clamp' circuit. Thisb prevents the input voltage swinging beyond 0.7 volts above/below the position/negative supply rails, as the diodes will conduct 'into' the supply rails, effectievely clamping the input voltage at these limits. Some integrated circuits (silicon chips) have these diodes built into the design to protect the inputs. If the positive rail voltage is open or high impedance to 'low' (such as being in the 'off' state), this current flow from the input can be enough to bring the positive rail up to a point that the circuit can start. The problem is that is normally these diodes are never intended to handle much power and simply burn out, removing the protection from the input. In the case of the Rover, the fan relays are powered permenantly, with the relays activated when the other side of the coil is grounded. The ECU pins that go to these relays are thus always at 12 volts, and IF diode clamps are used then a current could flow into the pins and power the ECU from the inputs. Due to the resistance of the relay coil and current demand of the ECU (not running the engine, that is about 50km away from the ECU atm), the result in only 8 volts at the ECU power rail. That's the theory anyway.

[Simon]

The attached PDF shows what is going on in this case.

Simon

[bootstrap]

Ha so a case of RTFM I guess. Thanks for the prompt reply dude.