NOT fom node30 (VBATT), good. So the FP and DME relay are in series (the fuelpump current goes across both).

* this means the low voltage flyback of FP relay (M52/69 output) cannot cause any undesired leak when ignition is off and DME is not powered up.

----

Maybe M52/9 fuel consumption output can cause damage to p259 chip ?

* this (normally ECU output) was marked as ECU input in BMW doc (believedly bogusly but not confirmed in an established systematic way).

* any note would be nice. Maybe 200mA PTC in series would be safer for this ?

----

OLD: Prelling - solved by adding diode, anode=ec36/25 >| cathode = mot88/65 cam sensor 12V (BMW88/54 left open)

The strange thing:

* mot88/54 (and mot88/65) cam12V seems to be fed from harness (+12V, via main relay) anyway

* might be unnecessary from the ECU (but feeding +12V eg. from EC36/23 flyback inside the ECU, if +12V is fed in the harness anyway, should not cause a problem)

----

Historical, only for the curious

(some confirmation, notes from install measurements would be nice; and would allow better cleanup).

We received a report of some relay prelling (when ignition off) that we cannot reproduce.

* powering up p259 outputs through a small 30A relay (or 2 in parallel) to +12V without powering the ECU.

** this is assumed (and verified) to "flyback-ed to inj flyback equivalent to EC36/27"

** not to ECU power mot88/49 (=EC36/25 switched 12V) - that would explain if the solenoid is strong enough to provide ECU supply current. Verified that no diode to ECU power.

What should be hooked to the mot88 (which pin) to simulate the "ignition off" state ?

* In the car the main relay controlled by mot88/73 supply is fed by mot88/26 VBATT constant 12V from the battery, the ECU is powered via the Ignition switch 12V+

** so if the main relay is ON and the ECU is switched off via the switch it seems to stick on when protection diodes are connected to mot88/49 (=EC36/25 switched 12V).

** The flyback=mot88/87 (not pin54 in the xls beware, that is cam sensor +12V!) in the car also comes from the main relay but could possibly be moved to a constant 12V (pin26) instead to solve this?

** good find: apparently the mot88/54 = mot88/49 (12V connection wire) inside the ECU is simply not needed, if the mot88/54 is supplied from the harness (weird, why is it connected back to the ECU at all ?)

** Can a failed diode in the relay cause this?

*** possible, but more likely the mot88/54 cam12V seems to be unnecessary from the ECU if the harness feeds 12V there (via main relay) anyway

OEM wiring information here

http://www.gbnetwork.co.uk/bmw/ftp/etm_e36_1998final.zip

Page 108 shows that flyback 12V mot88/87 = mot88/54 cam12V in the harness so how that same connection can cause any problem in the ECU ?

----

* EC36/25 = mot88/49 = 12V switched (ECU main supply)

* EC36/16 (p259/ch1) = mot88/73 main relay

** why not some output ? Which output should control mot88/73 ?

* EC36/23 = flyback 12V = mot88/87 flyback

** flyback 12V mot88/87 = mot88/54 cam12V (ALSO in the harness, see p108)

** flyback 12V mot88/87 = mot88/65 cam12V (see p114)

The bold, or last 2 are suspect to cause prelling, especially if switched 12V (ECU main supply) is connected to cam12V in the harness

* the ECU might be powered from the cam12V unintentionally

** this was not present in any of our previous experimental bench setup tests.

New Tests

* with mot88/54=65 connected to mot88/49 (= EC36/25) as in the BMW harness

* 83.2 Ohm 50A relay (or 2 in parallel ~41.6 Ohm) between VBATT +12V and mot88/73 main relay pin

No prelling at all. ECU does not power up.

If ECU forced ON with mot88/49 = 12V than ECU turns on, with minor 0.4sec delay the main relay turns on and no prelling: everything seem perfect.

Maybe the BMW has some very low resistance relay

Proposal:

* high voltage flyback to mot88/73 (main relay output), very easy to apply:

** either a 18V transient diode to GND,

** or a small (even 1n4007) diode, cathode to the 30V injector flyback rail.

** however, if we are not certain that the mot88/73 controlled relay caused the "prelling" in the BMW harness, we cannot be certain to have it fixed.

** Gunnar, do you remember that the "main relay" was prelling ? When main relay was removed, there was no prelling ?

*** Yes, main relay prelling and yes no prelling with main relay gone

----

Dave suggests:

The problem with the PNP relay prelling is caused by the

* main power relay being connected to vbatt permanent and the

* ecu power being connected to #15 wire

** which normally has many loads (low impedence like instrument lights) connected to it to gnd;

** this "somehow" creates a relay turn on path (slightly weaker than normal) through the p259 protection diodes

v3.8b

p259 diode >| mot88/87=EC36/23

not connected to EC36/25 (#15) inside the ECU.

Somehow mot88/54 is (suspected) involved in creating the path...

* but not if mot88/54 is connected to mot88/87=EC36/23 both inside and outside).

* beside the main relay flyback path changed to 12V+30V side of the transient diode (that alone should solve it)

* adding diode, anode=ec36/25 >| cathode=mot88/54 = mot88/65 solves it

Actually, perhaps the main relay itself connected the #15 to injector supply, EC36/23, and possibly causes inability to turn off main relay ? (reproducing failed because of some relay differences ?)