Re: SI reset tool information - more goods!

["Aaron Bohnen" <bohnen@xxxxxxxxxxxxxxxx>]

Hi John,

Thanks for the comments:

J> While a desire for more durability is laudable when designing other
J> things - bridges, for example :) - in this instance, a 600V 3A diode
J> is probably overkill. 

Yes, it is for sure. Just happened to be what I had on the bench and looked 
hefty enough for physical durability.

> Much of the car's electronics is already diode
> protected to at least 10 ma.  In 12v circuits, a 470 ohm resistor might
> allow too much current.

In which case a 1000 Ohm resistor was reported to work also, but to have 
been the highest value that did work.  Such a resistor would limit the 
current flow if the jumper were to be wrongly connected down to less than 
14 mA. 

Since I'm suggesting users install a non-conductive plug for the battery 
pin in the diagnostic connector I'm not too worried about this, but rather 
more about the +5 V that is coming off of pin 7. In this case, the 470 Ohm 
resistor would limit current flow to under 11 mA. Right in line with your 
observations about the car's electronics system.

If such a jumper were to be wrongly installed _from_ the battery pin rather 
than from the reset pin the max current flow would be under 30 mA. More 
than the 10 mA figure that seems to be more preferable.

Unfortunately, the paranoiac engineer's first choice, namely to use a 
higher value resistor (1.5k as you suggest), has anecdotal evidence going 
against - specifically that 1000 Ohms was the minimum value that Paul Reitz 
found to work. (the reference for this is the Service Indicator FAQ at:

http://cbsgi1.bu.edu/bmw/sil.html

Paul Reitz's observations are in the UPDATES section.

Now he specifically does not mention what was the next higher value 
resistance that he tried above 1k. Unfortunately for us these lights take 
some time before they need resetting, in fact even before one oil service 
light will go out. Scientific testing is tough with only one car!

Now if there were others willing to participate in this little venture 
and someone with the 15 pin connector and a car that needed a reset would 
sequentially take a 1.5k, 1.2k, 1k and 470R resistor to their unit and let 
us know the first one that works that'd be great. Same thing for someone 
with a 20 pin connector and car that needed a reset... Anyone willing to 
get in on this prestigious research topic?

> If you were to omit the diode and use a 1.5K resistor, your
> circuit would be simpler and might protect just as well against
> misconnections.

Well, I select against the 1.5k resistor for the reason that we have 
anecdotal evidence against it. Maybe a 1k resistor, but then if that was 
the marginal value (measured on the 20 pin connector) it may not work well 
on cars with the 15 pin connector. I just don't believe there's enough good 
information available to adopt the marginal value resistor at this point 
and I'm willing to take the somewhat less safe route because of the 
inclusion of the series diode.

> > Good luck and keep your eye out for more news on the mating connector...
> 
> Are you looking for one or are you asking us to help out? I 
> can suggest something for an E36. 

Both! :> What do you have in mind?

Thanks and best regards,

Aaron

p.s. - regarding your first reply to my posting asking where the 
concealment was, and what general parts would be required to just go ahead 
and build a timed jumper - I know that's a reasonably straightforward 
project. What I was looking for was better information on the car system 
itself - namely whether our understanding of the reset process was 
comprehensive enough. Specifically, is the BMW-designed reset mechanism for 
the Motronic system really just a timed short from the reset pin to ground, 
or is there more to it? Tx, A.
___________________________________________________________
Aaron Bohnen                     email: bohnen@domain.elided
- -Ph.D. Candidate, Civil Engineering Department, U.B.C.
- -Technicraft Engineering Services

------------------------------