Crystal change

[Marc Alexander]

I don't know the details of the Motronic ECU, but I can say with 99% certainty,
DON'T change the crystal unless you know what may happen!!!!!
If you do, pay very careful attention to the results.

If you put in a faster crystal, usually the following will happen. All the
system timers usually run from this crystal, and it changes them all.
Injection timing pulse width will be reduced. Will run leaner.
Ignition timing will probably go more advanced. Engine destruction risk.
The ECU will think the RPM is lower than it actually is, raising the
rev limit, and changing the map values for that engine condition.
The map values used now may well be richer too.
Any serial port/diagnostic link baud rate will change, meaning you may not
be able to communicate with the ECU diagnostics. ECU communications with other
in-car devices (like ABS / suspension control, etc) may well still work OK,
as the ECU often uses a 'clocked' serial port (SPI) for this.

The reverse will happen if you lower the crystal value.

Actually, raising the crystal value by the percentage rev limit increase
desired may work ok sometimes. Especially if the ECU has knock control,
and re-retards the ignition timing to suit. Then you may get a slightly
richer fuelling, also made richer as the ECU thinks the 
RPM is lower, and probably looks up a richer fuel value from the map.
The ignition timing would advance some, but an ECU with knock control may
compensate.  The ignition timing advance is the most risky result, as
the engine may start pinging/detonating more, and melt the pistons/blow up/etc.

If the crystal value is raised by say 6.25% (would increase rev limit 
from 8000 to 8500), it might just work without serious problems.
But I'm really not sure, and wouldn't recommend trying it without serious
consideration!!!!

I must throw in a large disclaimer here! The changes I mention are only what
would happen to the systems I know of, and the Motronic and many others
I dont know many details of. The info above would be a likely result on
the timer system for the 68HC11 processor, and I can only guess that the
timer systems on other CPU's will do something similar.

Raising the crystal frequency also creates a risk of CPU error/overheating
or damage, as they are usually run at the max. recommended speed already,
but I must admit there's maybe room for an up to 20% increase from most
manufacturers recommendations. Some gadget makers use this to their 
advantage to make those little 'power boosters' for Macintosh/PC's, they
just bump the crystal speed up a bit. Everything then executes just that bit
faster on the CPU, but with greater risk of failure.

Sorry, didn't explain why the rev limit increases. It increases because
the CPU usually counts the 'timer' (crystal driven) pulses between each
engine (crankshaft/camshaft sensor) revolution(s). More pulses = lower RPM, 
Less pulses = higher RPM. So by raising the crystal frequency, the CPU
counts more pulses between the trigger points, and thinks the RPM is lower
than it really is.

Marc

>A friend of mine, changed the crystal in his Motronic ECU
>and this, increase the rev limiter  -500-  rpms.
>Can anyone tell me, why did this happen?
>Is it safe to change crystal ?
 
_Marc Alexander m.alexander at sungear.mame.mu.oz.au ph 93448105  fax 93482108_
_CMC Mechanical + Manufacturing Eng. Dept. Melbourne University,  Australia_
_______To stay young requires unceasing cultivation of the ability to_______
_______unlearn old falsehoods.__________Robert A. Heinlein__________________