[Fwd: [M] Knock Sensor testor (longish)]

[Joe Boucher]

Here's a follow up:

UPDATE:   Since my earlier post, I built the tester, and tried it out on

the car ('89XR4Ti 5sp). It seems to work correctly. To connect up to the

knock sensor, I took an 8" length of 26 gauge insulated stranded wire.
Any
thin insulated wire would do. I pulled off the KS connector and inserted

the stripped end of the wire into the (+) socket of the KS connector.
The
(+) side can be determined with a voltmeter and the ignition on. The (+)

side will show about 2.5 volts. On my car the (+) KS wire seems to be
white
with a red stripe. I then pushed the connector back on, trapping the 8"
wire in the connector to the KS (+) terminal. Then I ran a longer wire
from
the end of my 8" wire to the inside of the car.

Inside the car I connected the circuit as per the previous post and
attached the 9V battery. The LED came on, showing that the KS sensor
terminal was below the 1V trip point of the circuit. I then turned on
the
ignition, and the LED went out, showing that the sensor wire was above
the
1V trip point. I then started the car.

I watched the LED while the car idled. Once in a while I saw it flicker
as
I briefly tapped the throttle, but mostly it stayed off. I then drove
the
car around on city streets, watching the LED. I saw sporadic flickers.
Sometimes when accelerating the LED would flicker on more. I tried to
get
detonation by applying throttle at low RPM, but didn't hear any
"pinging"
and did not get much activity on the LED. My static timing is set at 13
BTDC.

Anyway, the tester seems to indicate KS activity correctly. I found that

using a "superbright" LED is required, because the flickers are very
fast
and especially in daylight a normal LED can't really be seen very well.
I
used a superbright green LED, but a red one would work fine also. Radio
Shack has these. In any case, nighttime might be the best time to do
this,
and probably safer with someone riding along as well.

If you thought your performance were being compromised by the KS "coming

on" too soon, you could hook up the tester and see if at higher RPM's
under
load while accelerating the KS showed increased activity. This
additional
KS activity might be getting interpreted by the computer as detonation,
and
retarding the timing. Since we don't have access to the source code, we
don't know how the computer is programmed to react to the KS. Maybe a
lot
of little KS signals are disregarded, and only if the KS comes on for
longer periods of time is any action taken?

Anyway, there is it for what it is/isn't worth. If anyone wants to try
this
tester and needs some help, I'll be happy to help. email off list
please.

Steve

-------------------------------
Steve Roberts
sroberts at sunstroke.sdsu.edu
619 454 5234,  619 454 7008 FAX
-------------------------------