Many times not all graphics are stored by the WaybackMachine, so if there are graphics missing from the following VFAQ, please don't email me about them (unless you are emailing me to tell me you have a copy of the missing graphic you can send me).

If you know the original author of this VFAQ, and have their contact info, please have them email me if they have any missing info to add, have the VFAQ at another site I could not find, or wish to have it removed from my site.

Club DSM Knock Sensor Theory

Most of these posts are the result of a discussion initiated by Randy Nuckels. Randy is also responsible for providing the "collection" found here! SIGNIFICANT contributions have made by many others as well! Though it takes a while to read all of this, there is a LOT of good info about knock and DSMs. Once you've installed a Knock Sensor LED, you'll know whether the performance of your car is being reduced as a result of Knock Sensor activity - the contents of this page can help determine what may be causing it.

I have spent a great deal of time editing the formatting of these posts to help provide easier readability.... I'm not completely satisfied with the format (maybe it needs an outline?). If you have any suggestions about how to best format this data, (especially if you want to volunteer to do it :)), just say the word!

There is a TON of additional Knock-Related info to be found in the Club DSM archives -- If you'd like to use the Archive Searcher to add to this list, or if you consider yourself an "expert" on this topic and would like to add your $.02 to this discussion, please let me know!

Subj: Knock Sensor
Date: 97-01-30 17:22:31 EST
From: RandyNet@ticnet.com (Randy Nuckels)

THE QUESTION:
My car: 91 Eagle Talon Tsi FWD 92,000 miles,
2.5" Cat back exhaust, Cockpit variable boost controller,
Boost control solenoid cockpit LED light & Buzzer, No engine mods.

I need suggestions of what type of things, aside from true pre-ignition,
that would cause the detonation sensor to signal the ECU to retard the
ignition timing. Noisey lifters are the first quick answer, but what else?

As these DSM engines get older there are many moving engine parts that
could wear out and cause "knock" that the knock sensor would consider
pre-ignition. A WAIS search found many people that cured hesitation
problems by replacing noisy lifters. The old lifters were not causing
the performance problem, but instead the noise they created was causing
the ECU to think it had a knock problem and retard the timing. Over
the weekend, I replaced all of my lifters, and it did not fix my
hesitation problem.

In attempts to educate myselft, my DSM has been running with 2 working
knock sensors for months. One is soft mounted to the engine bay so
it cannot hear anything. I call this one my deaf sensor. The 2nd one is
mounted to the engine block where it should be. I call this one my live
sensor. I have switched the two locations to prove that both sensors work.

THE TEST: When the live sensor is hooked to the ECU, the ECU hears
"something", retards the timing, and causes a serious hesitation
problem, even at low RPMs. With the deaf sensor hooked to the ECU, it
runs like the wind. This test proves that the knock sensor singnal is
what is causing the problem, and nothing else. Note, you cannot just
disconnect the live knock sensor wires, because the ECU will recognize
it is not hooked up, and default to open loop mode, that is why you must
have 2 knock sensors for this test.

After I replaced all of my lifters, "THE TEST" has the same results. Why?
What else could the live sensor be hearing? The only thing I can think
of next is the timing belt idler pully. When I did the lifter job, I
decided it was "noisey" and had potential to fail, so I am going to replace
it next weekend. I definately do not need that part to seize up or
desinigrate. I don't feel like doing any more valve jobs!

I do all of my own engine work, and I figure that I can just keep
replacing parts cheaper than having the dealer try to figure it out. I'm
just not sure where to go next. There has got to be some other high
mileage DSMs with the same problem.

I am taking any and all suggestions. Any thoughts or ideas you have
are worth me researching at this point. Please e-mail any reponses directly
to me, if I solve the problem I'll be sure to post the ideas and solutions
to the digest so they will be accessable to others through a WAIS search.

Subj: Knock Sensor
Date: 97-01-30 17:22:31 EST
From: Tsitalon1@aol.com
To: RandyNet@ticnet.com

I read your posting on the digest and was astonished! You seem to be
experiencing the same problem I am. I to seem to be having hesitation
under acceleration, especially at lower Rpm's . It feels to me as if
something is holding the turbo ever so slightly every once -n-awhile
as it spools up. Now I hear the turbo spool up strong so I don't think
it's the turbo. In fact I took the turbo off many times and visually
checked for free movement on the compressor and turbine sides. Then I
checked for air leaks in the intake plumbing...Nothing!! Tried octane
booster, Replaced all sparkplugs and wires,Still same problem. I just
recently replaced all my valves myself so thats not it. Now I don't
have an extra knock sensor to install to test if thats my problem, But
reading your post gave me an idea.

Now something is setting off the knock sensor right? We don't know what
it is. Could there be a way to "turn down the signal from the sensor to
the ECU" say maybe with an in-line resistor that would limit the sensitivity
of the sensor? Who would we contact for that info? I'm thinking maybe
a Mitsu engineer?????? Someone who knows the voltages and currents
involved in the ECU system. I don't know if my idea is rational or even
feasible but it is an idea!!

Also, Another crazy Idea, How about synthetic oil? Isn't it quieter than
natural oil? I don't know I'm asking cause I never tried it.! How about
Slick 50??

How about using a real thin rubber washer when you mount the sensor to
the engine block?? would that work?? ........... I'm trying man .........

Please E-mail me back I'm very Curious

Subj: Knock Sensor
Date: 97-01-30 03:07:21 EST
From: bluegsx@accucomm.net (John Moore)
To: RandyNet@ticnet.com

Isn't it dangerous to have the ECU unable to hear knock? From what I
understand, the ECU constantly advances timing to the point of knock,
backing off right at the threshold of knock. When the sensor doesn't
sense knock (like your "deaf" sensor), the ECU radically advances timing
because it assumes everything's ok. I'm surprised you haven't blown
something. Check with Todd or DaveB before driving with the "deaf"
sensor. You could end up needing a new engine.

_/ <> /-/ /\/ /\/\
http://users.aol.com/bluegsx

Subj: Knock Sensor
Date: 97-01-30 09:00:28 EST
From: markp@Exabyte.COM (Mark Pilon)
To: RandyNet@ticnet.com

Great test !

I've long suspected an over-active knock sensor but haven't thought of
a way to get around it. lifters or an exhaust rattle could do it.
maybe a noisy water pump or alternator . . .

I think there's a torque specification for installation of the knock
sensor. too tight and I think it hears too well. easy enough to
re-install the sucker w/ less torque.

another option might be to install it w/ teflon tape -- the kind used
to seal threads on iron water pipe. this might filter some of the
noise, but it's hard to tell how much is too much.

keep me posted on your progress as this I'm having the _same_ problem.

thanks,
Mark Pilon

Subj: Knock Sensor
Date: 97-01-30 09:19:46 EST
From: Kyle.Zingg@RossNutrition.com (Kyle Zingg)
To: RandyNet@ticnet.com

>I need suggestions of what type of things, aside from true pre-ignition, that
>would cause the detonation sensor to signal the ECU to retard the ignition
>timing. Noisey lifters are the first quick answer, but what else?

Congrats on all of the work. It is impressive that you have sorted things out
this far. I have a very stupid suggestion, in case you haven't checked it
yet. Is the timing correct, and is the timing belt in the correct place. I
had my timing belt skip two teeth, and ran pretty well, but was a dog when I
stepped on it. The belt had skipped, so the cams were severely advanced,
causing preignition, and tripping the knock sensor. I found it by setting the
timing to 0deg, and noticed that the marks on the camshaft pulleys didn't line
up properly.

Dumb, I know, and you have probably already covered it, but I thought that I'd
try. I did the lifter thing, and tested every other component I could find, to
no avail. Through WAIS, I did find a comprehensive list of things to test
though. If you need it, Email me and I'll tell you where it is.

Kyle

Subj: Knock Causes
Date: Mon, 07 Aug 1995 23:47:50 EDT
From: STAV86A@prodigy.com (MR LUIS HADDOCK)
Subject: Re: detonation sensor cost
Message-ID: <#14>

> I'm curious as to how much the detonation sensor
> cost you. If they're cheap enough, I'll just go ahead and replace
> it without any further diagnosis.

Paul,
The sensor was replaced under warranty for me but I did see the prices. It was
$74 retail and $44 warranty cost. Now before you just go and replace a sensor
I would recommend the following:
1) Test your fuel pressure boost solenoid -- if it is not increasing the
pressure of the fuel rail you will lean out and cause knock
2) Test your inlet air temperature sensor -- if this is out of whack your
computer will miscalculate and again cause knock
3) Make sure all your injectors are working fine, try running a bottle (or 2)
of Techron through first -- deposits on valves,
or injector or even inside the cylinder can cause knock and you do have some
years on your car
4) Check your water temperature sensor (we have two, one for controlling the
EGR, the other for CPU; look for the one with an electrical connection mounted
near the thermostat) -- again bad sensor equal bad running engine
5) Check your O2 sensor -- same thing
6) If you live in high altitudes the barometric sensor should also be checked
7) Check your plugs and their wires -- a lot of old cars have bad wires causing
lots of hiccups
8) You say your car never been touched but you habitually wash the engine
compartment -- clean and check all wiring connectors to all electrical points
on the engine
9) Did your problems start after you bought gas from that new place? Just
checking...
10) Do you have any failure codes locked in the computer -- check and repair as
necessary,
11) Is your air filter still clean? how's your ignition timing? Are all your
vacuum lines connected correctly and not leaking? Is the EGR operating
correctly? Do you have the correct vacuum at your PCV valve? Is your exhaust
manifold in good shape?
12) Now, test the detonation sensor and replace if needed.

The idea here is to eliminate all other possible problems that may be causing
the detonation sensor to act up. Remember
some other problem might be causing the knock the sensor is detecting -- before
I tested/replaced the sensor I had eliminated all of these variables.
Normally, the detonation sensor will last longer than your car. In my case it
was obvious from a visual inspection that it could be part of the problem since
the mechanic had used an air wrench when he transplanted it to the new engine --
it had a damaged casing and twisted tight leads; but I still eliminated all
the variable first (I did find the timing out of whack by 10 deg). BTW, I only
used 15-20 ft-lbs to torque it, it only needs to stay on
the engine -- it does not hold anything up.

Subj: Knock Sensor
Date: 97-01-31 12:38:56 EST
From: xwing@execpc.com (Jack Tartaniun) <sp?)
To: RandyNet@ticnet.com

My understanding is that the computer wants to see SOME signal from the
knock sensor, so it knows it is not "soft mounted" to cheat it as you
have done. In my 3000GT I tried to do the same thing you did, soft mount
a new sensor not touching the block, but it was totally doggy like it
knew it wasn't getting ANY signal anymore, except a nonmounted sensor
being connected...

Jack T.

Subj: Timing [T]
Date: 97-02-06 01:14:36 EST
From: jfossum@prostar.com (Jim Fossum)
To: RandyNet@ticnet.com

>...Is the ECU capable of
>advancing the timing to make full use of 105+ octane fuel if that is what is
>fed to it? Some say these ECUs have a programmed "ceiling" of electronicly
>controlled performance based on =<95 octane fuel, because that is all that is
>commonly available. The performance based mindsets say that the ECU can
>advance the timing and other outputs to make full use of 105+ octane fuel. I
>tend to agree with this. So many cars are heavily modified with the stock
>ECU, that obviously the ECU's electronic "ceiling" is pretty high. I'm
>trying to learn all I can, so send me your thoughts if you have any!

There are two things to consider here. The first is the engines 'optimum'
timing curve. Any engine (turbo or NA) has a certain optimum timing that
varies with load, rpm etc. Advancing timing beyond this point will actually
start to decrease performance (and eventually damage the engine). This
optimum timing is essentially independant of fuel octane.

The second consideration is the knock limit. This is the maximum amount of
timing before the engine will begin to experience autoignition (or 'knock').
This is strongly dependant on fuel octane and boost levels. A turbo engine
is almost always limited by the knock limit. As boost increases, the knock
limit decreases (allows less advance), but the additional fuel and air
more than make up for power lost to less than optimal timing (up to
a point!). For this reason, boost levels are used which compromise timing.
This is why our engines run so much better on premium gas.

Back to the heart of your question, the ECU would certainly never increase
timing beyond the optimum level and is more than likely limited well on
the conservative side. As fuel octane is increased, the detonation limit
will be increased which will allow the ECU to increase spark advance closer
to its programmed limits. The benefits of this are greater at higher boost
levels since the engine is operating further from optimum timing (due to
knock limits). Therefore there is a maximum octane level that will benefit
the engine at stock boost, but this level increases at higher boost.

>The question is, how far will the ECU try to advance the timing? and
>will it advance >past the capabilities of 93-94 octane fuel?

I have heard that our cars can benefit from up to 100 octane fuel.
I assumed this refered to stock boost levels, since obviously they could
benefit from more at higher boost levels. This was quoted second or third
hand, but was originally attributed to Archer Racing. This would indicate
to me that the ECU would advance past the knock limit of 93-94 octane fuel
with a 'deaf' knock sensor (depending on temp, etc).

Jim Fossum

Subj: Knock Sensor torque.
Date: 97-02-06 11:45:17 EST
From: markp@Exabyte.COM (Mark Pilon)
To: RandyNet@ticnet.com

I wouldn't downplay the effect of torque on our knock sensors, even
if they're metal to metal. I had my mechanic loosen, and then
re-tighten my knock sensor. it took quite alot of muscle to break
it loose -- way over the spec -- and he tightened it just enough
so that it'd stay put. It's snowing today, so it may be a few
days before I get a chance to wring the car out. I'll let you know.

Mark

Subj: lack performance
Date: 97-02-07 10:17:49 EST
From: TalonTsi8@aol.com (Robert Sherman)
To: RandyNet@ticnet.com

I have a 91 awd tsi. It seems that I have the same problem you are
having. Mine has 88k on it. I do not know as much as you about these cars
as you, but i do know that my 91 is not performing to way it should!! I
have driven a 91 laser turbo, and 92 esclipse turbo, and both noticeably
out performed mine. I was told it might be bad gas, I was only running
87 octane at the time :( they told me that bad gas could make the knock
sensor kick in and that after using better gas in about 200 miles the
knock sensor would quit. no luck. I was also told that my turbo could
be going out. I took it to the dealer, told him my problem, he test drove
it and said the turbo was working ok. Dealer told me that the turbo works
or does not. He did not charge me anything, and he is a dealer so why
would he lie? and say your car is ok!!! I dont know what to do. I want
to start putting high performace modification's on it, but i dont want to
spend all that money and somthing really be wrong with it. I think that I
am going to force the dealer to make a complete check on it.

If you could mail me back with any suggestions i would
greatly appreciate it!!! I NEED HELP!!!!!!!!!!!

THANKS , Robert Sherman
TalonTsi 8@aol.com

Subj: Re: All DSM Turbos - Detonation Sensor problems
Date: 97-02-08 18:12:48 EST
From: paulk@chromatic.com (Paul Kalapathy)
To: RandyNet@ticnet.com

One thing that I've had ongoing problems with on my '90 GSX is that
if the oil level in the engine even gets down to the top notch
on the dipstick I get valve chatter at high RPMs. The knock
sensor seems to take this to be detonation and the timing bogs.
I have to fill my dipstick to about 2-3 mm above the full notch.
Hope this helps.
-Paul Kalapathy

Subj: Timing/underhood heat/LS
Date: Mon, 10 Feb 1997 16:15:42 -0800
From: luxjo@thecore.com (James Oxley)

>Is the ECU capable of advancing the timing to make full use of 105+ octane
>fuel if that is what is fed to it?

From my testing with PMS datalogging, the amount of octane required is
highly dependant on outside/air cleaner temps. In the winter, I can run
20 PSI boost all day long and never see any reduction in timing from the
knock sensor. I see 24-28 degrees total advance (this number includes 8
degree mechanical advance) near 6K RPM and under full boost. These
numbers are what the factory ECU puts out before any modification by PMS.
I currently have not modded factory timing with PMS. Intake air temps at
the throttle body generally follow air cleaner air temps. Winter (30-40
degree ambient temps) throttle body and air cleaner temps just crack 100
degrees after much beating at 20-21 PSI boost with 16G turbo. No knocking
and resultant retarded timing is present in the winter using any 92+
octane gas.

In the summer, I have tried 92, 93, 94, and 100 octane unleaded. At
80-90 degree ambient temps, even the 100 octane resulted in knocking and
the resultant 9 degrees (always retards 9 degrees for me) retarded
timing. Reducing boost to 16 PSI did not help at all. Throttle body temps
still followed air cleaner temps and would hit 175 F after a long full
boost run at 16-20 PSI boost. Moving air cleaner down in front of trans
helped air cleaner temps by about 10-30 degrees. The only thing that
helped reduce knocking and retarded timing in the summer was 115 octane
leaded race gas.

So in that sense, yes the knock sensor/ECU will take advantage of the
octane in the summer months, but the most timing I have seen from the ECU
under full boost is 29 degrees total advance, even with 115 octane gas.

Subj: Re: Oversensitive knock sensor. / oil pressure
Date: 97-02-14 09:10:25 EST
From: markp@Exabyte.COM

I have been overfilling my oil to the tune of 1/2 quart or so and
it does seem to help. bottom line here is that any noise from the
lifters will be heard as knock. I finally replaced mine and the
car is like (better) new.

btw, I had more valve noise w/ mobil 1 than w/ conventional oil.

I've now settled on durablend (valvoline's semi-synthetic).
come summer I may switch back to mobil 1 15-50 as I now have
new lifters.

If any of you have DSP experience, you may want to take note of
the frequency spectrum of valve noise, vs. that of knock and
consider a filter for the valve noise (or all noise that isn't
knock . . .) there's a filter in the ECU that preprocesses the
output of the knock sensor. another approach may be to apply
some type of insulation between the knock sensor and block
but I don't have a clue how to select such materials.

Mark

Subj: Knock and the BCS
Date: Mon, 24 Feb 1997 10:32:19 -0600
From: Robert.Arrowood@sullivan.sullivangraphics.com (Robert Arrowood)
Message-ID: <#33>

Caution: Long post

Randy Garner writes:

>With a 2.5" cat back exhaust, upgraded B.O.V., and K&N filter, my car would
>peak at 13 psi and then drop to 10 or 11. After removing the restrictor, my
>car will peak at over 20 psi (only for a split second), and then hold at 15
>to 16 psi until about 5,000 rpm, and then slowly decline to about 10 or 11
>psi at the redline. That decline is due to the stock turbo.

I am not contradicting what Randy says here. We all know that
the Garrett is smaller and will not flow as much as the stock
turbo on the first generation. But, let's look at possibly
another reason as to why your boost is falling off as the engine
approaches redline. As you rpm increases your engine WILL start
knocking. This knock will get worse as you approach redline.
This is dependent on the grade of octane you are using. I
always use 93, but I still get knock usually starting around 5k
rpm. You indicated you removed the restrictor in the BCS.
This, as you have found, does increase your boost by about 4psi.
When you get knock the ECU sends a ground signal to the BCS.
This signal is a fluctuating ground but will, if knock is severe
enough, remain on. When this signal is occurring the BCS stops
bleeding air. When it is fully closed (complete ground signal
from ecu) then you will effectively loose 5psi of boost. When
the ground is fluctuating, the boost will fall off steadily
until the BCS remains on. I think this is what is happening to
your boost as you approach redline. Bypass the BCS completely
to check this. Just unplug the connector or remove the BCS
completely and you will not loose that precious 5psi. Better
yet, get an LED and hook it up to indicate when you are getting
knock.

OK....Sorry that was long. Since we are on the subject of BCS
and knock, I have an interesting problem. I pushed the car
fairly hard the other day, up to about 115mph at full throttle
starting from 3rd gear. As I approached redline my Knock LED
started flashing and I shifted. It never came on in 5th gear.
Slowed down for a stop sign then nailed it after stopping.
Nothing too hard but hard enough for a good rush. Anyway, again
the LED did not come on steady, just flashing. Got to my
destination and allowed the car to cool as normal. Finally the
car shutdown. The next day, I start the car and the LED is on
solid. I think great - 100 miles or more before the ECU will
let go. I 'baby' the car for several miles. I was in Eastern
Kentucky and was approaching a mountain and started to
accelerate up the hill. I then noticed as my boost approached
10psi the LED started flashing and finally went out. Total
miles - around 20-30. My question here is - Why did this
happen? It has happened before as Kevan Riley can attest. I
thought I had a severe problem somewhere. This is the first
time it has happened in about 1000 miles. I thought the ECU
would not remember a knock scenario unless it was severe?
Everything was fine when the car shut down but when restarted -
the LED was on causing me to loose 5psi of boost. I wonder if
the timing was retarded also? Anyone have any suggestions as to
why this happened?

One more question: Looking at the timing marks on the 1990
model, there are three large marks with the one on the bottom
possibly marked as 10 degrees. I would really like to know what
the marks are. Anyone know? The service manual says to set
base timing to +5degrees. OK - easy enough. But the marks are
not labeled in the book so I am not sure my timing is set
correctly.. It is set for the middle mark - above the one that I
think is labeled with a number 10. Where is +5degrees?

Sorry this post was so long.

TIA
Robert Arrowood
1990 Talon AWD - XCLRATN

Subj: Detonation
Date: Mon, 24 Feb 1997 22:04:12 -0600
From: Smokin@csinet.net (Dave Flaherty)
Message-ID: <#10>

>Ive tried not to say anything but it is time somebody did.. A lot of you
>guys dont understand detonation, or how serious it is!!

I'll try and clear this up

>It means that your Air/fuel
>mixture has gotten sooooo lean, that temps have reached VERY dangerous (to
>your engine) levels. . I know mine would detonate
>at ANYTHING above 15PSI before I got my fuel pump.

Not necessarily, read on....

>eventually, I melted 8 valves, and managed to melt a hole in a piston..

I would think that you would fix what was causing the car to knock before
it got to this point, or at least turn down the boost

>Dont trust the Air/Fuel Ratio meter, those things arent reliable enough to
>trust with your engine..My Cyberdyne read FULL RICH at 18PSI when my car was
>detonating like crazy!!

They are plenty reliable, remember that a lean air/fuel ratio isn't the
only thing that will cause detonation, read below

Sounds like your knock sensor wasn't working or you were running 87 octane
or some other low octane fuel.

>If you want to run that high boost, go get a fuel pump
>upgrade, with that your are easily good to about 20PSI, 200 bux for a pump is
>much cheaper than 3k for a new motor.. Thanks for the time, and sorry if I
>stepped on any toes, just trying to wakeup some ofthe newer members..

Read below regarding fuel pumps, detonation, etc.

Hope I'm not stepping on yours.

Rusty: This is not a slam or anything but I think you have been misled. I
am sure that your post will generate a lot of response so let me add my two
cents.

The main reason for detonation is low octane levels in the fuel (poor fuel
quality), and/or lean air/fuel mixtures, assuming that your engine is timed
and tuned correctly. Detonation is when the air/fuel mixture ignites
before it is supposed to. This is kind of hard to explain succinctly but
I'll try. Diesel engines detonate. Diesel fuel mixes with air and is
heated by compression until it explodes. This is good in a Diesel but bad
in a gasoline engine. What happens in your engine when detonation occurs
is the following: Air and fuel enter the engine. They are compressed as
the piston approaches top dead center (TDC). At a certain point before top
dead center (BTDC) the spark plug arcs and ignites the mixture. When this
process goes wrong your engine may "detonate". This means that due to a
number of factors, the main one being low octane fuel and lean air/fuel
ratio, the air/fuel mixture ignites without the presence of a spark from
the spark plug, instead the mixture is ignited by the heat of compression.
Detonation may also occur from having very hot intake air (like when you
boost to 18-20 psi or higher, depending of course on your set up),
excessive heat being stored in the head and block of the engine (three
reasons to get a colder thermostat), improper timing, etc. What happens in
our cars is that when you have a low octane fuel that is succeptible to
knock, and/or a lean air/fuel mixture the knock sensor will "hear" the
detonation and retard the timing (does it retard or just activate the boost
control solenoid or both?). When someone blows up their engine, or less
dramatically damages it (burnt valves, holed pistons), it is because
they A.) Boosted so high for a certain length of time that the damage was
done before fuel cut occurred or B.) Your knock sensor isn't working
properly.

The trick to understanding all of this is to understand how the engine
controls on our cars work together. First, the protection of fuel cut is
activated by a preset or "prelearned" air mass ratio, when you are boosting
so much air into the car that the fuel system can't keep up with the demand
the fuel cut occurs. The knock sensor, however, monitors the combustion
process. The ECU will tell the injectors how long to pulse by measuring
the air mass coming into the engine. What I'm not sure about, and I hope
someone can answer is: When the knock sensor senses detonation and retards
the timing, does it tell the ECU to put in more fuel (assuming that
detonation occurred due to a lean air/fuel mix), as well as activate the
boost control solenoid?

[As far as I can tell from the first gen code, this is not done, although
it is something I've been thinking about doing. I don't really know enough
about the science involved, though. -talon mgr]

Also, can a DSM with a TOTALLY dead battery be bump started? I tried it
(unsuccessfully) and my theory was that because the ECU didn't have any
juice the bump starting method didn't work.

[That is most like correct. If batt voltage is below 8V, the ECU will not
trust any of the sensors, and won't fire the injectors. Also, I've found
the ECUs to be a bit on the flaky side with batt voltages under 9V.
-talon mgr]

Thanks,

Dave Flaherty
92 GSX
irc SmokinDav

Subj: Detonation and boost...
Date: Tue, 25 Feb 1997 08:12:14 -0500 (EST)
From: Btalont@aol.com (Brian Hood)
Message-ID: <#17>

Rusty Chappell was on the right track with his disertation on boost and
detonation but did not quite hit the target. It is possible to have enough
fuel and STILL have detonation. People forget about the octane value of pump
gas.

You can reach a point that even with adequate fuel, even an overly rich
condition, that detonation will occur because the pump gas octane won't
support the high cylinder pressures/temperatures generated at high boost. I
do not recommend very high boost levels for very long on pump gas because you
WILL detonate. Worst of all, you won't necessarily hear it.

There are other factors that will affect detonation thresholds. One big one
is charge air temperature. This is affected by the ambient air temp,
intercooler efficiency and turbo compressor efficiency. A stock car with
stock IC and turbo will detonate sooner due to high charge air temperatures
than a car with a 16G and aftermarket IC at the same boost levels with the
same fuel. This is controlled to some degree by a richer than necessary fuel
ratio to help cool the charge but this only helps so much and doesn't make
the most power.

There are many more things and I could go on and on but my advice is to be
careful and understand what and how things work and how higher boost will
affect your individual car. Ask questions. There many knoweledgable people on
this list who would be glad to help and answer questions. I hope this saves
someones engine.

Brian Hood

Subj: Re: wakeup call
Date: Tue, 25 Feb 1997 09:36:43 -0800
From: luxjo@thecore.com (James Oxley)
Message-ID: <#20>

> It means that your Air/fuel
>mixture has gotten sooooo lean, that temps have reached VERY dangerous
>(to your engine) levels..

Possibly, but not necessarily. During datalogging, I could see
detonation (or actually retarded timing due to ECU reading knock sensor)
without EGT's going through the roof or even changing at all. It was
much more dependant on intake charge temp and use of race gas than lean
running.

> You could EASILY detonate 20 times a day, and not >have
>any clue that you are doing.. A sign you might be, is that you are
>running over 15PSI of boost without a fuel pump upgrade.. Some say that 16PSI is
>safe with the stock fuel system, this is kinda misleading, it MIGHT be safe,
>and it might not.. it is kinda dependent on your car.. I know mine would
>detonate at ANYTHING above 15PSI before I got my fuel pump..

Why 15 PSI, just cause it was your crossover point. I do agree that a
knock sensor LED, Fuel pressure guage and A/F guage (or even better, a
real good volt meter tapped into O2 sensor) should all be used before
raising boost above stock.

>If you are one of these guys running 18PSI, unless you like buying new engines, you need to turn
>it back down.. I was a boost junkie once too, I had my Eclipse turbo hittin
>17PSI on the stock fuel sytem, even hit 18 or 20 at the track once or
>twice, and eventually, I melted 8 valves, and managed to melt a hole in a
>piston..

The stock fuel system is not that bad. I did not loose FP or have lean
A/F when my car was virtually stock running higher boost. If you run long
stretches at higher boost and/or don't have all the things to monitor
your A/F, then the FP should be mandatory. I do agree that the upgraded
FP should be near the top of the list for upgrdes.


>Dont trust the Air/Fuel Ratio meter, those things arent reliable enough
>to trust with your engine..My Cyberdyne read FULL RICH at 18PSI when my car
>was detonating like crazy!!

I don't agree, I've found the A/F guages to be quite helpful. The
upper most red LED corresponds to about .89-.9 volts. I checked this
against 2 different O2 sensors, both new. If you light up the second red,
you should be fine from a fuel standpoint. I've also found the O2 sensor
to be very linear. It may not correspond exactly to a certain A/F ratio,
but small changes in fuel addition will yield a small change in O2
voltage. The question remains, what is a safe low value for O2. Some say
..78, some say .85. If your running .9 or above your OK. Sounds like your
detination problem was due to air intake charge temp or lack of race gas
or both, not a lean condition. I honestly beleive that high effeciency
intercoolers are worth way more than most people realize. They are highly
underated and probably worth much more than a bigger turbo during street
blasts in high temp climates/summer.

>Use your common sense.. If you run over 16PSI,
>chances are, you are doing damage.. It may take a year for it to even
>effect the performance of your car, but once your engine starts to go downhill,
>it goes downhill FAST.. If you want to run that high boost, go get a fuel
>pump upgrade, with that your are easily good to about 20PSI, 200 bux for a
>pump is >much cheaper than 3k for a new motor..

Again I agree that the FP upgrade should be near the first step, but
the equipment to monitor various things going on in the engine should
come before even that. You can't make blanket statements about when
someone needs the FP upgrade. You also can't tell me that 20 PSI is fine
as long as you have the pump. I've proven thats not always the case. Get
as many guages as you fit in your car. Thats the only way to see what is
going on in YOUR car. This should be your first step.

OX

Subj: Re: Knock and the BCS
Date: Tue, 25 Feb 1997 09:51:47 -0500
From: rgarner@csra.net (Randy Garner)
Message-ID: <#22>

Robert Arrowood wrote:
>This, as you have found, does increase your boost by about 4psi.
>When you get knock the ECU sends a ground signal to the BCS.
>This signal is a fluctuating ground but will, if knock is severe
>enough, remain on. When this signal is occurring the BCS stops
>bleeding air. When it is fully closed (complete ground signal
>from ecu) then you will effectively loose 5psi of boost. When
>the ground is fluctuating, the boost will fall off steadily
>until the BCS remains on. I think this is what is happening to
>your boost as you approach redline.

Robert,

That's a good point, and I thought the same thing when I first noticed it.
It makes perfect sense too since 10 psi is where the waste gate actuator
will open the waste gate if there is no bleeding air. But after some more
testing I've found that even without the solenoid the boost will still fall
at the redline. The Garrett turbo in stock form just can't keep up.

This brings me to a question: Who has the fastest quarter mile time with a
2nd gen running a STOCK turbo? I'm just curious..... because if it's not
very fast, I'm going to quit wasting my time trying to improve the
performance of it, and go buy a bigger one. I'd like to get low 13's out of
it; High 12's even better, but that's probably not going to happen without
a major upgrade.

Thanks,
Randy Garner

Subj: Running rich with new fuel pump
Date: Tue, 25 Feb 1997 11:47:13 -0800 (PST)
From: wigglit@earthlink.net (Ron Sinclair)
Message-ID: <#30>

Hi guys!

Next topic: Upgraded Fuel Pump

I received my new pump yesterday and immediately installed it. Took me 2
hrs......that banjo bolt was being an anus! Vise grips didnt help....I
used 2 open end wrenches with long handles, for leverage. Then I had
problems with removing that little screw on the bottom of the pump
bracket......it was too tight and I ended up stripping it. Had to use vises
on it (should've done this before I stripped it).

After that, the install went smoothly.....everything went in fine (had
problems with putting the pump bracket back into the gas tank. The tank
level kept getting caught on something inside the tank.).

Now comes the fun part.....I go out and buy a multimeter to measure the O2
sensor voltage (wish I would've measured the voltage before the install).
After finally finding the open white connector under the dash on the
passenger side, I couldn't seem to get the alligator clip into the
connector.....so I settled for a half-assed connect. I thought I had
connected it wrong at first cuz the LED on the meter was going bonkers at idle.

Then i get to the street. I have the meter set to the 20 range (I lack a 10
range). I nail it at the light and look at the reading: 1.0! When I
measure at speed in the higher gears, I get .98-1.0. When I coast, it goes
to 0.0 and when I idle, it's all batty. Duz that sound right?

[Sounds about right, except it should also go bonkers when cruising
around. -talon mgr]

Well, I'm running WAY rich now and need to scale back the richness. I went
to work today and its only 10 miles to the workplace and I actually saw the
needle on the fuel gauge moving! I burned a quarter tank from the time of
the install (last nite about 7:30) to a trip to and from work!

Do I need to remove the honeycomb on the MAS now and play with the settings
on the MAS? Is there a page out there on how to set the MAS settings (a
usable page??)? I can't really scan now, as I'm currently offline. Are
there some good guidelines on fooling with the MAS? I shouldn't have any
probs arising with the MAS being adjusted since I have the upgraded pump, no?

I need to cut the richness down......I'm feeling like I still own the 'Stang!

Ron Sinclair
aka WIGGLIT

Subj: Re: Detonation
Date: Wed, 26 Feb 1997 09:18:28 -0500
From: scalbert@flwsoutheast.com (Steve Calbert)
Message-ID: <#18>

>>Dont trust the Air/Fuel Ratio meter, those things arent reliable enough to
>>trust with your engine..My Cyberdyne read FULL RICH at 18PSI when my car was
>>detonating like crazy!!

>They are very reliable.

The typical O2 sensor, be it a single or three wire unit, are only
ccurate at very close to stoich. Anything outside of around 14.7:1 the
sensors are not repeatable from unit to unit.

You can get a rough idea what's good for a particular vehicle, but it may
not translate to a different vehicle or if the sensor is changed.

Just wanted to add a bit into this,
Steve

Subj: [All]Detonation and Preignition are not the same!
Date: Wed, 26 Feb 1997 10:00:46 -0500 (EST)
From: CW31@LNS62.LNS.CORNELL.EDU (Curt Ward)
Message-ID: <#22>

I feel I must dispel a common myth: Detonation and preignition are not the
same!

Preignition occurs in a spark ignition engine, when the mixture ignites
before the spark occurs due to the charge temperature exceeding the flash
point of the mixture. This is the normal mode of operation for diesels
(compression ignition engines). Note that the flame front can proceed
subsonically in a preignition event, i.e. the flame is a deflagration wave
(normal combustion). Preignition is usually rare in a spark ignition engine,
although it was common on old, caburetted engines that had run rich for many
miles - the carbon deposits in the combustion chamber created hotspots,
allowing preignition to occur after the ignition was shut off.

A detonation occurs when the flame front is traveling faster than the sound
speed. It occurs in a spark ignition engine at the _end_ of a normal combustion
event when the pressure and temperature of the charge are highest.
In a knocking event the mixture is ignited by the spark, the flame front
travels normally until the pressure/temperature exceeds the Chapman-Jouget
point at which time the flame front accelerates beyond the sonic velocity. The
extreme peak pressures and temperatures of the detonation wave are what causes
damage to pistons, valves, etc. (and cylinders if you have a honda).
The onset of detonation is primarily related to combustion chemistry (this is
why the octane rating of gasoline is so important) and the rate of pressure
increase during combustion. Lean mixtures have higher dP/dt than rich ones,
hence their tendancy to have reduced knock.

Curt Ward

Subj: [All] Knock Sensor vs Boost Control Solenoid.
Date: Thu, 6 Mar 1997 00:04:18 -0500 (EST)
From: RandyNet@ticnet.com
Message-ID: <#12>

I have an LED hooked to my boost control sylonoid (BCS). RTP to me if you
want to know how to hook one up. The wires provide a signal whether you are
using your stock BCS or not.

Now, I am 95% sure that the ECU retards the ignition timing before it signals
the BCS to close and reduce boost flow. I'm interested in any opinions!

[As I said before, the ECU will open the BC Solenoid when it has had
to retard timing beyond a certain point. -talon mgr]

Subj: Solution found!!
Date: Sun, 23 Mar 1997
From: RandyNet@ticnet.com

It seems like there are dozens of differant cures for this, none of
which have worked for me yet. However, over the weekend, I think I might
have figured out my problem, so I thought I'd pass it on to you.

First, after 90,000 miles I replaced my fuel filter. I thought maybe I
was having a fuel delivery problem that was causing legitimate knock.
I blew through the new one and the old one, and they seemed to flow air
about the same (the "skin of my teeth test", not officially licensed yet!).
This theoretically could be causing problems for the rest of you.

Now here is the good part. I replaced my thermostat. Recently, the gauge
had been running a bit high. Nothing I was too worried about, it maintained
a temperature, but it was higher than normal. I took my old thermostat and
my new one and put them in a pot of water on the stove. The new one fully
opened at about 190 F, but my old one only partially opened at about 210 F,
and never fully opened even sitting in a roaring boil. I installed the new
one, and my temp gauge is now just a hair on the cool side of strait up
middle of the gauge. Now suddenly the thing seems to be running better.

The moral of the story is that there seems to be a reletively undisputed
theory that the temperature the ECU sees directly affects the cars ignition
timing advance and fuel delivery ratios. Keep in mind there are actually 3
temp sensors in that thermostat housing. One for the temp gauge, One for
the ECU, and one that controls your radiator fan. My gauge was running
"a bit hot", after my stove test probably about 200 F, and the ECU was
reading from a differant temp sensor, but was reading the same body of
water, so it probably thought I was running hot.

Now I don't know what the thing does when it thinks your hot, but my
retrospective theory is that it causes bad performance. I am running cool
now, and the thing runs much better.

More theory: A hot engine is much more likely to have knock. I know it
was running hot, so I'm thinking it really was knocking. In my previous
theories I was trying to tie the problem back to some kind of mechanical
noise such as noisy lifters.

Subj: Re: Still got knock sensor trouble?
Date: 97-03-24 09:07:24 EST
From: markp@Exabyte.COM

I replaced my lifters and am amazed that I lived w/ the car for as
long as I did -- that cured my knock/hesitation ! any mechanical
knock, clank, whack (...) on/in the engine will likely be heard by
the knock sensor. the other thing I did was loosen and retighten
the knock sensor -- according to my mechanic, the spec is for a very
low torque figure. he put it in just tight enough so it would stay
put.

Subj: Knock sensor
Date: 97-03-25 13:58:47 EST
From: Robert.Arrowood@sullivan.sullivangraphics.com (Robert Arrowood)
To: RandyNet@ticnet.com

Randy,

Could you send me the related info you were talking about for
the knock sensor. I have not had my main problem occur in
awhile, knock on wood. What it has done several times is set a
knock condition in that the ECU will shut the BCS down
completely, Knock LED is on at this point, not flashing. I
assume at this point the ECU is also retarding ignition. It
seems that if you put the car in a knock situation it will
correct this problem, although I have not tested this fully
since it has not happened in awhile. Currently though, like
others, I am getting knock, LED is flickering, starting slightly
above 4500rpm. This seems relatively low to me. I am running
about 18psi w/ DaveB fuel pump, stock turbo, etc., etc. I think
this may be a heat problem within the cylinders. I am currently
using the LED as a shift light. I will let if flicker for about
three seconds then shift. I made a WOT run the other day on the
interstate in fifth gear. At about 5k rpm @ about 120mph the
LED started flickering. My EGT at this point was 1200degrees,
probe mounted after flange in the downpipe.

Any ideas? Normal?

Robert Arrowood
1990 Talon AWD - XCLRATN

Please send your suggestions and comments to Damian_Sigman@homemail.com
My Home Page

Page created April 20, 1997 - Damian Sigman