Knock sum values & Ignition advance question w/ Datalogger

[turbo21psi]

I posted a similar question in the General section and I didn't get much answer so I'll try to post it here.

On the TMO datalogger, the knock sum values range from 0 to 43. I understand that it doesn't show actual detonation but it listens for noise that may be detonation. What is the highest value should I allow when tuning?
I know that 0 knock would be the safest but when I tune for no knock, I would have to run on the rich side and the car runs pretty slow.

I've also heard that 15degrees ignition advance would be the optimum point and anything more and you can afford to run more boost and anything less degrades performance.
My logs show that with knock sum as high as 20, my timing advance is still at 17 deg. The only way my timing gets retarded past 10 deg. is when the knock values are above 30.

Also, on 2nd Gen dataloggers, I've heard that they don't show the knock sums and tuning would have to be based on timing advance alone. So, would it be safe to ignore the knock sum values as long as the timing is not being pulled?

Please give me some ideas as I want my car to run at the edge without going over and blowing it up. Thanks

[Quadcylla]

I tune so that I get no knock... or only the occasional blip. I like to get the most advance I can and plus I bumped my base timing up 3 extra degrees so keeping knock just at 0 (on the edge) gets me plenty of power.

I run 15psi on the street (stock IC). At the track I put race gas in, turn up the boost and dial down the S-AFC till I find that 0 knock boarder again.

[dyezak]

Knock sensors from the factory arent very accurate. Try to think of your knock sensor like your O2 sensor. A good tuning AID...but we all use EGT's right, because they're better than our factory O2 sensor. When tuning look at the overall picture, knock sums, ignition advace/retard yada yada yada. A knock sensor is nothing more than a tuned mic, pics up certian freqs, therefore it pics up some things that aren't knock. If your knock sums are WAY high...and ignition is being retarded...then you know you need to back off (or get a new knock sensor, they wear out). Each car is different (as far as sum values go) so you need to tune your car to the point of not pulling ignition. You got $60 extra? Get a new knock sensor, cant hurt, you may be gettin false readings. Don't ignore A/F ratios, and EGT's also.

[nukefission]

I prefer more timing advance than boost. You can run umpteen lbs. of boost but if it results in yanked timing, the car will feel flat up top. It's very noticeable and it's no fun. You can end up running slower if there's too much boost. 15 degrees of advance is ok, but somewhere around 20-25 would be favorable.

One effective way of eliminating knock is to install water injection. This will lower your charge temps A LOT, and allow you to run as much boost on the street as you would on the track. A local club member with W/I runs 19-20 psi on pump gas on the street with little to no knock at all. This is on a Big 16G with supporting mods. I don't remember what his timing advance figures were but they were pretty high. Lots of timing + lots of boost + safe charge-air temps = GOOD!

[dsm4eva]

Running the highest amount of boost without negatively affecting the timing advance curves is the best way to tune for highest output from the engine. Both advance and boost add power, too much of either will cause detonation/retarding. The scientific approach is to note the air temp and barometric pressure as these can change how your system responds. Then continually up the boost until you find the timing being pulled back due to knock and then drop back to the last boost level you ran at and call it good as that combo will give you more power than the setting you're at now.

[FastGSXauto]

I also asked about this in the general section but no one seemed to have any ideas.

My Car: 93 AWD AT
To those that know anything about the ECU code.. What factors determine advance at WOT? I'm using the tunerstein datalogger and i have 0 knock count but only 10-11 timing advance throughout the rpm range. And when cruising at part throttle, I only see about 15 (17 max) advance where most people say they get over 25. I'm thinking one of my sensors are bad and causing the ecu to not advance. So does anyone know which factors come into play with timing advance?

[Quadcylla]

Durring a wot run the car looks at air-mass (air-flow compensated for density [temp & baro]) and rpm. It takes those values and runs to it's timing tables and comes back with an advance to use. Knock sum and the octane value (similar to knock sum but over longer period of time) are then monitored and when need be the ECU will index down on it's timing table to retard the timing to help control the knock. Once the knock is gone it will index back up to the original timing table output or until it see enough knock again.

10 degrees is very low for a wot run... does the knock sum work on that logger?? I get 22~23 degrees when I tune to the no knock point.

[Edited by Quadcylla on 06-13-2001 at 02:09 PM]

[Jehu]

This is off the digest archives, and should give some ideas as to what affects timing.

MMC uses Mass Air Flow type of system with a Karman Vortex type of Air Flow
Sensor on these Systems. Listed below are the sensors in order of importance
and what effect they will have (fuel and/or timing).

MMC SENSOR ORDER OF IMPORTANCE:

1. CRANK ANGLE SENSOR, engine CANNOT run without this sensor. First, this sensor
is used for an RPM input. With that information, the ECU will determine Ignition
Coil Dwell & Timing. Second, the ECU will use this sensor to start the Injector
firing. On the DOHC engines, the Crank Angle Sensor along with the TDC Sensor
will determine which Coil to fire and when.

2. AIR FLOW SENSOR, this sensor is the load sensor of this system. The Air Flow
Sensor will "count" the air going into the engine. The ECU will then add fuel
based on the Air Flow Sensor's output (Hertz). Higher Air Flow Hertz will result
in more fuel and lower Air Flow Hertz will result in less fuel. So in that way,
the Air Flow Sensor affects the major amount of fuel (Base Pulsewidth).

It is important to note that the Air Flow Sensor's output also has a lot to do
with the Ignition Timing. If the load is high (a lot of Air Flow Hertz) the ECU
will give the engine less advance than if there is a light load on the engine.
Load is based on Air Flow, TPS (Throttle Position Sensor) & RPM.

3. TPS (Throttle Position Sensor), can increase fuel by 500% (quickly increasing
TPS voltage) or decrease fuel by 70% (during decel). The TPS will lower the timing
advance as its voltage is going up. If the TPS voltage is steady, it will have no
effect on fuel.

4. COOLANT SENSOR, can increase fuel by 60% (at -22 F) and have no effect on fuel
(at l76 F). The Coolant Sensor can increase the timing (by 140 at -22 F) and will
have no effect on timing at 95 F.

5. BAROMETRIC PRESSURE SENSOR, can increase fuel by 12 1/2% (below Sea Level) or
decrease fuel by 50% (above 8,000 Feet Altitude). The Baro Sensor can also add 70
of timing (above 7,000 Feet Altitude) and will have no effect on timing below
1500 Feet Altitude.

6. INTAKE AIR TEMPERATURE SENSOR, can increase fuel by 23% (at -26 F Air Temperature)
or decrease fuel by 17% (at 185~F Air Temperature). This sensor will have no effect
on fuel at 730F Air Temperature. This sensor has no effect on timing.

7. OXYGEN SENSOR (02 Sensor), can increase fuel by 17 % (lean condition/too much
oxygen) or decrease fuel by 17 % (rich condition/lack of oxygen). This sensor has
no effect on timing. This sensors output will only be used when the engine is
in Closed Loop.

8. ADAPTIVE MEMORY, comes from the Oxygen Sensor switching activity over a period
of time. This value can add 5% fuel or take away 5% fuel. This value is always used
in the Pulsewidth (Fuel) Formula. It is only updated when the engine is in Closed
Loop. When the ECU's Memory is cleared, this value is set to 0% until the engine
goes into Closed Loop, then it will change based on the Oxygen Sensor's output.

[The code agrees with most of this except the adaptive memory. It seems to
me that it can kick out about +/- 20% on the DSMs. -talon mgr]

MMC INJECTOR PULSE WIDTH FORMULA

Base Pulsewidth =

Air Flow Hz/Engine Speed x TPS (decel) X Coolant X Baro X Intake Air Temp X O2
(Closed Loop) X Air Fuel Compensator (Open Loop) X Adaptive Memory X Detonation
Sensor (Turbo)

Total Pulsewidth = Base Pulsewidth X TPS (accel) + injector Lag Time

NOTE: The TPS Does Affect the Base Pulsewidth a Little Bit, But Not a Noticeable
Amount (checking With an Oscilloscope).

[Mainly, the PW gets affected by the *transient* TPS response. -talon mgr]

Brian Spray Chrysler Editor
spray@alldata.com Alldata Corporation

[FastGSXauto]

Quadcylla, yeah the knock count defenitely works becaues I had knock before but i lowered boost and played with fuel a little to reach no knock. I also reset the ecu so that put the long term octane value at max timing.

Thanks for the long post Jehu.. I saw that one before, but is it comprehensive? The only sensor in that list that can have a negative effect on timing is the air flow sensor. But I don't think i pushing enough air to make the ecu only give me 10 degrees at WOT. Also, at part throttle or crusing shouldn't my timing jump way up in the 20's or 30's to help gas mileage? I rarely get above 15 while cruising and i think that's why my mileage sucks. Something is defenitely missing here.

[Jehu]

Yeah, I'm not exactly sure what's going on there. Typically timing goes to the max. value during coasting. What is your timing when you coast,i.e. accelerate up to speed then let off gas and let the car coast. Should be 30+. Do a search on the digest and see if anyone else has had low timing with low knock.

[FastGSXauto]

Yeah my coasting value for timing is about 15-17. Should be way higher.

[Quadcylla]

Sounds like you are getting exactly 1/2 of the timing you should be. Maybe that's a clue to the problem. Are there settings so that that logger can be used on different cars? If so go over your settings and make sure they are right. Also does it just plug into the diagnostic port or did you have to hardwire any connections. If you did have to hardwire some connections list them here and check that they are right.

[FastGSXauto]

Well what is normal timing at idle? Mine says between 5-9 usually which I think is normal. This logger is specifically made for this car, and it's plugged in right, or else it wouldn't get any data. I think it's something wrong with my car.

[FastGSXauto]

Jehu, you say a low coolant temp can increase timing? I thought when the engine is below normal op temperature it actually retards timing.

[Jehu]

Like I had said, that was a post on the digest from Brian Spray Chrysler Editor at alldata corp. Not being privy to the ECU codes I'm not sure if that is correct or not. But since Todd didn't offer any corrections I assume that he didn't find anything wrong with that statement.

My idle timing typically jumps from 7 to 12. Cruising is around 26, coasting is at 33.5. WOT is usually at 22 degrees by 6500 rpm. Just more data points for you to consider. This is on a 2G by the way. 1Gs should dial in a few more degrees of timing at WOT.

Which datalogger are you using by the way? Also, did you inspect the base timing of the car? That might have been changed.

[FastGSXauto]

I'm using the tunersteing datalogger, but I verified with the maker that it's the same as TMO's. I have my base timing set at 5. Your data points seem pretty consistent with what is normal.. mine are defenitely low.

[blunttwister]

well i was logging last night and today. i did a 1st gear rollat about 10-15 through 3rd to about 80mph. at only 14-15psi. ambient temp was about 80 and this was at night, f'n phoenix.

1st gear 3500rpm-21.5degrees-.96//4800-12.5-1.02/
2nd gear 4700-14-1.00//5200-13.5-1.02//5800-17.5-1.02//6700-18.5-1.02
3rd gear 4800-18.5-.98//5000-13.5-1.00//5500-15.5-1.00//

why would my timing get better/higher with rising rpm's.

is this normal to be running so pig rich, .98-1.02, would this mean that i probably have a boost leak?

timing-
coasting- 33.5
idle- 7-9

is there any way that you can make the datalogger(alex pepper 2g) record faster? besides just selecting less catagories?
thanks

[jimmoltz]

Your timing is supposed to advance more as RPM increases. Even under boost it should still increase with RPM.

[dsm4eva]

There is a timing table that the ECU will adjust to based on engine load and air flow values. As you approach the higher RPM ranges, the tables automatically start dialing back the timing advance.

PS. FastGSX, I just got back from vacation so I'll take a look at those logs tomorrow.

[jimmoltz]

Does anyone have the alexpepper logger??? If so what data/sensors does it give you? How long can you record real time data?

[turbo=good]

Originally posted by jimmoltz
Does anyone have the alexpepper logger??? If so what data/sensors does it give you? How long can you record real time data?

I do. it gives: load value, engine coolant temp, short term fuel trim(have not figure that one out),long term fuel trim, mph, rpm, spark advance, air intake temp,MAF air flow rate,tps,both o2 sen outputs.

"My idle timing typically jumps from 7 to 12. Cruising is around 26, coasting is at 33.5. WOT is usually at 22 degrees by 6500 rpm. Just more data points for you to consider. This is on a 2G by the way. 1Gs should dial in a few more degrees of timing at WOT." --JEHU

for my 2g, I'm at 4-8 at idle, cruising at about 26, WOT at about 17 from 4K to whenever I let off. It doesn't go up with the my rpms. My cars feels slow and flat at high rpms

[jimmoltz]

I haven't been able to figure out a problem I have been having for a long time. I hope that the datalogger will help me to nail it down!!! Mine usually runs +32 degrees cruising +22 coasting, and +20 WOT 5K and above.

The problem is sometimes it retards to -5 degrees at 10psi on up. Even more interesting is cruising at 15"vac then add a little throttle to say 10" vac it retards to +16 from +32 even going down hill!!! I have replaced everything but the MAP, and Crank angle sensors.

Nothing has helped, this only happens for about 5 minutes of driving then goes away. Same conditions and when adding the same throttle to 10" vac retards to +29 degrees.

I have a thread going in the help forum, but no one has been able to figure this out. So I hope you guys are a little more advanced and can help me out. This is a wide spread problem. Who ever figures this one out will be a DSM Hero!!!

Sorry for the long post.

[FastGSXauto]

imagine having that problem all the time.. i never get past 16 while cruising, and 11 at WOT. I just can't figure it out. I have little or to no knock sum.

[dsm4eva]

Nothing in your logs indicate anything wierd either. Do you have anyone nearby that you can borrow their Cam Angle Sensor?

[jimmoltz]

Hey guys I already replaced the cam angle sensor, ECU, knock, swapped MAF with a friends, TPS, checked fuel, BOV, checked and cleaned EGR, all hose connections. NADA!!!!

What ever it is the ECU thinks it is a normal event. All I can say is Damn it!!!