Ignition Tuning - bizarre results [Archive]

Chuck Dean

11-08-2005, 09:36 PM

Our Suzuki GSXR600 was tuned on the dyno this semester and yielded some bizarre results.

Typically, ignition advance should increase as RPM increases. (lets say 10 degrees advance at idle to 40 degrees at 11,000 rpm) This was not the case for us. The engine liked 10 degrees of advance at startup and idle. However, our timing DECREASES as RPM increases. At 10,000 RPM we have about 0 degrees of advance.

We did this based on the engines power output. If we tried advancing the ignition, power would continually fall off. Where as if we backed it off to 0, it would make peak torque. And the engine 'sounds' like it likes it there best. If we tried to retard the timing past that, we would get small "hiccups" out of the engine.

This goes against everything I've learned about tuning engines. Right now, it is my belief that something may be wrong internally with the engine. Perhaps it's not physically synced? The students who put together were 'first-time' engine builders.

We made...
60 SAE hp
32 ft-lbs.
...off of the output shaft (water-brake dyno)

Fuel injection timing is correct (verified by calculation and testing)

CRIP angle is similar to other teams.

Please provide input on this problem. I am thoroughly confused.

Chuck Dean

11-08-2005, 09:36 PM

Erich Ohlde

11-08-2005, 10:48 PM

What electronics package are you using? My first guess would be bad trigger wheel setup (software thinks there are less/more teeth than there actually are).

First thing I would check would be to verify timing with a timing light.

Chuck Dean

11-08-2005, 11:28 PM

...using Motec...I can pull the numbers we are using for #teeth and CRIP tomorrow. But I believe they are correct.

Something worth noting...

Although our CRIP angle is similar to what others are using on their Suzukis, we experienced some wierd results when double-checking the CRIP angle with a timing light. The timing line alternated from one position to the next. It was consistant (ie 30 degrees,40,30,40...) and happening when it was supposed to (TDC compression). We eventually got it to just one line but the gun indicated it was 43 degrees off. We believe the timing light's pickup was picking up too much noise.

Frank

11-09-2005, 01:13 AM

you have wired the coil drivers and/or coils incorrectly,

the (switching off / switching on) functionality is reversed

it is usually accompianed by burning hot (eventually will fail) coils

Erich Ohlde

11-09-2005, 07:05 AM

Frank, I'm not seeing how wiring the coil drivers backwards would work, let alone throw off ignition timing. Can you be more specific?

Igor

11-09-2005, 07:38 AM

It has to do with excited electrons and is not easily translated into its magic smoke analogy. http://fsae.com/groupee_common/emoticons/icon_wink.gif

Igor

Kirk Feldkamp

11-09-2005, 11:12 AM

I recall we had similar problems when we first started the engine. Make quadruple sure you have the right firing order as the Motec sees it. Are you still using the stock 8-tooth trigger wheel? If you are, have you confirmed your CRIP is correct-ish? How are you checking the CRIP?

Another thing to check is the various ingintion tables to see if someone set some goofy compensations that you forgot about. That happened to me when we went to a turbo... I had set the MAP compensation table to WAY retarded when we were still running NA. When we got the turbo motor on the dyno, we were only making 65hp. After banging my head on the computer for an hour or so, I found that problem. Once I fixed the table, the motor immediately woke up and make like 87hp the next run. Look at your datalogging to see if timing is being added/pulled for some reason other than what you set the main ignition table to.

-Kirk

Colin

11-09-2005, 12:07 PM

if your using reluctor pickup's and the timing light is showing an inconsistent spark chances are you have the pickup sensor wired around the wrong way. Try changing the crank pickup trigger setting in MOTEC from leading to trailing edge or vice versa. Are you using the GRIP setup screen to test your angle?, if so why is it showing 30 deg? Check where TDC on no. 1 is carefully with a dial gauge and mark it on the trigger wheel then use the test advance screen to set your GRIP, set the test advance to 0 deg, disconnect the FI so it wont run then spin the engine with the stator motor and adjust the GRIP accordingly

Alexandre D.

11-09-2005, 01:14 PM

Falling/rising edge may be your problem. Plug a scope with both ref and sync and analyze it. Motec has a nice app explaining how to sort it out. The ref/sync of the GSX are mag or hall sensors?
Like turbotwig said, check the datalogging. If it's not the right value, you probably have a compensation table somewhere messing up your values. If you can confirm it's not a table problem, it's a position problem.

I doubt it's a CRIP problem. CRIP is an offset. He says that advance is reduced when RPM rise. Wrong CRIP would not cause that. Wrong CRIP would cause an offset of the timing table.

I seriously doubt it's reversed coils. Believe me; you won't have enough time to start your engine... They will melt from inside and a strong plastic smell will fill your dyno and probably the whole floor. Oh... Don't try to remove them without a thick insulated glove if it happens....

Chuck Dean

11-09-2005, 05:52 PM

Some settings on the Motec...

CRIP=465
#teeth=8
Ignition Type=100 (special)
Ref/Sync Mode=11 (multi-tooth)
Ref Type=4 (low-level magnetic)
Sync Type=4 (low-level magnetic)

Ignition type is set to '100' because we are using the motec ignition expander. You don't have to specify rise/fall.

Alexandre D.

11-09-2005, 06:39 PM

When i talk about rise or fall, I talk about the sensor trigger, not the ignition trigger (sorry it wasn't clear)

You have a 11 setting meaning rise rise multitooth. Have you checked the wave of the 2 sensor on a scope?

Chuck Dean

11-09-2005, 07:24 PM

The guy before me determined that both sensors were RISING EDGE sensors.

Frank

11-09-2005, 08:11 PM

not coils wired backwards...

the signal is working such that the coils are charged when they should not be charged (and vice versa)

we did this once

another question, on your motec log ... what is the %sync reading.., it should be constant, at somewhere in between 25% and 75%, or you should change mode....

if it is between 25% and 75%, and wandering, you have problems with hardware that needs be sorted

Erich Ohlde

11-09-2005, 09:18 PM

ah i see what you are saying about the coils frank

Also, what the heck is this CRIPS stuff?

Colin

11-09-2005, 09:24 PM

CRIP
Crank Reference Index Position

it's the amount of degrees between when the cam sensor is triggered to TDCC Cyl 1.

Chuck the sensor's are only Rising edge if they are wired around the same way as the guy before you wired them. Are you getting any errors while running?

Chuck Dean

11-10-2005, 12:34 AM

We only get ref/sync errors during cranking. But we believe that is due to low, fluctuating cranking speeds. If we attach a battery charger w/ cranking assist, we don't really get the errors any more.

Colin:
Are you suggesting that if you swap the leads going into the sensor it changes from rising to falling edge? I know very little about electronics and waveforms.

Colin

11-10-2005, 12:55 AM

yep chuck that's exactly what i'm saying, if you swap the leads it swaps the wave around and
you will always get error's when cranking.

Frank

11-10-2005, 02:14 AM

dude,

the motec unit varies the "dwell" with rpm

if you have the "logic" of the coil driver wrong, it will stuff your "timing"

do this test

set the ignition timing at a constant position
ie 10 deg BTDC regardless of throttle position, and rpm

motec>ignition>crank index position>test advance
(while the cursor is on the value, the ignition will be constant)

rev the engine, and (using a timing light) see if the timing is truely constant, if it is advancing, i bet the function of the coils is wrong, we did this once

sorry about the quality of the explaination, i grew up changing points

Chuck Dean

11-10-2005, 09:05 AM

Frank,

As said before, we tried checking the crip w/ a timing light. At first we were getting multiple signals. After playing with it for a little bit we got a single signal, but it was 43 degrees off. ??

-the dude

Frank

11-10-2005, 09:11 AM

the window you use to check / change crip...

motec>ignition>crank index position>test advance

with the ecu connected to the 'puter, start the engine, and get the (motec>ignition>crank index position>test advance) window up. Rev the engine up and down, the timing light should indicate constant timing..

is it doing that?, if not there is a problem with the "logic" controlling the way the coils are operating

2nd question / test, with the engine running normally, rev the engine, download a "log", is the percent sync constant?, if not, the sensors are not set up well

3rd question, what is the %sync value?, if it is not between 25-75%, you need to change mode

obviously, there is an individual crip value for each of modes 11/12/13/14, but they should all be with 100 degrees of each other, you need to set the crip ("time" the sensors) before it will start

changing modes (11/12/13/14) will change the percent sync, and perhaps; the engine may not start in one or more modes (you wont be able to find the appropriate crip value, the ecu will be jumping between teeth ie the %sync will be changing from 5% to 95%)

one more question, what is the "tooth" on the cam? is it a single "tooth", or are you using (the dodgy way, we've done it for ever) one carbon steel bolt, one stainless steel bolt method, a common fault is to accidentally have 2 CAM teeth, when you meant to only have 1

ok, after editting this post about 10 times, im finished the rant

Chuck Dean

11-10-2005, 09:44 AM

Frank,

1)No, with test screen active we get a constant value (43 degrees off) with the exception of when it was alternating between two different positions.

2)"download a log"?

3)I don't recall seeing %sync anywhere on the motec...elaborate. Keep in mind our motec setup parameters were setup by someone before me so I am unaware of what tests they did to determine these parameters.

4)We are using the stock triggers and sensors. The gixxer has a single tooth cam trigger.

Frank

11-10-2005, 10:15 AM

the ecu collects data ("data logging" or a "log")

you can display the data in motec "interpreter"

%sync is an output from the ecu, it is sensing where the teeth are in relation to each other

have you even read motec documentation?

Chuck Dean

11-10-2005, 10:49 AM

Frank,

I was asking you to elaborate because you are vague with your replies. I assumed you were talking about Motec's datalogging but I wanted you to clarify. We do NOT have that ability (datalogging). We haven't paid for that password yet.

Yes, I have read plenty of Motec documentation. You don't need to be a dickhead about it.

Alexandre D.

11-10-2005, 02:32 PM

Dwell time change only if you specify a custom table and with battery voltage, which should be constant once the engine has started( not always if you ask too much amps...)

I wouldn't recommend blindly trying the 11-14 settings. Considering you have mag sensors, it's critical to have the right setting. Hall sensor make square wave. So in theory any setting is ok as long as you adjust the CRIP. But you don't want ref and sync signal to be too close to each other. If they are too close to each other, the motec may mix up the timing. Luckily the motec has some protection that would allow it to reject such readings, because you don't want your engine to be offesetted 30? closer to TDC. For mag sensor, it's triggered by the 0 crossing of the signal, not a high or low signal. So if it's not the good setting, it will read the 0 signal that happens when it "discharge" itself. This signal is highly variable and I don't know what the motec uses for 0 detection, but if it's what i think, results will vary a lot as RPM change.

Kirk Feldkamp

11-10-2005, 07:53 PM

I think I'll chime in here.

You're going to want that data logging! You will save so much time and effort if you just pony up the dough for the logging. Loggin is probably the single most useful tuning tool. You'll use it on the dyno. You'll use it on the car. You'll use it for testing. You'll bang your head on the table for hours if you do not have it. Trust me.

Case in point, the problems you're having right now. If you could look at your logs you could see your ref and sync percentages. If you're stuck at 43 degrees off when you go to the test advance, it sounds like your crip is off by 43 degrees (or whatever that works out to depending on where our degree wheel is).

We used setting 11. I don't know what the difference between 11 and 14 is. Anyway, the key to making the evenly spaced trigger wheel work is to make certain the sync (cam) triggers somewhere close to 50% between two of the ref (crank) teeth. That's the ref/sync percentages we're talking about. Theoretically the Inscrutable Japanese would have done something similar, but it's something to double check.

Another issue that we were warned about was using less than a 12 tooth trigger wheel with the Motec. We never even tried to use the 8 tooth because the guy we were working with definitely knew his Motecs. I've heard that you can make it work sometimes, but he strongly advised against it. He actually described to me the exact problem you're having. I whipped up a 12 tooth wheel from some steel sheet and hacked up the stock wheel so that only the cam chain gear was left. Works like a charm. I'm not sure if Suzuki upped the number of teeth for more resolution in 2004 and on. What year is your GSXR? Are you using the coil-over-plugs since you're using the ignition expander? I've never used that box, but you might want to double check its settings to make sure you've got the right order of things (both in the Motec and in the wiring).

-Kirk

Frank

11-10-2005, 08:26 PM

well i think you can check % sync in real time with the laptop connected, and if you don't know what that is, then don't call me dickhead

its getting close to comp for me, so im a bit stressed, and im not replying as thorough as i should.. do you have local motec dealer support?

http://www.uq.edu.au/fsae/crip.jpg

Chuck Dean

11-10-2005, 11:14 PM

TurboWig:

1)The reason why we haven't paid for Motec's data-logging is because we have our own telemetry/data-acquisition system. But this is a seperate team from my own and they just recently got it working - so its still in the testing stage.

2)Keep in mind that I didn't setup the initial engine parameters so I don't know some of the procedures as thoroughly as I should. But, shouldn't the CRIP value be the same for each GSXR engine as it is a reference from cyl 1 TDCC.

3)It still doesn't make sense why advance decreases as rpm increases.

4)We are currently using massive independent coils (also used in 2000hp applications) that the previous team purchased but I am working towards using the stock coil on plogs with a Bosch igniter. I have tripple checked continuity and firing order.

5)'01 GSXR w/ 8-tooth crank trigger

Frank:
I honestly don't think %sync shows up on any of the main motec screens. It must be something specific to the datalogging side of Motec. I do know what you mean by %sync however. The guy before me did traces of both signals and was satisfied with the results. That's all I know about it at this point.

I'm familiar with the digram you provided. I have looked at Interpreter's CRIP animation/tutorial.

I have some contacts @ Motec but they are hard to get a hold of and I feel the problem we are facing isn't something that can be easily answered over the phone.

Chuck Dean

11-11-2005, 12:00 AM

So are you guys convinced that this is an electrical issue?

Is it possible that the cams could physically be out of sync with the crank due to a poor re-build. Perhaps, off by a tooth?

Frank

11-11-2005, 01:51 AM

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content"> As said before, we tried checking the crip w/ a timing light. At first we were getting multiple signals. After playing with it for a little bit we got a single signal, but it was 43 degrees off. ??
</div></BLOCKQUOTE>

so what percent sync are you getting?

for example, if its at 5% sync at idle, it might with increasing rpm, change over to 95% sync, therefor instantly changing your timing about 360/8= 45 degrees

whilst holding the engine at constant rpm and throttle (ie at constant ignition advance) :

a timing mark jumping around suggests this is the problem (ie you will we it at two distinct points, at 45 degrees of each other, not varying over a range of positions)

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content"> Is it possible that the cams could physically be out of sync with the crank due to a poor re-build. Perhaps, off by a tooth?
</div></BLOCKQUOTE>

FFS, check the cam timing then

Chuck Dean

11-11-2005, 09:11 AM

Good suggestion Frank. That does make sense. The engine I'm speaking of is currently in the car and the car is bouncing back and forth between different events so its hard to get access to the engine. But I'll definitely look into that.

Charlie

11-11-2005, 11:41 AM

Have you tried checking the polarity of the crank sensor yet. I agree with Colin that is the most likely cause.

Chuck Dean

11-11-2005, 02:20 PM

Frank:
I stand corrected. There is a %sync readout on the main sensor screen (by pressing v).

I took a look at it today. At 4500 rpm, the %sync gives a constant readout of about 67-69%. HOWEVER, if i rev the engine the %sync increases as high as 90%. I haven't seen how the %sync behaves under load yet.

So, why is the %sync increasing w/ RPM?

Frank

11-11-2005, 08:29 PM

sensor wiring/polarity/or setup problem

or..... heaps of slack in the cam chain (unlikely)

glad you got it sorted

BryanH

11-12-2005, 04:35 AM

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Chuck Dean:

I took a look at it today. At 4500 rpm, the %sync gives a constant readout of about 67-69%. HOWEVER, if i rev the engine the %sync increases as high as 90%. I haven't seen how the %sync behaves under load yet.

So, why is the %sync increasing w/ RPM? </div></BLOCKQUOTE>

50% sync means that the sync signal is exactly 1/2 way between two ref signals. if the % increases past 99% the ign will trigger off the next tooth which is 45degrees ahead of the intended tooth, advancing the timing 45 degrees!

the syn% is increasing because the mag signal is upside down. Motec is triggering off second edge of the wave which shortens V. rpm, so as rpm rises ref sig retards until sync sig "crosses" ref signal and timing events are triggered off next ref tooth, advancing actual ign event by 45 degrees. when you have it sorted check actual timing against screen timing at 8k to 9k rpm and adjust crip until they match.
datalogging is useless for sorting trigger/wiring probs. get a $200 Vellamen handheld scope and learn something instead of assuming.

MoTeC

11-13-2005, 10:49 PM

Hi Guys,
Just to clarify some points. Yes wiring a mag sensor the oposite way will invert the signal. Mag sensors polarity is something that must be checked with a scope. The signal is basically a sine wave going above and below 0 volts (positive and negative), if the signal pulls negative volts to startwith and "rises" to positive voltage before settling back to 0v it is a rising signal. Obviously the opposite applies to a "falling" edge signal.

As engine speed increases the length of the pulse decreases, if you have chosen the wrong edge this will result in the trigger point moving and will cause this type of problem. The zero crossing point in the wave form stays pretty much where it is which is why it is used as the trigger point.

MoTeC ECUs have definable trigger points so it can be offset from absolute 0v for noise filtering purposes, see help on sensor type.

Ok CRIP: Crip is the number of crank degrees the index tooth occurs before TDC compression number one cylinder. With evenly spaced teeth the index tooth is the crank (ref) tooth that happens straight after the cam (sync) tooth (Colin you obviously don't listen to anything I say in my seminars, haha).

There is no problem using eight teeth on the crank, you can use any number of teeth as long as it is a multiple of the number of TDCs there are in one revolution, 2, 4, 8, 12 etc. The more teeth you have the better the resolution of RPM. There are of course practical limits.

Hope this helps.