odd starting attempt

[gboone]

I've been struggling for several months to get my engine to run on MS2.
My engine is equipped with side draft TWM throttle bodies with 4 injectors and 4 throttle bodies (1 injector per cylinder on a 4 cylinder engine). The engine has not run until tonight after many attempts.
Tonight I sprayed some starter fluid in the intake with the throttle closed and it started immediately and it ran up to about 5000 rpm and kept running at this high rpm. I was using a remote starter and had to run around to the ignition switch to shut it down. I thought this was very strange considering the throttle was closed. I let the intake air out and adjusted the throttle plates to a more closed position, recalibrated the tps in megatune, and tried to restart. On trying to start again, the AF gauge was reading 19.9 (way lean)and would not even try to start.
I've been ramping up the crank pulses but I'm not seeing any improvements in starting attempts. While trying to start, the rpms surge alot. I've attached files.

Please help.
Gary

datalog200811102104.xls

megasquirt200811160002.msq

[FixItAgainTony]

Gary,
Glad to see the starter ether got it started => ignition okay. Looking over the log, I see:
1. Cranking RPM looks solid and looks reasonable.
2. The Throttle position Sensor (TPS) is reading 108
Possibly the throttle is floored or the TPS has not been calibrated. At that high value of TPS, the TPS flood clear has been activated and the fuel PW has gone to 0.3 for the duration of cranking => practically no fuel to the engine. That it spools up on starter ether also indicates no fuel to the engine.
3. MAP = 80kPa before cranking – are you at ~6000 ft altitude? During cranking, MAP = 79 or so, indicating the throttle plates are possibly wide open.
4. Looking at table in the msq, the tables look reasonable, except for the
Tuning > warmup wizard >
Pulse width at 0 degrees F = 5
Pulse width at 160 degrees F = 3

From: http://www.megamanual.com/ms2/tune.htm
• Cranking Pulse Widths:
o The cranking pulse width at -40°F should be about 3 to 5 times the 170°F value.
o If you over-estimate the correct values, you WILL flood the engine.
o To begin, set the -40°F cranking pulse width to about 88% of your "upper" req_fuel value, and the 170°F cranking pulse width to about 23% of your req_fuel. Those should get you 'in the ball park' for starting.
For your set up, req’d fuel = 12.8. Looking at table in the msq, the tables look reasonable, except cranking PW.
0.88 * 12.8 = Pulse width at -40°F = 10.9
0.23 * 12.8 = Pulse width at 170°F = 2.9

Because of the temperature ranges in your setup 0°F, 160°F these values would change to 9.4 and 3.3.

5. CooLant Temperature (CLT) 63°F => Cranking PW should be 7 ms after changing the Cranking PW to (9.4 / 3.3)

- Charles.

[gboone]

Thanks Charles,
Looks like you made some good finds with my set up. I think I need to determine my problem with the TPS calibration first. I recognized this problem last week while cranking and I thought I fixed it but obviously I did not. Along time ago with my first calibration for TPS, I had trouble with Megatune telling me the currents looked funny after calibrating. I was certain I had the polarity correct with the wiring from the TPS into the MS2 originally, but on the advice from the forum that I had the wires reversed, I swapped the wires into MS2. Last week during a start attempt, I noticed the TPS indication on the laptop was reading 108. So I changed the TPS wires back to my original orientation thinking it would correct the problem. Funny thing is after I do the calibrations for either wiring orientation, when I stroke the throttle, the indication on Megatune is correct. But when cranking the gauge on meagtune indicates 108 with the throttle closed. Any ideas why this is happening? I don't know what to do to correct this problem.
I'll post this problem on the Sensors forum too.
Thanks,
Gary

[gboone]

My TPS is a TWM unit that came with my new TWM throttle bodies.
I need to determine my problem with the TPS calibration. I recognized this problem last week while cranking and I thought I fixed it but obviously I did not. Along time ago with my first calibration for TPS, I had trouble with Megatune telling me the currents looked funny after calibrating. I was certain I had the polarity correct with the wiring from the TPS into the MS2 originally, but on the advice from the forum that I had the wires reversed, I swapped the wires into MS2. Last week during a start attempt, I noticed the TPS indication on the laptop was reading 108. So I changed the TPS wires back to my original orientation thinking it would correct the problem. Funny thing is after I do the calibrations for either wiring orientation, when I stroke the throttle, the indication on Megatune is correct. But when cranking the gauge on meagtune indicates 108 with the throttle closed. Any ideas why this is happening? I don't know what to do to correct this problem.

Thanks for any advice.
Gary

[Matt Cramer]

When you go to Tools -> Calibrate TPS, what ADC values do you get with the throttle wide open, and with it closed?

[Matt Cramer]

When you calibrate the TPS, what ADC values do you get at full throttle and at closed throttle? Also, measure the voltage with a multimeter - what readings do you get there?

[FixItAgainTony]

Matt, is there a easy way to completely disable Flood Clear ?
In the past I have changed the ADC counts manually in the Calibrate TPS to accomplish this along with setting
Tuning > warmup wizard > Flood Clear Threshold (%) = 70
to some high value (100 is the highest).

If a setup did not have a TPS I am not sure that would work, as there is floating input.
Is there a better more fool-proof method?

- Charles.

[gboone]

At closed throttle:
ADC value is 817
voltage between MS2 pin 19 (ground) and pin 22 is 3.94 VDC
voltage between MS2 pin 19 (ground) and pin 26 is 4.94 VDC

At wide open throttle:
ADC value is 135
voltage between MS2 pin 19 (ground) and pin 22 is 0.785 VDC
voltage between MS2 pin 19 (ground) and pin 26 is 4.94 VDC

These values are measured with a Fluke multimeter and with engine not cranking.

I appreciate your help very much.
Gary

[FixItAgainTony]

Those values look reversed to me, but the range looks good.
The closed throttle ADC reading (and voltage on pin 22) should be low, and the full throttle should be high.

On the TPS itself, there should be three wires:
#1 connected to pin 26 (Vref = about 5V, 4.94V on your board, close enough).
#2 connected to pin 22 of MS (TPS voltage itself)
#3 connected to ground
Wire #2 looks correct, but #1 and #3 probably need to be swapped. After that is done the closed throttle voltage and ADC count should be low and the full throttle setting should high.

When you look at the voltage / raw ADC count in Megatune do the voltages change smoothly as the throttle is opened?

It sounds like the wiring was set up this way at one point odd that it would have had problems then. Any chance that the TPS sensor is grounded through it's case?

- Charles.

[Matt Cramer]

Your TPS sounds like it is miswired. Try swapping the VREF and ground wires, and recalibrating it.

[gboone]

Charles,
I didn't answer your question regarding MAP readings. I am at 6,000 feet elevation which explains my 80 map reading.

Regarding my TPS wiring, I swapped positions with the VREF and ground wires to the relay board and it looks correct now after I calibrated the TPS. In my defense, the pin diagram labeling from the TPS manufacturer was wrong. So that is corrected now. But still the engine will not start.
Please see my datalog and msg files.
Thanks again for help.
Gary

megasquirt200811192102.msq

datalog200811192116.xls

[FixItAgainTony]

Gary,
I loaded your .msq into my set up and ran it through start up on the stimulator while data logging.
It looks like XTau settings are stomping on the After Start Enrichment and that the ASE % cold setting is low along with 0^F cranking pulse width settings.

I made the following change to your .msq (used a text editor - these were the only things changed)

1. Change Xtau to disabled for now.
On the stim, it seems to cause a short PW after start - about 1.2ms which would starve your set up.

Fuel Set Up > General settings > X-Tau Usage
Off

2. Changed
Fuel Set Up > After Start Enrichment
ASE Cold % 45 (was 20)

3. Changed
Tuning > Warm Up Wizard
Cranking Pulse Width
Pulse-width 0 degrees F = 9.4 (was 5)
Pulse-width 160 degrees F = 3.3 (was 3)

I then simulated the start up again looking at PW as the engine transitioned from cranking to running. I put this into one file so that you can see the changes. The 4 mark lines indicate the before / after transition.

Bon chance,
- Charles.

[gboone]

Charles,
Thanks for the effort you've made to help me. I burned your msq file to my MS2 but the engine still will not start.
I've attached files (the attached msq file name doesn't match the filename you named it but I verified it's the same according to the changes you made). Thanks for any further help.
Gary

megasquirt200811202052.msq

datalog200811202146.xls

[FixItAgainTony]

Gary,
The pulse widths look a bit better to me. From the log, it looks like the engine is starting to spool up and then it dies. Is that the case? If it is starting to kick off, then the cranking pulse width settings are in the right neighborhood. The After Start Enrichment, warm-up enrichment, required fuel / Gve values start to come into play as the engine transitions from cranking to running. If it does not keep running, it either had too much or too little fuel. I’d guess on the too little fuel side based on what I see in the logs from my car, which also has a 2 litre engine in it. Your injectors are slightly bigger, but if I take that into account, the pulse width (PW) after start still does not look right. Looking at the Ve table, Gve in the area of the table for cranking is 25. That is a pretty low value, mine is more like 60. PW will be directly related to Gve, so going from 25 to say 50 would double the PW.

I created a new Ve table using:
Basic Tables > Ve table 1 > tools > Ve specific > Generate table
Set engine size to 121 CID
Peak torque 145 @ 4600 rpm
Peak HP 160 @ 6000 rpm
And left the MAP readings at the defaults. Note that you will need to re-enter the engine size and RPM values each time you generate a table.

The attached “veTable1_145t_160h.vex” table can be imported into Megatune
Basic Tables > Ve table 1 >file >table Import

I also added this table to the .msq file, along with making a minor change in the warmup enrichment % settings for 0 and 20 degrees F.

The new Ve table has the Gve in the idle area around 75, which is probably closer than the 25 that was there before. This will give about 3X the fuel from the earlier setup. It may be a bit high, I was just guessing on the engine parameters – I figured you had modified the engine a little, 4 throttle bodies and all…

I also simulated the file for grins – notice that as the engine transitions from cranking to running, the PW tapers off slowly (ASE) from 8.5. On the earlier file I sent out, the PW went from 8.5 to 3.2 ms like a step during the cranking to running transition. At the end of the stim run, I changed the coolant temperature to give an indication of what that does to the PW. It can be pretty easy to flood the engine if everything is not set fairly close.

- Charles.

[Matt Cramer]

Getting an engine to start up is going to cover a lot of trial and error. You will need to get it through these steps:

1. Adjust the cranking pulse width until the engine starts, even if it dies in a few seconds.
2. Use whatever combination of after start enrichment, warm up enrichment, and throttle will get the engine to run long enough to warm up.
3. Dial in the VE table cells in the idling area with the engine hot.
4. Get those warm up / ASE settings set to their final values on subsequent cold starts.

You'll want to go over the section on "Getting the engine started and idling" in the MS2 Tuning Manual as it has a lot of hints.

[gboone]

Thanks Charles I now have been able to get the engine to start and run after I burned your msq file. I owe you so much and I thank you immensely. I’ve been working 12 hours a day including weekends and haven’t had much time to figure this out on my own.

On the first start, the rpm’s started ramping up too high, so I closed down the throttle plates and then I got it to run at a reasonable rpm at idle, but still the idle rpm’s surged a lot and the AF ratio was still way too lean at about 18 to 1. So is it time for me to use the Megatune Autotune feature? Should I use it only after the engine is at full after operating temp?

I’m thinking of closing the throttle plates completely and using the needle valve adjustments on the bypasses on the throttle bodies.

I’ve attached files after my last run in case you have any suggestions. Thanks for your continuing support.
Gary

datalog200811212232.xls

megasquirt200811212147.msq

[FixItAgainTony]

Gary,
Glad to hear that you were able to get it to run. I know that you have been at this for some time. In terms of next steps, I can’t really improve on the advice that Matt has provided.
Read:
http://www.megamanual.com/ms2/tune.htm

This:
http://www.megamanual.com/begintuning.htm
is also probably worth reading too – if for no other reason than to get a better idea of how different terms are used in megasquirt.

I have never used “Meagatune Autotune” feature for any rough tuning and would not use it until the tune of the engine is almost 90% complete. At that point, there is little need to use it. I have played with it, but decided, like the person who wrote autotune, that using Phil Tobin’s MeagaLogViewer (MLV) is a better way to go. I just reread some of the first link above – “autotune and mstweak” are mentioned at length for tuning, yet few people use those, most use MLV. The document would probably benefit from a minor update (introduce MLV, reference autotune and mstweak).

Get the engine warmed up, change the values in the Ve tables to get a solid idle. Trust your ear more than the wide band O2 sensor. It is probably worth recalibrating the O2 sensor in free air if that has not been done in a while. Once the warmed up engine idles solidly and runs a few different RPM’s with gentle throttle changes, take the car out for a gentle spin while data logging.

In MLV, View > Show Tuning Console open the .msq used to make the long data log and run the Ve analyzer. Look at the before and after results.
Read the links in:
viewtopic.php?f=25&t=21194

If you are unsure of how an option works, you can connect the stim, change the settings for the option with Meagatune and then tweak the knob on the stim that invokes the option and look at what happens to things like the pulswidth, etc. That was how xTau settings effect was checked. Easier than grinding the starter.

- Charles.

[a73camaro]

Gboone - You mentioned that you have a throttle body per cylinder. Do you have a common manifold for all the throttles or are they separated? Looking at your logged data there is a lot of MAP fluctuation and that tells me that your map sensor is connected to an individual runner/throttle body and not a collective manifold.

If that is the case you will need to add a “manifold” that takes pressure signals from each intake runner. That should be the place to measure Manifold Absolute Pressure (MAP).

[gboone]

a73camaro,
Thanks for your reply. My intake manifold has separate runners for each cylinder. There is a port on each of the 4 intake runners and I made a tubing manifold that connects to all 4 ports on the intake manifold which connects to the MAP sensor. You can see this in the photo below. It's the small tubing manifold closest the cylinder head, with all the blue fittings. This has been my set up since the FI conversion.

You are absolutely right that my MAP signal was fluctuating excessively. I verified this a couple of days ago by teeing into the MAP sensor hose with a vacuum gauge. The gauge needle bounced all over the place. So I fabricated a canister from a 6 oz can of tomato paste, soldered hose barbs on each end of the can, and put in the line. I tee'd into the hose again with the vacuum gauge, between the canister and the MAP sensor and it calmed the signal completely. No more fluctuation. I think I'll have better success with tuning now. I just completed the can installation last night so I'm looking forward to improved results.
Thanks for the help.
Gary

gboone_fi_2008_11_29.jpg

[FixItAgainTony]

Gary,
That a fantastic looking engine setup. Nice work.
- Charles.