Need help with starting my VW bus

[panel]

OK......battery is up to charge and video to see the spark is here. Does it look hot? And it still didn't start at all

http://www.youtube.com/watch?v=-Niiq8bl24s

Some of my constants are here to look at also.

[panel]

I think I might have noticed that at one time I had 2ms in my hot and cold (to try) cranking pulse widths and noticed in TS that it was higher than 2ms while cranking. Shouldn't it be 2ms or less? I can try this again if someone wants me to verify? Or what causes this? Is it a battery compensation thing or?

[panel]

I'm still having problems guys. I just can get it started at all Can someone verify if it's correct to have the cranking pulse width's match each other in the tuning software gauge and the chart? See below to see what I mean. While cranking I have 4.6ms in one of my boxes but the gauge on the right reads at 6.247. Is normal?

I've also disabled the priming pulse to maybe help not to flood it. I really smell lots of fuel but nothing seems to happen.

[Bernard Fife]

panel,

I think your engine is flooded, and you need to fix that before doing anything else. A flooded engine will act like it's getting no spark - you can crank and crank and there's just no life. It may also seem like it's getting no fuel (it might not even smell of gas at the exhaust, because after the additional fuel is pumped out by cranking, the spark plugs will still be wet and effectively shorted to ground, so there's no spark, and the engine won't run).

If you are very lucky, you *might* be able to leave the plugs in, but pull the injector fuses and crank the engine several times, wait a few minutes, crank the engine some more, then put the fuses back in and see if it starts. This 'flood clear' sometimes works, if the engine isn't badly flooded.

However, if that doesn't work the usual procedure is to remove the spark plugs and let them dry completely, put them back in (ideally blow compressed air into the cylinders through the spark plug holes to try and dry them out too, or alternatively crank the engine over a few times with the spark plugs removed to blow the cylinders out). Theres no real shortcut, and waiting any reasonable length of time won't usually help much either.

The cranking pulse widths you specify have the opening time, intake air temperature correction, and voltage correction added to them, so the actual values are often a millisecond or two longer than the table values.

You are getting a lot of advice. My advice, if you want it, is here: http://www.megamanual.com/ms2/tune.htm#howto I don't especially agree with setting your cranking pulse widths based on idle pulse widths, though. The problem is that the idle pulse width is highly dependent on idle speed, ignition timing, camshaft specs, and a million other things. However, you have to tune anyhow, so that might be a place to start.

One thing we likely do know is that the current pulse widths are too long (or at least that the pulse width at the current ambient temperature is too long). With current ambient temperatures, I would think your pulse width would be somewhere around 4.6+open_time+volts_adjust+... So if this was mine, and assuming the engine starts when hot, I would lower that value, and the ones around it. I would probably try something like:

temp --- old --- new
-40 --- 11.0 --- 7.5
-20 --- 10.2 --- 6.7
0 --- 9.4 --- 5.5
20 --- 5.9 --- 4.9
40 --- 5.0 --- 4.4
60 --- 4.6 --- 4.0
80 --- 4.3 --- 3.7
100 --- 4.0 --- 3.4
130 --- 3.7 --- 3.1
160 --- 3.0 --- 2.8

(after fixing the flood condition, of course).

When you have a few values at somewhat different temperatures optimized (i.e. good starts), then you will get a feel for whether the upper values or lower values need the most work.

Always start low and work up with cranking pulse widths, though, that way you won't flood the engine and can keep trying. If you start low you can increase the pulse width 0.2 at a time until the engine starts, and then you have a workable value at the temperature and a running engine. (But once the engine starts you can't do anymore cranking tuning for that day - except for the fully warmed-up value.)

Also, try to get your prime pulse as low as you can (1.0 is a start). The problem with the prime pulse is that is is activated every time you turn the key. If you troubleshooting, you might do this a half dozen or more times between start attempts. If the prime pulse is substantial, this will flood the engine.

Lance.

[panel]

I've pulled the plugs and I'll clean them up and dry them off. I'll also do the injector fuse pull and crank her over a few times. Thanks for taking the time out to reply. I'm glad you guys are patient with me and although I might be frustrated I'm glad I'm also a very patient guy. This will help me learn etc. I've read all the posts above and re-read them again to absorb as much as I can.

I've switched to tables in the tuning........should I put 1.0 'prime pulse' all those temp bins?

It must be flooded I guess........it does smell really gassy.

I'll report back.

[Heribert]

Not wanting to be a worrier , but saw the photo of your very clean engine with the IAT sensor,nice work.
Guess it is as good inside as on the outside and thought I would give a heads up.
Flooding an engine several times in a row can give some unhappy wear problems. Cylinder walls ,pistons, rings and wrist pins may get washed VERY clean
and wear will increase. You may also dump some of the unburned fuel into the crankcase where it will dilute the engine oil. To counteract the possible oil dilution that may aready have happened, change oil. You can use cheap but reliable oil while experimenting with PW´s and starts. Once you are close use better oil.


So , if you have indeed cranked cold for extended periods with overrich fuelling, try to go to unrealistically small cranking PWs. This will definitely stop the engine from firing ( start at 1 ms ) and creep up . This will give you a good ballpark figure for cold starts, but after it starts to fire in a semi-decent way , you will have to wait until the day after, so that the engine is cold. But do run it up to normal temp before calling it a day. Set ignition at 5-10 BTDC until you get reasonably close to a decent cold start.


Best regards

Heribert

[Bernard Fife]

should I put 1.0 'prime pulse' all those temp bins?

panel,

That's what I would do. I remember taking my Megasquirted Corvette over to Vancouver for an early meet (2004?). As you know, it's a 90 minute ferry ride over there. On the way, I listened to the radio several times (key on ACC), but shut the key off while not listening to save the battery in between. Inevitably, when I went to start it, it was flooded from the multiple prime pulses, leading to a minute or two of panic (a few hundred cars anxious to get going behind me!). Ever since then I always keep the prime pulse very low.

Lance.

[panel]

OK guys...........uploaded a couple of logs for ya. Pulled the plugs last night dried them off though they didn't look very wet. Got home tonight and cranked the motor with everything out pulled the fuses etc and still no go. There seemed to be a weird glitch I've never seen before that the Throttle gauge was out of wack all of a sudden. See if you can see something weird in the log maybe?

I thought the plugs were maybe shot so I video'd them while cranking and there was a spark for sure so I know it's sparking in the head.

Heribert......thanks for the kind words on the bus. Yes it's my baby for sure! Had it since 94' ish and really love to cruise around and blow by people and then they wonder what's in that thing. They always think it's a Porsche motor

[Bernard Fife]

Panel,

I don't see anything obvious in the logs. With the idle valve disabled, you will likely have to open the throttle slightly while cranking to get enough air to start at cooler temperatures. However, you shouldn't 'floor' the throttle while cranking (I see this in several places in the log - cranking starts then the throttle is floored after a couple of seconds). TPS values over ~75% will cut off *all* fuel regardless of cranking pulse width settings (so you only do this if the engine is flooded). Just open it slightly (less than 1/8 of the way), and let the cranking pulse widths work as designed.

Aside from that, there are also a few other things I would look at:

- it looks like you are running low impedance injectors. You may not have enough PWM% to open them once cranking is done and the engine tries to start. I don't think this is the problem (because I don't see your rpm hovering around the 300 rpm cut-off you have set in your MSQ). However it wouldn't hurt to try raising this from the current 20% to around 24% to see if that helps.

- We don't actually know if this is a cold start problem, or if something has happened. So if this was mine, I would:
--> reload the code,
--> reload your MSQ,
--> check to see if you are still getting spark while cranking (<edit>oops, I see you already did this!</edit>),
--> if you still get spark, then I would check the MegaSquirt on a stim to see if the injector drivers are still working (or if it's possible to pull an injector to feel if it's 'clicking' while cranking, that's even better).

Lance.

[panel]

- We don't actually know if this is a cold start problem, or if something has happened. So if this was mine, I would:
--> reload the code,
--> reload your MSQ,

Lance.

OK ....just did this and tried to crank her over again with no luck. After a couple of tries now the pump keeps going on and off again like it did a couple of days ago. This is the third time now in a few days. First was the pump on/off , second was the pump on fully and now on/off again.

Man do I have bad luck or what

[Bernard Fife]

panel,

It sounds like the fuel pump relay may be dying. That would explain a lot (if it's true).

Lance.

[panel]

panel,

It sounds like the fuel pump relay may be dying. That would explain a lot (if it's true).

Lance.

It's the same as last time where it looses it's connection and goes 'off line' and the laptop doesn't connect to the MS.

[Bernard Fife]

panel,

Okay, then you need to check if it is just the comms that are being lost, or if MegaSquirt itself is shutting completely down. I suspect the later, and the fuel pump cycling is a result of this <edit>looking back at your logs, I can see the reset almost every time you crank</edit>. The most common causes of this are either a D19 diode that's failing, or possibly the polarized capacitor C22 is failing.

There is more info in step 14.b) here: http://www.megamanual.com/ms2/V3trouble.htm

You can run without D19 (the V2.2 boards didn't have this diode), but I would replace it as soon as convenient if it was mine (but with a 6.2 Volt, 1 Watt Zener diode, such as 1N4735). D19 protects the 5 Volts circuits from seeing more than 5.6 Volts at any time (Most 5V components are safe to 6.5 Volts), so it is a kind of 'safety valve'.

C22 prevents spikes in the external Vref (generally induced voltages from spark plug wiring and that sort of thing) from reaching the internal components. So it's important, and if it's fried you should probably replace it before running the vehicle.

Lance.

[panel]

Got it running! Swapped back to an older laptop........re downloaded the s19 file,MS Download 2.00 etc and re-loaded one of my latest MSQ's and within a few minutes got it to start.When I first started out running the MS-II this year I was using Megatune, then borrowed a faster laptop so I could run the MLV and try Tuner studio. So now I'm back running MT on my original laptop. I know you said not to use the throttle but I did a bit.(you'll see it in the log )

I'll try it again tomorrow when I get home from work and see if it'll start again without using the throttle.

Let me know if you see anything weird? Up'd the PMW % also like you said.

I guess it's probably not a good idea to do too many changes like tuning software,computers etc but I was getting a little worried something was really a matter.

[panel]

panel,

Okay, then you need to check if it is just the comms that are being lost, or if MegaSquirt itself is shutting completely down. I suspect the later, and the fuel pump cycling is a result of this <edit>looking back at your logs, I can see the reset almost every time you crank</edit>.
Lance.

What does this look like in the ML Viewer?

And.....can I just add this (below) into my V2.2 board even though it didn't come with it?

You can run without D19 (the V2.2 boards didn't have this diode), but I would replace it as soon as convenient if it was mine (but with a 6.2 Volt, 1 Watt Zener diode, such as 1N4735). D19 protects the 5 Volts circuits from seeing more than 5.6 Volts at any time (Most 5V components are safe to 6.5 Volts), so it is a kind of 'safety valve'.

[Bernard Fife]

panel,

The seconds will normally count up to 255, then reset to 0 and start over. If you lose comms, but the comms come back, the seconds will count up to some number, but then jump to another number (could be higher or lower, but only has a 1/256 chance of being zero). If there is a reset, the count will drop to zero before reaching 255. A reset will always go back to zero. However, the best way to see the difference between a comms problem and a reset is to watch the LEDs. If they all go out when they normally would be on, then the MegaSquirt is shutting down.

One thing that can mimic this is if you are turning the key completely off between cranking attempts. You shouldn't do that for two reasons:
- it adds a prime pulse every time you start the power again, and
- it makes it hard for any one else to distinguish real resets from 'power cycling'. Of course, you know which are real resets - you were the one turning or not turning the key.

You can add a "D19" to the V2.2 board (between 5V and ground, banded end towards 5V), but I wouldn't.

Lance.

[panel]

Lance and others...

Just got home and thought I'd get at it while the bus is been sitting. Cranked for awhile and then slowly dropped the 'cranking pulse widths' down until she started. I did open the throttle a little and the engine kinda sounded like it was going to start so I dropped them down some more until it started.

I'll have to try with just straight cranking in the morning or when I get home tomorrow to see if those really low 'cranking pulse widths' and the 1.0ms prime pulse is making a difference. Could it be that I'm back using a different lap top and re-loaded everything that it seems to be starting better?

I also enabled the PWM valve to keep the idle up a little while warming. Helps a little but now with the CLT sensor in the head it climbs much faster than having it in the airstream. And I run out of coolant temp sooner to add a little more time on the valve. But I'd probably be driving it rather than just idling in the driveway.

I've posted my log and MSQ below if anyone is interested

[Bernard Fife]

panel,

Some serial ports are a bit 'funny' in that they seem to mess with the 5V supply on MegaSquirt. Sometimes this is a result of of problems on the board, sometimes it is the laptop itself. But it certainly doesn't hurt to try cranking with another computer, or with the serial cable disconnected to see if that helps (it seems to help you). Not having a laptop connected makes it hard to datalog the cranking, though. In general, connecting the serial cable once the engine has started seems to be fine for most users.

In the morning you might find that the lower pulse widths are a bit too low now, if you were working 'downwards' with the pulse widths. The reason is that there may be additional fuel in the manifold from the larger pulse widths used first. But try in the morning with the way it's set now, and then move up 0.2 millisecond at a time until it starts and you should be very close to optimum.

Also, I think I told you earlier that the start value for the IAC wouldn't matter for a PWM valve, but I now think I may have been wrong. I think this may be the source of the very high revs you are seeing on hot restarts. So you might have to experiment with that value to get it workable for your set-up. You can also try shortening the 'Crank-to-Run Taper Time' under 'Fuel Set-Up/Idle Control'. (I will check this on my bench when I get a chance.)

Lance.

[panel]

panel,

Also, I think I told you earlier that the start value for the IAC wouldn't matter for a PWM valve, but I now think I may have been wrong. I think this may be the source of the very high revs you are seeing on hot restarts. So you might have to experiment with that value to get it workable for your set-up. You can also try shortening the 'Crank-to-Run Taper Time' under 'Fuel Set-Up/Idle Control'. (I will check this on my bench when I get a chance.)

Lance.

When I have the PWM Ford valve at '0' or a bit of duty cycle on 'Cranking Position(extend)(steps)' when cold it seems to start ok. But when warm/hot operating temp the motor just screams like you have the throttle stuck open then dies down to the regular RPM for the 5sec (or whatever I set it to).....and to the duty cycle at that CLT temp.

This happens when you go into a store or something.......come back to the car and start it up and then it cranks I guess at the low % and then ramps down to the closed or nearly closed position.

There is no hot setting for having the valve closed while cranking. I don't really need it when the motor is hot.

For the Cranking Position(extend)(steps) I've settled on 38-40. Seems ok for hot start ups. But the 'Start Value'(retract) doesn't do anything as per Matt Cramer and yourself for my type of valve.

I might pursue a stepper valve this winter if it will do the 'closed' setting when hot cranking thing.

~Marc.

[Bernard Fife]

Marc,

I think if this was mine I would try a crank-to-run taper time of 0 or 1 seconds. That should keep the engine from screaming on a hot restart. Instead the PWM will (or should) go to the temperature table value very quickly.

Lance.