John_D.

iTrader: (2)

I'm going crazy. I blew my engine a while back. Finished putting a forged shortblock back in it, using the same heads (freshened up), same cam, same rockers, same 60 lb mototrons, pretty much same everything. Except for the shortblock, and a set of Morel lifters. Tried to start it last week. No compression. Flooded it. Found that the lifters were taller. Got my shorter pushrods in the engine at lunch today. Good compression. Tries to start now.

Still flooding. If I hold the pedal all the way down (flood control mode) it will start and flare to 3k then die right away. Tried many variations of pedal settings, and the only thing that worked was all the way down, and that was good only long enough to flare and flood. Ok, so to help cure the flooding, I unplugged the wires to the pump (at the connector behind the rear seat). After the 4th or 5th cough, it started. I figured it would run a second or two and die, after the pressure bled off the rail.

But imagine my great surprise, when the engine continued to run. For several minutes. I could hold the pedal down just a bit, for 1500 rpms, and it would run without a hiccup. (I got it up to over 160 degrees coolant temp this way). If I let it drop below 1500, it would flare from 1500 to 1000 a few times and die.

It would not restart unless I plugged the fuel pump back in and re-pressurized the rail. But would not run unless I unplugged the pump. And if I plug the pump in while it's running, it immediately dies.

Even with the pump unplugged, and my electric gauge showing zero psi on the rail, not only does the engine run, but it sets off the smoke detector in a matter of seconds, and my eyes are burning a few seconds later, because it is SO rich.

The gauge is working. It shows 70 psi at key-on. (not sure why it's so high) I upgraded to a walbro while I was doing the engine R&R. This car is running a '97 vette rail and fpr, I put those on back in '04 when I did the v6 to v8 swap, in order to keep my return line that the v6 cars use. This is a proven setup and worked fine. The fpr is not leaking fuel past the diaphragm.

So to summarize the major changes since the last time the car ran successfully: new short block, taller lifters, shorter pushrods, walbro pump, sts turbo, 2-bar hptuners tune, 2-bar cobalt map sensor.

How in the heck can the engine even run with the fuel pump not running? Does the manifold vacuum pull fuel from the injectors while they are being pulsed (and eventually pull/siphon from the tank...)????

And an even bigger question, how can it pull enough fuel to run so hugely rich?????


This is essentially the same tune I ran when the car was NA. The IFR table for the 60's, the VE for the cam, etc. I took timing out, pumped KR response back up again, fattened up PE, and turned on torque management again to help the auto tranny last a little longer.

I am so stumped on this thing.... Maybe the tune is holding the injectors open some huge pulsewidth at idle....????



The injectors are not stuck open. When I pressurize the rail to 70 psi, it holds the pressure.


I've never seen anything like it.

GOaT Cheese

Does the pressure hold with the key in the "on" position? I am only assuming that you are checking the rail pressure with the key off. If the pressure holds at 70 with the key off, turn it on without cranking the engine over and see if it doesn't dump.

Old Geezer

iTrader: (3)

w/ the rail and injs full of fuel, that pulsing the injectors, will allow the fuel to dribble out, and the engine will run.. to some degree, anyway.
After having said that, I ran a set of injs on the test bench, to duplicate the described occurance. The bench was run static, to fill the rail and injs. I then went to "pulse open" mode, and the entire rail drained, as did the injectors.[Gravity fed, as is your setup] However, it was a rather short period, due to the fact the rail is nowhere near the length of the engine rail, and the "pulse open" time is MUCH longer than what your ECM program would be providing @ 1500rpm. Given that info, I would say the engine could run, however poorly, until the rails were empty. At a very short commanded PW, it could be for some time.
The bottom line is: Find out why the FP is so high, and correct that first.
A. Regulator took a crap?
B. Return line kinked, plugged?
C. Gauge fubar?
Then, I'd pull the turbo to TB pipe off, and see if the engine will lean out, when run in N/A "mode".

I'd suggest that you also change the oil and filter. Fuel in the oil is a quick way to find yourself rebuilding your new engine!
In your testing to find the FP problem, I'd start at the rear, and work forward, 1 component at a time.

HTH,
Back to

John_D.

iTrader: (2)

Good question. I'll check that at lunch today.
Key-on, unplug the pump, leave the key on for a while.

John_D.

iTrader: (2)

Thanks for all the ideas, and the testing.

Let me add a couple of pieces of info that may help in the diagnosis.

The rear mount turbo is mounted in the exhaust, but it is not plumbed to the front of the car at the moment. I have a cone filter plumbed to the tb for this initial start-up testing.

The injectors are subject to manifold vacuum once the car starts, so that will pull fuel, more than just the gravity feed effect. Much like the idle jet below the butterfly on a carb setup. The engine ran for 2 minutes at a stretch, with no power to the fuel pump at all. And very rich at that.

So I have a super rich problem, with zero fuel pressure. I can work on getting the 70 psi down to 58 psi -- but -- if it runs rich at zero psi, getting it down to 58 is not going to help (at the moment). I need to isolate how it can run rich with zero fuel pressure first.

I think the next step is a quick test run at lunch to see what pulsewidth the pcm is commanding. If it's out of line, I can chase the pcm tuning angle. If the pulsewidth is low (1-3 msec) then I need to chase either a mechanical problem (slow close) or an electrical problem (injector grounded longer than the pcm commands).

The car sat for 17 months, so it's possible there is some gum - not enough for an injector to hang open constantly, but maybe enough to slow it down - giving it an effective pulsewidth much longer than commanded. I did completely drain the tank and put fresh gas in it, and flush the line and rail before trying the first start.

Old Geezer

iTrader: (3)

If the injectors sat open/with gas in them for 17 mo's, there's a REALLY good chance they are gummed up. I see that frequently with injectors I clean, and flow test.

John_D.

iTrader: (2)

I had a flash over lunch, might be a factor.... If the injectors need some pressure behind them to snap closed when the pulse signal is removed? Then my zero pressure scenario, while a problem in itself, may not indicate there is some other problem.

I probably do need to backtrack and address the 70 psi key-on pressure. The ideal fix is an adjustable fpr (which I don't have and would have to plumb if I did have one). The quick fix is to adjust my ifr table to account for the effective flow rate of the 60 lb Mototrons when faced with 70 psi of pressure. This would let me know if the injectors need pressure in order to function correctly. If they do, I'm chasing my tail with the pump unplugged....

John_D.

iTrader: (2)

While updating the IFR table to tell the PCM I really have 76 lb injectors (at 70 psi) I discovered something else that probably is having a really big effect. The last thing I did to my tune was turn on 2-bar. The VE table magically became 519 % in every cell. So that would have the pcm holding the injectors open quite a bit longer than normal....

Will load this corrected tune and try again in the morning.

John_D.

iTrader: (2)

That was it!!

Fixed the VE table, modified the IFR table, and connected the wire to the fuel pump.

Started first time, ran fine, let it warm up to 210 (fans unplugged), shut 'er down.

Lessons learned:
1. The engine will run for several minutes with no power to the fuel pump, if the injectors are held open longer than normal. Manifold vacuum will pull enough fuel out of the tank, through the line, and out of the injectors to keep it going.
2. When flipping a tune to 2-bar in hptuners, the VE table is filled with 519's (apparently to force you to fill in the new rows, since they have no way of knowing what you want in the boost region of the table). That's buried in some help text if you go prowl around enough paths for it long enough. Would be nice if the program would give you a notification, instead of just expecting you to know. Now I have fouled plugs, fouled o2 sensors, and diluted oil to deal with...
3. 60 lb Mototrons (acting like 76's at 70 psi) still are only down around 1.5 msec pulsewidths at 800 rpms. So I still have some room to drop the idle down more, without hitting the low end limit on the pulsewidths yet.
4. There's a short pulse adder table in the pcm, that still adds to your minimum pulsewidth. I had the minimum set to 1.1, but it would never have gone that low because of the adder table. I zeroed out that table. It was just one more variable that wasn't needed...

Stuff I haven't figured out (neither one are pcm issues though):
1. Why my fpr isn't holding the fuel pressure down to 58 psi, but is at 70 instead.
2. Why oil is coming out the cold side pipe. Maybe the diluted oil, maybe I need a smaller restrictor in the oil feed line to the turbo.

Just happy to have it running again, it's been 17 months...

98Z28CobraKiller

iTrader: (17)

Congrats. I know how it feels to solve a life ending problem such as this. I was not aware of the inj pulse adder table. That my resolve some of my issues and help me resolve some of my problems regarding low speed bucking/surging.

Thanks for following thru and posting the solution. Many others don't bother to wrap up a thread and leave more questions than answers behind.

John_D.

iTrader: (2)

Thanks!!

I was looking back over my notes from last night, and it looks like I just lowered the cutoff point for short pulse adder. Instead of changing the table (because I though I might need it someday) I just dropped the threshold from 4 msec to 1msec. So now it shouldn't reference that table unless the pw drops below 1msec, which should never happen.

I also found something else that probably feeds into the bucking/surging cycle. There is a table for pulling timing on overspeed, and adding timing on underspeed. The overspeed numbers are spread over pretty small rpm increments, and are pretty big. So I halved most of those, in the smaller rpm part of the table, but left them at 8 degrees if the overspeed gets into the upper part of the table. Looks like a slight overspeed could pull a bunch of timing fast, with the default tables. (which could lead to the following problem)

The underspeed table had values only to -200 rpms (to add in some timing). If you drop more than 200 rpms, that timing that it had been adding, suddenly goes away, so the rpms will drop even more... I extended that on out, since idle flaring can start shooting well past 200 rpm swings.

98Z28CobraKiller

iTrader: (17)

I zeroed out the adder table and it thru my whole VE table out of wack. I was like +23 in all of the lower rpm error scanning with the WB. I fixed it and now I have all my values in line. The problem is the same . I have already done all kinds of work to the overspeed and underspeed spark and airflow. It is much better now than it was. I am running OLSD so not having a mass air does make it worse and since I got the 4.10's it is much much better. It actually feels more like a vibration now than bucking while under steady speed. It is only at very low speed acceleration that it bugs me. I just really WISH that it would be smooth as silk. I may have to get my hands on some 42lb SVO's as there really is no reason for me to have 60#rs.