You highlight the text you want to italicized and choose the I button which is in between the B and U buttons just below the Fonts pull down menu :)

Bad power side or loose/dirty injector connector.

Well, I don't have a fonts menu. My screen has internet connections on top along with some other stuff, but nothing by way of editing. Then there is the Pelican screen, whose top doesn't have editing stuff on it either. Not like my E-mail or Word or the like.

I tried the 9 volt battery. It clicked the #5 injector just fine. When I powered up the fuel pump (about 44 psi static) and applied 9 volt power via a spare plug with wires exiting it I could clamp onto, I got a nice spray.

Then I used the regular 12V+ power from the regular plug, connected it to one of my spare plug's wires, and grounded via the spare plug. Same spray.

Then I plugged the regular plug back on, pulled the connector at the ECU, attached a wire to that plug's connection to #s 2 and 5, powered the pump, and grounded there. Same spray.

So, per these tests the wires are all fine, the plugs are all fine, and the injectors are all fine.

It was getting late at night, so I didn't try to start the engine after all this, not that any of it was likely to fix what ails it.

I've pretty much run out of things to try short of plopping carbs back on to try to prove that it isn't anything mechanical in the head/cam/valve/piston seal arena.

Bob

Wondering if there were some obstruction in the fuel delivery, I pulled the rail. It is fed from both ends, and the two injectors which flank the #5 work fine and those cylinders fire.

But the fuel rail was clean as a whistle. No possible obstruction. If there could be some odd resonance which allowed pressure from one end of the rail to cancel that from the other end in the middle, where the #5 is, you'd think that would also happen on the left side, as the rails are identical.

There is still one test. With a running engine, I will use the 9V battery to actuate the #5 injector by tapping one of the wires at the battery while the other is connected. And see if that causes that hole to fire. Kind of like squirting the starting fluid in.

So I'm guessing your PP screen does not look like this:

http://forums.pelicanparts.com/uploa...1374061481.jpg

Be careful with the tests not to back feed the ECU circuit. Sounds like you're fine but just wanted to state to obvious in case. The dead time for this injector may end up being greater than your idle pulse width from the ecu or the volume of spray is lower than the other injectors. Could still be faulty injector even if it sprays with 9v test...

Walt...all things considered in the fuel rail...one of the rules of hydraulics is liquids for the most part do not compress...so pressures at one end of a pipe will be the same at the other end...and hence...pressure should be constant throughout the rail.
What I was suggesting is that the fuel may be entering the intake passage...but because the #5 is the last one to fire on the cycle...the charge might be dissapating before ignition.
The only reason I can come up with is that (for whatever reason)...the charge is being expelled back out the stack of the ITB.
Like a back up of air blowing the charge out...sounds wierd to me even as I type it.
I know you have conducted a compression test that turned out OK.
If I am right...the valve opening of the intake is somehow to blame.
I can't think of any reason really...too tight a setting on intake?
Cam holding intake valve open at wrong time?
Piece of grunge on backside of lobe? (causing second opening).
I know that when you find the answer...we will all sleep better.
Bob

Maybe the ECU itself is the problem. Could the part of the ECU that fires that batch not have enough current capacity to run the two injectors? Can you switch with another batch?

-Andy

Andy - that was sort of the basis for what James Bricken suggested. It was the best fit suggestion I had heard or come up with, so I was excited. I pulled the #2 injector plug to see what happened. Alas, the #2 stopped firing without the #5 deciding it would. Same result with the plugs for each bank swapped over.

Bob - these injectors fire twice per cycle of each individual cylinder. The "first" firing is at a suitable interval related to when the intake valve opens. You can set this (though only for all injectors). I don't recall what it is, except that it is a standard setting I didn't fiddle with. The duration of the injection is what varies with RPM, throttle position, and load (MAP). You can fiddle with that, too, but it is the same for all and I haven't fiddled with what worked before.,

The second firing happens at 180 cam degrees (or 360 crank degrees) out from the first. This is the one which "prefills" the intake space between the injector and the backside of the intake valve.

The "twinned" cylinder (#2 paired with #5 in this case) gets the same signal, and thus about the same amount of fuel from the same timing and the same duration. Except that what "prefills" one is the "just in time" squirt for the other.

There are various other ways to set up EFI. Sequential, so you have only one big well timed squirt for each opening of the intake valve. Staged sequential, so you have the main timed squirt, plus a second high rpm injector which kicks in as RPMs and fuel need go up. These require a cam position sensor, which I don't have, so I don't use these. I think there is another (throttle body injection?) which I also don't use.

This "two squirt per cycle" system is a bit like the CIS, which works fairly well for street motors by squirting all the time, though if you use hot cams with these motors, the reverse waves play hob with things.

I am familiar somewhat with fuel blowback (or fuel standoff), typically from intake/exhaust overlap. The bottom side of the air filter shields on some of my carbureted motors used to have spotting on them from fuel being blown upward before it could be sucked back down, even if this did not noticeably affect running. If the motor wouldn't run, I'd be wondering if my cams were way off and something like that were happening. But it is just this one cylinder. so things can't be that far off in this regard. At least not from any mechanism yet identified.

Here is an observation, though. Using the synchronizer (comparing air flow into each throttle body, I can't get the pith ball (it is that older kind, not the dial Unisyn) to rise as high in #5 as in the others of this bank. If I ignore the compression and leakdown tests, I might think this showed the vacuum in #5 wasn't as high as in the others for some reason, and maybe that had something to do with something. However, this is EFI and not a carb which depends on a pressure differential to get the fuel to spray. And spraying starter fluid makes the cylinder fire and thus isn't affected by any such effect. And I think what explains this is the effect of exhaust extraction even at idle increasing the vacuum and thus air flow in a little bit.

I don't have vacuum ports to directly read intake vacuums overall with a gauge. The TEC3r, when hooked up to my computer, will tell me in KPa what manifold pressures overall are, based on taking vacuum from #s 1, 2, 4, and 5. But that's overall, and can only be further refined by blocking off either the 1/2 or the 4/5 line to the central vacuum accumulator which connects to the vacuum regulated fuel pressure regulator, and to the MAP sensor.

So this is another thing to try - hook up computer and block off first one side and then the other and see what the KPa readings are and how they change, if at all.

My continuing thanks for help in head scratching.

Any chance something dropped into the intake port and is holding the valve open intermittently. I had a brass pipe plug find its way into a VW Type 4 and do the same thing.

Initially, there certainly was a chance (along with a long list of other possible blunders on my part).

But a) #5 is a straight shot into the port, so I can see most of the valve's backside, and don't see anything, b) I ran a bore scope into the combustion chamber through a plug hole, and didn't see anything, and c) one would think that at least one of the compression and leakdown tests I have done would have let the foreign object intermittently cause suspicious test results. Oh, and d) if it were really intermittent, the cylinder should fire sometimes. Alas, it only does when I spray starter fluid or its like.

But I can see I'm going to get out the borescope anyway, and look down there.

Walt...I use the same type syncro unit...the one with the pith ball.
You say the vacuum reading on #5 is down from the others.
When I do a comp test...I jack open the throttle with a wedge or other suitable instrument so that all throttles are open the same amount.
I also remove all 6 plugs so that the engine will spin at the same RPM during the test.
Assuming you follow the same procedure....#5 being low on vacuum has to be the key here.
How about a cracked intake manifold? ... or a bad gasket between the ITBs and the mainifold?
If the setup is basically the same as with MFI...you have the head...then a gasket...then the manifold...then another gasket...then the ITB.
So...you have at least 3 places for a leak.
You are using the old ITBs and I assume the same manifold...so perhaps the gaskets are the culprit.
Unless...some kindly assistant dropped a manifold and didn't tell you?
Is it possible to swap left to right manifolds (checking the gaskets at the same time)..?
Obviously if the problem moves to #2...you have the baddy.
Bob

^I like that idea. An air leak under the butterfly would cause a lean condition which would mean higher egt though... not much to lose if we're exhausting all options however.

Well, I've squirted the starter fluid all around each of the two gaskets in question. No response. They were in reasonable shape when installed. I think I dressed the gaskets around the heat insulators (each ITB has three gaskets, two for the engine to insulator to manifold, and one for manifold to actual ITB).

I guess I could put my palm over the top of #5 so the vacuum is max, and squirt again to see if anything coughs.

The syncro testing was with a running engine. I've never tried that with just cranking. Is that worthwhile? I have had the plugs out so many times once more wouldn't matter. If you do that, would you screw even more closed the adjuster? These are good only for relative readings (adjust so the balls rise the same amount on each intake bell), unlike the kind with numbers which I think is set for gauge readings.

Can't switch manifolds, as they are set up to hold the throttle cross bar and throttle bell crank.
Plus they are PMO aluminum, and you can't crack them short of dynamite.

ITBs also can't be reversed - the butterfly levers would be on the wrong end, as would the TPS. And the ITBs simply sat quietly hanging from their hoses and a bungee or two for the better part of a year. Not dropped, and they, too, are sturdy aluminum castings with no signs of cracks and such.

Even CIS and 3.2 Motronic motors with dried out and slightly loose gaskets seem to run on all six, just leaner than desirable, no?

Walt...I have never tried a Comp test with a running engine...new one on me.
I usually jack the throttles open fairly WOT so that it stays as a constant and is not a factor in checking the compression numbers accross the engine.
With 6 plugs out and only one cylinder being tested at a time...you get a more consistant reading of cranking compression numbers.
Running comp numbers can be all over the place because of air induction/turbulance/ring seating/etc.
I have found some engines that build compression with RPM something fierce (Dykes type rings) they need some RPM and back pressure to work properly....so crank comp compared to running comp is way different.
Still can't come up with a solution for you...but brain is working away at it (keeping me young at the same time...ty)...LOL
Ooops...re-read your post...old eyes saw comp test not sync test....silly me.
Yes Sync test is correct...but also...check for consistant throttle openings (plates resting on stops at idle)...and yes...you can try the palm on ITB test...but I suspect the high vacuum caused will mask the results.
Bob

Have you tried increasing the injector pulsewidth at idle with the engine running? I'm thinking injector dead time... ie, the signal to open is too short for a damaged injector. It would still pass the 9v test and the prime pulse is often longer than the idle pulse.

Bob - I too do compression checking with a compression checker screwed into a spark plug hole. All plugs out, one hole at a time, cranking only. And since I have so much interest here, I remember to block the butterflies wide open (I'm using an SC rear brake pad, as a used one was close at hand).

What I was ruminating on was use of the Uni-Syn or its ilk, which I believe pretty much has to be used with a running engine. I was wondering if differences in its readings, if they can't be adjusted with the air bleed screw, were of some significance. Or whether a running cylinder produces a better air flow than one which is at the same RPM but not firing.

I'll just test that by pulling the injector plug with the device on a running cylinder and see what the difference, if any, is.

Walt...yup uni-syn when idling is the way to go...and yes...you should get the 6 ITBs balanced ...if they are all working properly.
If #5 does not show vacuum at idle...my question would be..how is it supposed to pull in the charge?
Is it possible that the butterfly on #5 is misadjusted?
It might be tightly closed instead of sitting on the stop screw.
Or (happened to me once) the throttle plate shifted because of loose screws on the shaft and that messed up the vacuum.
On an ITB...is it the same basic bypass of air passageway around the throttle plate...controlled by the air bleed screw?
If so...could the passageway be blocked?
Still thinking .
Bob

Andrew - I'll see if I can do that. The software has startup enrichments, which are for set durations. Otherwise, things at idle get adjusted by the engine temperature sensor (which I don't use - it doesn't work on air cooled motors, which either heat up too fast or not fast enough, depending on what temperature you are measuring), and adjusted by the intake air sensor. Which generally seems to mean that after things heat up for a while, idle speed drops some.

I ran the motor while manually operating the #5 injector using the 9V battery and tapping one wire onto one battery terminal. This did perk up the #5. If anything, it seemed easy to add too much fuel, as the perking up tended to take a bit of time after I stopped spraying from the injector.

Next is finding my notes on injector on times from when I had the computer hooked up to the TEC. That is adjustable, both as to a fixed amount and as a percentage.

I think you've got it narrowed to the injector or harness connector that attaches to the injector... perhaps the pins are loose or have light corrosion. Electrical contact cleaner in there might not be a bad idea incase it's an oil layer that's holding the show up. Otherwise I would replace the injector. Never hurts to have a spare or two anyway so if you have one spoil during a race you don't lose the whole weekend.

Walt,

Last weekend I had everything adjusted right through the VE table. At the time the engine was close to normal operating temperature. Last night it would almost would not start with a cold engine, after a while it stay running long enough so I could make it to the throttles and open it to raise the RPM 500 to 1000 RPM, so I'm also need to do some startup enrichment as well so but have not figure out which part of the software and how.

If # 5 injector is the main suspect, you could switch it with # 2 (you may have already done that)

Bob - #5 has vacuum (air flow, anyway) at idle. Just a bit lower than others, and unresponsive to idle air screw adjustment. When I had the ITBs off the manifold I squirted carb cleaner through the passageways for that. Made no difference. It gets enough air (and spraying starter fluid makes it run, so that shows air also).

If the butterflies were culprits for air leak, I'd think that would go away at 2,000 rpm or above. Doesn't.

I'm about to send the ECU off. The only connection not checked is the male on the ECU itself which holds the female on the big plug from which the common 2/5 injector ground comes. Those connectors all look to be the same, fine, etc. And good enough to fire 2. They are basically sealed, and have no oil/grease on them.

I do have contact cleaner, and I guess I could spray that on that part of the big plug. I've already done that on the injector connectors.

And since this is the 3d injector not to work in #5, I'm not suspecting injectors as such. Something is wrong with the ground, and that only happened inside the ECU.

Off to DE the SC and forget this problem for a weekend if I can.

Thanks all

Walt...depending on the circuitry inside the ECU...you could have a heat induced problem.
If a semi conductor is not properly heat-synced...it can run into overload or runaway very quickly.
When it cools down it sort of "heals" itself again.
I don't know the inside of the unit..so this is a general possibility.
I understand that all the injector fire at the same time...but you have stated that it is possible to use this same box as a sequencial injector firing device...which would suggest that each injector is attached to its own trigger inside the box.
So...if the firing device (transistor/SCR/Triac/ or whatever) is faulty on #5 only...no matter what you do outside the box will not fix it....connectors/wiring/etc.
I think we have narrowed it down to the ECU...now just a matter of having a competant tech open it up and give you the answer.
Hopefully, they fix yours...and tell you what they found...rather than just exchanging the box and you never know what the problem was.
Bob

#5
 

Walt,

Have you or can you check the fuel pressure at the #5 injector? All of this sounds to me that the cylinder simply is not getting fuel. You have eliminated problems with compression, leak down and spark to all that is left if fuel. Could something be blocking or partially blocking the fuel flow to the #5 injector? Also, have you checked the electrical signal going to #5 injector? I know that you said #2 and #5 were fed from a common source, but you could have a break or partial break or high resistance in the wire to #5 beyond or at the wiring split. Maybe you could put a test light on the #5 wiring harness plug and let the engine run to see if that lead is firing consistently?

Another idea (thinking out of the box again). Is it possible the cam lobes for #5 are worn through the case hardening and the valve lift has been compromised. Seems to me you have exercised most other symptoms. Could explain the compression anomalies and the reversion cause by an incomplete combustion cycle will affect manifold vacuum. A vacuum gauge connected will have a unstable reading if valve seating or lift is compromised while engine is running.

Well I got the same problem; The engine was not running smooth so I unplug the injectors plug one by one and when I got # 5 no sound or idling change...

Now I got to read this thread again and go through the faultfinding did. As 2 & 5 are fired from the same DFU coil, I'll switch them around.

Well, I think that is where I'm at. Not to hijack the thread but to add on.

Last week I took the first "3 minutes" ride (first one since December 12) to do data logging after finishing installing the TEC/GT with PMO EFI on my 3.2.

It did not sound right and mention to some of my Porsche buddies that if feel like I was running on 5 cylinders but brush it off as it needs tuning. Got back to it this weekend.

So this morning when it was running at a quasi steady idle, I unplug the injector connectors one by one starting with #1 and on, every time the idling would go down but when I got to #5 there was no change when unplug and plugged back. #6 was as the other working 4 cylinders. I also confirm with a mechanical stethoscope that #5 was not "clicking" while the other 5 were.

Next, I took a temperature reading at the headers of each cylinder and # 5 was half the temperature of the others.

Borrowing on Walt faultfinding, I switch the spark plug wire #2 and #5 on the DFU coil C. The problem remain with #5.

After putting them back to their original position, I replace #5 spark plug wire with the CE one that I was using before. No change # 5 is not firing.

Checked and confirmed that the 12 Volts is present at the injector connector (as mentioned in previous post, the ground to energized the injector, is supplied by the ECU).

So to eliminate the injector itself, I removed the spark plug wire on #5 (to avoid ignition) and put #4 injector connector on #5 injector. Listen to it with the stethoscope and it was clicking like there was no tomorrow so the injector is good.

Next, I ohm out the "ground wire for the #5 injector", pin 4 from the 35 pins GRAY Connector to the injector connector and it showed perfect continuity.

As per Richard suggestion at post #47, I checked the tightness of the connector. Pin 4 female receiver is as thigh as the others one. The pin itself on the ECU is OK and lined up with the others in both plane.

So I'm not getting the ground pulse from the ECU for cylinder #5.

Not being sure if a software glitch could do that but I did a new program, loaded it but #5 injectors still not getting the ground pulse.

I’ll contact Richard tomorrow to see if there is anything else I should do.

JJ - misery loves company! Hijack away.
I hadn't thought of attaching the 4 or 6 connector to 5, but I'll try that. However, in my case the #5 injector clicks away when connected to the #5 connector, which increases the bafflement. It seems to click but not squirt. Not the injector itself, as shown by switching injectors around and other tests. Likely not the wiring, as shown by swapping sides with the wiring, and various other tests. Starting to point more strongly to the ECU, despite improbability there.

Thumbdoctor - visual inspection of cam lobes (in situ, but you can still see much of the lobe and watch the rockers rock) shows #5s seem like all the others. I believe that a combination of no oil squirted on a lobe, plus wearing through hardening, leads to very rapid wear. Which makes the oil look gray. My oil looks like fresh oil should look. Plus artificially adding fuel (starter fluid) or "manually" causing the injector to squirt (9V battery and tapping lead to battery) would seem to rule out mechanical issues.

Fred - no convenient way of checking fuel pressure only at the #5 port on the fuel rail. A guy would have to machine a sort of injector like fitting with a pressure fitting to do that. But physically there seems simply no way for pressure to be different there - the fuel system is looped, so both rails on both sides are fed directly from a T fitting, and both come together separately at the regulator. I've physically inspected the fuel rail.

I've also checked the electrical signal using an LED (with a resistor) as a low buck NOID (don't know where that name came from). It flashes like it should. And also by measuring resistances in the wiring, plus swapping sides with all the injector wiring. And, as mentioned, the #5 injector clicks just like the others do.

But still fuel isn't being squirted out of that injector.

Bob - the TEC3r has extra outputs which I don't use. At least one of them is so it works on 8 cylinder automobiles. The others are for use with full sequential. I've toyed with the idea of maybe using one of those to make the #5 fire, but haven't memorized/internalized the manual on all that enough to figure out if that could be done.

This doesn't seem heat related, as the lack of fire is the same with the engine dead cold and just started up as it is with it at least warm from running at idle speed for quite a while as I futz around with this or that scratching my head.

But the box most likely is soon to be sent back to Electromotive for them to check out.

Brandon told me of a similar issue with a carbureted engine, although his problem had been only at idle. One of the butterflies got out of adjustment, so it was not cracked a little when the shaft was on the adjustable stop, but the others were. Thinking what effect having a motor blow up with these ITBs attached might have on the problem, this at least got me thinking, as debris from the exploded #6 was blown up into the air filter, and at least some of it and some of the oil was sucked down the #5 on that engine. It had to pass through the #5 ITB to do so. But a) the #6, which had the most debris flow, seems to work fine, and b) this problem does not seem to go away above idle, and c) unlike the Webers, this model of TWM has a single solid rod shaft holding all three butterflies. Also, the air bleed is supposed to let the motor idle with a fully closed butterfly, I believe.

But I'm going to check butterfly closures. He told me how: crank the idle stop up so that a small drill bit will just pass, say, the #4 butterfly. Then see if it will just pass the other two. As long as you get it at the 90 degree point away from the shaft, it avoids worrying about the fact that feeler gauges are kind of wide.

Walt

And it sure point it that way in my case. Success at last but it does not feel good...

We were able to make contact with the ECU pin while the connector was on and hook it up to a LED light which light up signalling that the ECU was putting out the ground pulse that did not reach the injector. Even though the "grabbing fit" of pin 4 was as good as the other, it was loose enough not to make contact with the ECU pin.

While the Tech Tip does mention that a loose female pin receiver may generate some bad sensor reading, this is a brand new kit so I'm baffle to why that would happened.

Once pin 4 was tightened up, #5 injector start clicking and the idling smooth out. As the plugged connector was move around the engine start loosing rev but the LED on #5 was flashing steadily!!! It only meant that another connection was loose.

So we move the LED light around all injectors and pull the ECU connector in all directions (left, right, up and down) and shake it all around. When we got to inject #3 the LED would stop flashing as the ECU connectors was move around. So pin 3 had also some loose in it. Pin 3 was also tighten up and every injector was running OK while the ECU connector was pull in all direction. It might be coincidental but the two loose pin connection happened to be side by side.

Having work with MIL spec connectors all my life, its the last thing I would expect from a new connector use in a sport car.

Walt,

While your #5 injector is clicking, you might want to scope the pulse and compare it to one that is working.

Hang in there, it will get solve. In the meanwhile enjoy a cold one;


http://forums.pelicanparts.com/uploa...1374545042.jpg

Ecu sent in yet?

With the engine running and 12 volts applied to one side of the injector, ground the other side and see if it floods. WOT usually grounds the injectors momentarily. This may support the bad ECU theory.

Walt, Did you ever figure this one out? If so, what was the cause?

Kind of funny I was thinking the same on Sunday!!!

Mine turn out to be a connector problem at the ECU plug (brand new).

After the not so good experience, I decided that I wanted to be able to monitored every signal anytime anywhere, so I did a "giant" breakout box.

Me and a buddy re-did the whole plug from new parts. Kind of time consuming but the Winter hang around for 4 months here so I got to finish it just in time as the snow disappear :D:):D


Before

http://forums.pelicanparts.com/uploa...1372153565.jpg[/QUOTE]

Borrowing the idea, of being able to "monitor any signal" from Trog, here is the result of my procrastinating almost finished winter project...

Started on the second try, after plugging the 12 volts to the DFUs :rolleyes::rolleyes::rolleyes:

http://forums.pelicanparts.com/uploa...1399510708.jpg

http://forums.pelicanparts.com/uploa...1399510741.jpg

Picture 2 is quite an improvement over picture 1!

I've done something sort of similar going crazy with fuses and screw clamp wire connector blocks, although mine are mostly up front. I had to make up a wiring diagram, as I've forgotten what is what.

My puzzling motor sits in the garage. Maybe it will heal itself?

My latest brainstorm is to put a vacuum gauge on the hole and see if it holds vacuum with the valves shut. I could try that first at the throttle trumpets, I suppose, being careful to find and block normal vacuum lines. Then with the TBs removed. Then with manifolds removed. It was suggested that I try to swap sides with the TBs, as another way of chasing a leak. But that would call for some jiggering, though the idle stops should substitute for any throttle linkage.

I've been busy chasing issues with my SC backup race car. Doing my PCA race officiating thing. Got a leak in the roof to find and fix.

If I find a cure (or at least a cause), it will promptly appear on these pages.