2seater

Thank you for sharing the fix with us. The sort of funny thing about the ECM failures is it doesn't really seem to matter what the miles are, just the age of the units?

It does sound like a good clue it may be heat inside the ECM or very slight chance the wiring/connectors are now in a different position. The connection is pretty tight, as I am sure you know, so I would think that is unlikely.

You can operate the car with the ECM laying on the floor, which will allow it stay cooler. I have seen sites that describe modifications to the ECM's to operate low impedance injectors, but can't find the link <img src="http://forums.aaca.org/images/graemlins/frown.gif" alt="" /> It sounds like you know about the workings of the electronics, so maybe something will be apparent? Bad solder joint or somehing physical? ECM's aren't terribly expensive and Jim Finn could probably help you with a good takeout. I would bet on the ECM at this point also.

You probably know more about how the actual circuit works than I do <img src="http://forums.aaca.org/images/graemlins/grin.gif" alt="" /> The impedance sounds right, used in what they call a saturated circuit. Current draw about 1 amp or so to operate the injector. Good enough response time for most street engines. Low impedance in the 2-4 ohm range are used also, useually in the high performance/high flow types, although not a universal rule there. Quicker response but a lot higher current draw and as far as I know, the hot side is switched using different drivers in the ECM.

No event would cause one injector to stop operating. Constant B+ to all injectors with switched individual grounds in the ECM. Only wiring, harness connection or the ECM itself would cause loss of injector pulse. The wiring is nothing special. You could run an external bypass wire from the #1 ECM connection around to the injector itself. There are a multitude of grounds used in the ECM and I imagine you could simply probe the correct connection at the ECM to see if the ground is turning on and off. The injector and other harness connection is at the right rear of the engine, sort of behind the power steering pump. If it is okay cold but acts up as it warms up, it does sound like the ECM is getting warm enough to have the injector driver fail. The other point Padgett made about the connection and harness being disturbed in an engine change is certainly something to look at too. It shouldn't happen but the engine comes out toward that side of the car so something could get pinched or crunched.

It shouldn't quit or stumble when you tap on it. Generally it will quit if you disconnect it while running but will start and run with it disconnected. It will usually quit or stumble if connected while the engine is running also. It sounds like it is ill and no amount of cleaning will help.

Are you certain the new injector is actually working? Even without a noid light, you should be able to hear a steady tapping of the injector with almost any rigid material held to your ear. It would be unusual for it to stick part time, they work or they don't (generally). I have had cleaned injectors be stuck from sitting around. That shouldn't be a problem if it was brand new, but there are lots of remanufactured injectors in the pipeline. If you suspect a problem with the injector wiring or ECM signal, it may be beneficial to swap harness connections for adjacent injectors if the harness will allow it? It may not run as smoothly since the injector timing will be off, but it would give an idea if the problem follows the wiring. I have heard of injector wiring connected incorrectly causing a rough idle.

I believe the '87 system is slightly different, but I know I have heard of situations where the crank sensor can be temperature sensitive. I have heard of people pouring water over it to cool it to see if it makes a difference. It is generally true the sensor either works or it doesn't, but my guess is the single ring sensor possibly fitted to the '87 may act somewhat differently? If the requirements are the same for the '87, only the crank sensor and ICM work to create the initial spark to start the engine, and of course the electrical feed to them. I think part of the problem with lack of help is the slightly different system used for the the '87 system. Very similar and yet different.

I understand the urge to "stir your own", and it certainly gives you the control you desire. I don't know if it would make it easier, or more difficult, to tune the hybrid engine you have?

None of the issues you mention sound unusual. The torque converter clutch will unlock when you lift off the throttle, it is not shifting out of overdrive (4th gear). This is normal operation. The ECM sees the throttle is closed via the signal from the TPS and assumes you are slowing or stopping so the the engine doesn't stall as you come to a stop, analogous to coming to a stop with the clutch engaged on a standard transmission. The feel of the transmission shifting down to the next lower gear does vary some from car to car, but it usually is noticeable, particularly the shift to first gear around 14 mph. This may be related to the harshness of the reverse engagement also. The primary external item that affects shift feel is the vacuum modulator. It senses engine vacuum to determine the approximate load on the engine and uses this signal to vary the line pressure inside the transaxle, reducing it under high vacuum to soften and slow the shift, and raising it under low vacuum, like when accelerating, to speed the shift and firm up the clutch engagement to better absorb and handle the power applied. A defective modulator or a vacuum leak will make the shifts harder than normal. As for the reverse engagement specifically, I have seen articles about a modification to the servo to modify the engagement, but I have found that shifting into neutral first and then to reverse helps smooth the engagement sometimes.

Hard to beat a good moderm auto trans. Power capacity is probably somewhat of a problem in this particular case, but I know the GN guys have experimented with stick trannies and generally have gone slower as a result. These are not high revving and peaky engines. Their strength is a flat torque curve and a boosted engine just enhances that even more than the horsepower. I imagine you have seen the guy on the B'ville forum with the turbo install and a 4T80 transaxle?

I know you have been checking for vacuum leaks but have you checked around the PCV valve at the right rear of the intake manifold? Padgetts explanation is spot on and trying a reset to make the ECM start over is a good idea. In a stock system 150 is about the limit of how far the BLM can adjust, and while the number difference from the desired 128 looks large, it is really only a few percent on the actual fuel delivery. Two other things can tell the the O2 sensor the engine is lean when it really isn't is a crack in the exhaust manifold, especially the rear one, or a misfire. You would think the cylinder that doesn't fire would read rich but that isn't really true, the oxygen isn't consumed even though there is excess fuel present. My personal experience is that the O2 sensor can get too cool after an extended idle and it will "go out", particularly if the engine is running rich (the opposite of what the BLM is saying). If you bring the idle up to about 1500 rpm for 30 seconds or so should get the cross counts back. One other thing is I have not had very good luck with Bosch O2 sensors. It isn't that they don't work, but they just seem lazier than a Delco or even a Neihoff or Standard brand? They may even be made by the same companies, but that has been my experience. No joy with Bosch platinum plugs either. I don't know why, but perhaps it has to do with the waste spark system we use that fires half of the plugs with reverse polarity?

The knock signal is a number that increases over time if knock is, or has been, present. Essentially it keeps a record of the knock counts over time and will increment upwards until it resets, or rolls over to start at the beginning again. It is not unusual to get some knock counts on start up so it is normal for the counts to increase when you take a look at it from one day to the next. It is not normal for it to increase with the engine idling in gear. Substantial knock counts will trigger the ECM to retard timing, reducing engine performance. In many cases you can cause the engine to stumble by giving the engine a sharp rap with a hammer or even a large wrench while idling which will trigger the knock sensor. If severe enough it will retard timing momentarily. The easiest would be to drive the car and monitor a few sensors to see what might be happening. Knock counts, spark retard or the spark timing can be monitored to see if there is something going on when the engine acts up. If nothing turns up, engine knock is probably not the problem and the solution is elsewhere.

Looks like a heater hose steel line connection as F14 said. It is in a somewhat different location on the manifold as compared to the LN3. It looks like there is a tapped hole right next to it for the retainer on the o-ringed steel pipe.

"Had nothing to do on a Saturday morning and threw it together with fiberglass and polester resin. Uses stock mounting points." Uhh, pretty nice work for "throwing it together" 2seater; you probably could get it into your driveway. I suspect that I could. It requires a leap of faith...you stab the gas just before you think it will drag and that lifts the front. There are some serious rub marks on the front lip. Yep, I understand the technique, just not too good at the timing <img src="http://forums.aaca.org/images/graemlins/blush.gif" alt="" />

It seems to help air flow through the radiator although I never really had a temperature problem. It was more of an experiment than anything else. I just made spacers with longer screws. Air dams and such only help when moving, so urban driving in traffic doesn't really do much. The water manometer test I mentioned will show you what effect speed, wind direction and following other cars does to air flow. That chin spoiler in the other post looks decent from the front although I amm curious how it looks from the side. I wouldn't be able to get the car in my driveway with that installed, but it is probably pretty effective.

There are some 3800's that use two sensors, one for the guage (or idiot light) and one for the fans, but the Reatta does not. One sensor is used by the ECM to perform multiple functions. Manuals that cover multiple cars lines can lead to this type of confusion.

I you want to measure the pressure differential from the front to the rear of the radiator a simple water manometer hung inside the car works pretty well and is interesting to see what happens. I lowered the piece referred to above by 1/2" and it did make a difference in the pressure differential from front to rear of the radiator, which will help flow. The piece does serve a mechanical purpose and does serve a cosmetic purpose as well.

What sensora and ECM are you using to run the new engine? The EGR should be a different style but the rest of the stuff is very similar. I would guess F14 might be able to help out. The MAF is a different style but the flow range output is the same and should plug in and read fine with the stock ECM. The ECM and BCM interact to display the data, and this sounds like what Padgett describes as the data stream being incorrect?

Post deleted by 2seater

Only way to tell on brightness is to replace just one and see what it looks like relative to what you have. Personally I use Silverstars in my vehicles with composite headlights, the type that are part of the body design, and generally like them. Where I have whole assembly that can be changed, I prefer to use a Euro style beam shape. They have a very sharp cutoff to the light pattern. Right now I have Hella brand with standard wattage bulbs in them, but have used Carillo's and Cibie's in the past. All work well. The one substantial downside to the flat cutoff is rolling terrain at night. When on the downside of a hill, there is essentially no illumination of the road far ahead until the car flattens out at the bottom of the hill. It is a little disconcerting. Operating on high beams helps this, but oncoming traffic may prevent that.

Make sure you adjust on high beam only. The low beam will take care of itself automatically. The eyeball method at 20-25ft from the wall works fine. Just place the center of the high beam spot directly in front of the headlight and sligtly down from centered on the point on the wall, maybe 2" or so. This can be adjusted a little to your taste. The low beam should move down and to the right when switched from high to low. A good euro beam type of headlight with a very sharp horizontal cutoff can be adjusted slightly higher without annoying oncoming drivers.

No problem with standing your ground <img src="http://forums.aaca.org/images/graemlins/grin.gif" alt="" /> I have done more than one, and prefer the hole saw, but I never tried it under the car. I can't imagine that would be a lot of fun to do, and my hat is off to any that do it that way. The O2 sensor only has a few threads on it anyway, so the remainder of the bung that isn't used is just in the way. You are right that there is no way to get rid of everything, it's just the way it has to be. Any improvement is good.

Hole saw is best, die grinder is acceptable. The O2 bung hangs into the opening, same a little bit, some a lot. The O2 sensor seals like a gasket type spark and only needs a few threads for installation. If you do use the grinder, you can smooth and taper the O2 bung for better flow. A canoe sort of shape is best.

I think that is Dave Buckshaws work? I spoke with him a couple of times several years ago when I saw the photo you posted. I would bet that installation was done 10+ years ago or so. I don't remember what model car it was in, but I was primarily interested in the ECM he used to control everything. If I remember correctly, it was an early '90's ECM that would accept the programming for the early supercharged engines and higher flow reading capability. He did his own programming too. He was familiar with the Reatta and the limitations of the air flow reading capability, a problem that exists today. That is definitely not one of the factory turbo prototypes. I think he is the same guy that pushed a stock 3800 LN3 block on a dyno to see how much power they could produce before destructing. It was about 500 hp before the bottom end failed. He was looking for viable replacement blocks for the GN type of application. He is an interesting and knowlegeable guy to talk with if you can get in touch with him. He is into all sorts of things and he may (likely) have moved on to other things. I think I have seen his name pop up in other areas and forums.