1991 SDV Ignition Timing

[Matt Innocenzi]

Hello all, I have a new challenge with a much newer Cadillac than I am accustomed to.  I have a 1991 Sedan DeVille with the 4.9.  The previous owner put in a new distributor and ever since then, never got it running right.  It throws codes E52 and E53 which I figured out are trouble codes with between the PCM and ECM and distributor signal interrupt, respectively.  I learned (the hard way) these cars are timed with the computer.  If my research is correct, you have to "teach" the computer where the distributor is at 10 degrees BTDC while in Set Timing Mode on the computer.  Despite my attempts to try and "train" the computer where the distributor is, I get rough idle, stalling, misfiring, and complete poor performance. 

I am old school and 1991 is NEW in my book.  I am used to setting timing, dwell, and carburetors the old fashioned way.  Can anyone tell me if I am on the right track with this set timing mode and a briefing on how to properly do it? 

Give me a '55-'66 Cadillac any day over this! 

Thank you in advance!
Matt

[TJ Hopland]

I don't have a 91 manual but would expect a 91 to be the same, yes 4.5 vs 4.9 but both sequential multi port.

90 manual says code 52 is "ECM memory reset" caused by a power loss.  This would be normal if the battery was disconnected and before engine has been started but says if it show up other times you should check for a power issue.

53 is the distributor signal loss for more than 4 seconds.  The signal its looking for for this code is that extra pickup you can see under the rotor.  If it doesn't have that signal it runs the fuel in batch mode rather than sequential injection.

Timing setting procedure is put a jumper wire in the diagnostic connector between A and B and not be in diagnostic mode on the climate control.  Start engine and set to spec given on the label.   The trick is the jumper,  that tells the computer you are setting the timing so it won't try to adjust it.

On the 90 the timing indicator the points are 0-4-8-12-16-20.  valleys are 2-6-10-14-18 

[Matt Innocenzi]

TJ, thank you very much, this was extremely helpful.  Code 52 makes sense, as the battery died and I had to remove it to give it a charge.  Code 53 is throwing me for a loop.  The sensor you are referring to udner the rotor - does this sensor basically replace the old vacuum advance and tell the computer how to advance and retard the timing?  It is a brand new distributor, so I would not expect it to be bad.  However, the code makes sense because when the code comes up, it runs very sluggish and pours black smoke out the exhaust likely for running in batch mode. 

I tried the jumper before and setting the timing to 10 degrees BTDC, but when I did it I had the car in diagnositc mode on the climate control.  You are suggesting use the jumper but NOT in diagnostic mode, correct?

Sometimes the car runs like brand new and then, out of the blue, it just falls apart and can't stay running.  Whatever the problem is, it is completely intermittent.

[TJ Hopland]

They 90 manual does specifically say to not be in diagnostic mode when placing the jumper and setting the timing.

There is nothing that physically moves to adjust the timing.  Its all done by the computer which is why the computer has to be in the correct mode and you have to set it physically to where the computer is expecting it.  If its not all how its expecting it to be it gets confused and will tend to hunt for the correct setting which it will never find.  The sensor is just there to give the computer a specific cylinder reference.  Without it all it knows is there is A cylinder at TDC.  The sensor lets it know a a specific cylinder and knowing one of them will let it know where the rest are.

You can see the sensor under the rotor.   Its just got a slot in it and there is a tab that hangs down and passes through the slot.   When I got my 90 I could see the sensor was cracked and I could see marks on the rotor base where it had been hitting but it was still working.  At some point it quit so I had to replace it.  Sensor on its own was harder to find and not cheap.   

I ended up getting a complete new distributor for less money than just the sensor, it even came with the coil.  Price of the complete assembly was about what any of the major components cost and there are several of those in there so I was really suspect of the quality.  Cap and rotor were complete garbage and didn't last more than a couple hundred miles before it was running like crap.  When I was changing the cap I noticed the coil smelled funny so I ended up installing a different coil too.   Seems to be doing OK now but I don't have a lot of confidence in the quality.  Any more issues that is for sure the first place I will be looking.  I still have my original so if needed I can maybe transfer parts between them and make something that still function.   

I don't remember the details now but there was something wrong with the wiring harness on mine.  I can't remember which wire or what the issue was but I know I had to scavenge a connector from some junk and splice some wires.   I will try and go look at it and jog my memory later.   I remember thinking it was a pretty dumb defect so maybe yours was the same brand and they all had that issue?

[Matt Innocenzi]

TJ, thank you so much.  I will give it shot trying to set the timing without it being in diagnostic mode and see if that works.  It all makes sense logically to "show" the computer where #1 cylinder is and it will take it from there. 

Given your experience, I am also suspect of the quality.  Like you, it was cheaper and more logical to replace the entire distributor.  I still have the old one, which completely quit and would not give any spark.  I will take a close look at that sensor below the rotor and see if there is any obvious signs of damage. 

If you happen to recall what your issue was with the wires, please let me know. 

Thank you again for all your help.

Matt

[TJ Hopland]

Not sending any spark would have to be the coil which is really easy to swap and a completely dead one would also fail the basic ohm test or the main module which is also mounted below the rotor.   Historically speaking the module would be the most common sudden failure. 

I wonder if you could end up with the plug wires one off?   Historically it doesn't matter but since this is kinda an oddball setup where the distributor does know which is which I wonder if there could be something there.    I will try and go look at mine after lunch.   

[TJ Hopland]

Looking at mine I think it was just a badly damaged during installation pin in that 6 pin connector.  I think it just would not plug in but it was in the correct place so if it had not been damaged it would have been alright.

Here is how mine is oriented and where the numbers landed.   Is yours similar?

[TJ Hopland]

If you pull the cap and align the timing marks it looks like the rotor should either be pointing to that gap between the bolt and that cable hold down nub on the air filter housing or 180 off which would be #6.

[Matt Innocenzi]

TJ, I thought the same thing without sending spark.  I put in a brand new coil and still nothing.  It is as if the old distributor had a pick pick-up sensor such that it was not sending the correct signal to the coil to fire. 

The orientation of my distributor looks exactly like yours with the latch part of the distributor cap at 4:00 between the air breather and the valve cover. 

I am going to try the process of setting the timing from scratch tomorrow.  I will put a jumper between A and B on the ALDL without it in diagnostic mode and report back.

[TJ Hopland]

There is still a pickup coil in there.  Its the same as any other HEI its just it doesn't move via a vacuum.   Its got 2 wires that attach to the ignition module which is under the rotor.   The pickup coil has the 8 points on it so one signal every 45* of cam / distributor rotation or 90* of crank. 

That that signal goes into the ignition module which then is presumably translated into some sort of data stream and sent to the main computer.  Main computer then based on operating conditions sends a an offset signal back to the module as to what the timing should be.   Using that info or a default value if that data is missing electronically switches the power on and off to the coil which creates the spark which theoretically and hopefully will happen when the rotor is pretty close to the desired terminal in the cap. 

The extra sensor in these which came with sequential port injection doesn't interact with the module, its got a connection directly to  the main computer.   Without it all the computer knows is there is a cylinder at TDC.  It doesn't know which one or if its exhaust or compression.   On a distributor less system this sensor would be the cam position sensor. 

Difference between this and the original HEI is the modules on the originals didn't do any timing.   They just took the pulses from the pickup coil and did some quick processing to optimize the dwell and current limiting based on rpm and then pulsed the signal to the coil.   The timing was still done mechanically either by moving the pickup coil with vacuum or the weights and springs that moved both the rotor and reluctor.   The reluctor is the thing with the points on the distributor shaft inside the pickup coil. 

The electronic timing distributors the pickup coil is fixed as is the rotor and reluctor.  Computer just knows what the fixed relationships are and comes up with an offset to get the correct timing.  This is why its critical to set them to factory specs in the proper modes.   The computer has no control or knowledge of the cap and rotor's relationship so that is up to you to follow the instructions.

The twist to this system I had not thought about before is I bet the overall orientation of the distributor and what terminal the wires are on is critical because the computer is using the distributor as the reference to know what cylinder is what for the injectors.    In my photo #1 is at what like 3 o clock?   So that is when #1 injector will be firing.   What would happen if we move all the wires over one terminal and rotate the whole distributor the same distance?   Spark should be fine but now or injectors will be out of time.

Most engines don't have this possible problem because they use a crank trigger and a cam sensor to know the position of the engine.  If they even have a distributor all its there for is to route the spark.  These were unique in that the distributor was doing everything.     

[TJ Hopland]

Just to be sure my car and photo was correct I just checked the manual and it does say that the big nub should be at approximately 45* angle to the rest of the engine.    It also confirms that #1 is on the top of the nub and #2 is below it so my photo and numbers on the cap are correct.   The really smudged number is 5. 

You should verify that yours is that way too.   I could see if the spark plugs and injectors were out of sync how you could get smoke and a very poor running condition.  The injector would be spraying the fuel on a closed valve when there was no air flowing.  Sure that was kinda how they used to do it but when the computer and rest of the design is expecting the valve to be open that could explain why it runs like crap.   

If you had 8-1 or 2-7 on either side of the nub it may be possible to get the spark in time but the injectors would be off by 45/90 degrees.    If it wasn't a sequential system it would not matter where the wires were as long as they were in order and in time but with this setup its likely pretty critical they are in the correct spot.

[Matt Innocenzi]

TJ, you just made a light bulb go on.  When I replaced the distributor, I marked the postion of the rotor, but I must have been off a tooth, as I ran out of travel on the cap to set it at 10 degrees BTDC.  So, I moved all the wires one position to allow for the travel to get the car in time.  I never thought that doing so would get the ignition in timing, but could offset the fuel timing!  I will reconfigure my wires to look like yours, re-time and report back tomorrow.  Your point makes complete sense if the computer thinks it is firing on cylinder 1, but the distributor is in the position for number 8 in the firing order.  The spark is right, but the injector is off.  Brilliant point!

[Matt Innocenzi]

TJ, I removed the distributor and positioned it such that the nub of the distributor cap was at 45 degrees and cylinders 1 and 2 on the top and bottom of the nub, respectively.  No change in performance.  In fact, it is probably a little worse.  Heavy smoke from the exhaust and it smells of fuel.  Still misfiring and any attempt to give it throttle just bogs it down to the point where it won't stay running.  I put the jumper between A and B terminals in the ALDL without diagnostic mode on and it makes no difference, like I cannot teach the computer where the 10 BTDC is.  Any other ideas?

[gkhashem]

How many miles on this car?

[TJ Hopland]

If it was me I would maybe try and ohm out the injectors to see if there was any that seemed out of place.   If its like the 4.5 you can't get most of the connectors off at the injectors because the rail is in the way but if you follow both harnesses to the drivers side they end up at a big connector there and you can probe around in that connector and find the injectors.

I would also be trying to get a hold of a scan tool that will read live data and just look at the data the computer is seeing.  I would be looking for something unusual or out of place.  When I bought mine it would usually start and run sorta ok when cold but after a couple minutes was very hard to keep running.   

I ended up finding a used scan tool and started looking at the readings and after a few minutes of just key on several of the readings like manifold pressure and coolant would suddenly just go pretty extreme,  not far enough to trip the light but far enough from reality that the engine just could not stay running.   Most of the sensors in question were on the same harness in the same general area so at first I was looking at connections and grounds and for physical damage. 

Many of the sensors have basic voltage and ohm tests and I tested them back to the connector on the main computer so I was pretty sure it wasn't a sensor or wire issue.  Turns out I actually had a bad computer.   In all my testing I also found out I had one injector that was almost shorted out too so went to ebay for an injector and a couple computers.

First time in my life the computer was actually to blame but it took a fair amount of troubleshooting to get to that point.  It was all still a gamble because I just bought used ones online but they (like the whole car) was priced low enough it was worth the risk and it worked out for me.   If I had not found that one bad injector what would that have done to the new computer?  Was that bad injector stressing and maybe caused some area of the computer to overheat?   What if I didn't carefully check the sensors and harness?  What if I got a dud used computer?  You just have to do all the testing you can and calculate the risks when it comes to replacing stuff.

My scanner is an OTC Genisys.  I got lucky and found one locally for $350 and it had the cables I needed.  I see now looks like there are several to choose from for $400 and some for a lot more.  Most of those don't go past 2004.  I think there were upgrades but they cost vs getting something new a lot of shops just got something newer rather than sink money into an old tool and you have to wonder how often most shops ever see something pre 96.   I think all the new tools go back to 96.    If I remember later I will go out and look at which cables you need for OBD1 GM so if you are looking at one you can ask or be sure you know where you can get the proper cable. 

[Matt Innocenzi]

George, it has 139k miles.

TJ, I am leaning towards a computer problem, reason being I also have code F32 which is a loss of communication between the ECM and BCM.  As a side note, I also have no blower motor for the heater, defroster, or A/C.  The relays and fuses are good, it just won't come on.  It is my understanding the blower motor is controlled by the BCM, so I am having serious thoughts the computer(s) are not working.

[TJ Hopland]

Did this car ever work right for you or did you get it this way?

Most common reason the blower doesn't work is the blower control module has failed.    Its kinda on the firewall on a ledge just forward of the cowl.  There may be a plastic cover over the whole area so you can't see it.   Its just to the left (standing in front of the car) of the blower which is kinda in the middle.  3 screws.  2 plugs.  1 plug just has one fat wire in it and the other plug has 3 wires one is fat.   If you make a jumper between the fat wires blower should run.    That at least tells you you have power and a working blower.    Shop manual has a test where I think it says to put a 20 amp fuse as a jumper and if it runs for something like 3 minutes without blowing its fine.   There is also a test to see if the command is coming to the module but I would have to look that up. 

Engine should run without the other computers but maybe the issues are not letting it give you proper diagnostics?  Computer isn't too bad to swap, its kinda under the glove box.   Looks impossible to access but then you figure out its only held with one screw and then kinda slides out and the cables that are not long you can access while its still in place and the others are long enough you can slide it out then get to them.

It appears that something in the eprom has specific programming in it to talk to the rest of the car.   Both my used computers came with proms in them.   I think both ran the car with my original prom.  One ran the engine apparently OK with its prom but nothing else worked,  I had no climate control,  no trip data, and no digital dash.  I don't remember what they did but I know what ever it was wasn't useful in any way.   The 3rd prom would not run the engine it just kinda sputtered like it wanted to start.   I don't think that one did anything with the dash either.

I think the one that didn't work at all supposedly came from an identical year and model to mine, even supposedly was a coupe and had a digital dash.   The other one I just took a chance on because it was the correct model number computer, seems like the ad said it came out of an 92 Allante.  I think that is the one I have been running for a couple years now. 

Its all a crap shoot but at least when I was messing with mine they were not too hard to find and were pretty cheap.   One I think was just over $50 with shipping and the other I think was closer to $25. 

[Matt Innocenzi]

TJ, thank you for the advice.  I got the car this way, several other "mechanics" got their knuckles in it, so it is hard to figure out what they "fixed" versus what they screwed up more.  I fooled around with it for about a year, going over and redoing almost everything the last people did and checking everything twice.  I thought it was something easy, but I am rapidly sinking time and money into it and not getting it fixed.  I did check all of the connections in the computers behind the glovebox, nothing obviously wrong, but still running rough, lots of smoke pouring out of the exhaust and running on about 6 cylinders.  I didn't realize the computers were that cheap, it may be worth a shot.

[TJ Hopland]

I wonder if the computer has already been changed and maybe you don't have the correct prom in there?   I think as long as the colors of the connectors match up you have a compatible computer but the prom is a much more complex mess to figure out.  I think back in the dealer days they had charts or someplace to call to find out what was correct or even have a correct one made up but not days I doubt any such chart or tech officially exists. 

The prom or 'chip' is what has all the specific info for your specific car.  Its got basic stuff like the hardware config of the engine as well as a base tune but its also got what I think amounts to channel assignments to talk to other computers and systems in the car.   The prom is literally unique to your specific car.  An identical same year and model may not be close enough and a similar model our year could be way off.   

May be worth a gamble on ebay for a couple more computers just to see if you can get lucky and get closer to something that works.   Look at several listings for just 'used' ones,  not rebuilt.   Rebuilt won't come with any prom and usually cost more.   Look for listings that actually show the actual unit for sale.  Sometimes there will be writing on them saying what car they came from that may differ slightly from how they are listed.   You can also pull yours out and search using its number.   See what turns up,  if your number comes up with things that fit a different year or model than yours that may be a clue there is an issue there.