Intermittent No Start

[Samnation]

My 89 944 NA has an intermittent no-start condition. A little back story: A few weeks ago, I had a low battery, so I used a jumper pack to get it started before replacing the battery. I later realized that the jumper outputs about 16V, and that it could potentially damage the DME. It ran fine for a week or so, until one day it ran horribly: cutting out and almost dying at idle as well as higher rpms. I limped it back to the parking lot, but later couldn't get it started again. I checked spark, injector spray, tach bounce, cam timing, S&R sensor resistances, and for major vacuum leaks, but saw no issues.

At some point in all of this, I realized that the fuel pump was running constantly with the ignition to ACC, although I don't remember it doing that before.
I ended up towing it home to the shop where I could do vacuum leak checks and swap out the DME with a spare from an '88 NA. The spare seemed to make no difference other than the fuel pump no longer ran at ACC: no sign of ignition whatsoever. I swapped the original back in and it started right up. It ran pretty good, but stumbled hard if accelerated from idle too quickly. I swapped back to the spare DME and it ran the same.

Since then, I have tested for vacuum leaks with a smoke machine and found nothing major except smoke was coming out of the oil fill port. Is this an issue?
I removed the throttle body to check the TPS, it was glued shut but the ohmmeter checks were correct. I verified these at the DME connection as well.

Just today, it once again wouldn't start. I tried swapping DMEs again as well as a spare DME relay with no difference. Because I was out of things to try, I made a jumper for the relay, and it started up, running like before. Swapping back to the relay and it still runs, but not perfectly, as before.

The only other factor I can think of is that I have had a battery charger on and off to keep it topped up from extensive cranking. Maybe it only runs when the voltage is above a threshold? Before towing it back, there were periods when the battery got below 12V, although the engine still cranked. I didn't think to take note of any correlation between applying the charger and the engine running.

Sorry for the long post, I hope the details help. Any guidance is much appreciated.

-Sam

[Samnation]

Update, I let the battery charge overnight and this morning it measured 13.2V (a bit high?). I tried starting with the relay installed, then installed the jumper wire, without ignition either way (it was starting when I left it last night). I tried swapping the DMEs with no change. It seems that the variables I’m changing have no effect on the intermittent issue. What else could be causing this? I have spark and fuel…

[wwdwgs]

OP,
if it runs fine with the dme jumper, and bad with the dme relay, one may think that the dme relay is faulty. If you have a spare, hopefully, working dme relay, try that. I myself have an "updated" dme relay (from Uro, I must admit), an original one and another spare (original from a donor car). When I disassembled dme relay, it looked line new. Many people on this forum state that the problem is within the solder points on the board. Due to the hot-cold cycles, the solder cracks. I inspected them on my dme relay and found no visible evidence. Nevertheless, I re-soldered them.
Also, sometimes is a good idea to inspect the bottom of the relay/fuse box. Yours, if I'm not mistaken, under the hood. Check for corrosion. I could be wrong with my assumptions.

[walfreyydo]

The two things that come to mind are DME relay or ignition switch, especially ignition switch (you mentioned it stays on ACC and fuel pump keeps running, which is suspect of switch but also DME relay). Start there.

If you look at the link below regarding the DME relay you can see that it gets its power from the ignition switch, and often when ignition switches start to fail, they fail intermittently. This can often mask itself as a DME relay issue, especially when it works sometimes and sometimes not. Additionally, often "spare" DME relays can still be bad because analog relays, especially old ones, are very unreliable and often bad as well. (as a side note I recommend you take a look at the solid-state relays sold by F9Tech as these are much more reliable and highly recommended-I and many others are running them). DME relay is also suspect because you said it runs and starts fine with it jumpered but not great when not jumpered. This could be an issue with the DME relay and whether it is sending power to the pump consistently (bad connection inside the relay).

Also, many folks come on this board and say "I have fuel" but they never measure fuel pressure. This needs to be done so you can assess whether you have an issue with the pump/filter or FPR. You other assumptions such as vacuum leak would not be the cause of a no start issue (rough running yes), and the DME going bad, while it does happen, is usually the last thing to swap after you have diagnosed everything else and is not overly common. It shouldn't be the first assumption in any no-start diagnosis.

Im sure you have found this page, but in case you havent, please read through the entire engine troubleshooting section and start at the top with the checking for spark and checking for fuel steps and go from there (after you have done your DME relay replacement/tests and ignition switch tests):
https://www.clarks-garage.com/shop-manual/ts-01.htm
https://www.clarks-garage.com/shop-manual/fuel-05.htm
https://www.clarks-garage.com/shop-manual/elect-25.htm
https://www.clarks-garage.com/shop-manual/elect-17.htm

The only way the car wont start - if you truly have fuel and ignition - is if you lack compression/timing, and that is rarely the case unless you broke or incorrectly installed a timing belt recently.

Again, I strongly suspect ignition switch or DME relay so go through the testing procedures for ignition switch (above) and repeat the test a number of times, jiggle the switch, try to see if it is intermittent or you can get it to fail the test-or just replace it, along with a solid state DME relay and go from there.

[Samnation]

Originally Posted by wwdwgs

OP,
if it runs fine with the dme jumper, and bad with the dme relay, one may think that the dme relay is faulty. If you have a spare, hopefully, working dme relay, try that. I myself have an "updated" dme relay (from Uro, I must admit), an original one and another spare (original from a donor car). When I disassembled dme relay, it looked line new. Many people on this forum state that the problem is within the solder points on the board. Due to the hot-cold cycles, the solder cracks. I inspected them on my dme relay and found no visible evidence. Nevertheless, I re-soldered them.
Also, sometimes is a good idea to inspect the bottom of the relay/fuse box. Yours, if I'm not mistaken, under the hood. Check for corrosion. I could be wrong with my assumptions.

I don’t believe it can be the relay, as it now won’t run even with the jumper wire in place.

I will check fuse box, I’ve had issues with that before. Any pins in particular I should check?
Thanks

[wwdwgs]

OP,
check all contacts in fuse box. Also, check the ground points.

[walfreyydo]

Originally Posted by Samnation

I don’t believe it can be the relay, as it now won’t run even with the jumper wire in place.

I will check fuse box, I’ve had issues with that before. Any pins in particular I should check?
Thanks

Ignition switch, as I mentioned above. Follow clarks garage engine troubleshooting for a no start condition, starting with identifying lack of spark or lack of fuel (that includes a fuel pressure test) and following the steps from there.

[Monkey Wrench]

I once made the mistake in a volvo of momentarily connecting jumper cables or a charger backwards and immediately heard a pop from the ECU, that fried some component.

it sounded like the contacts of the fuel pump (DME) relay were stuck when you said it was starting in accessory but maybe there is a point in the post above suggesting key switch issues.

the contacts may weld themselves closed due to the arc and then recover only to cuae more issues later. the newer relays with no contact points are a good idea and then they have alreday put soem indicator LEDs in that you can monitor.

occasionally Ive had a downed car ( or machine) and found the contacts in bad shape and as a temporary measure I fold a bit of fine sandpaper double , into a strip about the size of a paper match , put that between the contacts and pull it out a few times.

Ive also tried unsoldering the wire and unhooking the tiny spring to adress them with them more exposed. a paperclip wihth a Ushape at one end of thewire works good to unhook the tiy spring

this can work but Id still just use it for the interim until I had a new one. i have saved a tow that way before. the DME relay does have more contacts..

the solid state ones can fail and cannot be repaired this way, I still dont think they are a bad idea.

one approach I might take is to connect some small bulbs the little grain of wheat ones seem good for that , they have wires. for example you could monitor power to the DME and the key switch positions and fuel pump power in real time as you struggle, that may show some intermittent or somethign like a key wire being hot when it should not I find doing that helps if Im confused as I can keep my eye on power at various points. then just remove the little lamps once you have a solution.

35 year old electronics contain electrolytic capacitors and they do fail with age so you could look at replacing them or having an electonics bench tech do that if it comes down to needing a DME..
I did open the one in my ford van bosche and the same era , I could see some slght staning on the board beneath the electolytic capacitors a sure sign of failure. they can bulge or they can leak the electrolytic grease or PCBs or whatever it is.

as capacitors fail they may do this slowly rather than suddenly and you'd have to be a lot smarter than I am to go down to board level and troubleshoot that. an ESR meter can be used on the board to test them without unsoldering.

I think failing electolytic capacitors is as likely to cause issue as failed solder or what is often coined as "cold solder joints" I dont think you should jump to replacing the MCE as it is pricey and you do not have proof that this is the issue.

true that some do fix issues onboards without much electronics knowledge simply by re-melting soder hopin g to hit upo a bad joint and that can be successful. It is also possible to go in there and cause damage.

it is not uncommon for other eelctronic boards, fo rpeople to jsut go about carefully replacing all the electrolytic caps and that could have a chance of being sucessful but there is no guarante that you wont get into ttroubble and need a new board.
I noted that the maximum working voltage rating for thee electolytic caps in my 88 ford van which is also bosche was 16 V
electrolytic caps are polarity sensitive the stripe on the side of the can is negative, if you ever unsolder one make sure to take note of it's polarity first.

Id perhaps do that as a last ditch effort if I know a board is bad and im at the pont where there is nothing left to loose.

Someone taking the part in as a core for reconditioning may not like to see that it was mucked with.
E-proms or chips can be damaged by electrostatic shock so its common to wear a grounded wrist strap so you don't shock them that way accidentally.

Do not let me stray you from the post above which was very good guidance in the process of systemtic troubleshooting.

[Monkey Wrench]

id take tr ground post off the battery, make sure y9u have good pics then go ahead and pul them all , if you can see screws holding the fuseholder assy maybe you can pull it away a little, then clean each fuse contact and the corrresponding spade terminals.

if you have trouble wiht one blowing , it can cause heat and that can sometimes remove spring tension. if you heat brass and cool it quickly that will alter the springiness of the metal strip.

less springiness, means less tension..against the fuse.

this in turn can cause further issues with bad contact against the fuses.. If you ever feel them hot, do not put a wet cloth to cool them.

similar for spade terminals on wires.. if they feel loosew try to squeeze them up a tad without crushing them.



id put a dab of electrolytic grease on each once its nice and clean to slow further oxidation. new fuses are cheap.

fuses have some resistance, thats how they work. when there is a lot of current they blow, so by measuring the voltage at either side you can detct a voltage drop across the fuse, that proves the circuit is conducting current.

there are some cheat sheets around, youll find them if you just search, they can help determine the amps that are being drawn as it is relative to the voltage drop across the fuse.

all fiuses are not alike so make sure if you do that that the reference sheet and the fuse type corresponds.

example you woder if your pump is running , so you find the right fuse for the pump. check voltage at either side. when the pump is running there would be a voltage drop across the fuse.
with the fuse installed

measure with the same ground reference, the voltage at either side, if it is the exact same it is not conducting much current. you know that because you cant detect a voltage drop across the fuse.

if you start finding warm fuses or blackening you are probably looking at an issue, that maight be a result and not a cause. if any fuses have less tension , maybe someone bent it, or maybe the little prong has lost tension due to it being heated and cooled rapidly at some point in time.

to restore tension you can heat and cool slowly changing it back to being springy metal..

the fusholder may also have some plastic parts so you don't want to melt it then.

when you plug in a relay, it is difficult to feel if one prong has a loose spade terminal as the block has several pins.

so get a little piece of wire with only one flat spade push that in each of the holes that the relay uses, now with only one wire , note if each has a good squeeze..

If one becomes a loose fit it may be a sign it was overheated and youll wanto correct it so when you push the block in ALL the spades have a good grab.

ok so you could check this on the plug where the fuel pump relay sits..


ok another situation you could find, lets say the negative battery to body is weak, it has oxidised, it is still making contact but now there is resistance..
similar if the fuel pump has a bad ground to body. similar if you have a bad contact at the fuel pump plug in.


now what you have is TWO voltage drops in series. the pump is no longer seeing 12 V , thi sis because two loads in series each have a voltage drop, the bad ground is sharing the voltage and you have a voltage drop at the bad ground' now you have a pump that is designed to see 12 V but it is actually only seeing perhaps 8 volts, the other 4 volts are lost due to heating a bad contact.

it may still run but things arent; right. maybe you'd detect fuel pressure is lower than normal.

the load on the fuse is reduced, why because the two loads are causing higher circuit resistance

the circuit starts and ends at the battery that battery to ground is the part of EVERY circuit.

if you have a circuit with two loads, where there should only be one load the voltage is divided and the two voltage drops will add up to 12 volts. in this instance you have more resistance than normal and because the resistance is higher you now have LESS current flow than normal through the circuit. in this scenario the pump fuse is seeing less current than normal and not not as much voltage drop. ( less than 12 volts available for the pump)

if you put one long wire on your meter youcan connect that to either the hot or ground terminal of the battery in this way you can check the voltage at verious points with reference to your battery, and learn from doing so.


example you find resistnce between a wire that should be grounded and the negative terminal. then WHY?

another example , you have 12 volts supplyng currnt to the pump you check the hot lead to the battery negative and find 12 V, you check between tha same point and the body and find less, well why? because its not grounded very well.

bad grounds can hide because they dont fail sudenly they may increse in resistance even for years, until a fail point.

that time in between is called "confusion" ;-)