Cold Start: Runs for 20 seconds, then dies. Cause?

[maxm50]

Hi all,
I'm bringing an '81 Spider that I just bought up to daily driver running condition, and I'd like some help in diagnosing a cold start issue:

On a cold start, the engine fires up immediately with no throttle input. The car has a brand new starter and a starter relay installed - it honestly fires up faster than my 2011 daily driver car. RPMs hover around 1100-1200 for the first 20 seconds. Then, the engine RPMs drop, stumble, and it dies. I've done this many times and it always happens around the 20 second mark after startup.

If I give it throttle input within the first 20sec to raise the RPMs to ~1500-2000 range and HOLD it there until the temp gauge crosses 120, I can then release the throttle and it will hold an idle of around 900 RPM. It does not idle terribly well - it has some sputtering and coughing. But it never dies while warm. If I turn it off and start again when the engine is warm, it goes back to it's 900RPM somewhat sputtery idle without dying.

I'm working my way through Brad's FI tuning guide. So far I have set the idle, set the throttle position sensor, cleaned and verified AAV operation, checked the coolant temp sensor resistance at room temp and while hot (both are in spec). However, I was only able to check this at the coolant temp sensor itself, not the ECU. This car has a "Factory Aire" AC system installed in the early 90s that is located under the glovebox. It is either blocking the ECU or the ECU has been relocated somewhere that I have yet to find.

My best guess is that the problem is NOT the cold start injector system, because Brad's guide says that only runs for 1~8 seconds, and my engine starts up fine for about 20 seconds. Does anyone know of a sensor / algorithm / whatever that kicks in about 20 seconds after startup? For instance, is this when the ECU starts reading coolant temp and compensating fueling accordingly (perhaps I have a break in the sensor line to the ECU)?

Any help would be appreciated! I plan to rip out this awful AC system and locate the ECU, and continue through Brad's guide (along with hunting for any vacuum leaks) but would love to hear from anyone that's had a similar problem in the meantime.

Cheers,
Max

[RRoller123]

Could be many things, but an easy one is to check the tightness of the Intake manifold nuts/bolts/studs, whatever you have there. Adjust the TPS switch at the throttle plate (the black plastic housing) so that it clicks just as the throttle is applied. Could be so many things.

Good FI manuals:

FUEL INJECTION DEBUGGING
http://www.hiperformancestore.com/Ljetronic.htm
http://www.njfiats.org/joomla/images/st ... BoschL.pdf

FUEL INJECTOR CLEANING:
WWW.OKINJECTORS.COM
www.witchhunter.com

FUEL INJECTION MANUAL
http://www.mirafiori.com/faq/fiatFI_may2002.pdf

FUEL INJECTORS CHART
http://users.erols.com/srweiss/tableifc.htm

L-JETRONIC MANUAL
http://www.hiperformancestore.com/Ljetronic.htm

L-JETRONIC
http://mastertechmag.com/pdf/1988/11nov ... ronic2.pdf
http://jenniskens.livedsl.nl/Technical/ ... tronic.pdf

L-JETRONIC FI MANUAL (FROM VW)
https://www.type4.org/manuals/ljet/index.html

BOSCH L-JETRONIC MANUAL – BEST ONE:
http://foxed.ca/rx7manual/manuals/BOSCH ... Manual.pdf

http://dtec.net.au/Tech%20Articles.htm

[maxm50]

What a treasure trove of info - thank you!!

I did find a vacuum leak this evening by spraying carb cleaner at every vacuum hose or leak location I could find. All hoses were OK, idle speed screw was OK, but the throttle shaft (on the spring side, not the TPS side) seems to be leaking air. The idle raised noticeably and even smoothed out a bit when I sprayed carb cleaner on this shaft right where it enters the throttle body. Funny enough, I had the exact same leak on Keihin carbs on my '71 Honda motorcycle... those carbs used felt seals soaked in grease to form an air seal around the shaft.

What type of sealing does the FI throttle body shaft use? Is it an O-ring, felt seal, or none of the above? Any advice on fixing a leak in this location?

Thanks!
Max

[RRoller123]

This is one of those oddities on these cars, where the throttle plate shaft does not appear to have any serviceable seals. Here is a video of my spare plenum top/throttle housing. No seals that are visible nor can be added, as far as I can see. But I haven't disassembled it, possibly there is a seal underneath the Cir-Clip on the TPS side, and something on the Spring side?

The video won't show here for some reason, maybe too large, but it is on my Imgur page.

https://i.imgur.com/kAYhPPz.mp4

[SteinOnkel]

The irony that you specifically went out of your way to get an FI car to avoid these issues is not lost on me

First and foremost, check for vacuum leaks.

[maxm50]

The irony that you specifically went out of your way to get an FI car to avoid these issues is not lost on me

Ha! Indeed! The irony isn't lost on me, either

[Spider951]

Max,
Keep us posted if you resolve this. My '81 has a similar issue with cold starting - I have to feed it gas until it warms up or else rpms drop below a couple of hundred after initial vrrooomm (your 20 seconds ?) and then engine dies. Once warmed up it's good to go and idles at 900 or so with no more sputter. I have never made a serious effort at resolving the issue because it runs great after the initial warm up - I have chalked it up to "just the nature of the [old] beast", but I bet when new it did not do this. And yes, I have replaced injectors, rubber parts, gaskets, etc. My guess is it's some type of air/vacuum leak. Not trying to hijack your thread, I'm just hoping you find a resolution and will let us know.
Thanks

[maxm50]

Max,
Keep us posted if you resolve this. My '81 has a similar issue with cold starting - I have to feed it gas until it warms up or else rpms drop below a couple of hundred after initial vrrooomm (your 20 seconds ?) and then engine dies. Once warmed up it's good to go and idles at 900 or so with no more sputter. I have never made a serious effort at resolving the issue because it runs great after the initial warm up - I have chalked it up to "just the nature of the [old] beast", but I bet when new it did not do this. And yes, I have replaced injectors, rubber parts, gaskets, etc. My guess is it's some type of air/vacuum leak. Not trying to hijack your thread, I'm just hoping you find a resolution and will let us know.
Thanks

Absolutely, I'll keep this thread updated. I'm still working on this. Here's what I've found so far:

- My idle air bypass screw on the throttle body was only barely 1/8 turn out... ie. almost fully closed. That's further evidence that I have an air leak somewhere.
- Vacuum lines going to the charcoal canister and brake booster are original fabric covered lines and though they did not show signs of leakage when sprayed with carb cleaner while engine was running, they basically disintegrated in my hands when I removed them. Could have been contributing to leakage.
- I'm missing the cap on my charcoal canister, which I now understand is a metered orifice. So, there's another too-large air leak. Replacement in the mail now. These seem impossible to find so I'll make a 3d model of it when it arrives and could 3d print them for others who are missing (which seems to be quite a few).
- Carb cleaner sprayed directly on the exposed end of the throttle body butterfly shaft (in the area of the throttle spring) causes the idle to jump up. It's not the vacuum tube just below that goes to the distributor. It is indeed leaking into the throttle body around the butterfly valve shaft.
- After finding basically no documentation or other posts about a leak here, I pulled the plenum/throttle body off the car, cleaned it out, and disassembled the butterfly valve and shaft assembly. Just like an old motorcycle carburetor, if you've ever done that... I did indeed discover an O-ring on each side of the shaft which was not cracked, but rather hard and a bit deformed. It did not look like it was making a good seal anymore on either side. It's quite possible I had air leaking in at the throttle position sensor also, but it's obviously not so easy to spray carb cleaner on that side. New O-rings arriving Wednesday in Buna-N rubber and Viton, which resists high temps better. If these work I'll also provide them out to the community, because they came in big ol' packs...
- Coolant temp sensor reads OK, but looks original, so I replaced it with a new one that reads a little more in the center of the suggested ranges.

So, that's where I'm at. Basically it seems like I had numerous possible air leaks, and I'm trying to fix them all at once. If I'm still having the issue, my next step is the AFM, because it had its rubber plug removed and there's evidence someone took a a screwdriver to its bypass screw sometime in the past.

More later this week!

[76was124]

My 81 behaved similar, no one thing fixed it, but taking the idle up to 1000 rpms was one of the cheats to eliminate the stall. Other things to check:
- check brake booster diaphragm, not just the hose. -clamp off a good hose to the booster, does the idle change?
- similar for distributor vacuum advance diaphragm (remove vacuum hose at advance and plug), does idle change?
-confirm you have the correct AAV; previous owner might have replaced it with one for an X/19 or the Bosch FI system...should be a stamped part number on it.
-be sure distributor is set correctly and no loose parts inside, nor warn bearings causing axial play.

[RRoller123]

Really interested in those O Rings for the throttle body butterfly shaft. Following this. Will check the manual tomorrow and see if they are indicted there in the stack up.

Pete

[maxm50]

I had some great success with the throttle body shaft rebuild!

To get to the guts of the throttle body shaft, you need to:

1) Remove intake plenum.
2) Remove throttle cable drum. There's a 10mm nut holding it on. Mine was corroded / seized in place pretty good, so I had to use a drift punch to lightly tap tap tap it off the shaft after the nut was off. Clean it up so it goes on easier next time.
3) Remove 2 screws holding throttle body butterfly to shaft. USE A SCREWDRIVER THAT FITS. These are slotted screws and have a fairly deep but narrow slot, so don't strip them with a too-big screwdriver. I had to use a fairly small one and use pliers on the handle to get enough torque to break the screws loose. The screws are swaged after installation at the factory, so they're a bear to get out. (just look at the backside of the screw before trying to remove, you'll see it's mushroomed). But it CAN be done! I applied some PB Blaster first, waited an hour, then slowly went at them with the screwdriver and pliers. It takes a lot of torque to undo the swaged tip. When you reinstall these, use some blue loctite unless you have some clever way to swage them again yourself.
4) Now that the hard part's over, just slide the butterfly plate out of the shaft slot.
5) Remove throttle position sensor if you haven't already.
6) Remove circlip on shaft at throttle position sensor side. Use circlip pliers with small tips. I got some 6" Channellock ones off Amazon that work GREAT.
7) Make sure there are no burrs on the shaft in the throttle body bore (hit with some 2000 grit if so), then gently pull shaft out from other side. You don't want burrs on the shaft to scratch up the bushings that it rides on.
Use a toothpick to pick the dust shield and O rings + carriers out of each side if they didn't come off with the shaft.

Here's a pic showing the O-ring + carrier (little white ring inside the O-ring) in the bore, with the dust shield removed (This pic was taken during reassembly, hence the lack of gunk and nice clear grease):


Here's a pic of the shaft components disassembled at this stage:


And here's a closeup of the O-ring and carrier:


There is an O-ring on both sides. It rides on a tiny plastic carrier. It can be carefully separated from the carrier with a steady hand and a dental pick or toothpick. It's easiest and safest to do this while the carrier is ON the shaft - this prevents it from distorting too much. Just outboard of each O-ring/carrier is a thin washer / dust shield that offers some protection from gunk.

My O-rings were not visibly cracked, but made audible cracking noises when flexed and were no longer circular in cross section. They definitely weren't making good seals. Good luck finding out what the original size of these was. I measured the bore and shaft carrier and ordered a few sized from McMaster to try to get one that would fit, and I'm happy to say that I did: 1.9mm thick x 8mm ID. Here's the link: https://www.mcmaster.com/o-rings/system ... llimeters/

They seem a little larger than the stock ones, but still fit seem to fit well and I'm really not sure if the stock ones have shrunk as they've degraded. Who knows.

I have a big pack of these now, so if you need some and don't feel like ordering from McMaster, just PM me and I'll give you two if you send me a self-addressed stamped envelope.

So - now, clean everything up, load your new O-rings on the carriers, apply some grease to the shaft and O-ring bores, then start reassembling. You can pull the shaft to seat the new O-ring on the throttle cable side, and a 10mm deep socket works GREAT to seat the O-ring, carrier, and dust shield in on the TPS side. Otherwise it's rather hard to press the O-ring in evenly on that side.

Reassemble the throttle plate to the shaft, paying close attention to the correct opening direction. It CAN be installed so the throttle plate opens backwards - ask me how I know. Use blue loctite on the screws for this. I accept no responsibility for little screws making their way into your intake valves. I did this same operation on my motorcycle carburetors and the old Honda guys recommended blue loctite on those screws, so, hopefully it's good enough here too.

Clean up the mating surfaces of intake and plenum, use a new gasket if needed. Might as well replace the cold start injector seal while you're at it, because it's only a few bucks.

Did this fix my cold-start problem? Not sure yet! But I'm sure that at least this one vacuum leak is fixed. My new intake plenum gasket arrives Friday, then I need to replace my heater core and refill coolant before I fire her up. I'll report back. I hope this helps others, because I couldn't find this procedure documented anywhere, and I surely couldn't find the size of those O-rings documented!

Cheers,
Max

[RRoller123]

Excellent write-up, thank you for sharing that!

[Spider951]

+1 in the thanks department. Special appreciation for the details and pics. I will undertake this in the future (sometime when the dewpoint here drops to a number less than my age). I hope this fixes your starting problem.

[maxm50]

Thanks, guys! If you want a pair of O-rings, please do send me a PM and I'll tell you my address - you can send me a self-addressed stamped envelope, then I'll drop some O-rings in there and send them back to you! Saves you about $20 vs. McMaster.

[maxm50]

Update: I finally got her started up again after replacing these throttle shaft seals and all coolant hoses. As expected, my idle was now way, way off. Good! That means I fixed some leaks for sure. I got it roughly tuned in again, enough where I can check timing, replace plugs, and then start marching through Brad's FI manual. So, success so far!