starter run-on

[Roadtrip]

Taking the ignition switch apart? . . . Good luck. I think you're probably a prime candidate for installing a couple of relays (one for the starter solenoid and one for the SPICA CSS. At least that way you know it won't be the CSS pulling current off the starter solenoid. It'll also make your ignition switch last a lot longer because it won't have high current going through it. Imagine the load that those small contacts in the switch have to carry actuating 2 high current draw solenoid? Installing a couple of relays is cheap and easy . . . . . . sure a lot cheaper than arc welding a starter solenoid ($100) or an ignition key switch ($100).

[SILVERBULLET]

well guys, the cold start solenoid it is! rolled it out of the garage
and popped the hood so i could familiarize myself with where
everything is...relay was right where it should be..on the firewall
next to the coil(thanks jim..made it easy). then i made sure
i could get at the css, and put the
tools i might need next to the battery to pull the cable if need be.
car started fine three times in a row. the fourth time it started
fine, but while it was idling, it stalled. restarted and it ran fine
then stalled. then it would not stay running. seventh try the
starter hung up. i turned off the key and it kept turning. i got
out and pulled the css wire first. starter stopped and it ran
perfectly...no stalling. i then tried it with the css wire pulled...
no problem. soooooo, now what do i do? wire a diode in line
as per alfajim's suggestion? what kind of diode and where do i
get one? and why is the css seeming to stall the motor?
thanks for the help
john d

[alfajim]

SILVERBULLET
Glad to heat that was the problem I just don’t remember where I used to get the diodes but they were big. Another way to do it is to use a relay to energize the solenoid. The only reason I can figure they do this is because the cold start solenoid is on the edge of drawing enough power from the starter solenoid that the field does not collapse and keeps the starter solenoid engaged.

[SILVERBULLET]

well, i searched "starter cold start solenoid" on the alfa digest and
i found an owner of a 71 spider with the same problem. his
solution, for which he gives credit to wes ingram, is the following:
cut the wire to the css at the spade terminal on the starter. then fit a ring terminal end on the wire. now attatch the ring terminal
to the lower post on the starter which has a braided wire going
to it from inside the starter body. "the css can now only be
engaged when the starter is engaged, thus preventing the
feedback loolp from occurring". so, before i start cutting and
splicing, what do you gurus have to say on the subject?
thanks again
john d

[papajam]

Good show, John!! Congrats.
I've never tried Wes' repair (I like relays) but it is brilliant in it's simplicity. The mod results in the CSS being powered from the starter motor instead of the ignition switch so the feedback can't occur. Now THAT"S a macgyver! I'd go with Wes' mod! Should take only about 15 mins. I'm curious if the stalling will then disappear.
I like alfajim's idea of a diode too. But like Jim said, it must be big. On the order of 20 or 30 amp I'd guess.

[Roadtrip]

Diodes to fix this problem? . . . . Pleeeeze. That sound like something the Brits would think up. It's called the "Hardest Way Rule" . . . . . take a simple problem, get a brainstorming group together, get 10 alternative fixes . . . . then pick the hardest one.

OK, guys from the UK . . . . JUST KIDDING.

I don't see anything wrong with powering the CSS off the starter motor terminal (bottom) on the starter solenoid. All the CSS cares about is getting 12v to energize the solenoid and snap the mixture rich in the pump. The starter solenoid switch is a really big hunk of copper flatstock that should be able to handle the relatively minor increase in current necessary for the CSS. A relay will do the same thing . . . . isolate the high current draw of the CSS from the ignition switch.

At the risk of "building you a watch when you just want to know what time it is" . . . . here's a short explanation on the CSS and how it can stick and cause starting problems.

Do be aware that CSS's can "stick" in the "on" position (pulling up on the rod and piston internally) and case stalling due to too rich of a continuous mixture.

There are two enrichment mechanisms, the thermostatic acutator which is governed by coolant temp, and the Cold Start Solenoid (CSS) which has just a momentary "get-the-engine-to-initially-fire" function. They are inter-related in that when the coolant is cold the T/A allows the CSS to enrichen the mixture A LOT, but when the engine is hot, it governs down the CSS down to a small amount of starting enrichment.

You can ops check you CSS by just running a jumper wire from the postive battery terminal, momentarily (1-2 seconds) touching the CSS terminal and listening for a nice audible snap as the solenoid acutates. If no snap, then the mechanism may be binding . . . . or more specifically the little hydraulic piston at the bottom of the CSS rod. The function of this hydraulic piston is to release the mixture slowly (1-2 seconds) back to "non-start" enrichment. Over the years, the plastic piston can expand a little due to the engine oil it's bathed in and occasionally stick in its bore. The emergency fix to a hopeless stuck piston is to disconnect the wire to the CSS and tape it over to prevent a short. Then remove the side inspection plate, get a pair of pliers or needlenose, and pull DOWN on the CSS rod. It should release fairly easily and snap back down. The long term fix is to remove the injection pump, take the CSS rod and piston out the bottom of the injection pump and remove a very slight amount of the swelled piston with fine sandpaper.

Here's a picture of what the CSS mechanism looks like behind the inspection plate:

[alfajim]

Well roadtrip since you were so far off base on what the problem was I don't see how you can put down a repair that I have used for years to cure this problem. The diodes that I used were a simple plug in in line with no modifications. So Pleeeeze get off your high horse.

[Roadtrip]

Sorry Alfajim, didn't mean to belittle the diode fix, but I'm totally in the dark on the theory of why this would work.

Since the CSS doesn't produce electricity and only "eats" it, the only thing I can see that would cause this problem is an excessive current draw current draw that sticks the contacts in the ignition switch or an arcing of the main solenoid contacts (due to undervoltage arcing. Disconnecting the CSS reduces the current draw thereby reducing the load on the key switch. My understanding of diodes is that they only allow current to pass one way through them (as used in converting AC current to DC). Do they also restrict current (thereby reducing the load on the key switch)? I'm at a total loss to understand how the CSS can "backfeed" current since the only pathway for current to the starter solenoid and CSS is through the key switch.

Someone illuminate me . . . please.

BTW, I had this happen to me one time years ago in a 71 Spider, on a very cold Denver morning. Car wouldn't start at first crank attempt, released key from start position to "on", still cranking, turn to off and removed the key . . . . still cranking. Sitting there in the drivers seat looking at my key in my hand and listening to the starter crank . . . . I was about the most confused man on the planet at that moment. Luckily the negative battery terminal was loose enough to wriggle off. Never happened again, so I caulked it up to bad Karma that day.

Last year I had an intermittent starter problem and was lucky enough to get a rebuilt unit off ebay cheap. After the swap out, I removed the back cover on the starter solenoid to see what the contacts looked like, and they looked fairly burned and arc'd. Unfortunately, my temptation to further dismantle it resulted in a broken shaft that the copper switch plate attached to, thus ruining the solenoid. The starter motor itself seems good. I'm pretty sure that if I had just cleaned the contacts (and left the rest of it alone) and maybe installed some relays, that I would have saved myself $100 on a replacement starter.

Sometime I learn hard, but I usually learn good.

[alfajim]

What happens with the css and the starter solenoid is the power being drawn from the starter solenoid from the css allows the field in the starter solenoid to stay energized. There is just enough power being drawn thru to keep the plunger drawn in and this keeps the contacts connected. The reason we used the diode fix was it is cheap (Alfa owners demand this). This repair did not require any mounting of relays so everything stayed original no holes drilled in the inner fender well (another thing Alfa owner’s like). This problem seems to be most prevalent in 69 thru 72 Alfas.

One thing to add to your post on the sticking css plunger is that in 77 thru 79 the nylon plungers were slightly larger in diameter and when hot (I have never seen it happen on cold starts) would stick in the bore causing the plunger to stick down flooding the engine. The cure was to remove the plunger and decrease the diameter by a couple thousands.

[Roadtrip]

But if the power is cut (key switch off), why doesn't the electrical current stop and de-energize both the CSS and the starter solenoid? How can a simple solenoid like the CSS "produce" current that you would need a diode to stop? Do I fundamentally mis-understand a diode's function?

[alfajim]

The power is only cut to the terminal supplying power to energize the solenoid. The battery is what powers the contacts in solenoid. The diode was just enough to keep the starter solenoid from staying energized allowing the magnetic field to collapse and the plunger to disconnect the contacts.

[SILVERBULLET]

well guys, since no one thinks this is going to fry my car, and if it's
not raining here on long island tomorrow, i'm gonna roll the car
out of the garage and give it a go. i'll let you know how it all
works out. and thanks again, all of you. this could have been
one of those things that drives you nuts and makes you hate
a car.
john d

[SILVERBULLET]

it worked!!!!! took about 25 minutes to do the refit. part of that
time was waiting for the liquid electric tape to dry so i would not
have to worry about corrosion ruining the connection in the
future. had one of the "OH MY GOD, WHAT HAVE I DONE?"
moments when i tried to start the car...nothing. redid all the
connections, checked the battery, etc. not even a click when i
turned the key. after about 10 minutes, i realized, DOH!, that i
had not reattatched the spade connector from the ignition to the
starter. that done, it started and idled perfectly! can't thank you
guys enough for your help. i never would have thought the spica
could have an effect on the starter.
thanks
john d

[Farace]

Quote:

Originally posted by alfajim This problem seems to be most prevalent in 69 thru 72 Alfas.

[Roadtrip]

I've been putting this off myself. My car has the standard wiring. Given that this is a general design weakness, it makes sense to me to do the mod as PM.

Weight the problem and large expense caused by a burned ignition switch and/or starter malfunction (probably in somewhere real inconvenient) vs. the minimal time and expense to put a ring connector on the CSS wire and install a relay between the ignition switch and starter solenoid . . . . . I think the mod wins that contest. I don't enjoy electrical work . . . that's why I've been putting off:

Relaying the starter
Switching the power supply to the CSS to the starter terminal
Relaying the fuel supply pump and installing heavier gauge wire
Installing an inertia switch in the fuel supply power