grease deez

 The new alternator arrived yesterday, and I installed it. I subsequently discovered a few things. Firstly, the car still does the zombie run-on thing. Secondly, all my alternators are junk, but the original one with which I started is still the best one. The new one charges at 13-odd volts, where the original one just barely did 12.5, and the other two just lay there gasping. So worn brushes on them all, and the original problem of not shutting off with the headlights off still persists. I think this is a wiring problem, those butt connectors are connecting some things that I don't think are supposed to be connected, but everything works fine with the lights on. So that's how it's going to be. 

I have been through pretty much the whole car at this point, except the front wheel bearings, which I have never examined. I popped off the driver's side dust cap and found grease, but it was pretty old, with a big old cracked cakey pile on the inside of the cap. I pulled off the whole hub and got to cleaning with some doomed rags and acetone. The spindle on the driver's side looks really good.

I do not have new seals, and apparently you also are supposed to use new nuts, which I will explain shortly. 

But working heavily on the old grease with acetone and air disposed of most of it, and then it was time for the fun part. 

Back to the nuts. Fiat uses a "preload for dummies" approach which involves torquing to 14.5 ft lbs to seat everything, backing off to 5 ft lbs, then loosening the nut 30° by scribing a mark on the thrust waster midway between corners of the nut then rotating the right corner of that face of the nut over to the scribe mark. Then using a special tool, the nut is staked down over two grooves on the end of the spindle

 

Usually this involves a slightly more sane approach using a castle nut and a cotter pin, but one works with the engineering with which one has been presented. Lacking new nuts or the special Fiat tool, I commenced to hammering with a ball peen and a punch (which I suspect is all the special tool actually is), and after a long while of hammering and having the nut move, I managed to get it staked down to what appears to be an acceptable level, and added a couple of big divots on each side for good measure. Pondering the troubles here though, I held off on side 2 and I think I'm going to order some new spindle nuts. Poking around on Amazon, they say a set meant for Subarus fit a 78 Spider with a 5-speed, so I guess we'll see.

 

There really isn't time to do anything else. I'll repack the passenger side, use the old nut, then just swap them out when the new nuts arrive, and I'll be driving those nuts and not these nuts.  

 

UPDATE: The new nuts didn't fit, oh well. I did the passenger side after dinner today, and found it to be in not great shape. There was red grease squirted in on top of a residue of old black grease, and precious little of both. The seal is probably not working too well any more. I cleaned it out and found a chip out of a roller in the rear bearing as well, but, other than that, things looked ok, and as a bonus it had a very newish-looking nut on that side. So I cleaned everything thoroughly, repacked it all in fresh grease, then found to my utter annoyance that my little torque wrench only clicks one way, and since this is a left-hand thread, no click. (I realized I don't even know if they make double-sided torque wrenches.) I estimated it by tightening it until it just barely took a click to get it off again, and I hope that's close enough for Fiat work. The hub felt ok after that, similar to the driver's side. Surely, he said hopefully, it will last for 1800 miles more. 

 

This concludes the time I have to fix anything, and the car will henceforth have to fend for itself. I feel pretty good about it though. Next is packing and logistics, then the game is afoot. 

sport mode on

 This is the Spider:

 This is the Spider WITH RALLY LIGHTS!!!!!

 It was totally worth all the work. For years I had a single 5-inch Cibie fog light in a box, and one day it occurred to me I should mount a set on the car, in the style of the old 124 Rally cars (well not exactly in the style, but close enough). I poked around on Ebay and saw a single Cibie light for sale cheap, got excited and bought it, and realized when it arrived that I'd purchased a 7-inch light, instead of a 5-inch to match the lonely one I had. So there I was with a set of mismatched lights and a lot of annoyance, but then I found another more-or-less-matched set and picked those up. I had to make a couple of brackets out of angle iron to mount to the bumper, and ran a new circuit and relay from my new power distribution box. 

I got a few cheap round single-pole switches, one of which happened to fit exactly width-wise into one of the unused holes in the console, which i think was originally where power window switches went, and saved me from drilling holes. 

 

 The lights bolted down just fine on the brackets, I wired up the power and switching circuits, and awesomeness ensued. The new alternator still hasn't shown up yet, which is annoying, but I'm going to see if I can order some brushes for one of these broken ones which I think might fix it (but really at some point I should get an oscilloscope). 

feedback

 This is the last bit of time I have to work on the car, and while it's in pretty good shape, the zombie engine problem persists. It only took me a few minutes with the multimeter to find that there is 4-5V remaining on the ignition run terminal of the switch when the key is off, enough voltage to keep the relay activated on the infamous pink wire. (I found that the wire is actually red, just somewhat faded in its run through the engine compartment. Hence it shall be called The Red Wire.)  

The battery light in the gauge cluster remains on while the engine runs, but I found if I pull that bulb out, problem solved and the engine stops on command. (Or if the headlights are on.)  Some Internet chatter indicates some people solve this by putting a diode in the charging circuit, but I think that is just masking the problem.

I am pretty sure this means the alternator has a blown diode and is feeding low voltage back through the charging light into all the unswitched electricals, which includes the ignition circuit due to another butt connector.  Sure are a lot of these things.

I have a couple of spare alternators, but swapping those out only demonstrated that none of this garbage works at all, so I see some Bosch rebuilding experience in my future. 

By some kind of miracle, Autozone had a single remanufactured unit in stock. This is a chancy thing, but it's the only one anyone in town has, so it's arriving tomorrow, and we'll see what we see.
 

the new on/off switch

 Drove it to work today and all is largely well, except the not-turning-off problem persists. Unplugged the now-prominent pink wire, and plugged it back in to start it up and go home. When I got home, I jiggled some things with no pleasing results until I happened to turn on the headlights while the car ran on its mysteriously-powered coil, and the engine died. I found that it won't quit if the lights are off, but if the lights are turned on, it acts normally. I could almost understand it if it's the other way around, but turning the lights on to kill the engine is pretty bizarre. I only have a barely-half-baked theory that there is a voltage leak through the ignition switch that is enough to keep the relay energized, and turning on the light draws it down enough to de-energize the relay. The weather precludes me working on this much until next week, so we'll see what happens. 

bubbles

 The Fiat twin cam cooling system evolved over the years from a thermostat in the head to a three-way setup external to the engine. I don't know the engineering behind it, but when I put the 1972 engine in this 1978 Spider, I decided to convert it over to the external setup since it was then the same as all the other cars, which is useful at times. Once set up correctly, these cars have excellent cooling capabilities - I've never had any issues with it in the summer heat, not even when racing. 

One oddity is that I am not sure where the radiator in this car was originally from because it doesn't actually fit in this car. It's about an inch too narrow.  Someone had (very kindly) given it to me and I had already recored it before realizing with a good deal of aggravation that it seemed to be from a different car. Since the only real problem was that it didn't quite reach the mounting bolt on the passenger side, I just made a spacer for it to bolt it on. I also converted over to a larger (and lighter) fan than the original Fiat metal-framed stocky yellow 4-bladed unit.  I don't like how those aftermarket fans just ziptie to the radiator so I used some more flat stock to make mounts for it. The mounts are ugly, but they work. I might pretty them up at some point, but I doubt it. 

One of the things to be aware of with Spiders is that the radiator is lower than the engine, which leads to problems if you haven't fully bled all the air out of the system. My simple solution is to have a flushing tee in the top heater hose that is the uppermost point of the cooling system. I fill, burp all the hoses, and then top off through the flushing tee, and that does it. 

 The first problem with this new overheating issue was that I, like a big dummy, had installed the thermostat the wrong way round, and it was not in fact stuck closed. There's an arrow pointing at the nominal outlet tee on this thermostat, but turns out that's not actually how it goes. That was easily fixed, but I still had a problem with overheating, and it did not appear the thermostat was opening to allow coolant flow from the lower radiator hose back into the pump.  After a few rounds of careful bleeding, the problem persisted. I theorized that new thermostats (with which I have little familiarity) perhaps seal far too tightly, causing an air bubble in the lower hose that prevented coolant contact with the thermostat, which in turn does not open as it should. 

There's a trick for this, drilling a tiny hole in the thermostat plate to allow air to escape (some other engines have this by design, some even with tiny little float valves to allow air but prevent liquid).  I commenced to drilling, and on the next round everything worked just fine.  

I also fixed the thermoswitch wiring to the fan that was falling off, tidied away the wiring harness, flushed the whole thing and refilled with the green stuff and distilled water, and that little problem appears to have been solved. 

The car has not repeated the ignition malarkey, but I moved the new ignition relay over behind a brace further away from the exhaust, in case heat was causing it to seize up or something. It is more likely that there is some stray voltage through the pink wire, or the ignition switch is not disconnecting cleanly. The power usage of the relay is so much lower than that of the coil, it might be causing some latent electrical shorting issues to be more apparent.

behavioral problems

 I drove the car to work today, and it ran pretty hot, but performed in a considerably more lively manner than it has in the past, with its new ignition relay and some carb tuning.  I arrived at work pondering the growing temperature problem, parked, and turned the key off which to my utter astonishment had absolutely no effect. The car kept on running, in the manner of my old Mercedes diesel with a vacuum leak. Embarrassed, I opened the hood while curious coworkers looked on and made ribald remarks. I tried the first solution I could think of which was to pop off the pink trigger wire to the relay. The engine died and I went on about my day pondering the significance of this event. I could make no sense of it, as the trigger wire powered off the relay when physically disconnected, just as it should have done with the key off. At the end of the day I went out, reconnected the wire, and drove home with the engine still running too hot. The same thing happened when I arrived, but after disconnecting and reconnecting the pink wire, everything worked normally. 

 Mysterious, but I spent a few minutes troubleshooting the overheating problem and soon found (in addition to some slightly janky fan wiring I had done previously) that the new thermostat is stuck closed. The sensation of the fan blowing hot air at the top and cool air at the bottom was a strange one. I'll have to swap that out and see what's up with it, and at this point I might as well flush the cooling system completely.  

the blue vs the yellow

I spent a few minutes attempting to start the car today with no success. You may notice in the previous post that the picture of the new ignition relay shows the yellow wire connected to the positive terminal of the coil. This turns out to be wrong, and it's actually the blue wire that was supposed to be connected, as yellow is normally closed (NC) on that relay and blue is normally open (NO). Fortunately I had the battery disconnected so the coil didn't bathe itself in a warm glow of fire during the overnight chill. I swapped yellow for blue, and the engine started right up. It sounds considerably better with the exhaust leaks fixed (or at least greatly reduced), and the voltmeter now shows a perfectly normal 13.5V with the headlights on. I expect the ignition switch is really happy not to have all that load through it as well. 

While the car warmed up I took the time to fix the burnt-out starboard-side license plate lamp as well. I didn't drive it, but the clutch feels a whole lot better too - it used to be somewhat graunchy which I imagine was the threaded end of the cable scraping over the edge of the throwout lever. 

I'm pondering how to pack the ever-growing list of tools and supplies I'm generating. The Spider has a decently-sized boot for a 2-seater roadster. It also has a weird little rear bench seat, the provenance of which is completely unknown. No one could actually sit there except sideways, as there's no legroom. With snacks, liquids for car and man, parts, and tools, every little nook will be filled. Possibly some bags may be strapped onto the luggage rack, as is right and proper until it rains. 

There is an additional problem.  According to the specifications, the maximum cargo weight of a wooded and watered 1978 Fiat Spider is around 500 lbs. I like ice cream, and my co-pilot is of the same ancient stock, so just the weight of meatbags in the car is north of 400 lbs. I have therefore abandoned my idea to fit the floor jack into the trunk. There isn't any carpet in the car though, so there's that.