Well, I have been wondering why my car will not run well under load. I'm beginning to think, after all, that the latest issue has to do with a bad generator and/or small race battery.

Info: engine is a customized 3.2L, twin plug, Dougherty racing cam with high overlap, JE pistons with high compression (10.5:1), and PMO ITBs. It is EFI via EDIS, controlled by a Megasquirt system. It has been tuned using TPS (alpha-N I believe it's called) instead of MAP due to the difficulties with ITBs and a good MAP signal.

The guys at PMCI (I do believe them) said it was working fantastic when they had it, albeit a bit rich. I took it home two weeks ago and had these issues immediately. Called them, they are quite busy, but suggested that the battery might be a problem - they used an Optima temporarily to tune it, and since that's the only difference between when they had it running and my problems, I'm thinking that might be the source of the problem. More evidence: before they had it for tuning, the Generator light would come on after starting and not go away without a briefly sustained rev to about 1200-2000 rpm.

After logging and looking at the logs, I've noticed that the pulse width (PW) bounces (dropping to 0 very briefly) under load and the AFR can spike super lean when this happens. The car bucks, pops, and shows all the signs of having a messed up fuel mixture under any more than 10% load, so I can putter around but cannot accelerate or even go up a steep hill without making a helluva racket and being bucked around in my seat. I noticed on the log that my battery voltage drops at these times ... from 13.5V to around 12 V ... not much, but shouldn't this not be happening at all? Isn't the generator (if it were working properly) capable of keeping up with the minor load the injectors are putting on it under throttle?

Anyone have any experience with this?? I'm thinking of a new battery anyway, since the shop's Optima worked out well, but I don't want to mask the problem by buying an expensive new battery if the generator is the problem. If I do get a new battery, I'll probably go a bit larger than the Braille B2015, so any suggestions there are welcome. It will have to have a lay down bracket available so I can keep it basically where the Braille is now ... I'm thinking a bigger Braille or an Odyssey model will avoid the issues I've heard about Optima's.

So I'm pinging the World here, looking for experiences you may have had ...

Tack!

After a conversation at my local FLAPS now I'm thinking that it might be the alternator and/or the voltage regulator. With a 1986 engine I don't think I have a generator. I'm tempted to upgrade the battery anyway just to see if I can drive it.

I had a very similar experience twice, and it turned out to be my coil connections both times. They looked good, but weren't.

Update: It's not the battery, that much is for certain. I'm still thinking voltage regulator and/or alternator. I've a good feeling about the connections, they were very thoroughly gone over three weeks ago. Hey ho, the saga continues...

EDIT: Well, looking at the log shows that the RPM is all over the place. Something ain't right here. No lost syncs according to the log. But something is definitely wrong with what the computer thinks the RPM is and what I know it was when I was driving it. Megasquirt forum time! I'm now thinking the computer is fuel-cutting since it thinks the RPMs are spiking to well over 3x redline! Yes, I turned off the "100X" multiplier in the log ... still, the log recorded RPMs up to over 40,000.

What's providing the RPM signal? A 36-1 wheel? Distributor points, or hall effect sensor? I'd start with that...

I'd also go over the connections again. Especially the grounds, and especially the connections at the battery itself. (Both + and ground.)

And yes, you have an alternator and not a generator. I'm pretty sure that none of the 9-series cars ever had a generator. Not sure when the 356s changed over, though.

--DD

I had a very similar event with my Megasquirt system. I was on a local PCA drive and I noticed that my power seemed down. Then my engine started to ping under load. I finally looked at my volt meter and noticed that I was at about 11 volts so clearly I was not charging the battery.

I turned around at that point and nursed the car home - over an hour away. I was down to about 10 volts bt the time I got home. When I got home I noticed that my generator light was on even though the car was turned off but not when the car was running! Fortunately I had another alternator that I swapped in (not fun on a 914/4). Problem solved. I haven't seriously looked at this failure mode for Megasquirt but clearly there is a voltage sensitivity.

Check the voltage at the battery with a VOM across the battery terminals:

1.) Battery voltage not running .... 12.6 VDC

2.) Battery voltage running .... 13.6 VDC

More than likely you have an internally regulated alternator. Regulator is on the bottom side of the back of the alternator. When the engine gets hot, the regulator can get overheated. Regulator is really close to the case.

The MS rpm value is directly computed from the "tach" signal from your crankshaft or cam sensor. If you have a hall sensor it does rely on a +5 volt (supplied by the MS ecu) or more often +12v supplied indirectly by the cars electrical system. If you have a a VR sensor, it operates independently of any source voltage. I don' understand how you could be getting a high rpm reading without dropping sync.

MS will never work well if the tach signal is not absolutely reliable.

The voltage issue is definitely still there even with the new battery, and I've heard from more than one person that MS is sensitive to voltage fluctuations. The voltage is great for a bit then inexplicably tends to drop to ~12V or lower when the engine is running and warmed up. Since the injectors tend to shut off when the computer is seeing these very high RPMs, I don't want to run it like that, so she sits for now.

I did notice that the computer tach is spiking when I plug in the laptop to it, but the analog tach in the dash - which I've been told gets its signal from the computer - is reading the correct RPM, or at least it was when I was driving it while it was logging. Very strange. I'm wondering if I'm imagining things or if the log is actually showing what I think it's showing...

The car was recently at the shop where they replaced and upgraded a lot of the connections to the battery and computer, so I'm pretty confident it's OK. I can ask them to check them over again since I paid for all the work to be done and it's clearly not working.

For info, I do have a 36-1 toothed wheel and a VR sensor (I believe) that has been replaced from Clewett along with a better mounting bracket ... here's a picture of the old one that became "groovy" ... the shop found that the old one was not seated well in its bracket and had drifted too close to the wheel.
(Sorry about the lack of focus, iPhone camera isn't focusing in on near field objects for some reason. Might be time for a new one.)
Click to view attachment

The MS gets the signal for the RPMs from the cam sensor input, I just verified. There still might be an issue with the replacement they made so I'm also going to ask them to check over it again. I have been wondering if this might be contributing to the issue of the RPMs going crazy and the injectors closing or remaining closed when they should be squirting fuel, or if they're behaving as programmed and the computer is being fooled somehow into thinking the RPMs are ridiculously high. Then again, with a VR sensor, the voltage irregularity should not be causing or contributing it seems, so perhaps it is a failed repair.

If the second one were the case, I think I should be seeing lost syncs in the log, which isn't there. Then again, maybe it's not being reported (like it was before when the sensor got grooved) in the general log file from the MS ... which seems unlikely. In hindsight, I should have logged the initiation / timing events (so-called "composite log"). Maybe I'll start it up and just idle it to get a composite log. I'll look at the general log again to make sure I'm not crazy, and that the RPMs were really being recorded at Ludicrous Speed.

All of that thinking and logic leads me to conclude that it's one of two things: a failed repair of the old VR sensor causing RPMs to be reading wacky and the computer is fuel cutting in the mistaken belief the engine is way over redline, or some other voltage irregularity from a bad alternator/regulator is causing the major confusion somehow and the VR sensor repair is fine. I'm returning it to the shop, for sure.

Thanks for all of your input folks! Will post in this thread again in a couple weeks when I can get the car to the shop next. I will say that, for the very brief moments when she does run correctly, the thing is a monster!

As a minor update to this, the shop guys are as befuddled as any of the rest of us. I took one them for a ride with the laptop and he was astonished to see the RPM jumping all over the place, just like I said it was - apparently it wasn't doing this as often when they tuned it. It seems to be temperature related. As the car gets warmer, as in after the oil temp is up around 215F and steady for about 5-10 minutes, the issue shows up. Before that, it runs pretty well. A couple quick checks with a voltmeter seemed to indicate that there's a bad ground somewhere, perhaps with one of the Ford EDIS modules - I have two since this is twin plug. That was Friday. I'm just waiting to hear from them now.

We shall see what we shall see.

356s & 912s (except the '76 912'E') all had generators. 911s have always had alternators.

I had a similar problem and, like Ansbacher, it turned out to be the wires to the coil. The hotter the engine compartment got, the more resistance in the already-stressed wires. Changed it out and no more issues!

In this case, the new "cam sensor" (really a toothed wheel VR sensor) the shop put in had the wire's shielding poorly or just not grounded ... or so I believe, they didn't say much and didn't charge me much - I should've put up more of an argument about that, but I was too eager to drive the car in full up reliably running condition. It was causing all the chaos. Bad grounds are bad, mmmkay?

I am now slightly hard of hearing ... ITBs will do that ... but as I said in another thread, you couldn't knock this grin off my face with a sledgehammer!

I'm thinking of getting a GoPro now...