328 ECU Troubleshooting - Expert Advice Needed

If you have the 396/85 OM (1986 & 1987 US 328), the injection system schematic is also shown in Section 3 (page 73, Fig 65) -- might help you identify the pin numbers better.

Pin 8 is the +12V power supply to run the whole injection system -- relative to ground, it should read whatever voltage the alternator is putting out. This is also connected to the "+" side of the frequency valve. You might try measuring this pin 8 relative to ground and see if it is also "X" when engine runs OK and "Y" when engine runs poorly. (If this voltage was "going away" for some reason, like a flaky protection relay, this would also cause the pin 8 to pin 15 voltage measurement to go to 0V.)

Pin 15 is the "-" side of the frequency valve (and, relative to ground should be alternating quickly between ground and whatever pin 8 reads IIRC)

Couple of questions:

1. When you are making this pin 8 to pin 15 voltage measurement (which is effectively measuring the voltage across the FV) are you measuring VDC or VAC?

2. Does this pin 8 to pin 15 voltage measurement behave exactly the same (3~4 volts when engine runs OK and less than 1V when engine runs poorly) with the O2 sensor plugged in or not plugged in?

 

Curious since I never authored, co or otherwise a document on the subject.


What are the cars symptoms? I assume you speak of the lambda control unit.

 

You have a very smart multimeter . One thing to check is if there is a small "AC" or "DC" that appears in the display window while making a voltage measurement to indicate what it thinks it is measuring. But check the manual if you've got one -- there are many instances where measuring the AC component of a normally DC signal is important -- like measuring the output of an alternator -- where you want a large DC signal (~14VDC) and a small AC signal (less than 0.1VAC). If a multimeter can't be forced to measure only VAC, you could never make such a measurement, and I'd guess it very unlikely that a multimeter manufacturer would make this impossible to do (although you may have to read a manual the size of a phone book to figure it out ).

The good news is that you've been able to actually confirm something measures "bad" when the engine is running "bad". As indicated in Carl's .pdf file, the higher the VAC measured across the FV terminals, the more fuel gets delivered for the same airflow plate opening. So when your voltage between pin 8 and pin 15 goes to ~0VAC, it's much too lean (although your 3~4V when it's running OK seems a little lowish and might indicate you could use a small tweak). The tricky part now is determining if:

1. You're just losing the +12V power to run the system (which holds the FV closed because there's 0 VAC across the terminals and the A/F goes lean),

2. The O2 sensor is wacky and its occasional "bad" signal is causing the ECU to "think" it needs to hold the FV closed and forces the voltage across the FV terminals to be 0 VAC, or

3. The injection ECU is just flaky and isn't working right occasionally.

Good hunting! -- will await more info if you can get the Gremlin to repeat.

 

Yes -- this would be the cheapest and easiest fix.

The two wire plug is just an internal +12V electrical heater, so when the O2 sensor is already warmed up from the exhaust gases, it will make no difference whether it's plugged in or not, just as you've found.

The single wire is the O2 sensor output signal, and unplugging it is not the same as having a "bad" signal come out of it. When unplugging it, the ECU can detect this condition, and it sets the VAC at the frequency valve to a fixed mid-range default value. Since your running changes a little when you unplug this, that is also an indication that your manual mixture tweak is not quite right (but it's obviously not that horrible -- one check you can make is, at warm idle, the average VAC at the FV should be about the same whether the O2 sensor single wire is plugged in or not plugged in if the mixture tweak is right). However, if the signal coming out of the O2 sensor is stuck at ~0.9V (regardless of the O2 content in the exhaust gases), the injection ECU would think "it's too rich" and would intentionally decrease the VAC at the frequency valve trying to lean it out.

Agree -- we don't want to go here

 

Gene,

Recently, we ran into a intermittant rough running problem with a late 1985 308QV. As it turned out, this car has the same FV set-up as on a 328, which revealed itself after going through the 328 parts manuals. There's a small thermoswitch on the front side of the engine, near the oil vapor recovery drain hose. As best as we were able to determine, this switch shuts off the FV circuit when then engine is cold...and then allows the FV circuit to come into play once the engine comes up to temperature.

As the switch is of the simple bi-metallic type, the resistance should be either zero or infinite resistance...Our was always at 300-400 ohms, regardless of temperature.

See #18 on the diagram:
Image Unavailable, Please Login

 

Gene,

As I recall, we couldn't find this switch referenced on the 308QV wiring diagram...I didn't check to see if it was on the 328 diagram. As a quick test, pull the two wire connectors on this switch and short them together. We did this...and all of the driveabiltiy/rough running quirks went away.

Sorry I can be more help here...

David

 

Oh c'mon, Gene...It's a Ferrari. Of course the switch (only) acts up when the vehicle is hot. LOL! It's easiest to access the switch when the vehicle is on a lift. You could still test the switch, in situ, cold. It's either zero or infinite ohms...but not a value in between, I just can't recall which.

David

 

No, a "bad" signal doesn't necessarily equal "lack of a signal", and it's probably unlikely the O2 sensor signal is intermittent, but it's not impossible. It is true that if you have the bad running occurring, and unplug the O2 sensor, and the bad running stays present, the O2 sensor can't be causing the problem so that's the easiest way to rule the O2 sensor in or out as the cause of the problem.

 

Not a good sign for the O2 sensor, but there's always the danger that just wiggling things around is having an effect. If you can make a VDC measurement on the O2 sensor single wire relative to ground during the same sequence you could be more sure that the O2 sensor is wacko. For example, if you had for that sequence:

1. engine runs poorly with O2 sensor plugged in; O2 sensor single wire measures a constant 0.9~1.0 VDC relative to ground (and the FV measures ~0 VAC)

2. unplug O2 sensor and engine runs better; O2 sensor single wire still measures a constant 0.9~1.0 VDC relative to ground (and the FV measures the default 5~6 VAC)

3. plug O2 sensor back in and engine runs poorly again; O2 sensor single wire still measures a constant 0.9~1.0 VDC relative to ground (and the FV measures ~0 VAC)

4. unplug O2 sensor and engine runs better; O2 sensor single wire still measures a constant 0.9~1.0 VDC relative to ground (and FV measures the default 5~6 VAC)

5. plug O2 sensor back in and engine runs fine; O2 sensor single wire now measures 0.5 VDC average relative to ground as it should normally during closed-loop warm idle (and FV measures 5~6 VAC or something close, but not ~0 VAC).

Then I'd condemn the O2 sensor for sure.

 

Yes, http://www.partsamerica.com/productdetail.aspx?MfrCode=BOS&MfrPartNumber=13957&PartType=291&PTSet=A $144.99

I replaced it on mine.

 

When you unplug the O2 sensor, the (green) wire on the ECU side will default to 0.5 VDC -- so that's all OK. The output of the O2 sensor is the wire coming from the O2 sensor -- "no voltage" on this wire seems a little strange to me, but when something is "bad" could be something strange.

I'm just going by what Carl stated in his paper -- that it should be in the 5~6 VAC range when everything is working closed-loop and well. I'd still be more suspicious that you have an instrumentation problem (and are not measuring VAC), but it's really more important that you can detect "something" when running good and "nothing (near 0V)" when running bad.

Definitely, a good sign that you're finding the troublemaker. During cold-running, the ECU ignores the O2 sensor signal like it is unplugged so not ever having the problem when cold is a good sign for the other components (protection relay, ECU).

 

Would still like to see some voltage measurements on the black wire coming from the O2 sensor when warm and unplugged (and when engine runs poorly if plugged back in). If you only made this measurement at initial light-off, it isn't that meaningful as the O2 sensor only gives a meaningful signal when warm.