Abnormal 328 frequency valve duty cycle measurements

A little context: I've been the happy owner of an 86' 328 w/ 40k mi for the past year and a half. Over the past several months it has developed a slight miss/"burble" at idle. The car starts uneventfully cold or warm, pulls well through full range of RPMs, and warm idle consistently settles at just over 1000rpm. I started my investigation w/ a timing light and thought I found the problem in a failing spark plug wire. Ended up replacing all the wires/extendors/plugs, cleaned the distributor contacts/rotors, cleaned all the coil connections and had the ignition modules bench tested. Unfortunately, this did not fix the problem. Continuing down the ignition pathway, I ran the microplex electrical test procedure (service bulletin 80-28) and found that that TDC sensor resistance was out of spec presumably due to mild corrosion on the TDC sensor connector. After cleaning the connection and coating it in Stabilant 22, resistance was found to be in spec at 860 ohm. For good measure, I pulled the TDC/tach sensors and cleaned off all the gunk but non of these maneuvers fixed the problem. A short while back I had replaced the fuel accumulator to correct a warm start issue (internal diaphram had failed). As part of a Bosch fuel injection promotional internet sale, I also purchased and replaced the fuel pump, fuel filter, and injectors/injector seals. Come to think of it, the miss at idle started sometime after swaping these parts out leading me to believe that I likely had thrown off the Air/Fuel mixture. Pending the arrival of an exhaust gas analyzer, I thought I'd try to adjust the A/F mixture using FV (Frequency Valve) duty cycle measurements. Using instructions from this site and others, I attempted to adjust the A/F mixture by aiming for a FV duty cycle of 50% w/ the car warm and the O2 sensor disconnected. Interestingly, at start up w/ the O2 sensor connected , the FV duty cycle was fixed at 72% which seemed a bit high to me. After full warm up with the O2 sensor still connected, the FV duty cycle fluctuated between 75-80%. I then disconnected the O2 sensor and the FV duty cycle changed to a fixed value of 74%. At this point I made slight adjusments in the 3mm mixture screw attempting to achieve a goal FV duty cycle of 50%. Once I reached a duty cycle of 70% the miss at idle resolved, however attempts at adjusting the mixture screw to get the duty cycle closer to 50% caused the engine to run very rough and nearly die. I should note that the oxygen sensor is new as of one month ago as it was replaced along with all of the exhaust gaskets during the fitting of a Tubi exhaust. So finally the question . . . Any ideas why the FV duty cycle is so high and how it may relate to the above described symptoms? Also, why can't it be adjusted to spec (50%) in open loop without the engine dying? For what it's worth, subjectively, the engine exhaust smell suggests that the system is running rich. Hopefully, the arrival of an exhaust gas analyzer should shed some light on the situation and assist with adjustment. Thanks in advance for any help that you can provide.

 

If the O2 sensors are disconnected, how are you evaluating your mixture changes? Do you mean frequency valves disconnected. I'm going to be diving into this pretty heavily in the very near future. I'm an agressive student in C.I.S. University [/QUOTE]

Nope, O2 sensor (just one on the US version 328) disconnected and frequency valve connected with motor warm and running. I pulled back the rubber boot on the FV connector to access the wires then inserted the leads of my duty cycle meter to access the exposed wire portion of the connector (there are two wires going to the FV, I put one lead on each wire) to obtain a reading. Per instructions that I obtained elsewhere on the board, I was attempting to adjust the mixture by measuring the duty cycle of the frequency valve. With car warm and oxygen sensor disconnected I took a duty cycle reading at the frequency valve, then incrementally adjusted the 3mm CIS mixture screw (plugging the hole between readings to avoid vacuum leak) trying to adjust to a duty cycle of 50%. The problem is I couldn't get the duty cycle down to lower than 70% before the car ran rough. My understanding is the default frequency valve duty cycle with the O2 sensor disconnected is supposed to be 50%.

 

Thanks Brian, your input is always greatly appreciated. Does the fact that the system doesn't default to a 50% duty cycle with the car fully warmed and the O2 disconnected suggest a primary fuel injection ECU failure or are there other components of the system that can malfunction and cause this? During this process I exposed the jetronic ECU and checked/cleaned all connections including the protection relay/fuse all of which seemed in order. My next step was to check cold/warm pressures (gauge/testing kit coming in the mail this week along with the exhaust gas analyzer).

 

Good suggestions. I've already checked/tested the coolant switch and throttle switch and both appear to be working as they should. The car's going on the lift in the morning and I'll take a look a the oil temp switch next. Can't thank you enough for your help

 

So put the car on the lift today to take a look at the oil temp switch. Pulled the leads off the connection to the sensor and found one stained w/ slight blue hue. ? prior coolant leak? The area was completely dry w/out evidence of active leak, coolant or otherwise. Cleaned off the connectors, slathered them up w/ stabilant 22 and plugged them back in. Working the night shift this month so wasn't able to start the car up before work. Hopefully will find some time tomorrow and I'll report back on whether this had any affect on the default frequency valve duty cycle measurement. As previously mentioned, I did pull the TPS (throttle position sensor) and found it functioning according to the following protocol:http://ferrari.cdyn.com/carl_rose_docs/Ferrari%20328%20TPS.pdf. I also attempted to check the TPS via instructions on 328 jetronic ECU testing (http://ferrari.cdyn.com/carl_rose_docs/Ferrari%20328%20Fuel%20Injection%20ECU%20Testing.pdf), however on page 12, after removing the cold start injector diode and inserting a fused jumper, it doesn't state which terminals to bridge with the ohm meter to determine idle/off idle TPS resistance. Maybe terminals 5 & 12 as these are used in the next section to determine wide open TPS resistance? Anyone else complete this diagnostic test before and have the answer?

 

Update: Unfortunately, cleaning off the connections to the oil temp switch made no change in the FV duty cycle measurements with or w/out O2 sensor connected. I tested the oil temp switch directly and it seems to be operating appropriately: switch open at room temp (infinite resistance between sensor leads), closed after car warmed (0 ohms between sensor leads). I also checked the coolant temp switch directly and found that switch closed at room temp (0 ohms betwwn sensor leads) and open after car fully warmed (infinite resistance between sensor leads). I finally completed the Electrical test procedure for the jetronic ECU (service bulletin 80-29) also described w/ pictures by Carl Rose here: http://ferrari.cdyn.com/carl_rose_docs/Ferrari%20328%20Fuel%20Injection%20ECU%20Testing.pdf and found very similar variations as described by Carl at the end of the document. I've sent him a PM/email to see if he ever sorted out his issues as I suspect my car is suffering from the same probleme. Finally the LM1 gas analyzer showed up today and initial testing revealed a lean mixture 33:1 which would correspond to the increased FV duty cycle measurements that I found. Using the mixture screw I was able to get the ratio down to 27:1. Any further attempts to richen the mixture (screw clockwise) caused the car to run poorly w/ strong gas smell from exhaust. I guess I'll start looking for vacuum or other air leaks that could cause such a lean mixture, any further advice/suggestions would be greatly appreciated.

 

Before the cat. I've since sorted out the A:F ratio. Turns out the T piece in the AIR injection system (328 parts catologue Table 9, part #8) had disintegrated and the two check valves had failed thus resulting in a constant flow of air into the exhaust with subsequent lean readings down stream. Replaced the T piece and check valves. A:F ratio now spot on. Initially set idle at 14.7:1 but car seemed to like a little richer mixture so now idle is set at 14:1. Still sorting out the strange frequency valve duty cyle measurements, however, I think I have that narrowed down to the oil temperature switch. I previously mentioned that I had tested the switch warm and cold and it seemed to be operating apppropriately (switch open when cold and closed when warm). However, on two subsequent checks, the switch did not close at operating temperature. When I close the circuit with a jumper the FV duty cycle decreased to 55% (open loop w/ O2 sensor disconnected). Still not the fixed 50% that it's supposed to be much much better than the higher readings that I was getting during previous open loop recording. New temp switch is on the way and I'll report back once it's all back together. I'm currently in the process of replacing the various hoses in the blow by sytem and I just cleaned out the oil vapor separator. Thankfully no rust but full of gunk. Many thanks to Rifledriver for setting me on the right path!