Probable tach signal error?

[Mike D.]

Got a problem with my '74 Benz 280C with M110 engine. It has a GM 4.3L TBI with adapter plate in place of the old Solex carb. MS2E V3.57 PCB, "Hot Spark" ignition points replacement set-up (same thing as the first Pertronix set-up), the relay board from B&G, etc. Fuel only set-up using the Hot-Spark system to power the coil and provide the tach signal.

Installation is documented over at the DIY-Autotune site.

The car had been running fine up until a few weeks ago. Now it misses randomly and the amount of missing is beginning to become constant. You can see the injectors stop fueling when the miss begins. They begin to pulse and feed fuel erratically. The engine will begin to run normally again but soon repeats. It seems to be more noticeable when the RPM's drop as it shifts to a higher gear but that might just be the fact the engine vacuum increases, the fuel feed is lessened and the load is increased.

Things done so far:
Replaced fuel filter (of course, always the first thing to do)
Drained tank, checked strainer, filled with fresh fuel
Replaced fuel pump and rubber lines (didn't really think this was the problem but since I had the tank drained and a new pump available...)
rebuilt the GM stock regulator
Installed an in-line adjustable fuel pressure regulator
Checked, double checked, triple checked the wiring harness, timing, vacuum lines, grounds, power leads, etc.
Changed the ECU to three different units from three different working cars, doing a fresh install each time.
Checked the injectors (1.2 Ohms, perfectly normal for the GM low-impedance units)
Set the PWM to 30% (recommended setting from the MS forum)
Checked, cleaned distributor cap, rotor, coil, plug wires connections.

No difference.

Ran a data-log on the system and the readings are all over the map.

Those GM systems are pretty basic and fool-proof so the fluctuations have got me baffled.

I'm leaning towards the ignition system as being the culprit but I'd like someone to take a look at the MSL and give me their opinion.

[Mike D.]

I have a suspicion the solid core spark plug wire set is causing the problem but I am not adept enough at reading the logs to be sure. I know the Pertronix ignition conversion specifically states to NOT use solid core wires due to EMI.

[billr]

By "solid core" wires do you mean non-resistor? Why, if both the ignition system manufacturer recommendation and the consensus here is to use resistor?

Get a transducer for fuel pressure, even if you only used it temporarily installed for trouble-shooting. Logging FP can save a lot of time and money by avoiding all that pump, filter, regulator, etc. fussing.

[Mike D.]

Mercedes specs call for solid core wires. They are what I had on the car and had previously given no problems, or at least none I was aware of. 3+ years of use.

[Mike D.]

Fuel pump and/or pressure is not a problem. I direct-wired the pump temporarily while trying to figure out the problem. I had a manual gauge connected to check. Fuel pressure never fluctuated. Constant 13 P.S.I. Bumped it to 20 P.S.I. with no difference.

No ideas from the log file?

[kaeman]

change the plug wires to resistor wires...... the factory uses resister wires (because they have to warranty the vehcile) on all the American made vehicles because the solid wires can and will eventually create problems or breakdown electrical components from the emi. my guess is that you are having a grounding problem somewhere and the net result is that the ecu is seeing false signals generated by the emi bleeding into the processor and disrupting calculations...

[Mike D.]

Made new wires. Changed distributor cap and rotor. Waiting on delivery of the Pertronix system.

Wires and cap/rotor had no effect.

Anybody had a chance to look at the file?

[Mike D.]

Installed Pertronix. No change.

[Mike D.]

Reworked all the grounds, connections, touched up the solder joints on the relay board, etc., again. No change.

Set EGO control to "disabled" and changed to 100% MAP acceleration, problem seems to have disappeared.

Can't figure out the original cause.

Don't want to say the problem is cured without knowing what created the situation to begin with.

Any theories?

[billr]

Post your .msq

[Mike D.]

280C3sqrtnoo2.msq

Here is the latest hacked up version which runs stable. I haven't had a chance to do any further tuning on it. The O2 sensor is completely out of whack because I disconnected it during the trial by elimination process.