Megasquirt Support Forum (MSEXTRA)

My car has developed a odd RPM loss during the first 90-120 seconds of initial cold start. The Rpm shows actual RPM, or exactly 0 or 152rpm (142rpm in another instance) every second or so, which throws in Cranking timing & PW which makes the engine surge at high idle. After the 90-120 seconds the RPM signal is fine and idle and running are great. I don't see anything else in the data log that coordinates with the RPM loss timing.

This is on one of the DIY MS3PNPPro SW0205 units, so I don't have many specifics on the MS3 Pro's interface with the OEM sensors.
It was not doing this during my prior tuning, not sure what could have changed to give me this wierdness.

What firmware are you using? There is no sync loss reason 97 listed in the 1.4 setup guide, p 135:
http://www.msextra.com/doc/pdf/Megasqui ... Up-1.4.pdf

Battery volts look good
Seems to be running a lot of advance for idleish speeds

What type of sensor is the CAS?

I don't know the exact firmware, It was 1.3.x something as delivered from DIYAutotune.
The idle timing should be in the low twenties/high teens, but with the oscillations it's getting into higher numbers.

The OEM CAS is a two wire polarized sensor, I assume a VR/reluctance style unit, as is the cam sensor.

The RPM MAP Scatter plot of the whole datalog shows the odd staged rpm data

Upgrade the firmware to the current stable 1.4.1

Updated my firmware to 1.4.1, things are much the same, maybe worse as I now don't trust that the calibrations and settings carried over with my .msq through all the firmware/definition/serial changes involved. Those I have checked seem to be the same, but I did not have a hard copy of the in/out settings to check.

Takes forever to start, runs a few revs and dies over and over, the AFR is cycling, still showing the 97 sync error and discrete rpm levels in the log until it's had its time out or warmup then the sync loss goes away. I tried locking in steady timing and fuel cells at at start and idle, changing the cranking timing to closer to table timing etc, but still won't start until it's warmed up from a dozen or (many) more attempts. Then it wants to cycle in increasing loops of MAP and RPM until it dies. Once its really warm I can get it to idle realtively steady and run fine, but the next cold start is an ordeal.

I'd post another log showing it running, but it would be a mixup of idle advance turning on and off, PWM idle on/off, manual VE and advance changes as I tried to get a steady idle. The starting attempt RPMs scatter out the same as before.

have you looked at the crank and cam signals using the tooth and composite loggers?

Not yet, because the car ran fine initially for several days on the controller before this issue arose, and for years before that with a UTEC piggyback on the OEM ECU. I'll give it a look though.

I was working on boost control setup when I got this bug, adding a different solenoid and trying various settings. At one point I tried to change the frequency range from slow to mid while the engine was running, which it most decidedly did not like. I also grounded the PWM3 output which should not matter on a low side output. The boost control was working on open loop with the new solenoid on the original HC2 output by the time this happened. I would not of been aware of the onset if the car had been running at the time, as it manifests on a cold start..

A composite log will help me see what signals are making it through to the MS3 processor; if you can post one, that will be a big help.

Sorry for the delay. I got frustrated and moved on to another engine for a bit-
Here's some cranking tooth and composite logs. I don't know the OEM cam or crank patterns for the Subaru at the moment.
Lots of double pulses...

It appears to be on the cam signal - to confirm this, can you change the number of coils to wasted-COP and the injection from fully sequential to semi-sequential? If the problem goes away with these test settings, that narrows it down to an issue with reading the cam sensor. If it doesn't, please also post a tooth log; they are useful for looking at the crank sensor on its own.

No start with the cam input disable changes, here's a tooth composite and trigger log with the changes. I'm going to try tracing all the crank cam signal wires looking for interference or damage.

looking at the cranking lost sync reason datalog points , I see it is now oscillating between 97 and 30 "36-2-2-2 tooth 14 error" where it was 0 & 97 before.

Checked cam and crank sensors at ECU plugs, both showing 1.9 kiloohms. Checked connectors through to ECU.
Wastegate wiring is in the same OEM harness, tried starting with solenoid unplugged and with Boost Control disabled. Still no start.

Tried semi-sequential and wasted COP again and it starts! Turned it off and set to sequential and COP and it still starts. RPM cycling is from lean mixture auto-tuned into VE table during hot running. Fixed it manually. We'll see how it goes next cold start.

-Now my LC-2 is sticking at full lean, maybe it does not like all the time spent powered up with no exhaust flow attempting starts and waiting for datalog downloads? grumble/grouse/growl.....

Let me know if the issue comes back.

On the Innovate sensor, try a free air calibration and see if that helps.

Still doing the synch loss from a fully cold start, eventually starting after a lot of tries. Maybe the weird rpm readings are just from such slow rpms as the motor tries to start with the wrong mixture?

But I think I've got bigger problems from the knock I got with the MPX4250 setting error in the base tune giving more advance than the actual MAP required. Maybe I lost a ring as I'm burning some oil at idle and I think that's sooting up my O2 sensor and I don't know my mixture. Calibrating the LC-2 got the meter running, but it just reads leaner and leaner as you idle longer in spite of boosting the idle cells higher and higher. I'm afraid I'm going to have to rebuild the motor before I can get the starting issue back to normal.

I have also seen a starter on its way out create enough electrical noise to prevent sync. If it is cranking slowly, it may be worth checking the condition of the starter.

The #4 piston ring land broke out due to knock from MAP sensor mismatch excess timing. Subaru's really don't like knock.
Rebuilding the daily driver, glad I have another vehicle and tools!

Closing this thread out, as after the rebuild the motor is starting fine with no further changes.

Well, to soon to close this out. Now that I'm tuning cold starts at around 0 C, I'm back to getting synch loss 0 and 97 when cold. Engine cranks, starts, starts to gain some RPM for a sec then dies like the key was turned off. Repeat 2-7 times until car starts, then no synch loss from then on. Logs show RPM dropping straight down, not tapering off. Not losing synch while I'm cranking.From a warm start everything is fine, the fast idle can coast through the synch losses if they occur and they're gone after 30 seconds or so, with the CLT coming up a degree or two.

I forgot about this prior issue and chased my tail for quite a while with cranking pulse, ASE, WUE etc before I saw the loss in the logs and realized I was not flooding the car or getting misfire leaness on my WBO2.

I'm going to try pre-heating the Cam and Crank sensors with hot air before a cold start and see if I lose the loss. Gives me something to do while I wait for a new set of sensors. Still don't know what a 97 is...

Can you post up the logs, including tooth and composite logs?

Error 97 isn't very descriptive - it's basically "I don't know what this pattern is, but it's not what a Subaru ought to have."