[BombshellBikes]

I am a long time Speeduino builder, tuner hacker etc. I currently own 3 vehicles running on various versions of boards I have designed myself. All of mine are loosely based on on the 0.43 version.

My current conundrum is this:

1994 Geo Metro, 3 cyl, 1.0l, turbo, custom manifolds, 330cc injectors, LS (smart coils)... been running and daily driving for about 18 months and about 10k miles. Recently made some changes to intercooler piping and raised the boost level from 12 to 17 psi.

After my last bit of tuning, I had the car dialed in really well, made good power, AFRs were spot on through out the range, everything was great. Next morning went to drive the car to work, would backfire, miss and generally run like crap as soon as I tried to go above 5psi. Naturally, I grabbed the laptop and tried to get some data logs. Plugged it in started logging.... and the problem disappeared. Tried unplugging the laptop and everything seemed to be back to normal. Called it a fluke and went back inside. Next morning, same problem! Plugged in the computer, problem gone.... I have now determined that when the car is started without laptop connected, the tune reverts to some incorrect version. If the ecu is allowed to connect to TunerStudio, even for just a second, it almost like the ecu "wakes up" and works fine. Cycle the key and its back to barely drivable again.

Has anyone else experienced this and what did you do to fix it?

Thanks

[LPG2CV]

Check the 12v/GND supply to Speedy.

12v is supplied via the loom to other engine components. It it fails to supply it to Speedy, it can run from USB power.

If the supply to speedy appears ok, then check the voltage reg inside speedy.

[BombshellBikes]

Voltage was the first thing I thought of. To rule that out, I plugged my USB cable into my laptop but didn't start TunerStudio. The problem persists in that condition.

[LPG2CV]

It wouldnt be TS that was running it! It would just take 5v from usb.

[BombshellBikes]

That's my point, the problem only goes away when I actually connect TS. USB power has no effect.

[PSIG]

While I've seen odd stuff, I haven't seen that specific issue. My first reaction would be basic IO/power tests as-suggested. Then create a new project in TS to see if there is an issue in TS. No joy, I would then wipe EEPROM in the Mega and reload FW and tune on a new project. At some point I would hope it again functioned "normally", indicating where the issue originates. Good luck, and let us know what you find.

[BombshellBikes]

I guess I was thinking along the same lines. This morning, I did a fresh install using the latest firmware, and then loaded my old tune. No change. I'm going to try replacing the Arduino tomorrow and see if that makes a difference.

Thanks for the suggestions!

[BombshellBikes]

Okay,
Started with a virgin Arduino, burned 202207.1 firmware. Went menu by menu matching all settings, copied and pasted AFR, Spark and VE tables. Swapped it into the ECU and the problem persisted. I am attaching a copy of my latest tune. If anyone has the time to look it over and can possibly see why the car only runs well when it is connected to TS, I would appreciate it.

[LPG2CV]

Is there any indication that TS is updating the tune in the mega when you connect.

There are OPTIONS with regard to how TS will react if the mega tune differs from the TS tune.

I am wondering if the mega tune is corrupting when you power off, and TS is automatically re loading.

Check out the OPTIONs menu.

[BombshellBikes]

I am very familiar with the "differences" dialog box and there is no evidence of that when connecting.