I had a look at the tables, but it's difficult to see what's happening in a dynamic way, e.g., looking at different, possibly interacting, parameters over a period of time.

What are you using to gather the data (i.e., tool, program) and in what form is the raw data (.csv, .xls spreadsheet)? If it's in a file format (e.g. .csv spreadsheet) that can't be attached here, "zip" it -- zipped files can be attached. With the spreadsheet we can select parameters and graph them to better see what's happening and what might appear abnormal. Not saying there will be a solution from that, but it might help.

Also, what did the new, zero time engine include, and not include. In other words, what equipment was carried over from the old one? If the exact same codes appear now, it sort of suggests that the problem is being generated by something that was carried over, and not the new engine itself. (Is it a brand new engine, or one that has been rebuilt but has zero miles?)

 

Sorry about that. I had the data in an Excel spreadsheet but wasn't sure how to upload that. I've attached those files on this post. It originally is a .csv file that I have saved as an excel file so I can open it. I collected the data with OBDlink lx scan tool and obdlink app. As for the engine, everything in the block and heads are new except the crankshaft and pistons. I purchased it from SSI Subies in Traverse City, MI. They replaced all of the timing components and timed. I pulled the torque converter and replaced the oil seal and the orings on the input shaft. All of the other components were transferred over including everything attached to the intake manifold including ignition coil and wires as well as all sensors. I've included a picture of the engine below after attaching the harmonic balancer and crossover tube. My thinking is that it isn't the new engine since it is showing the exact same symptoms as before the swap. I hope this helps. Thanks

 

That's a bit better. With the spreadsheets it's easier to graph data and have a visual view of what the parameters are doing. However, I looked at the timing drop-offs in the long log that you identified, and noticed that the drop appears at, say, 650 rpm, but at the next line the timing is recovered and the rpm could be above 2000. That seemed remarkable until I realized that the lines are about 2.5 seconds apart, meaning that from the line before the drop off to the line after there's something like five seconds. That's a long time when trying to identify what might be causing the timing to drop away. I wonder if the OBD Link scanner has a setting for faster polling.

That said, I don't see anything glaring in the idle log. Perhaps others might spot something.

 

swap #2 wire with another one to see if the misfire moves with the wire.
swap #2 injector with another cylinder for the same.
install a used coil as a test

 

plain OM

I wonder if the OBD Link scanner has a setting for faster polling.

I checked the settings on the OBDLINK and I have it on fast polling with the dwell time at 0msec. I just tried cutting down the number of PID being recorded to two to see if that would help and it cut the time down to 1 sec. What time intervals are we looking for when recording? The scanner will be no good if I can't get it to record data fast enough.
Thank you

grossgary

swap #2 wire with another one to see if the misfire moves with the wire.
swap #2 injector with another cylinder for the same.
install a used coil as a test

I'll try the wires tonight and the injectors tomorrow
thank you

 

plain OM

I checked the settings on the OBDLINK and I have it on fast polling with the dwell time at 0msec. I just tried cutting down the number of PID being recorded to two to see if that would help and it cut the time down to 1 sec. What time intervals are we looking for when recording? The scanner will be no good if I can't get it to record data fast enough.
Thank you. . .

That's very slow. I use Romraider and BtSsm, and both usually have Time intervals between 100 and 200 ms, with ~120 being typical. I have several Romraider "benchmark" logs where I recorded over 50 PIDs at the same time and the intervals are less than 200 ms. I'm not familiar with the OBDLink, but the website certainly plays up the fast timing, and I've read about the MX version being used with polling at less than 50 ms. There might be another setting or selection that's slowing down the adapter or the App itself.

In the meantime we'll await the results of the tests that grossgary suggested, because the continued symptom does suggest the cause is in something that was carried over.

 

grossgary wrote:

swap #2 wire with another one to see if the misfire moves with the wire.
swap #2 injector with another cylinder for the same.
install a used coil as a test

Had a chance to swap the wires. It took a few days to throw a code. It was the wires. I guess I had discounted the wires simply because they were pretty new. Replaced with a new set of NGK wires and runs great. No hesitations. Guess I shouldn't have overlooked the simple. Thank you.

Plain OM:

I use Romraider and BtSsm, and both usually have Time intervals between 100 and 200 ms, with ~120 being typical. I have several Romraider "benchmark" logs where I recorded over 50 PIDs at the same time and the intervals are less than 200 ms. I'm not familiar with the OBDLink, but the website certainly plays up the fast timing,

In the meantime I contacted tech support for OBDlink scanner and this was what they said

Your 01 Subaru uses ISO 9141 which is a very slow protocol.
Read mj0151s post in the following thread..

https://www.scantool.net/forum/index.php?topic=5833.0[/URL]

As Jason mentioned, the fundamental problem is that the protocol that your vehicle supports (ISO 9141) is very slow. However, you can get the fastest response for one sensor by doing the following:

1. Make sure you are running the latest version (2.12 or later).
2. On the Setup->PID Monitor page, make sure Fast Polling is enabled and the dwell time is set to 0 msec.
3. On the Setup->Fuel Setup page, select "None" as the Fuel Calculation Method (this will turn off all fuel calculations).
4. On the Setup->Connection page, click the "Decrease Port Latency" button, if it is enabled.
5. When you go to plot the sensor, select only the sensor(s) that you care about seeing.

I was able to get it to around 1 sec. intervals with 2 pids. So I'll keep looking into that and see what I can find on the forum with other 01 OB owners using scanners. I'm sure I can do better than 1 second intervals.
Thanks for the help

 

Excellent.