Gen 1 2002 CEL OBDII Code P1636 HV ECU Malfunction

Bob,

I fat fingered 8 as 13.47 instead of 13.74 in my initial post. Corrected in a later post and updated chart in original post. Will edit for clarity.

Rename .ZIP to .CSV. Forum doesn't allow .CSV.

Steve

What Brad says, but also try to record the charge as well.

You need to run the car to the highest SoC it will allow - typically 80%. The only negative I see with the charge test is the car seems to taper the charge current as it approaches max SoC, so the block voltage disparities aren't as large as when discharging.

Actually, I think I just figured out the sampling thing. I was working with a reduced data set, so I wasn't sampling as much data. You can reduce the data logging by specifying less data. Include only SoC, Current and 19 block voltages. That should speed up the sampling a little bit.

 

You did it, but you didn't do it aggressively or long enough. The peak charge was about 12A, and you never got higher than about 55% SoC. You need to mash both pedals... floored in "D" right up until it won't charge anymore.

Here's an example of a Gen2 charge from 61 to 83.5% SoC (42.4A average current, 770mAh input):


SoC is blue line, charge current is magenta, and their values are on the right Y axis. The rest of the lines are the 14 blocks, and their values are on the left Y axis.

Note that this was a relatively healthy pack, it just suffered from severely reduced capacity due to voltage depression. The jumpiness of the data near the peak suggests there was no significantly "worst" module. The Min/Max block numbers were jumping all over the place, which is generally a sign that they are performing in sync with each other with no outliers.

Where the voltages peak at a current of 45.7A, the min-max delta was only 0.24V. The max delta V of any data point in the test was 0.31V - those are pretty good numbers.

When you have an issue, you'll see the same 1-3 block numbers coming up as a min or max depending on the test. On my Gen1 rebuild with Gen2 modules, during the discharge, near the end of the tests, blocks 1 and 19 came up exclusively as they were the lowest capacity blocks. Max voltages during the discharge test were most frequently 10 and 11 as I put the highest capacity modules in the middle.

The numbers mean the most with the current flow is highest and you are closest to the min or max SoC thresholds.

Steve

 

Never mind, I figured out the file and am loading it into excel. This is what I think happened:

1. Techstream has a ".TSE" (or something like it) that is the internal representation of the data. I think this is what we have. Fortunately, it has the data saved as text values with some rather strange delimiters including a whole bunch of ";".
2. By renaming the ".zip" to ".txt", I was able to read the data into excel and then use formulas to parse out the data.

The attached spreadsheet has the graph.

 

If you do the test again, please see the instructions on how to set the capture interval to unlimited. This will make a much easier file.

Also, after I save the file, I reload it and then save the CSV. This solves the formatting problem.

Finally, you can use the 'data editor' (lower left, two pages with lots of lines on left page and smaller lines on the right) to capture just the interesting stuff. I will hack some more on the original at home now that I see what happened.

Bob Wilson

 

Ok, this is the final version. Some technical details about the conversion:

• Used text editor to replace all ";" with null "".
• Loaded into excel as delimited text, just <TAB> character
• Copied "TesterSnaps<n>" to a second sheet
• Removed the "MM/DD/YYYY" from date and time to get start time for each series
• Added start time to each sample row
  
• Replaced long text with shortened version for first header
• Copied transposed abreviated and original text to right most columns
  
• Removed redundant header text rows and sorted data
• Plotted traction pack amps, highest and lowest voltage modules
• Adjusted scales and data series to make it easier to read

About the data:

• Lowest was always #16 pair but others are close
• Highest switched between the end pairs #1, #19
• Voltage offset > 0.30, enough to trigger the fault
  
• Battery internal temperatures were warm enough, 30-35 C, the fan was coming on low
• Insufficient data to call the capacity of modules

From a cold-start (i.e., over night off)

1. Block the tires (SAFETY!!!), set parking brake
2. Start car and configure miniVCI to record battery controller data
   
3. Set user snapfile length to maximum
4. Trim data to SOC%, traction current, temperatures and 19 module pair voltages
5. Start recording
6. Holding all the time with foot brake, shift into "D", floor accelerator and hold while traction battery takes a charge to at least 70%
7. Release accelerator, shift into "P", see if battery fan is running, hold here for 1-2 minutes
8. Double check the wheel blocks and make sure the rear is clear
9. Shift into "R" and monitor SOC to 45-50%, you may increase 12V electrical loads but keep them constant so we get a clean slope of the discharge. No air conditioner but maximum fan is OK with heater set to coldest (to avoid cooling engine); headlights on brights; e.t.c. You want a constant load. This is for Ahr capacity measurement.
   
10. Shift into "P" and let engine normalize SOC, recording from here is optional
11. When engine stops, let car sit as fan cools pack and then shutdown car, save snap data, disconnect and head into house

Load the snap file and export as CSV. Load into Excel (or spreadsheet of choice) and calculate the SOC change as a function of amps in and especially out. Use partial fractions to calculate Ahr capacity of traction battery pack. Advanced, use voltage swings to calculate individual module pair capacity. It will be the weakest module in the pair. Alternatively, post data file here.

FYI, anyone have additional data they'd like to see? Perhaps fan voltages?

GOOD LUCK!
Bob Wilson

 

Quick bump, I've added details I was too tired to post with the file. In particular, how to do a better traction battery test that can provide Ahr capacity for the pack.

Bob Wilson

 

I would include high/low voltages and block numbers. Those can be hard to pick out of a graph, and it's useful for identifying trends.

 

During Draco's load test Techstream reported the minimum voltage as 13.98. Seventeen of the 19 blocks read BELOW this voltage. There is a sampling delay in reporting the min and max voltages. I see this all the time. It could screw up data collection.

Brad

 

Bob,

What is the trick to force charging the hybrid battery?

I tried my Gen 3 as described above and it would charge to 60% and stop charging. I held it for a few more minutes with accelerator floored but it would not charge any further. The battery amps dropped from a high of -98 amps at the beginning at 52% to just switching to slightly positive when I reached 60%, and then remained a positive 1 to 2 amps.

If you hold it longer will it eventually begin charging again?

Thanks

 

The 2001-03 Prius will continue to ~80% when the engine will stop automatically. I have not tried the full, forced charging on the 2010. The primary reason is I had no need and realize charging puts a strain on the traction battery. I have peak charged the battery descending a 525 ft. hill but this is something I try to avoid.

Bob Wilson

ps. Santa brought me an industrial quality, recording OBD system. I'm expecting to have a lot of fun!

 

I was going to ask if there is any benefit to test the HV Battery to get a baseline trend now. If it is going to harm the battery I sure do not want to do this.

Enjoy your new recording OBD system.

 

If you do your tests on a 'cold soaked' car, which means over night off, no problem. The battery will be at the coldest temperature for the day and one test cycle is not going to raise the temperature to the point of risk. Remember, 'heat is the enemy' and do your testing with that in mind.

Bob Wilson

 

Great thread!!!! I always learn so much from you guys.