So I have had ebl for quite some time and never really got into it, until just a couple days ago. At first, I was very nervous to drive more than a block or so, but after a couple velearn, I'm courageous enough to drive into my two adjoining neighborhoods.

Anyways, my velearns keep removing fuel, but it seems like I have a surge. It is only at very small loads with little throttle, if I go just a hair deeper in the throttle it seems to clear up.

Also, when I do a velearn the newly created file doesn't show up as a bin file; however, it works normally when opened as a bin in TunerPro. Is this typical?

This is the link of my latest datalog in zip format, the first 7 and half minutes is just warm up so there is not much to look at.


Edit: Forgot to mention that I disabled async mode, by reducing the INJ- Async Transitions to 0. Before, I did this I was unable to go a block with the engine wanting to die at idle and surging like crazy.

 

For the light throttle surge reduce the PRP - Gain Multiplier vs Airflow table. Mostly in the lower air flow areas. The log is showing excess prop gain in the 25 gms/sec area under light to no throttle.

This surging is caused by the closed loop routine forcing the AFR to oscillate. This is done by changing the injector PW. Which when changed too much causes surging.

Not sure about the newly created BN file (from VE Learn). It should have the extension of '.BIN', which will show up as a BIN file. The only way it wil not have a .BIN extension is if it is overwritten by the user. This can be done when the WUD requests a 'learn-to' file name.

The default extension is .BIN, and if left blank should end up with a .BIN extension. Please let me know if it isn't working this way.

RBob.

 

I reduced the PRP - Gain Multiplier vs Airflow lowest three values by 20% (of orignal value) and I will try that when it is light out again.

When I do a ve learn, the output file extension shows "file" instead of "bin". It is not a big deal, but I always have to change from File name: "bin files" to "all files." The files always work as bins, so I can't complain there.

 

I made two runs today. One the first run, I had slightly reduced the values of the PRP- Gain Multiplier vs Airflow. I could not tell much difference in the surging, except that my ve learns showed the first lean cells since I had begun tuning (which could just be the nature of things or other).

Second run, I changed the PRP- Gain Multiplier vs Airflow even slightly more and still did not see much change; however, it felt like the surges were worse at the beginning and then tappered slightly. The most notable surges are around 2000 rpm in second gear. Also, I had even more lean cells showing up.

Here is my current chart

37.65 128
37.65 112
37.65 96
37.65 80
32.94 64
27.84 48
18.04 32
9.80 16
7.45 0

I do not know if that is too little or too much change, or if I need to change other tables as well.

In other news, I purchased a LC-1 today, so I'm growing up in the world of diy tuning.

Link to my latest datalog.

 

I missed the oscillating INT in the first log. Thought about it this time. Need to slow down the INT update rate to allow the proportional gains some time to work. Start by increasing (edit: was reducing) the values in the 'INT - Delay vs Airflow' table. Note that the INT is even oscillating at airflows of 70 to 80 gms/sec (cruise).

Try raising (edit: was lowering) the values by at least 20%. Then check the next log to see if the INT has stabilized. It is OK if it moves, but not OK if it is constantly rising & falling during steady throttle.

RBob.

{edited to increase the delay values, reducing them will quicken how often the INT updates}

 

I tried moving the down 'INT - Delay vs Airflow' 20% at first and did not notice any difference so I moved it down about another 20% and still no difference.

Current 'INT - Delay vs Airflow'
msec GmSec
125.00 128
150.00 112
175.00 96
200.00 80
250.00 64
275.00 48
325.00 32
350.00 16
400.00 0

Latest datalog, the largest surging is book marked. Also the surging is strong enough that it is quite audible in the exhaust tone.

 

Gladstoneiroc, I goofed. Need to increase (not lower) the values in the 'INT - Delay vs Airflow' table. Increasing the values will slow down how often the INT updates. From the original values increase them by 20%, then see if the INT is still oscillating.

I had reducing the surge in my mind and kept thinking in that direction. I also edited my previous (incorrect) post of this.

RBob.

 

I bumped the values in the 'int - delay vs airflow' table about 30% and have not seen too much improvement. To me it seems like the surge happens more right after start-up than after an extended drive, but that could be my body learning to avoid the surge.

Also does this seem like a reasonable way to set up my wideband?

Latest highway run datalog

I'll try and post more information later.

 

I haven't been able to download the data log. Something isn't quite right with the upload.com site.

Can you email the log to me: bobr at dynamicefi . com (just remove the spaces and change to @).

RBob.

 

Took a look at the logs you sent. At airflows below about 80 the INT is still oscillating some. Above that the INT is fairly stable. I checked the values in the PRP- Gain Multiplier vs Airflow table you posted above. They seem to be a little high. Here is a good starting point for a stockish setup:


If your current INT delay vs airflow is greater then the values I listed, stay with it for now.

If your current the PRP - Gain Mult vs Airflow values are greater then the value I listed, use the lower set of values.

Then try again. Basically need to get the INT to stop oscillating and let the proportional gains do the work. If the prop gains are too high then the INT attempts to make up for it. To get the INT to steady down you can even double the INT delay values I listed.

Then work on the prop gains to stop the surging. Here is a graphed portion from a log. Note that O2 high/low swing. In the first half (before the upshift) the O2 swing is about 800 mV to 100 mV. In this area the INT is barely oscillating and the airflow is in the 90 to 110 gms/sec.

After the upshift the O2 swing increased dramatically. It is swinging in the 880 mV to 35 mV range. Which you can probably hear in the engine. The airflow is in the 60 to 64 gms/sec area.

Then once the surging is under control can speed back up the INT update rate (INT - Delay vs Airflow).

RBob.

 

First, I would like to say that Rob, you have the best customer service ever.

The attached picture should sum up my problem, but it is too late to go for a spin now to test it.

My 'INT- delay vs airflow' values were in assending order instead of desending order. Which makes sense because after changing the values before only the idle got a little funky.

I should have some results posted tomorrow morning.

 

Changing the values to the two tables, Rob posted made the surges much smaller. They is only now a hint of surging (small enough that even an observant passenger probably wouldn't feel it).

The only worry I have is that my ve learn showed a lot of high values (-10) in the low map/ low rpm section. I'm not sure if that is normal or not and I have already had a lot of change in those cells.

The next thing, I need to work on is the unstable rpms while coasting downhill out of gear. And possibly making the coast rpm slightly higher, so the power steering doesn't seem like so much load at 550 rpm.

Latest datalog, if it doesn't work I'll email it to anyone.