Im in the middle of tuning, my blms are not right on 128 but getting close. The problem is I have a mild surging problem, my Int goes up and down 10 counts at Idle or part throttle in closed loop, you can see the affect on my wide band 02, rpms and hear it with the engine. This does not happen in open loop at all. My 02 fluctuates from 70-875 but my 02 cross counts just slowly count up from 1 to 255 and start over again no matter what rpm. Is this normal because of my blms are not at 128?

 

The O2 cross counts counting up is normal. No worries there.

The INT doing the steady up/down is a problem. This is caused by incorrect proportional gains. They can be tricky to get right. Usually, when the INT oscillates the proportional gains are too low. Need to increase them in the area that the engine is running with the oscillating INT.

Need to look at the prop gain table(s) and see what the ECM is using. Usually it is an airflow value. Then lok at the data log in the area of concern and get the airflow value. Then increase the prop gains in that area of the table(s).

Go slow, make small changes at a time.

RBob.

 

Thanks for the reply Rbob

Prop gains & prop gain table?
Are you reffering to the ve tables? and saying I have lean spots around the cells where my car is running?

I'm using Tuner Pro RT with the S_AUJP_V3.ccp on a 730 ecm

 

This is most likely the table that requires changing:

Although this table is nearly max'd out, so the table at L84DE may need to be changed.

See chapter 5 in the tuning guide book sticky. ECM '8746/'7747/'8063, there is a fueling logic thread. That explains prop gains and the other closed loop correction stuff.

RBob.

 

Ok, I had to switch to the 8D_SAUJP_V4a.xdf to find that table, the values in that table are 1.0 across the board from 08 - 64. The Int oscillating happens at all rpms, what value do you think I should try?
I'll try and find that fueling logic thread also.

 

OK, I have read the thread.
Using SAUJP_V4a.xdf, I'm not sure what these tables your talking about are under. If I right click the table I was looking at and click on "edit item xdf info" It tells me address 04DE, is that the table you were talking about? If not, what constants or tables in this xdf file should I adjust or how do I create the tables your thread is talking about. From what I read, one thing I should adjust is "Int Delay vs. airflow" at L84BD, because I have changed to a heated 02 and moved it down the exhaust behind the super comp header. But I cannot find the table/constant to do so.

Sorry I'm new and need someone to hold my hand right now lol

 

In the table I posted above is the line:

gain = (L84CF[gms/sec] * L84DE[err]) / 256

This shows that the values from two tables are multiplied together then divided by 256 to create the gain term. One table (L84CF) use the airflow in grams/second for the look up. With the second table (L84DE) using the O2 error value for a look up. Basically the airflow and error level define which row in the table to use.

Since your one table is all at 1.0, which I'm going to guess means that they are all max'd. Then halve each value in that table. Then go to the error table at L84DE and double each row in that table. In the end the resulting gain will be the same.

But you will have head room in the airflow table for adjustments.

RBob.

 

You will never get all of the BLMs at 128. If they are between 122 and 134 consider yourself very well tuned. When I watch my O2 cross counts, they run up quickly to 255 and then start over again at 0. This is normal. If your cross counts do not do this, you have a prolem. You are probably running lean or rich nearly all the time, and the BLMs would most likely drop to their minmum value or increase to their maximum value allowed in the ECM programming.

 

Thanks so much for the Info so far RBob

I did a bunch of reading, found the hack with all the tables and now I know what I'm looking at in the Saujpv4. I changed the values in table L84CF to .5, doubled the values in L84DE, started the car then moved the values up little by little to .8, no change, maybe I should take it further?

Is there a way I can tune the 02 tables by watching the s graph in the ALDL Monitors. I noticed my curve does not have a sharp turn at the 800-900mv side. It kind of flattens out at the top, but the curve is sharp and small at the 60-70mv side. Ideally should it be like a perfect S??

 

Thanks for the reply Doc

My Blm is 125 at hot idle and the car runs strong and good. Even at Idle I have the oscillation, but it’s not noticeable in rpm, you can see it in the wide band and in the INTs. About 10 count’s back and forth in the INTs and 13.8 - 15.1 in the wide band, but doe’s this pretty slow, with rpm it speeds up, then you notice it in the rpm.

So the cross counts are supposed to count up fast?
How fast? How long should it take to get to 255?
My cross counts count up really slow, and when I was making the adjustments Rbob offered they would speed up, pause for a second then speed up again. I'm guessing the counts should count up in a linier fashion, no jumps or pauses?

I'm doing all this testing at Idle

 

14.7:1 AFR is 450mV on the narrow band O2 sensor. Each time the O2 value crosses over the 450mV value, you get one cross count. So as the ECM corrects the AFR to stay near stoic (14.7:1), the O2 sensor will bounce above and below 450mV. Each crossing is one cross count. Since the various sensors which supply info to the ECM are constantly changing slightly, the initial fuel calculation will lead to an AFR slightly above or below stoic. If the AFR stays on one side of stoic for a second or so, the INT and BLM will adjust, and throw a correction into the fuel calculation which should bring the fueling back to stoic, and so on, and so on. The cross counts usually count up fast, but at times can stay put for a few seconds also. This is the best that I can explain whats going on with the cross counts, the BLM and INT relationship. Hope this helps.

EDIT: I think that I have seen the cross counts go from 0 to 255 in one to two minutes easy. I can monitor the cross counts and they can biuld up quickly to 255 and start over at 0 again.

 

something else to keep in mind is how often the ECM reads the O2... i think it's either 40 or 80 times a second?

and to "cross" it has to go above AND below 450mV, so figure the maximum rate it could ever accumulate cross counts is either 20 or 40 per second.

 

Mine are moving much slower then that
Are there some things I can tweak to speed it up?

Also to update:
I maxed the min enable temp for closed loop to run the car in open loop. I did this because I have been getting prom error 51, which I thought was due to the Int oscillation, but the car also gets the error 51 in open loop. Also found a lot of lean spots in my tables watching the wide band, so using the wide band I'm going to try and correct those in open loop. (I'm guessing the blms are not correcting like they should in closed loop because of the oscillation). I have not changed much in the saujpv3 ccp, other then injector constant, cylinder constant, max spark, initial spark, idle rpm, spark tables and ve tables. So I'm wondering if you guys might know why I'm getting error 51?

Thanks for all the info so far

 

i thought about it again, and the maximum cross rate would be 40 or 80 per second, since going 800-200-800-200 would be 3 crosses assuming 800 was the first sample.

as far as speed: from what i understand, the O2 will react slowly if it's "worn". i also expect it to react slower at idle.

 

Ok thanks
My 02 is brand new, I put a Bosch 13077 heated 02 in the car

 

Error code 51 for my '87 is a "faulty memcal". I assume its the same for you. My manual states to "check that all pins are fully inserted in the socket and tha memcal is properly latched. If OK, replace memcal, clear memory, and recheck. If code 51 reappears, replace ECM." This looks a bit grim.