Rebuildman

I am tuning a 400 sb with the HSR using a MAF ARAP bin in open loop. One problem I run into is at PT with low load on motor watching the WB it will intermittently go very lean, like from 14:1 to 20:1. The inj PW drops by 1/2, the MAF LV8 and MAF grams airflow also drop alot, MAF RAW input does not change nor does the TPS.
I replaced the MAF and the ECM with used units no help. Checked the egr and ccp tables and they are disabled, so is the Highway mode. My next step is the wiring to the MAF, unless someone else has run into this problem and can offer some insight? Suggestions?
Thanks!

1MeanZ

iTrader: (9)

I think my car may be doing the same thing, but much less severe. I've added some fuel (BLM wasnt at 128) and I've added a bit of timing. My problem has gotten better but is still not gone yet. I'm going to play with timing some more because it may not have enough advance in those cells. Make sure it has enough fuel and try throwing some timing at it and see if it does anything. If there is not enough advance your piston can be running away from the flame front, then in extreme cases you blow still burning fuel mix out your exhaust which can cause the O2 to do funny things.

RBob

iTrader: (1)

It sounds like the ECM thinks that the MAF is bad and is creating a default value. If the MAF raw input is OK, but the LV8 and Gms/Sec airflow drop then the ECM isn't using the MAF input.

This may be being caused by a borderline MAF 33 condition. This is where the ECM checks some parameters to see if it can then check the MAF airflow. If the airflow is too high it will go into default mode. With a modified 400 this can happen.

Check the data log for these values (stock ARAP values):

less then 3200 RPM
less then 14.8% TPS

now if the airflow is > 45 gms/sec, there is an issue

This airflow value is stored at $C210, increase it to a value greater then what the data log shows.

RBob.

Rebuildman

RBob, yes I have the parameters you speak of. <3200 rpm, (1750), TPS is at .94 volts in data log (%?) just before it leans out the MAF is 47.28 grams/sec.
So now you lose me with the HEX speak, sorry. But if it's not a table in TunerPro or Tuner Cat I have no experience in how to change the hex code.
This certainly seems to be my problem, now I need to figure out how to change the code. Same issue with the SA Latency table you replied to.
I appreicate the help!

Jeff

Rebuildman

I also checked my fuel pressure while driving, had a gauge taped to the windshield and noticed no change in pressure when lean out occured. So fuel supply is ruled out.

Rebuildman

I tried the 'find tool' in tunerpro and used $c210 in the hex address box and the search shows nothing found? Same issue I had with the SA Latency table, I will play with it some more to try and find my problem with location.

RBob

iTrader: (1)

Drop the $C000 portion off of the address and use the remainder. In the PROM that address would be $210. Once the PROM is inserted into the ECM the address is offset by $C000. Because of this the BIN software uses the address with the $C on the front as $C210.

When working with the BIN we need to skip that part. (Broad statement warning) all f-body 3rd gen ECMs have an offset of some nature. Some are $8000, while others are at $D000, and another at $3000 (not inclusive).

RBob.

1MeanZ

iTrader: (9)

My engine is similar to yours in that I have about that TPS voltage at that RPM, but your MAF is programmed to 47gms/sec at that rpm? Holy Cow! My 383 is about half that at 2000rpm. I'm just amazed at the difference. My engine is doing something similar, it just "trailer hitches" above 2000rpm at a light load, but based on the info here I highly doubt we have the same problem!

Rebuildman

I don't know that my MAF is 'programed' at 47 grams/sec but that is what the data log shows is flowing? (for lack of a different term)
RBob, thanks I will try to locate it. I am also guessing the $ is not really a figure to use but just away of saying 'Hex' location. I actually remember somethings from reading about hacking bins several years ago! I also remember it made my head hurt! But I will try again. Go easy on me.
Your help is greatly appreciated!

1MeanZ

iTrader: (9)

If you've modified the MAF tables, you've modified the gms/sec the ECM is using to calculate fuel requirements. When you change the tables or scalars, you are re-calibrating the MAF sensor. The computer sees a MAF voltage output, in the MAF tables, you are telling the ECM how many gms/sec airflow is associated with each output voltage.

In theory we should never have to re-calibrate the MAF. It simply measures air, so if a modified engine draws more air, the ECM should add enough fuel to compensate. The problem is that it's not that simple. Especially in engines with reversion in the intake caused by camshaft overlap etc. I dont know all the reasons why we have to re calibrate, all I can say is that we dont live in a perfect world and it takes tweaking to the tables to supply the fuel that the engine needs.

I'm curious to see your MAF voltage output at that point. I'm not saying 47gms/sec is wrong, but it is double what my 383 is using.

Rebuildman

1meanZ - O.K. I agree with your statement. Here is what I have, @ 47 grams/sec. the MAF table #2 shows about 2 volts, maybe a little more. Target AFR in data log is 13.54, my WB is actually showing a little richer 13.3 - 13.1 so I need to lower the MAF table a smidge. Now this is at a load of 132 LV8 very little throttle .94 V TPS @ 1700 rpm 40 mph. I also probably need to adjust my target afr a little leaner and then drop the MAF table even more at this load. Sound right so far? Your thoughts?
RBob - Revelations! I found the code address of 210 and bingo it highlights a table in TunerPro to make the change to the MAF high diag. threshold which was set at 45! Holy Crap, today is going to be a good day!
Your help is very appreciated!

1MeanZ

iTrader: (9)

at 2.01v my maf gms/sec is set at 42.19. Why do you have your target afr set one full point rich? just curious. As far as my thoughts, I think you should set your AFR for 14.7 and tune for your BLMs. You could probably avoid the MAF threshold if you brought your MAF tables down a little to not run so rich. This would also bring your LV8 down, which is a lot higher than mine as well. My opinion is an AFR of under 14.7 at part throttle loads is just adding to cylinder wash and wasting fuel.

Rebuildman

Thanks for the input, yes I need to lean it out. Reason it's rich is because I didn't want to be lean on a new motor, and then I ran into the 20:1 lean out problem and it really threw me. So now that is resolved, haven't actually tried it yet but I have a good feeling about it, I can lean it out some.
Being that this engine is new and it's a 400 with the HSR and vortec heads I was kinda at a loss for what bin to use and how to modify it, so I was quite happy that I got it to run at all, let alone drive it around town. I tried to error on the side of rich, I also am trying for the first time to tune open loop so the BLM's are out of the picture. Using a WB but it's not being recorded in TunerPro as of yet, need to figure that one out also. I just wired the output of the WB into the the O2 sensor wiring today, not sure what the results will be seeing that I have the closed loop disabled, but will find out soon.
I appreciate the input, it does help. I will work on leaner AFR now that I know whats happening.
Thanks!

RBob

iTrader: (1)

The commanded (target) AFR and the MAF airflow value are two separate items.

The commanded AFR has nothing to do with the calibration of the MAF sensor.

The MAF sensor reports airflow.

The commanded AFR is what the ECM targets.

In the PW equation both items are taken into account.

As for each of your MAFs reporting a different airflow at the same voltage, this has to do with reversion, cam selection, intact tract ducting, PCV valve selection, TB bore dimensions, and a host of other variables.

A MAF sensor isn't the end all to be all. Same as a speed density setup, they both have their pluses and minuses.

RBob.

1MeanZ

iTrader: (9)

My thought was that in order to meet his rich commanded AFR, he has the MAF tables richened up to hit that value. I know the 2 are not linked in the code, but he's got to be adding extra fuel somehow to achieve his commanded AFR. Once he switches back to 14.7 commanded AFR, he'll remove fuel using the MAF tables.

PS, I fixed my problem last night, highway mode was coming in and throwing 9.8* of timing in, obviously that was a problem. I disabled it and will tune the main table for now. As I progress further I'll re-enable it and start tweaking for fuel mileage.

RBob

iTrader: (1)

When switching back to 14.7 AFR, that in itself will cause the PW to shorten and the AFR to go leaner. Shouldn't need to change the MAF tables to achieve it.

The inverse AFR (FAR) is used in the PW calculation. Which is a variable.

If the MAF calibration is off, this can be caused by the injector constant and/or the injector compensation values (short PW comp) also being off.

That is a bit much timing even for hiway mode. Was the fueling also being commanded leaner? Some code bases don't tie highway spark to highway fuel, which doesn't really make sense.

RBob.

1MeanZ

iTrader: (9)

Ok so to re-state that in my words to make sure I understand, the commanded AFR is used as a factor to calculate PW. if you change the AFR, you're just changing one variable in the equation. The result is longer or shorter pulsewidth depending on what you set that variable to be. Got it. That makes a lot of sense now. thanks.

My highway mode timing is not what I specified, its what was in the base .bin that I'm working with. I didnt realize what was even happening until I began to diagnose my problem further last night. My datalog caught the problem and made it pretty easy to figure out. My highway mode afr is 15.5. I'll mess with highway timing once I get my fuel polished off, i'm pretty close now.

Rebuildman

Tried the new bin with the corrected MAF high diag. threshold changed. No more 20:1 lean out! Thanks RBob! Now I feel safer leaning out the mixture.

RBob

iTrader: (1)

With the default airflow value so low you may want to change some of the default values. This will allow the car to be safely driven in the event of a MAF failure. What was happening is that the ECM ignored the MAF and went N-Alpha to create a default airflow value.

There is a table that is RPM based along with a TPS% and IAC factor. These values are in the "ERR # 33/34 Params" section of the hac.

RBob.

Rebuildman

I would love to do that! But it's beyond my capabilities at this time. I need to learn more about the HAC's and how to change the code. This is probably not the place in the forum to learn. I need to go to the 'source code for dummies' class. What I do know about this I found in the HAC for ARAP on Moates website, opened it and found a LOT of INFO! Just for the heck of it I picked the first thing I came across and pasted it here. I guess I would need to find the part in the HAC that addresses the items you mentioned. Even if I find it, I wouldn't know what to do with it. So for now I will have to leave it as is. Hopefully the MAF will not fail too far from home! Yeah right!
So anyway, am I looking in the correct place? I guess I need to take this and compare it to my ARAP bin at address 015 and see if I can make any sense out of it.
Thanks for all your time and effort RBob! If you ever need a 757 or a 767 worked on, let me know, I'm better at that!



LC015: FCB $04 ; %0000 0100 2nd AIr Flow Mode Word
;
; Bit 0 =
; 1 =
; 2 = 1 = 2 AiR VALVES, 0 = 1 VALVE
; 3 =
;
; 4 =
; 5 =
; 6 =
; 7 =

RBob

iTrader: (1)

The parameter you chose is bit mapped. Meaning that each bit (of the 8; 0 - 7) changes an option in the code. Note that not all bits may be used.

LC015: FCB $04 ; %0000 0100 2nd AIr Flow Mode Word

The LC015 shows the location in hex ($15)

The $04 shows the value that is in the BIN in hex

The %0000 0100 shows the value that is in the BIN in binary

%0000 0100 = $04

As for the option that is listed, it tells the ECM that the AIR system has either two valves or one valve. With bit 2 set (which in this cal it is), the vehicle will have a two valve AIR system (note that the hac may be incorrect, there are errors in them. Just need to keep this in mind if changes don't do what is expected).

For the MAF default parameters, see my previous post for the quoted string. Search for that in the hac (ALT->Edit->Find).

RBob.

Rebuildman

You have an amazing talent! Being able to convert your knowledge to terms a lay person can understand. I am going to go and check out that hac right now. I don't like the statement about hac's not being correct all the time, that's a little unnerving.
Is your head bigger than most peoples? Has to be, to store all that info.

Thanks, one more time!

Rebuildman

O.K. I found the location of 33 and 34 error coding in the hac. It shows values in decimal. I opened my arap bin, changed the values to decimal and I found the following at lc210 50, which is what I changed my arap to in tunerpro, the hac shows 45 which is the stock setting. So far so good.
Now in the section you referenced in the thread err 33/34 the hac shows lc21E 167, lc21F 21, lc220 1024. My arap bin shows 167, 21, and 4. I see the explanations but they don't mean to much to me. I also see how the values were arrived at ie: 167 / 2.56 = 65.2, 21 / 256 ?, 1024 / 256 = 4.
My arap shows 4 at lc220 the hac shows 1024. Is this one of the mistakes you speak of? Any clearer explanations for these locations?

I sorta feel my head getting bigger. But nothing noticeable.


;--------------------------------------------------
; ERR # 33 Params
; >> MAF Sensor Hi <<
;--------------------------------------------------
LC20F FCB 38 ; If TPS >= 14.8%, set ERR #33
; CALIB = Arg * 2.56
;
LC210 FCB 45 ; If Air Flow <= 45 gms/sec, disable ERR #33
;
LC211 FCB 96 ; If > 0.6 Sec, Set code #33
; CALIB + Arg * 160
;
LC212 FCB 10 ; 1000 Msec ERR 34 Min time req.
; CALIB = Arg * 10
;
LC213 FCB 128 ; If RPM > Disable #33
; RPM/25
;
LC214 FCB 100 ; 10.0 VDC PUMP VDC FOR ERR 33 ENABLE

;--------------------------------------------------------
; ERR # 34 Params
; >> MAF Sensor Lo <<
;--------------------------------------------------------
LC215 FCB 0 ; 0049 COUNTS, IF ANALOG CNTS x 7 or
; PP2 DLT LT 0 SET ERR 34B
;
LC217 FCB 10 ; ERR #43, 1 Sec Min time req.
; cal + Arg * 10
;
LC218 FCB 24 ; If RPM > 600, Ck ERR #34B
;
LC219 FCB 16 ; If TPS > 6.25%, Ck ERR #34B
; cal = Arg * 2.56,
;
LC21A FCB 0 ; If LV8 >= 0, Ck ERR #34
LC21B FCB 130 ; If LV8 <= 130, Ck ERR #34B
;
LC21C FDB 0049 ; 0049 COUNTS, IF ANALOG CNTS x 7 or
; PP2 DLT LT 0 SET ERR 34B


;--------------------------------------------------------
; ERR # 33/34 Params
; >> MAF Sensor Hi/Lo <<
;--------------------------------------------------------
LC21E FCB 167 ; If TPS > 65.2%, then use for MAF Default
; CALIB = Arg * 2.56
;
LC21F FCB 21 ; Gms/sec Scale factor
; gms/sec/IAC Count
; CALIB = Arg * 256
;
LC220 FDB 1024 ; 4 Gms/Sec default air flow offset
; CALIB = Arg * 256

;--------------------------------------------------
; Default air flow offset per pct TPS vs RPM

RBob

iTrader: (1)

The difference is in a value that uses one byte and a value that uses two bytes.

Take these as an example:

The FCB defines the value as using a single byte (an 8-bit number).

The FDB defines the value as using two bytes (a 16-bit number).

For the two byte value need to get the byte (value) at $220 and the value at $221. The value at $220 being 4, and the value at $221 being 0. Then need to do some math. Mutiply the first byte by 256 and add the value of the second byte to that result.

4 * 256 + 0 = 1024

Another example: first byte is 20 and the second byte is 12:

20 * 256 + 12 = 5132

RBob.

Rebuildman

O.K. now here is where I get lost. So all FDB values use two bytes, correct? The two values being the first one at the FDB location example: LC220, do you just automatically take the byte value at the very next location? In your example you give LC221 and a valve of 0, but in the hack there is no LC221. it goes from LC220 right to LC222.
Also why would the ARAP HAC use a FDB at LC220 (1024) and my ARAP use a FCB (4) at LC220?
All these little things that you need to know about are what makes changing the codeing for a guy like me just about impossible without starting from scratch and learning how to do it.
But I appreciate you taking the time to answer and trying to explain it. I enjoy trying to learn how to do it, but the learning curve seems pretty steep and I can see how it would be very easy to screw up!

pandin

The lean issue I have with the arap bin and all the $32/$32B bins is the O2 sensor going cold at PT (55 mph). What this does is as the unheated NB O2 (tried 3 different ones, 2 used, one new, all tested good) drops in temp, it also drops it's voltage output.

The ECM reads this as a lean condition and raises the BLM to 160. Raises it so far that I can smell the richness at idle. I will install a heated O2 and a WB in the near future. Adjusted BLM limit to 140 for temp fix.

This leanness shows up even with 24# injectors @ 50# fuel pressure, with a stock 23# injector constant and stock new reman MAF. O2 is in the stock location, factory vette shorty header, in the exhaust pipe, just past the 4-1 collection point (1 ft from the head).

Shows up in the data log as low volts (.5>), low cross counts (dead spots), and a slowly rising BLM til it hits the max 160.

tequilaboy

Pandin,

I suspect you really are lean. You may need more low pulse width offset for your injectors. You mention you can smell it at idle. Can you somehow maintain 55 mph in trim cell 0 or 1 that would affect the idle?

Do you have any fuel trim cells running strangely rich or lean compared with other nearby cells? Like cell 14 running at 108 while 11 and 15 are fine, or maybe 14 and 15 are fine and 7 or 11 running strangely lean? You may have some injector linearity issues at low (under 2 ms) pws.

My car drops into open loop pretty quickly with a cold O2 sensor (like within 1 second) when the O2 is continually within the 200-700 mv hysteresis band.

pandin

At idle injector pulse widths are at 3 ms should be at 2 ms or just under (earlier datalogged runs), the injectors (ford 2nd design 24#) MINPW (ms) 1.053, is not likely the cause (from the factory injector setup sheet).

All the cells travel together weather rich or lean, richer at lower BLM's (0-1) and leaner at higher ones (14-15).

When the O2 goes cold it affects all the cells. 55 - 58 mph is cells 6 -10 depending on the load. I run with no cat and mufflers 2-1-2. Just changed to 2 1/4 duals w/h pipe and magaflow mufflers and it got worse. Before I was running a weak 187 peak MAF, ran fine all winter, but the pipes rusted clear out (under the rear end).

When the O2 stays hot the BLM's ran at 160 idle 0-1 but all the rest were 128. This was with stock the 32 chips. When I tried to raise the injector constant to 25# from 23# everything went rich, BLM's were at 108 and the engine wouldn't hardly run. I thought higher injector constants leaned the mix and lower riched ????

The fuel pump relay went bad, then found the pigtail is bad, the insulation gone just below the relay connector. When the new pigtail comes in, I will splice in another relay for the heated NB O2 and the WB. I added two extra bungs for WB and right side for testing/tuning.

I wanted to settle these lean issues before trying the simulated NB, don't want too many things changed a once. (exhaust, arap bin, prominator,new maf, and WB).

tequilaboy

3 ms pw at idle. Wow, I bet its rich.

I would expect to see something like 1.5-1.6 ms. Has the battery voltage compensation been increased or the low pulse width offsets? What's causing the high pw???

Is the MAF flow reasonable (around 7-9 gm/sec) at idle or 16-20 gm/sec. If the latter, that would explain the 3 ms pw.

Increasing the injector constant to 25# should have made it leaner and shorten the pw. Doesn't make sense that it would get richer.

I assume that you changed both single fire and double fire constants and that the car is running double fire mode.

Something strange is going on. Good luck in your efforts.

RBob

iTrader: (1)

Also why would the ARAP HAC use a FDB at LC220 (1024) and my ARAP use a FCB (4) at LC220?

This is where the errors come in. But easy enough to get past. This is why: a single byte value will range from 0 through 255 inclusive. Once a value goes over 255 then another byte is required.

With two bytes the value can range from 0 through 65,536 inclusive. {edit: should be 0 through 65,535 inclusive. The 65,536 is incorrect.}

There are other ways to tell the length of a data value. tha tis to look at the CPU instruction. If it loads a 16 bit value, then without a doubt it is a two byte value. This would be instructions as such: LDX or LDD.

RBob.

Rebuildman

O.K. RBob, I give. Now I know this is above my current ability. I think I will leave the code hacing to you. I will stick with the TunerPro and TunerCat software to do it for me. At this point I would do more damage than good to try and alter the code on my own.
But thanks for the input!

----------
I solved any O2 sensor problems by going open loop. No narrow band sensor installed. I have a wide band just to monitor my tuning, when thats done it comes out and goes to the next project.

pandin

The idle grams/sec are at 6 up from 5, this is at 575-625 rpm, load is same 50-55.

Everything in the chip is stock as to fueling only change was to lower max blm to 140 from 160. It does this same thing with stock memcals.

The ECM doesn't go back into open loop, what it does is stop the learning (BLM's don't update up or down). Then after a time the integrator resets to 128 from 180 and that is where it runs till the O2 heats up again.

When the ECM transitions into, closed loop at start up, the engine makes a noticable change to the good. I suppect the open loop is rich do to the low injector constant in the program.

Also when the engine runs too rich, the O2 soots up, and until this soot gets burnt off the O2 runs cold longer. A good cruise at 70 mph really improves things for a little while.

This engine is set up pretty mild, stock intake, exhaust, and only running a 206/212 @ .50 cam .480 lift 1.5 RR. Performer heads 10:1 flat tops.

pandin

Default air flow offset per g/sec vs RPM, the hac doesn't explain this very well.

I seem to remember the "Math" being explained but searching is not finding it.

RBob any words of wisdom as to matching this table to the recorded air flow.

I have even disconnected the Maf and ran on the default values. Manually recorded the Maf's volts out by running the Maf on straight battery power.

tequilaboy

The default airflow calculation is pretty simple.

There is a 4 gm/sec base offset (too low) which represents the aiflow through the closed throttle, an IAC gain factor of 0.082 gm/sec/count (too high) which represents the gm/sec/count and a tps% vs rpm table which represents the additional airflow as a function of throttle position and rpm.

There is also a max tps% limit value 65% (too low) to limit the actual tps% input to the table. I would increase this to 100% for starters.

If you look at the table and then think 100% throttle, the table then represents the open throttle airflow for any given rpm at 100% throttle.

The table will be basically rescaled as a function of throttle position. 50% throttle cuts it in half, etc.

The whole shebang is then limited by the max maf flow vs rpm table.

Increase the base offset to roughly equal your min air rpm/100. If min air rpm is 600 rpm, use 6 gm/sec. If 800 rpm use 8 gm/sec.

Reduce the IAC gain factor to about 0.0547 gm/sec/count to dial in the IAC contribution.

Use the tps/rpm table to dial in your blms. Use the max maf table as a final check to avoid going too rich at any rpm.

pandin

tequilaboy Thanks for the explanation.

The attached file:

Has at the top, running a 383 engine at idle, with Maf unplugged and power up on the battery (MAF DEFAULT AIR FLOW).

Default Grs/sec readings using Auto Xray Scanner

Maf's at wot
SIA is a Bosch 240 grs/sec
CB is a Microtech 255 wot
205 is a Bosch 205 wot
187 is a Bosch 187 wot

Notice how the volts and grams a sec get less. This makes each Maf run abit leaner at idle, the SIA and CB reverse at wot, CB is the richest and SIA is next.

The bottom table is set up the same, engine running on Arap default table and Maf unplugged, powered by battery. The readings were taken by slowly raising the idle speed to the next grs/sec and recording the volts out of the Maf.

tequilaboy

pandin,

I'm curious how you're measuring the MAF voltage with it unplugged and powered by battery?

As far as I'm aware, you need the 5 Volt reference signal provided by the ecm to be pulled down by the MAF to result in a meaningful voltage signal.

In this regard, the MAF behaves similarly to the coolant temp sensor or MAT sensor.

With my Ford MAF set-up (with the signal input into MAP2 instead of VMAF), I can monitor the MAF voltage while running in default mode, by first setting a MAF failure and then reading my Ford MAF/MAP2 A/D input channel instead of the fuel pump voltage with corresponding bin changes to make this possible

The same could be done for a stock style tpi MAF if you want to log the real MAF voltage along with the default airflow.

pandin

I was using a Fluke 87 Digital volt meter, output to gnd. The volts out to gnd seemed to vary "reasonably" close to the change in grs/sec.

This file attachment is using a shop vacuum cleaner for air flow. The Maf's were hooked in "series" with the same air flow through both. Both were powered by a DC power supply to 14 volts. Air flow checked with a anemometer.

The difference in the volts out is apparent in this test also. The reverse of the wells vs the rebuilt Bosch is evident here also.

I tested these Maf's at idle on my 325 LS motor, results were similar, but couldn't get a reading on the 325 grs/sec at idle but idle was lower like 5 grams/sec (lower volts output from tested Maf)

I wondered why the outputs changed alittle when the maf were both powered from the common DC power source.