TPS% value 1227747 ADX

[EagleMark]

Well I finally got the vacuum gauge working as well. Dulicated the MAP kpa, changed the name, changed the values and found the correct calculation of:
29.9-(X*.349)

[EagleMark]

Well actually what I meant was if you make a conversion for that as follows:
0.00V = 0% TPS
0.05V = 1%
0.10V = 2%
~~~~~
0.55V = 11%
0.60V = 12%
~~~~~
5.00V = 100%

Then for instance at idle the TPS sensor is reading .56 volts, this is what the ECM sees. But your .adx would show (0.55V) 11%. Right? That can be worked around but.......

somewhere in the '7747 there has to be a conversion written to put out a %TPS, there are tables and constants that are defined as %TPS in a column or row. So somewhere in the bin there is a conversion already. It is just matter of feeding it out through the data stream.

For instance:

;
; CK FOR LOW TPS ERR TEST ENABLE
;
D6D8: LDAA LD50D ; ERR ENABLE WD 1
D6DB: BITA #$04 ; b2, ERR 22, TPS lo
D6DD: BEQ LD6F8 ; BR IF NOT b2
; ... else
D6DF: LDAA L0048 ; TPS, VDC
D6E1: CMPA LD523 ; TPS limit, 0.20 VCD, (ERR 22)
D6E4: BCC LD6F8 ; IF TPS > .195 BR
; ... else

So here, as the code is running when it reaches D6DF the instruction is LDAA L0048 which to paraphrase "load from register A L0048" ; comment says TPS, VDC, this is also the register that the datastream spits out TPS Volts from.

F0FF: LDAA L001A ; RPM/25
F101: ASLA
F102: BCC LF106
; ... else
F104: LDAA #255
F106: LF106 BRA LF120
F108: LF108 LDAA L0049 ; CURRENT TPS VALUE
F10A: LF10A PSHA

Now here is a portion of the TCC routine. Here at line F108, the instruction is again LDAA but this time it is "from register L0049" ; and is commented TPS Value.

My point is there are many references to TPS% in both of the hacs, and the .xdf posted here. It is just a matter of getting that value to be put in the data stream.

I found a $42 hack and went through it and your right there are many instances where TPS% is used and some of them are numbers lower than 5% which could be confused with voltage that can not go lower than 5. So the ECM must take idle voltage and use it as 0% or 1% and use it in it's routine. So you are exactly correct.

But I have no idea how to find TPS % coming from the ECM. Or even if it is there to begin with, PCM like the $OD have sepreate address for voltage and percent. if I knew where it was I could set it up in the ADX. May try to figure out how to use the lookup table and create one from voltage like you showed above.

[gregs78cam]

Ok. so we are on the same page now.

the number that gets spit out into the data stream is a 0-255 number and that number is is stored at a location on the EEPROM. L0048 for TPS VDC, and L0049 for TPS%. so we just have to tell the ecm to put out the % instead of the VDC. because really the only useful time for seeing VDC is settign minimum TPS, and trouble shooting, right? If you look through the hac and look at the aldl data stream section which starts at $04E7, and scroll down to $04F5 you see this:

LD4F3   FDB $001C   ; 7.  RPM/25
     ;   RPM = N * 25
     ;----------------------------
LD4F5   FDB $0048   ; 8.  TPS (A/D, 0-255)
     ;   VOLTS = N * .0196
     ;----------------------------
LD4F7   FDB $00A4   ; 9.  INTIGRATOR, (0-255)

this is the part of the code that tells the ecm where to look in the RAM area of the chip to find the data to put into the outgoing data stream. If you open TP5, and open the hex editor scroll down to $04F5 (LD4F5 above) and change the two bytes there from 00 48 to 00 49. that should tell the ecm to send the data from the tps% address instead of VDC.

[EagleMark]

Not sure about the same page... but I think I am on page one of a book you read!

Going back and looking at these I was wrong and most of these settings are TPS%. The reason they have been driving me nuts for years is there is no TPS% to read in data. The old XDF (mask) have conversions that do not go low enough to be close to these settings!

Look at other ECM data and you have both TPS voltage and TPS Percent worded one way or another. Look in the ADX and find TPS Voltage at 0x08 and percent at 0x09 or differant in other ADXs.

I see what your telling me will work but only if someone changes there hex to begin with, so that won't work for all. What I was trying to do was just fix the ADX so it has an accurate reading for TPS voltage and TPS percent. So although you have found a way to fix my issue, I am trying to find a way to fix the ADX for everyone. But from what your telling me there is no TPS % being put into the datastream to begin with?

If you look at MAP in the $42.ADX it only has one address for voltage and KPA, but differant conversions and they are accurate.

No matter what I try for a conversion on TPS I can't get it to work, it will lower or riase both ends? Probably because of my lack of math skills... So I tried doing a lookup table to convert the numbers like it does for coolant temp but have not got that dialed in, or even close yet... I think it's the only way I will get this to work in the ADX?

So now that you have identified the problem with your hex/hac skills I know that the information is just not coming into datastream to begin with.

So it's back to how to fix the ADX ?

[gregs78cam]

I see what your telling me will work but only if someone changes there hex to begin with, so that won't work for all. What I was trying to do was just fix the ADX so it has an accurate reading for TPS voltage and TPS percent. So although you have found a way to fix my issue, I am trying to find a way to fix the ADX for everyone. But from what your telling me there is no TPS % being put into the datastream to begin with?......

So it's back to how to fix the ADX ?

OK. I see your point about wanting to make the ADX work for everyone instead of just fixing an issue in one particular instance. But that was my point in an earlier post. Have you looked through the XDFs and ADXs put out by 93V8S10. He put 'read me's in those to give the individual the ability to change their XDF and ADX to what they want. I ended up making a constant in the parameters that lets me change the .bin easily without the possibilty of screwing up the .bin. Something similar is what I was suggesting we put in the XDF and ADX. Basically we make a parameter in the XDF that with just a few keystrokes can change what gets put out in the datastream. But the XDF and ADX have to be used together.

I can write the XDF and ADX parameters in no time I just need to have something running to verify results. I can get a '7747 powered up on the bench next week, and see if I can get it to work. Or if you already have one set up I could drop by next week and we could get it done in no time at all.

[gregs78cam]

So it's back to how to fix the ADX ?

Try this conversion

= (X - 25.6) * 0.435920

this should be basically removing the 0.5V (=10% * 5VDC) minimum TPS reading from the equation. Actually it is subtracting the the first 25 steps then dividing the remaining 229.4 steps by 100, before converting the 0-255 value to %

[EagleMark]

Yeah I was looking at his stuff but those have a lot more to offer.

My bench is in the middle of rewiring a harness, but I do have a running 7747 in my Suburban and we don't even need that because I have been checking these with datalogs!

I'm going to try your conversion now and see what it does. If not back to seeing if I can make the lookup table work...

[gregs78cam]

well on paper when I convert various 0-255 values to % TPS it works out. It still won't be as accurate as the data from the ECM, due to MIN AIR adjustments, and MIN TPS position, but should be pretty spot on.

[EagleMark]

Pretty spot on it is! I went out and adjusted my TPS to .53. Did a datalog, not running but idle to WOT and all spots in between. Your conversion came out to .05% so I don't think you could ask for anything better.

But the high end I only get 4.26 volts which shows 75.2% . That more than covers all parameters in XDF for things like PE. I checked and it is WOT. So I looked at other datalogs on vehicles I worked on and I know were set to .54 on TPS and verified I was getting WOT and never got more than 4.48. I know there's a 5.0 volt reference there but I don't think a TPS uses all of it.

Could you redo that math to only use 4.5 volts? Or would that be dumb? It would look better in datalogs if it were closer to 100% when guys are looking at WOT settings for PE etc...

Next time you stop by you have to teach me some of these basics so I can start learning this stuff!

[EagleMark]

Wait a minute! That conversion changed from the first time? Anyway now I get .06% at TPS .53 volts and 83.9% at TPS 4.26 volts so it got better!

[gregs78cam]

Sweet. The ECM converts the analog 0-5V signal to a digital 0-255 and stores it as a hex number, then our ADX converts that 0-255 hex number to what ever we tell it. I wonder if the ecm is not using all 0-255 'digits' for the TPS.

If you set the ADX to simply show raw hex number, what does it show at WOT?

[gregs78cam]

Ok, that makes sense, because at 5V it would be around 100%. I have looked at my datalogs and usually see around 4.89-4.95 VDC at WOT. Have you actually made sure that butterflies are open all the way and read what the voltage is? I think you may have something keeping them from opening all the way, if you are only seeing 4.26V.

[EagleMark]

Shows a bunch of weird numbers and letters but at 4.26 volt TPS WOT the TPS% shows 53

[EagleMark]

Ok, that makes sense, because at 5V it would be around 100%. I have looked at my datalogs and usually see around 4.89-4.95 VDC at WOT. Have you actually made sure that butterflies are open all the way and read what the voltage is? I think you may have something keeping them from opening all the way, if you are only seeing 4.26V.

But are your datalogs from a 1227747? Mine are! Your 16197427 has a seperate datastream for TPS%

[gregs78cam]

Sorry I hit the wrong button and edited your post...