$EE Code for electric water pump on AIR output

Hi!

I'm trying to patch $EE code to use AIR pump output to control electric water pump on an LT1.
Here are parts of E-side disassembly that control the AIR pump:

Code:

;~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ; ; 0x531F some AIR Pump code (probably enable at startup ???) (part of hc11_irq_00) (executed when L0023 b0 is clear) ; ;~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 531F B6 20 28 ldaA b_2028_sw_b3vats_b2spddens_b0tcs 5322 85 04 bitA #%00000100 ; test AIR Pump switch bit (E-side 0x2028 b2) 5324 27 1B beq @8 ; Bra to continue if bit clear 5326 B6 01 92 ldaA l_0192_CoolTmp_fltrd_$f0 ; Load Coolant Temp 5329 B1 23 B2 cmpA L23B2 ; Compare to AIR pump disable temp threshold (E-side 0x23B2) 532C 23 10 bls @7 ; Bra if Temp <= threshold 532E B6 01 28 ldaA L0128 ; Load ??? 5331 81 AB cmpA #$AB ; Compare to 171 5333 24 06 bcc @6 ; Bra if L0128 >= 171 5335 CE 18 00 ldX #$1800 ; Load Hardware output address 5338 1C 00 02 bset 0, X, #%00000010 ; Set AIR Pump output bit 533B 14 28 10 @6 bset L0028, #%00010000 ; Set AIR Pump ON flag 533E 7F 02 44 @7 clr L0244 ; Clear AIR Pump Timer ;~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Code:

;~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ; ; 0x4962 Update AIR Pump State (part of loop_c_00) ; ;~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 4962 B6 01 28 @12 ldaA L0128 ; Load ??? 4965 81 AB cmpA #$AB ; Compare to 171 4967 24 29 bcc @14 ; Bra if L0128 >= 171 4969 12 2F 01 2D brset l_002f_fts_2_stat_byte, #%00000001, @15 ;Bra if T-side AIR request 496D 12 28 20 1E brset L0028, #%00100000, @13 ; Branch if AIR ON flag is set 4971 13 28 10 1D brclr L0028, #%00010000, @14 ; Branch if AIR Startup Request flag is clear ; Increase AIR Pump timer 4975 B6 02 44 ldaA L0244 ; Load AIR Pump Timer to A 4978 8B 01 addA #$01 ; Add 1 to A 497A 82 00 sbcA #$00 ; Subtract 0 from A with carry 497C B7 02 44 staA L0244 ; Store AIR Pump Timer 497F B1 23 B3 cmpA L23B3 ; Compare increased AIR Pump Timer to AIR pump timer threshold (E-side 0x23B3) 4982 25 1F bcs @16 ; Bra if A lower than threshold 4984 14 28 20 bset L0028, #%00100000 ; Set AIR ON flag 4987 15 28 10 bclr L0028, #%00010000 ; Clear AIR Startup Request flag 498A 7F 02 44 clr L0244 ; Clear AIR Pump Timer 498D 20 03 jr @14 ; Here if AIR Pump should be turned OFF (T-side AIR request clear and AIR ON set) 498F 15 28 20 @13 bclr L0028, #%00100000 ; Clear AIR ON flag ; Here if L0128 >= 171, AIR Pump timer threshold met or AIR Startup Request flag clear 4992 CE 18 00 @14 ldX #$1800 ; Load Hardware output address 4995 1D 00 02 bclr 0, X, #%00000010 ; AIR Pump OFF 4998 20 0F jr @17 ; Bra to continue ; 499A 14 28 20 @15 bset L0028, #%00100000 ; Set AIR ON flag 499D 15 28 10 bclr L0028, #%00010000 ; Clear AIR Startup Request flag 49A0 7F 02 44 clr L0244 ; Clear AIR Pump Timer ; Here if AIR Pump timer threshold not met 49A3 CE 18 00 @16 ldX #$1800 ; Load Hardware output address 49A6 1C 00 02 bset 0, X, #%00000010 ; AIR Pump ON 49A9 7E 4E 38 @17 jmp L4E38 ;~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

So far I've managed to get it working most of the time, pump swithes on when engine is running and turns off when engine stalls, but sometimes it turns off while engine is running. It does not report that it's ON in EEHack, but I can still enable it using EEHack control and then the AIR ON flag is set.
Here is my code patch:

Code:

;~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ; ; 0x4962 Update AIR Pump State (part of loop_c_00) ; ;~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 4962 B6 01 5F ldaA l_015f_RPM_2.5/bit ; Load engine speed 4965 CE 18 00 ldX #$1800 ; Load Hardware output address 4968 12 2F 01 10 brset l_002f_fts_2_stat_byte, #%00000001, @14 ;Bra if T-side AIR request set 496C 81 A0 cmpA #$A0 ; Compare to 400rpm 496E 24 0C bcc @14 ; Bra if higher or same 4970 81 78 cmpA #$78 ; Compare to 300rpm 4972 22 0E bhi @15 ; Bra if higher 4974 15 28 20 @13 bclr L0028, #%00100000 ; Clear AIR ON flag 4977 1D 00 02 bclr 0, X, #%00000010 ; AIR Pump OFF 497A 20 06 bra @15 ; Bra to continue 497C 14 28 20 @14 bset L0028, #%00100000 ; Set AIR ON flag 497F 1C 00 02 bset 0, X, #%00000010 ; AIR Pump ON 4982 20 25 @15 bra @17 .... 49A9 7E 4E 38 @17 jmp L4E38 ;~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Maybe some of you may get me on the right track here, there must be something I don't see in the code that turns the AIR output OFF. Also, do you know what is in L0128 RAM address? I can't figure it out.

I think the air pump is not good candidate for the patch. It will be over complicated and too long There are some unused pins like pin b9 to run the waterpump.
It is software address is byte_1800 $01. You can use engine on flag byte_23 $80 for trigger, and add other stuff like ignition on or Coolant temperature corrections.

Byte_128 is ad Ignition voltage.

If you still have troubles I will compile a quick patch for you

Thanks!

So, the most basic solution would be to change this:

Code:

4935 CE 18 00 ldX #$1800 4938 12 2F 02 05 brset l_002f_fts_2_stat_byte, #%00000010, @7 493C 1D 00 01 bclr 0, X, #%00000001 493F 20 03 jr @8 ; 4941 1C 00 01 @7 bset 0, X, #%00000001

to this:

Code:

4935 CE 18 00 ldX #$1800 4938 12 23 80 05 brset L0023, #%10000000, @7 493C 1D 00 01 bclr 0, X, #%00000001 493F 20 03 jr @8 ; 4941 1C 00 01 @7 bset 0, X, #%00000001

And if I wanted some more functionality, then I'd insert a jump at 0x493C to additional subroutine in some unused space to test conditions to run pump w/o engine running.

Could you point me to some free space I could safely use?

You found very basic elegant solutions, since b9 is already set in the code.
Just change at
4939

2f 02 to 23 80

and wire the water pump relay to pin b9. For safety reason you can check polarity and functionality of the pin.
As I can remember when the bit is set, the pin provide ground path for the relay.

OK, so the subroutine would look like this:

Code:

;~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ; ; Routine to enable Electric Water Pump when engine not running ; ;~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 9000 1D 00 01 bclr 0, X, #%00000001 ; Disable WP (X already set to $1800) 9003 B6 01 28 ldaA L0128 ; Load A/D Ignition Voltage 9006 81 30 cmpA #$30 ; Compare to 48 9008 23 0A bls @1 ; Bra to return if Ign Voltage <= 4.8V (Ignition off) 900A B6 01 92 ldaA l_0192_CoolTmp_fltrd_$f0 ; Load Coolant Temp 900D 81 B9 cmpA #$B9 ; Compare temp to 99 deg C 900F 23 03 bls @1 ; Bra to return if Temp <= threshold 9011 1C 00 01 bset 0, X, #%00000001 ; Enable WP (X already set to $1800) 9014 7E 49 44 @1 jmp L4944 ;~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

And:

Code:

4935 CE 18 00 ldX #$1800 4938 12 23 80 05 brset L0023, #%10000000, @7 493C 7E 90 00 jmp L9000 ; jump to new subroutine 493F 20 03 jr @8 ; 4941 1C 00 01 @7 bset 0, X, #%00000001

Now I just need 23 bytes of program memory to add my subroutine. Kur4o, can you point me at some usable address?

[EDIT] I've found E-side 0x9000 - 0x93FF marked as free memory in EEXtra.xdf, so I guess the subroutine goes to 0x9000 :)

Half the bin is empty.
You can start at the lower section to keep all patches nice and tide.
I suggest to use F000-FF00 memory range.

also use subroutine bd code.

at 4935 put
BD F0 00 20 0A

at F0 00 write your subroutine and add the return instruction [39] at the end

just curious, why do you want ECM control of the electric waterpump?

and if so, why not just use the fuel pump relay to trigger a second relay if you only want it on when the engine is running? (although you'd get a prime cycle, that shouldn't be a big deal...)

it's a cool project, just curious

Quote:

Originally Posted by steveo

just curious, why do you want ECM control of the electric waterpump?

and if so, why not just use the fuel pump relay to trigger a second relay if you only want it on when the engine is running? (although you'd get a prime cycle, that shouldn't be a big deal...)

it's a cool project, just curious

Because... why not? :)

A feature to have the pump work on hot engine while it is off might get useful sometime.

Did you ever consider PWM control and remove the thermostat? By remove, I mean disable it in the full open position or modify the pump to do the equivalent.

Then, you'd cool more than just down to the thermostat closing temperature by running the pump with the engine off.

Quote:

Originally Posted by lionelhutz

Did you ever consider PWM control and remove the thermostat? By remove, I mean disable it in the full open position or modify the pump to do the equivalent.

Then, you'd cool more than just down to the thermostat closing temperature by running the pump with the engine off.

I didn't think about it. Removing the thermostat is not a good idea, as the heat-up using the main circulation (through the radiator) would take much longer. The heater would also be less effective without it.
The PWM control might be useful to shorten the warm-up time, and maybe to free up a fraction of horsepower lowering PWM at WOT momentarily, but not much more. The pump in the car I'm working on consumes only around 7 amps, so almost nothing for the alternator.

The modern cars have thermostats with electric heater integrated in the viscous fluid to vary the opening and closing temperatures some. Anyway, why would you want to cool the car to temperatures lower than thermostat temp?

The idea is that the PWM runs the pump slow during warm up. Decently implemented, a PWM pump and PWM fan will maintain a fairly constant engine temperature. Some OEM's are doing PWM instead of letting the thermostat regulate the engine temps. They probably build the system to have flow to the heater core first though using some other method.

You don't want to keep cooling your engine to thermostat temperature or even below it when it's OFF?

1 Attachment(s)

If anyone else would want to use this patch, here is XDF with the code patch so it can be applied easily.
Code is tested in car and all works fine, so enjoy ;)

I would like to add this to my PCM. Too many close calls with just a toggle switch...

Is there a tutorial or guide of how to add a patch to a XDF ? I see in Tunerpro there is a 'Control + N' which allows me to create a new XDF Object with an option of a patch.

Then what?

Quote:

Originally Posted by el Jefe

I would like to add this to my PCM. Too many close calls with just a toggle switch...

Is there a tutorial or guide of how to add a patch to a XDF ? I see in Tunerpro there is a 'Control + N' which allows me to create a new XDF Object with an option of a patch.

Then what?

You don't need to do any of that; you can just use the XDF that dzidaV8 posted without any changes.

If you have some other XDF you really prefer to use and want the water pump patch he provided, then just open dzidaV8's XDF in a text editor, find the sections relating to that patch, then copy and paste them into the appropriate place in your favorite XDF. I wouldn't even bother using TunerPro to do that; I'd just use Notepad++.

Thanks for the reply NomakeWan.

Maybe I am making this too hard for myself. I have been working with $EE-16188051-V3.xdf from here. Before I posted my question above I had already opened the file provided by dzvidaV8 in post #12 in windows Notepad. The file appears to follow a programming routine similar to C++ and is fairly small.

Based on your advice I opened $EE-16188051-V3.xdf in Notepad and found the file much larger and gibberish, which I suspect is Notepad translating binary to ascii text. I don't think what was posted by dzvidaV8 is a complete XDF file.

Then I loaded the XDF from dzvidaV8 into TunerPro and it literally has only 1 Scalar and 1 Patch, both relating to driving an EWP off of pin B9.

To me this logically means I must somehow "import" the contents of dzvidaV8's subroutine into $EE-16188051-V3.xdf.

I am sure this is something simple I am missing and again, thanks for the help.

1 Attachment(s)

If you opened an XDF file and it appeared as gibberish, that's not normal. I'm on Windows 7, and tried opening the $EE XDF I commonly use (EEXTRA v0[1].001[MOD_V0005_03_02]V6_3.xdf) in both Notepad and Notepad++. In both cases, it opened as a normal text file, as is to be expected. I have attached a screenshot for reference. My XDF is ~1MB in size.

The file dzidaV8 posted is just the water pump patch. If you select it as your XDF in TunerPro, you will only have two options available--installing the patch, and setting the minimum coolant temperature threshold to shut off the electric water pump after the engine has been shut off.

If you want to add those two items to your existing XDF, you would first need to figure out what's going on with said XDF to prevent you from opening it like a normal text file. Perhaps Notepad++ will work better for you. I can't say.

However, if all you want to do is add the electric water pump patch to your bin, you can just use the XDF that got posted to do it, then switch back to your preferred XDF when you're done.

I have made progress. It helps to open the XDF file and NOT the Bin file.

However,

Whenever I right click the patch and choose 'Edit Patch In BIN' and select 'Apply Patch' I get the following error: "Patch aborted. Base data does not match."

Here is a section of the file provided by dzidaV8 that I am wondering if it has a typo. I looked to send him a message but, I see he has not been on since 2020.

<XDFPATCH uniqueid="0x6551">
<title>Electric Water Pump on B9 pin - Engine off cooling enable</title>
<description>Patch will enable electric water pump control on pin B9 (ground side of relay).
The pump will work always when engine is running, and will also be on on ignition on, engine off when Coolant temperature is higher than set threshold, or if coolant temperature sensor failed.</description>
<XDFPATCHENTRY name="Add E-side 0xF000 Subroutine" address="0x1F000" datasize="0x1C" patchdata="1D0001B601288130230F12304008B60192B1F01 C23031C00017E4944" basedata="0000000000000000000000000000000000000000 0000000000000000" />
<XDFPATCHENTRY name="Modify E-side 0x493C code" address="0x14939" datasize="0x6" patchdata="2380057EF000" basedata="2F02051D0001" />
<XDFPATCHENTRY name="Engine Off Enable Cool Temp" address="0x1F01C" datasize="0x1" patchdata="B9" basedata="00" />
</XDFPATCH>

Now, I have never worked with C++ or these ECM files and the last time I programmed in binary I did it on punch cards. but, it looks to me in line 4 the subroutine is being stored at location 0001F000. I have confirmed the bits starting at 0001F000 are changed to the string listed in line 4. Line 6 also makes a change at 0001F01C and that is also confirmed looking at the BIN editor in TunerProRT.

Line 5 changes start at 00014939.

My question is that I believe E7 000149C is a jump command that is directing the stack to 0000F000 when it should be directed to 0001F000. Can I change the address for the jump without breaking something or should I write the 0001F000 string to 0000F000 instead?

I accept that I may be completely off base here and may just switch to using an Oil Pressure Sending Unit to provide a ground for the EWP relay if this is too much for me. I would like to keep the process elgant though.

you have 2 cpus, each using half the bin file. The addressing per each cpu is 0-FFFF. Anything outside that is a brick disaster.Not that you can set 3 bytes address on 16bit cpu.

I just checked the patch against my 94 automatic Corvette and my 95 manual Corvette. On the 95 manual, the patch worked perfectly. This makes me think it was written for a manual car.

On my 94 automatic, I had to click "remove patch" then patch, then "apply patch" and patch again. This bypasses the "data does not match" error message you were getting. As long as you have a backup of your original BIN, I see no reason why not to give that a shot.

Can this patch be upgraded to turn on the pump when a certain temperature is reached? And turn off after it cools down.
I want to use an additional external electric pump with an extra radiator connected to a lower left hose. (Number 16 oil cooler bypass hose)
http://shbox.com/1/93-94_hoses.jpg
I want it to operate when the t-stat is opened only.