Author Topic: Adjust Timing using Barometric Pressure Circuit  (Read 1434 times)

0 Members and 1 Guest are viewing this topic.

Rusty

  • Lightly Modded
  • **
  • Posts: 354
Adjust Timing using Barometric Pressure Circuit
« on: June 24, 2004, 05:30:55 PM »
Adjust Timing using Barometric Pressure Circuit

CAUTION - don't try this unless you have data logging (e.g. VAG-COM) and the good sense to interpret it.  And be CAREFUL working around the ECU...

I thought I could use the Barometric Pressure sensor to "trick" my AEB-engined car into giving me a richer mixture at WOT. So I with the help of Gary, an EE at work, I designed a circuit to adjust the voltage that the ECU sees, making it think the car was at a higher or lower altitude. I was wrong - it did nothing for A/F ratio.  But it apparently changed the base timing curve.  I stopped there because right now, I don't need timing adjustment, I need FUEL.  But I offer this writeup as a basis for further experiments, if anyone is interested.  For example, one might have really good A/F ratios with no knock retard - one could advance timing and get that extra "nth" HP.

The primary altitude-adjustment mechanism is MAF.  I knew that from the start, I just didn't know what the B.P. might do in addition. In summary, it looks like the ECU retards timing a teeny bit with lower altitude, and advances timing a teeny bit with higher altitude.  It probably does more than just a timing adjustment, but I couldn't/didn't measure or detect it.  In order to get the average 1.4 degrees of advance that I saw, I had to adjust high altitude to what must be Mt. Everest.

The B.P. sensor is inside the ECU's protective box, held in by a single phillips screw, and connected to the ECU harness via a 3-blade connector.  Pin assignments are:
    1 = variable resistance governed by altitude
    2 = reference resistance
    3 = ground

These pins are clearly marked on the B.P. sensor.  The ECU is connected to pins 1 and 2, and puts 5 volts on each pin.  As the resistance from pin 1 varies, the voltage difference between the two changes, and this voltage difference is what the ECU reads.  By wiring a pot (potentiometer) at the right place, you can cause the voltage difference to change.

Here's the raw B.P. sensor data, logged on my trip from ABQ to the ND show back in May.  Every stop required me to open up the ECU, disconnect the baro sensor, and measure the resistance between pins 3 and 1.  The resistance between pins 3 and 2 was always constant at 2.39K ohms.
    Albuquerque NM - 5300' 3.25K
    Kingman AZ ----- 3449' 3.25K
    Apple Valley CA - 3019' 3.19K
    Barstow-Dagget - 1927' 3.04K
    Ontario CA -------- 944' 2.83K
    San Jose CA -------- 56' 2.52K

Yes, the ABQ reading is the same as Kingman.  Probably a humidity difference.  But I'm looking for trends, and the trend is that the resistance lowers with decreasing altitude (increasing baro press), and rises with increasing altitude (decreasing baro press).  I drew up a little chart and figured out what resistance values I'd need for the new circuit.  I did this in two stages using one pot, because it has to be wired differently depending on whether you want to raise or lower the resistance. I assume a clever person could figure out how to do both with one knob.  Also, you may want to use a "10-turn" pot, which can be expensive but it allows very fine adjustment.

To raise resistance/advance timing:

Connect a 5K or larger pot in series between pin 1 and the ECU.


I logged data at 0, 1.5K, 2.5K, and 3.5K added resistance:
    002 RPM, load, speed, status
    005 RPM, load, fuel inj, MAF
    011 RPM, coolant temp, IAT, timing
    020 knock retard
    031 O2 sensor voltage

The timing and knock retard were the only things that noticeably changed, and the trend showed increased timing advance. At 3.5K (total resistance 3.25 + 3.5 = 6.75K) there was average of 1.4 degrees of timing advance. At lower RPM, there was as much as 4 degrees, leveling out to even at higher RPM.   Knock retard varied directly with timing (i.e. increased advance = increased knock retard), so I'll only list the timing data from block 011:
    RPM - 3.5K---- 0K
    1720 - 24 ---- 24.8
    1840 - 23.3 -- 21.8
    2000 - 16.5 -- 16.5
    2120 - 20.3 -- 15.8
    2280 - 20.3 -- 16.5
    2440 - 18 ---- 15.8
    2600 - 18 ---- 13.5
    2800 - 12 ---- 14.3
    3000 - 9.8 --- 6.8
    3240 - 11.3 -- 12
    3480 - 12.8 -- 12.8
    3720 - 13.5 -- 14.3
    3960 - 18.8 -- 16.5
    4200 - 17.3 -- 17.3
    4400 - 20.3 -- 19.5
    4640 - 24 ---- 17.3
    4840 - 21 ---- 18
    5080 - 19.5 -- 21.8
    5240 - 24 ---- 21.8
    5440 - 19.5 -- 19.5
    5600 - 24 ---- 23.3
    5760 - 24 ---- 24
    5880 - 24.8 -- 21.8
    6040 - 25.5 -- 23.3
    6160 - 24 ---- 24
    6280 - 24.8 -- 23.3
    --------------------
    avg -- 19.7 -- 18.3[/list]
    To lower resistance/retard timing:

    Connect a 5K resistor in series to a 100K pot, and connect this assembly in parallel between pins 3 and 1.


    Adding resistance in parallel between pin 1 and ground "bleeds" the voltage to ground, reducing the total resistance that the ECU sees from pin 1.  For the experiment, I actually used a 50K pot because it was free from Gary's parts bin, but this value won't let you get a "zero" point.  With the pot full on (50K + 5K), I had a 3.08K reading, as opposed to the nominal 3.29K at my altitude.  A 100K pot should have let me get really close to 3.29K. A zeroed pot (0K pot + 5K resistor) gave a total resistance of 1.99K.

    I did several runs at different settings, logging only O2 sensors,  fuel trims, and knock retard (and I didn't actually log the data, I was just watching the display).  Knock retard was the only thing that changed, and I saw decreased retard.  At the extreme settings, the car felt slower and boost was down from 15psi to 13psi.  So this must be retarding the base timing curve, by 2-3 degrees based on the displayed knock retard.
     
    Note that the 5K resistor value is somewhat arbitrary.  You want to have at least some resistance in the circuit for when the pot is at zero. A direct wire to ground is otherwise known as a "short-circuit."

    These Results are Relative:

    The results are relative to 5300 ft.  If you live on Mt Everest, you're not going to get any more advance, but you might get 3-4 degrees retard.  If you live at sea level, you probably can't retard, but you might get 3-4 degrees advance.

    Connectors:

    So that I wouldn't have to modify existing connections, I used the following parts to make a plug-in "insert" harness.  Plug the harness in-between the baro sensor and the ECU, and do your mods and testing on the harness.  Remove the harness for back-to-OEM or troubleshooting.



    Parts List:

    443-906-233 Housing, female 3-blade (plugs into baro sensor) - qty 1
    000-979-133 Wire/termi, female (inserts/snaps into above) - qty 3
    443-906-247 Housing, male 3-blade (plugs into ECU connector) - qty 1
    000-979-129 Wire set, male (inserts/snaps into above - NOT - see note) - qty 3

    NOTE: The connector housings mated properly, but the male wires wouldn't insert into the male connector. It was not molded to allow the wires to insert and snap-in, and also had no VW part number on it. Re-ordering resulted in the same non-numbered, half-functional part.  I suspect one of VW's suppliers goofed and is supplying the wrong part.  I finally dremeled-out the offending plastic, and silicone-sealed the wires into proper position in the connector.

    The wires are double-ended, so you get extras in case you goof.  Two each of the wires is actually all that's necessary.

    Pots and resistors from Radio Shack or wherever.  The pot I used is a really nice 50K 10-turn pot with the spiffy built-in dial.  It lists for $92 :shock: but Gary had three of them in his salvage parts bin.
    Wire nuts for testing.
    Solder, heat-shrink and/or electrical tape for final configuration.

    CAUTION - don't try this unless you have data logging (e.g. VAG-COM) and the good sense to interpret it.  And be CAREFUL working around the ECU...

    Hope this helps somebody.  Have fun!
    i]AEB Engines have better-flowing heads and a real throttle cable...[/i]

    jayryan

    • Supercar
    • *****
    • Posts: 6,211
      • JayFriesen.com
    Adjust Timing using Barometric Pressure Circuit
    « Reply #1 on: June 25, 2004, 10:34:48 AM »
    You never cease to amaze and wonder the common mind...ie. mine.
    '03 A4 1.8TQM
    '68 MG C
    '06 4Runner

    Rocky Mountain Club B5

    Adjust Timing using Barometric Pressure Circuit
    « Reply #1 on: June 25, 2004, 10:34:48 AM »

    ColoradoB5

    • Global Moderator
    • Supercar
    • *****
    • Posts: 3,311
    Adjust Timing using Barometric Pressure Circuit
    « Reply #2 on: June 25, 2004, 03:18:18 PM »
    Nice writeup, Rusty.  I'm sure someone will get some use out of this.  Keep up the experimenting you mad scientist!
    You can't rehabilitate a fire ant.

    RobD

    • Modded
    • ***
    • Posts: 682
    Adjust Timing using Barometric Pressure Circuit
    « Reply #3 on: June 26, 2004, 09:42:06 AM »
    Ditto to what JR said!

    kt037

    • Modded
    • ***
    • Posts: 510
      • http://www.edition719.com
    Adjust Timing using Barometric Pressure Circuit
    « Reply #4 on: June 27, 2004, 01:19:00 AM »
    woah... very impressive.
    OT DISH?-----> nope!
    and YES I KNOW I CAN NOT SEPLL!!!!