Versatile Engine Management Systems - VEMS wiki

The trigger hardware is not particularly complex, but there are some options and things that shouldn't be missed

The trigger signals are sensitive. Especially the VR signal at low-RPM ( <300RPM, thus cranking). VR is not really sensitive from noise via capacitive coupling (although long wires that drive ignition coils arranged parallel might inject noise) but rather to noise injected through ground loop (switched power-currents that partially sneak through GND instead of GND5) if not installed properly.

• If GND5 is not connected, you get smoke from the ECM.
• If only 1..2 GND5 wire connected instead of 4..5 you get more ground-noise.
• If GND-GND5 bridge (a soldered "arch" left to the 4 FETdrivers, 30mm under VEMS logo on top side) is still soldered (and not cut), you get the ground-loop that injects noise into VR and might make startup problematic

The following precautions highly recommended:

• use shielded cable. The cable must have at least 2 wires besides the shield: the GND and the signal itself (VR- and VR+)
• The shielding must be connected on the ECU side only, so the shielding is NOT actually used for connecting the GND signal to both ends
• use GND (for VR-) and not GND5. The GND5 signal would make the signal noisy (when FETs and IGBTs switch).
• connect sensor GND only near the EC36 connector (not at battery and not at other places). Search the web for "ground loop" if you want to understand in more depth.

Primary trigger (trigg1) (on v3.1 ?)

For Hall setup:

• R30=10k (pullup resistor)
• C88 = does not matter
• short SJ1
• R55=10k
• do NOT short SJ2
  • remove R56(=10k, next to SJ2)
  • and solder SJ2=150k (or at min 47k resistor; this makes RC filter with the IC1 line's appr. 10..20pF - which comes from AVR input capacitance + trace capacitance)
• do NOT short SJ6
• The other parts near the LM1815 do not matter

Adaptive hysteresis: for use with VR sensor (LM1815 involved). (for v3.1 ?)

• R30 pullup resistor NOT soldered (unsolder it if there). A few mm left from the 2x3 trigger pinheader. R30 is the horizontal, 0805 package; further away from the EC36
• C88 = does not matter
• R55=10K
• short SJ1
• do NOT short SJ2
• short SJ6
• D14 is optional (3.9..5.1V zener diode)
• U11 (LM8515)
• R56 (10K) resistor of input RC lowpass
• C30 capacitor of input RC lowpass (1..3.3nF for multitooth, upto 100nF for coiltype-only-for-sure setup)
• R57 (75K)
• R87 (1 .. 1.6M) resistor of peak-detector
• C38 (0.22uF) capacitor of peak-detector
• R88 (10K) output pullup resistor
• C31 (1nF)
• C39 (0.22uF)

• measure with DVM 200kOhm range between primary trigger (trigger1_conn) signal and LM1815 pin3. If >21kOhm than you don't have VR input selected. Check around SJ1
  • follow the path on the GenBoard/VerThree/Schematic, from the inputtrigger to the LM1815 input pin: Identify the components that are in the path
• measure with DVM 2kOhm range between LM1815 pin12 to AVR's pin29 (IC1). If > 10 Ohm, than SJ6 is not shorted
• apply self stim, it only requires a 1..10uF capacitor and a few simple menu-commands (see OutputTrigger) and it can save you hours. Don't skip this unless the trigger setup works immediately.

---

setting up VR on v3.2 - most parts are soldered. Still check at least the SJ6 and R30. THIS IS THE MOST COMMON SETUP - assembled boards come this way unless customer specifically requests otherwise

The reason for the 2 missing bits: other option would be HALL type: therefore this is only done before shipping when the user declares (eg. on project wiki page, or order note, preferrably both) that he needs VR type.

• R30 pullup resistor NOT soldered (unsolder it if there). A few mm left from the 2x3 trigger pinheader. R30 is the horizontal, 0805 package; further away from the EC36
• short SJ6 (right below LM1815) by soldering a very short wire
• connect primary trigger input to the internal signal directly (no capacitor needed, just a very short wire). Either
  • put a jumper: connect the upper-middle pin to the upper-right pin of 3x2 pinheader (viewing from the "Top" side)
  • or rather solder on bottom of the board to short JP2 and JP7 : connect upper-middle to upper-left pin (viewing from the "Bottom" side)

The 3x2 pinheader is between the EC36 and the LM1815, it has an additionl GND pad besides the 3x2 pins.

Make sure you view the board in a good orientation, so econoseal points down and "Top" or "Bottom" text is readable.

setting up HALL on v3.2

trigger #1:

• connect JP2 & JP7 (jumper top right of the 2x3)
• do NOT short SJ2
  • remove R56(=10k, next to SJ2)
  • and solder SJ2=150k (or at min 47k resistor; this makes RC filter with the IC1 line's appr. 10..20pF - which comes from AVR input capacitance + trace capacitance)
• change ioinit.c Line55 to: PORTD = 0xE0; // pullup on IC1 (0x10) off (it is commited in STABLE1_0 CVS but might take some days before it is released)

trigger #2:

• connect JP5 & JP13 (jumper bottom right of the 2x3)
• do NOT short SJ5
  • recommended SJ5=150k (resistor; this makes RC filter with the IC3 line's appr. 10..20pF - which comes from AVR input capacitance + trace capacitance)
• and define TRIGGER_PULLUP_OFF in my_make and recompile or change ioinit.c Line61 to: PORTE = 0x00; // pullup on secondary trigger PE7=IC3 off

[Picture of 2x3 header, SJ6, SJ2 on Genboard v3.2] R56 and C30 are right after SJ2 (LM1815 pin3).

---

Special setup: when input does not go below 0V

You only need a C88=10uF (1210, 35V) when the signal otherwise does not go below zero. SJ1 must not be shorted in this case.

Useful, for example when stealing the 0.65V based VR signal of an alien ECM (as on MembersPage/Gabor).

Note that R30 pullup resistor is not recommended for VR sensor. If you already have an R30 pullup, that produces an offset (as a voltage divider with the VR sensor's internal resistance) and make the input NOT go below 0V at low RPM (cranking): fix it by removing R30 or following this section ( apply C88=10uF and not short SJ1).

---

Primary and Secondary setup for Hall on Genboard v3.2

I've set up two v3.2 Genboards with primary and secondary triggers configured for Hall sensor input. I've gone by advice from Emil and Jörgen here, which conflicts with what is recommended above.

• Bottom view [Picture of trigger area]

• I've added 1n capacitors between ground and each trigger signal. I've shorted SJ5 in this picture.

• Top view 1 [Picture near jumpers]
  • The jumpers are in place for the trigger signals, they might be replaced by solder jumpers in the future if I have issues - the jumpers won't budge easily.
  • There are two pull-up resistors (103) near the jumpers, R30 and R48. These are replaced with 102=1k to make the pull-up stronger and lessen the chance of signal problems.

• Top view 2 [Picture near LM1815 chip]
  • I've shorted SJ2 in this picture. I've also removed R56 (103=10k), the resistor next to SJ2, it would pull down the signal if left in place.

This has been tested to work on the primary trigger only! The secondary trigger is set up the same way, I have no doubts using this method on an engine that needs both crank + cam sensors of Hall-type.

//Mattias (MembersPage/MattiasSandgren)

---

TODO: v3.3 (similar to v3.2 but secondary trigger prepared to support 5cyl audi engines without AudiTrigger board)

---

Secondary trigger (Trigg2):

For Hall setup: R91 10k, SJ5, (SJ10 or C89)

For use with LM1815: U12, R58, R59, R89, R90, R91, C32, C33, C34, C40, D24, D27, SJ7 & SJ10 (shorted)

V3.2, Solder the LM1815 + SJ7 for VR setup, only SJ5 for Hall

---

See also

• GenBoard/Manual
• GenBoard/VerThree/Schematic
• [very good introduction] to VR and HALL sensors

Document last modified 2006-03-16 04:43:08

Search: