Ignition (output, ignadv and dwell) page for MembersPage/EricN
Note that trigger is on MembersPage/EricN/Trigger
Even-fire 6-cyl
using 6 LS1 coil near plug ignition transformers
This is what I know so far: 4 wires per coil
- 12v
- ground
- ground to the ecu
- trigger wire
Factory look is eventually the goal, but for now, I want to practice on this engine the same setup that will be in my racecar when it is actually built, then I will revert back to more simplistic setup.
I have done some various measurements on a running vehicle. Trigger is by a 4 Volt signal (I didn't check the trigger time but the scale was set to 1 second and it was just a vertical blip) that does not change at all with RPM. I suspect it is extremely low amp, logic level signal.
How to output a logic-level ignition signal
Use the same driver as the original IGBT is driven from, just solder a jumper wire in place of the IGBT (don't solder IGBT for given channels), between the gate and drain the 2 pins closest to the EC36 connector, and use ign_out=70 in your config /Emil
We usually use ignition channel 2 and 5 for these (recent tradition, that assembled ECMs are setup with ign channel 2 and 5 this way unless ECM goes for V8 with COP: very handy when IgnitionPage/TransformerWithAmplifier is to be used), you need 4 more logiclevel channels (for total 6).
MembersPage/MarcellGal/EngineSwap also uses similar logic-level ign setup for IgnitionPage/TransformerWithAmplifier
Ignition sequence
- cyl order 1 5 3 6 2 4
- wiring
- EC36 pin 35 goes to cyl 1
- EC36 pin 33 goes to cyl 2
- EC36 pin 34 goes to cyl 3
- EC36 pin 36 goes to cyl 4
- EC36 pin 11 goes to cyl 5
- EC36 pin 12 goes to cyl 6
... TODO: fill this in, there is just too many combinations with a 6-cyl fully sequential to get it done with "try and see" method.
Cant seem to find the description for this but I used the injector one, if I am wrong, please point me to the page that has this info so I can learn it.
h[2]= 08 02 20 04 10 01 00 00
Dwell - for LS1 ignition transformers
I have measured dwell:
- dwell length is around 59 degrees at all engine speeds and loads.
- Duty cycle was at about 95%, stayed constant.
Please correct the above, obviously there must be some misunderstanding as 59 degrees is not equal to 95%*120 crankdegrees.
- I probably screwed it up. I stuck my Snap On Vantage (an overpriced but extremely nice multimeter/scope. I think more like a graphing multimeter)
There really are many ways, eg.:
- scope
- soundcard
- DVM in DC voltage mode (20V): duty = measured_voltage / (VBatt - 2.5V). Knowing RPM and duty, you can calculate dwell
- DVM + on VBatt
- (min.600V) diode (eg. ES2J or 1n4007 in GenBoard/VerThree/RescueKit ) used for the same purpose as on TimingLight
- cathode on ignition channel (ECM output that goes to primary)
- diode anode connected to DVM -
- optionally 270 .. 1000 Ohm parallel with the DVM. A LED in series with 270 Ohm is the best, basically same as on TimingLight (except that lower brightness is used for indication than for a real timing-light).
-
- Pulse width changed dramatically but if I understand it correctly: if either of the above (59 degrees or constant 95%) is true, that is natural consequence of engine speed ( pulse period decreasing as RPM increases)
I never mentioned coil near plug, I said LS1 thinking that was a good description and I keep forgetting that this is world wide and most of the world has little interest in north american V8