Megasquirt Lighting the fire (ignition upgrades)
05-12-11, 06:08 AM
#1
Lighting the fire (ignition upgrades)
Hi Ken/All
So there are plenty of upgraded ignition system for sale out there (compared to stock FC style system).
I'm blowing out my spark at my current power level / state of tune / boost. From these mods I wish to be able to enrich my tune a further 10% from where it is and be well clear of the risk of misfires.
My peak power lands at ~7400rpm. 4.0mS cycle to cycle for charge and discharge (wasted spark). I'm currently using dwell around 3.2mS something like that, leaving me ~0.8mS for the ignition event. I'm using the factory FC leading coil pack.
The solution as I see it is to EITHER:
#1 CDI box
#2 Coil near plug inductive system (full sequential would make 8.0mS available at 7400rpm and give me upto 6.0mS dwell and 2mS discharge and rest time).
The CDI system could be as simple as getting a single crane Hi-6 in between the FC ignition module and the leading coil pack. That would be messy but it should give me closer to 100mJ ignition energy with plenty of volts. This setup *SHOULD* be better than what I have and *SHOULD* be guaranteed to work out of the box once wired in. No firmware changes or testing required.
I don't know if I could piggyback 2x Hi-6;s off the leading output ignition module and have it still work. That would allow me to replace the twin post coil with any of the CDI coils (and a few non CDI coils that are proven to work). This would give even more energy.
The M&W CDI (very similar to motec and a few others) seems to do 4 channels. It will require 4 specialised outputs from the megasquirt however.
Past that there is dynatek, autronic 500r, MSD DIS/2 or DIS/4.
PS #1 The crane Hi-6 is documented to need to be triggered off a distributor pickup OR the injectors output that normally drives the coil.
#2 The M&W/Motec is proven to need the logic signal from the ECU (a guy I know had big problems trying to run it off the ignitor).
So that gets me to getting the M&W/Motec CDI or the IGN-1A or LS1/LS2/LS7 coils to work.
I think they will all require a true non wasted spark 100% sequential ignition output. I think from reading documentation they may require as much as 12-20mA drive current to work. I would not be surprised if I needed to use 270R pullup resistors on the LEDs to trigger the LSx or IGN-1A ignition coils. Then there is also the question of what waveform they want. The rotary code feels like its modified slightly to give EXACTLY what the factory FC modules want and it does a brilliant job of that. I just don't know how much hardware needs to be changed to use all the other ignition setups.
So Assuming timing at TDC.
They will require ignition events at:
TDC rotor 1, split degrees later than TDC rotor 1, 180 degrees later than TDC rotor1 and split +180 degrees later than TDC rotor1.
The megasquirt documentation has options for FD and RX-8 mode. My feeling is they are 100% untested at this point in time.
I worry there are code problems with all the messy modes (by messy modes I mean barely/if ever used FD and RX-8 mode and whether either modes work correctly with the direction set to either going high or going low.
Car is being daily driven.
So there are plenty of upgraded ignition system for sale out there (compared to stock FC style system).
I'm blowing out my spark at my current power level / state of tune / boost. From these mods I wish to be able to enrich my tune a further 10% from where it is and be well clear of the risk of misfires.
My peak power lands at ~7400rpm. 4.0mS cycle to cycle for charge and discharge (wasted spark). I'm currently using dwell around 3.2mS something like that, leaving me ~0.8mS for the ignition event. I'm using the factory FC leading coil pack.
The solution as I see it is to EITHER:
#1 CDI box
#2 Coil near plug inductive system (full sequential would make 8.0mS available at 7400rpm and give me upto 6.0mS dwell and 2mS discharge and rest time).
The CDI system could be as simple as getting a single crane Hi-6 in between the FC ignition module and the leading coil pack. That would be messy but it should give me closer to 100mJ ignition energy with plenty of volts. This setup *SHOULD* be better than what I have and *SHOULD* be guaranteed to work out of the box once wired in. No firmware changes or testing required.
I don't know if I could piggyback 2x Hi-6;s off the leading output ignition module and have it still work. That would allow me to replace the twin post coil with any of the CDI coils (and a few non CDI coils that are proven to work). This would give even more energy.
The M&W CDI (very similar to motec and a few others) seems to do 4 channels. It will require 4 specialised outputs from the megasquirt however.
Past that there is dynatek, autronic 500r, MSD DIS/2 or DIS/4.
PS #1 The crane Hi-6 is documented to need to be triggered off a distributor pickup OR the injectors output that normally drives the coil.
#2 The M&W/Motec is proven to need the logic signal from the ECU (a guy I know had big problems trying to run it off the ignitor).
So that gets me to getting the M&W/Motec CDI or the IGN-1A or LS1/LS2/LS7 coils to work.
I think they will all require a true non wasted spark 100% sequential ignition output. I think from reading documentation they may require as much as 12-20mA drive current to work. I would not be surprised if I needed to use 270R pullup resistors on the LEDs to trigger the LSx or IGN-1A ignition coils. Then there is also the question of what waveform they want. The rotary code feels like its modified slightly to give EXACTLY what the factory FC modules want and it does a brilliant job of that. I just don't know how much hardware needs to be changed to use all the other ignition setups.
So Assuming timing at TDC.
They will require ignition events at:
TDC rotor 1, split degrees later than TDC rotor 1, 180 degrees later than TDC rotor1 and split +180 degrees later than TDC rotor1.
The megasquirt documentation has options for FD and RX-8 mode. My feeling is they are 100% untested at this point in time.
I worry there are code problems with all the messy modes (by messy modes I mean barely/if ever used FD and RX-8 mode and whether either modes work correctly with the direction set to either going high or going low.
Car is being daily driven.
06-17-11, 09:11 AM
#2
OK I ordered,
4x IGN-1A's + connectors
2x BUR7EQ (leading)
2x BUR9EQ (trailing)
Looking at getting some Magnecor R100 10mm or KV85 8.5mm leads.
Looking at making up a coil on plug bracket.
Car is currently being driven 200+ miles per week and it is dependable, its just it blows the spark out at full revs and full boost because there is no dwell time. Install is thoroughly tested, no noise problems whatsoever, motor has held together for 4+ years hundreds of pulls.
The install is currently using these settings,
Generic Type Settings
Spark A output pin: D14
Number of Coils: Wasted Spark
Spark Output: Going Low (Normal)
Rotary Mode: Rotary on
Overdwell Protection: off
Accel Compensation: 0
Dwell: 3.4mS
Minimum Spark Duration: 0.1mS
Dwell Type: Standard
Use Prediction: 1st Deriv Prediction
Trailing Settings Page!
Output Mode: FC Mode
RX-8 Mode: Disabled
Allow Negative Split: Allowed (5 negative above 4000rpm lower than 30kpa load)
I currently have 3 4k7 pullups on the 3 LED's.
From reading diyautotune manual I need to put 270R pullups, presumably for a single coil, they want around 20mA drive current.
My problem is I need to know what settings to get FD mode working, and also what settings to get RX-8 mode working.
For EG, is FD mode considered wasted spark or coil on plug. What about RX-8 mode?
If I want to use RX-8 mode what do I set for FC/FD mode. It says to set it to either FC mode or FD mode, but neither is RX-8 mode. Or does RX-8 mode only care about trailing?
Is there updated documentation about where the spark outputs come out and which ones are which. This sort of stuff could cost me a motor (and this is a properly built motor) so I need to know what it is.
4x IGN-1A's + connectors
2x BUR7EQ (leading)
2x BUR9EQ (trailing)
Looking at getting some Magnecor R100 10mm or KV85 8.5mm leads.
Looking at making up a coil on plug bracket.
Car is currently being driven 200+ miles per week and it is dependable, its just it blows the spark out at full revs and full boost because there is no dwell time. Install is thoroughly tested, no noise problems whatsoever, motor has held together for 4+ years hundreds of pulls.
The install is currently using these settings,
Generic Type Settings
Spark A output pin: D14
Number of Coils: Wasted Spark
Spark Output: Going Low (Normal)
Rotary Mode: Rotary on
Overdwell Protection: off
Accel Compensation: 0
Dwell: 3.4mS
Minimum Spark Duration: 0.1mS
Dwell Type: Standard
Use Prediction: 1st Deriv Prediction
Trailing Settings Page!
Output Mode: FC Mode
RX-8 Mode: Disabled
Allow Negative Split: Allowed (5 negative above 4000rpm lower than 30kpa load)
I currently have 3 4k7 pullups on the 3 LED's.
From reading diyautotune manual I need to put 270R pullups, presumably for a single coil, they want around 20mA drive current.
My problem is I need to know what settings to get FD mode working, and also what settings to get RX-8 mode working.
For EG, is FD mode considered wasted spark or coil on plug. What about RX-8 mode?
If I want to use RX-8 mode what do I set for FC/FD mode. It says to set it to either FC mode or FD mode, but neither is RX-8 mode. Or does RX-8 mode only care about trailing?
Is there updated documentation about where the spark outputs come out and which ones are which. This sort of stuff could cost me a motor (and this is a properly built motor) so I need to know what it is.
06-17-11, 09:40 AM
#3
So I've got
SparkA D14
SparkB D16
SparkC D15
SparkD JS11 (and build another LED/transistor circuit in proto area).
So for FC3S its well documented
SparkA D14 - Leading (wasted spark)
SparkB D16 - Trailing
SparkC D15 - Trailing (coil-select)
So for FD3S it is????
SparkA D14 - Leading
SparkB D16 - ????
SparkC D15 - ?????
So for RX-8 it is????
SparkA D14 - Leading
SparkB D16 - ????
SparkC D15 - ?????
SparkD JS11 - ????
SparkA D14
SparkB D16
SparkC D15
SparkD JS11 (and build another LED/transistor circuit in proto area).
So for FC3S its well documented
SparkA D14 - Leading (wasted spark)
SparkB D16 - Trailing
SparkC D15 - Trailing (coil-select)
So for FD3S it is????
SparkA D14 - Leading
SparkB D16 - ????
SparkC D15 - ?????
So for RX-8 it is????
SparkA D14 - Leading
SparkB D16 - ????
SparkC D15 - ?????
SparkD JS11 - ????
06-19-11, 12:03 AM
#4
Rotary Enthusiast
man tonight I revived my megasquirt project and I was thinking of upgrading the FC ignition system to individual coils (Yukon or LS1) ....humm I'll keep an eye on your setup.
06-19-11, 05:43 AM
#6
talking head
So I've got
SparkA D14
SparkB D16
SparkC D15
SparkD JS11 (and build another LED/transistor circuit in proto area).
So for FC3S its well documented
SparkA D14 - Leading (wasted spark)
SparkB D16 - Trailing
SparkC D15 - Trailing (coil-select)
So for FD3S it is????
SparkA D14 - Leading
SparkB D16 - ????
SparkC D15 - ?????
So for RX-8 it is????
SparkA D14 - Leading
SparkB D16 - ????
SparkC D15 - ?????
SparkD JS11 - ????
SparkA D14
SparkB D16
SparkC D15
SparkD JS11 (and build another LED/transistor circuit in proto area).
So for FC3S its well documented
SparkA D14 - Leading (wasted spark)
SparkB D16 - Trailing
SparkC D15 - Trailing (coil-select)
So for FD3S it is????
SparkA D14 - Leading
SparkB D16 - ????
SparkC D15 - ?????
So for RX-8 it is????
SparkA D14 - Leading
SparkB D16 - ????
SparkC D15 - ?????
SparkD JS11 - ????
in the OEM application,, FC is leading,, trailing and flop command
essentially two igniters
the FC leading coil is 0.4 ohms,, but is ballasted under the baseplate with a 0.5 ohm resister
( 0.9 ohms total, various years have the resister move to the - side of the coil )
[ trailing coils are each 0.7-0.9 ohms ]
if you wanted a higher rate of saturation at rpms ,,
a simple approach for those with duty dwell settings would be to use the auc output to swing a relay to bypass the resister after a threshold rpm
( or just bypass the igniter entirely if you are running a fixed dwell )
on OEM FD the igniter is a triplex unit,, and works as leading wastespark with dedicated trailing 1 and trailing 2 channels
( 3 igniters )
if the megasquirt rx8 has 4 channels then obviously it must be the sequential ignition you seek
but i expect you will now need the funky rx8 chopper wheel to run the logic
best bet i see is the FD mode using the trailing channels only and having leading and trailing firing together with no split
ie two channels driving 4 igniters
( much like some primitve motec modes )
the issues i see is FD mode is 12/1,, FC 24/2 ( same same ),, but begs differences in the trigger tooth and angle settings
Trending Topics
06-20-11, 03:10 AM
#8
in the OEM application,, FC is leading,, trailing and flop command
essentially two igniters
the FC leading coil is 0.4 ohms,, but is ballasted under the baseplate with a 0.5 ohm resister
( 0.9 ohms total, various years have the resister move to the - side of the coil )
[ trailing coils are each 0.7-0.9 ohms ]
if you wanted a higher rate of saturation at rpms ,,
a simple approach for those with duty dwell settings would be to use the auc output to swing a relay to bypass the resister after a threshold rpm
( or just bypass the igniter entirely if you are running a fixed dwell )
on OEM FD the igniter is a triplex unit,, and works as leading wastespark with dedicated trailing 1 and trailing 2 channels
( 3 igniters )
if the megasquirt rx8 has 4 channels then obviously it must be the sequential ignition you seek
but i expect you will now need the funky rx8 chopper wheel to run the logic
best bet i see is the FD mode using the trailing channels only and having leading and trailing firing together with no split
ie two channels driving 4 igniters
( much like some primitve motec modes )
the issues i see is FD mode is 12/1,, FC 24/2 ( same same ),, but begs differences in the trigger tooth and angle settings
essentially two igniters
the FC leading coil is 0.4 ohms,, but is ballasted under the baseplate with a 0.5 ohm resister
( 0.9 ohms total, various years have the resister move to the - side of the coil )
[ trailing coils are each 0.7-0.9 ohms ]
if you wanted a higher rate of saturation at rpms ,,
a simple approach for those with duty dwell settings would be to use the auc output to swing a relay to bypass the resister after a threshold rpm
( or just bypass the igniter entirely if you are running a fixed dwell )
on OEM FD the igniter is a triplex unit,, and works as leading wastespark with dedicated trailing 1 and trailing 2 channels
( 3 igniters )
if the megasquirt rx8 has 4 channels then obviously it must be the sequential ignition you seek
but i expect you will now need the funky rx8 chopper wheel to run the logic
best bet i see is the FD mode using the trailing channels only and having leading and trailing firing together with no split
ie two channels driving 4 igniters
( much like some primitve motec modes )
the issues i see is FD mode is 12/1,, FC 24/2 ( same same ),, but begs differences in the trigger tooth and angle settings
I thought about bypassing the ballast resistor. I could have done that and dropped the dwell about 0.2mS and probably still ended up with more energy at high RPM.
Well its done now I have coils on the way. I also thought about firing the ignition like a Wasted Spark 4 cylinder with Rotor 1 leading and trailing tied together. That would give me my desired dwell time 5.5mS/2.0mS rest @ 8000rpm without any software restrictions, and with these new coils should make for a tasty spark, but then I lose leading/trailing split and I do not know how to tune that...
I do have a scope at my disposal so I can unplug injectors and coils and measure what actually comes out of the ignition driver while cranking.
Ideally I will use RX-8 mode.
Nicer again would be RX-8 mode with an option to use wasted spark below 3500rpm to clean up idle and cruise.
I looked at twin power, and I do read good things. But I don't think its guaranteed that a twin power is enough to light 1.2+ bar with water injection and E85 @ .71 lambda. I really want to step over to 4 independent channels/ modules to make future upgrades easier.
06-20-11, 05:58 AM
#10
I decided to brave the cold and test the megasquirt settings against the scope.
I think I may have gotten a partial cycle out of RX-8 mode but it was resetting mid cycle, presumably because it is coded to expect the 36-2-2-2 wheel? Is that right ken? I still had my FC CAS settings in the trigger input page.
You can see from the scope traces FC_MODE and FD_MODE the outputs are performing as intended.
For cranking I have 5.0mS dwell set with 5BTDC advance.
I think I may have gotten a partial cycle out of RX-8 mode but it was resetting mid cycle, presumably because it is coded to expect the 36-2-2-2 wheel? Is that right ken? I still had my FC CAS settings in the trigger input page.
You can see from the scope traces FC_MODE and FD_MODE the outputs are performing as intended.
For cranking I have 5.0mS dwell set with 5BTDC advance.
06-20-11, 06:07 AM
#11
This gets me back to the point where I am still dwell time limited. FD mode will still have have an ignition to ignition time of around 4.0mS @ 7400RPM.
Ken is it possible for you to modify MS2extra source code to leave the FD mode trailing as is and just separate the wasted spark FD leading signal into 2 single signals and allow 2x dwell time.
My only other option would be to tack another microprocessor ecap module onto the end of your ignition outputs and molest the ignition waveforms that way.
Ken is it possible for you to modify MS2extra source code to leave the FD mode trailing as is and just separate the wasted spark FD leading signal into 2 single signals and allow 2x dwell time.
My only other option would be to tack another microprocessor ecap module onto the end of your ignition outputs and molest the ignition waveforms that way.
06-20-11, 12:05 PM
#12
MegaSquirt Mod
This gets me back to the point where I am still dwell time limited. FD mode will still have have an ignition to ignition time of around 4.0mS @ 7400RPM.
Ken is it possible for you to modify MS2extra source code to leave the FD mode trailing as is and just separate the wasted spark FD leading signal into 2 single signals and allow 2x dwell time.
My only other option would be to tack another microprocessor ecap module onto the end of your ignition outputs and molest the ignition waveforms that way.
Ken is it possible for you to modify MS2extra source code to leave the FD mode trailing as is and just separate the wasted spark FD leading signal into 2 single signals and allow 2x dwell time.
My only other option would be to tack another microprocessor ecap module onto the end of your ignition outputs and molest the ignition waveforms that way.
That's what rx8 mode does... rx8 mode + FD mode is the same thing as saying "COP for a 2-rotor"
Front rotor leading is A
Front and rear trailing are B and C respectively
Rear rotor leading is D
Ken
06-21-11, 09:42 AM
#14
MegaSquirt Mod
If it's having triggering problems when set to RX8 mode, it's being caused by something other than the code.
Ken
06-23-11, 07:37 AM
#15
Well I tested FD Mode with RX-8 disabled and had no triggering problems. I then tested FD Mode with RX-8 mode enabled and had triggering problems.
I then switched back to FD mode and had no problems.
I then switched back to RX-8 mode and had problems.
How could you say its not code? Or do you mean it is something else possibly hanging off a microprocessor pin causing the problem?
06-23-11, 09:48 AM
#16
MegaSquirt Mod
I'm saying that in the code, RX8 mode does not do ANYTHING with the wheel decoder. It only changes the outputs.
The wheel decoder code is independent of the output code.
If you didn't have the outputs wired properly for rx8 mode, then I could see it causing all manner of problems.
Ken
The wheel decoder code is independent of the output code.
If you didn't have the outputs wired properly for rx8 mode, then I could see it causing all manner of problems.
Ken
06-24-11, 07:02 AM
#17
I'm saying that in the code, RX8 mode does not do ANYTHING with the wheel decoder. It only changes the outputs.
The wheel decoder code is independent of the output code.
If you didn't have the outputs wired properly for rx8 mode, then I could see it causing all manner of problems.
Ken
The wheel decoder code is independent of the output code.
If you didn't have the outputs wired properly for rx8 mode, then I could see it causing all manner of problems.
Ken
Using either FC or FD mode the car will crank for 5+ seconds with no start condition and still maintain sync. My car has absolutely no triggering noise or earthing problems. I can maintain data logging connectivity for above 45 minutes with no losses of sync or ECU resets recorded. And I have done that a good 5 times in the last few weeks.
Unless there is a conflict between the Spark D pin and another auxilary function I am using? Or are you suggesting Spark D is physically connected to either the CAS Ne or Ge inputs?
06-25-11, 03:06 AM
#19
This one!
Code:
;-------------------------------------------------------------------------------
; $Id: base.ini,v 1.1.2.11 2008/11/09 18:22:52 culverk Exp $
; configuration defines
;Normally these next lines should be UNset
;This enables the trigger logger page. At this time no simple software exists to interpret the data
#unset TRIGLOG
;This enables the "full" CAN/serial commands so this INI can be used anywhere on the CAN network.
#unset CAN_COMMANDS
[MegaTune]
MTversion = 2.25 ; MegaTune itself; needs to match exec version.
#if CAN_COMMANDS
versionInfo = "r\x00\x0e\x00\x00\x00\x3c" ; Title bar
queryCommand = "r\x00\x0f\x00\x00\x00\x14" ; Verify against signature.
#else
versionInfo = "S" ; Put this in the title bar.
queryCommand = "Q" ; Verify against signature.
#endif
signature = "MS2Extra Rel 2.1.0d" ; MS-II sends a null at 20th byte.
; 123456789.123456789.
;-------------------------------------------------------------------------------
06-25-11, 10:58 PM
#22
How about I send you my MSQ file in FC mode, and you scope trace that. Then you change the tickbox to FD mode and scope trace that. And then change the scope trace to RX-8 mode and see something bad happens.
Would you still have the source code from that build in 2008?
Would you still have the source code from that build in 2008?
06-26-11, 08:19 AM
#23
MegaSquirt Mod
How about I have done that many times, including when I wrote the feature, and I have never seen a problem.
If you want my help, try the latest firmware. There have been a huge number of changes since the 3 year old firmware you are using.
If you want my help, try the latest firmware. There have been a huge number of changes since the 3 year old firmware you are using.
06-27-11, 07:07 AM
#25
I can't risk the fallout from a firmware update until the weekend because my megasquirted 7 is my trip to work every day.
What would be the chances of a newer release of 2.10d or whatever was the release code fixing the RX-8 mode bug? This is my prefered solution at this stage.
But I'm not buying more ECU related hardware until I know for sure that we can get around the ignition dwell time issue.
EDIT:
Accidentally edited this instead of responding to it.
What would be the chances of a newer release of 2.10d or whatever was the release code fixing the RX-8 mode bug? This is my prefered solution at this stage.
But I'm not buying more ECU related hardware until I know for sure that we can get around the ignition dwell time issue.
EDIT:
Accidentally edited this instead of responding to it.
Last edited by muythaibxr; 06-28-11 at 10:26 AM.