ALFAGTV6.COM

Once I made some wiring mods to power the unit correctly, connected it up and it literally started on the 1st twist of the ignition key!
I was actually pretty shocked since I was expecting at least a few more technical glitches. The car is running Steve R's msq file which should be a pretty good match for my engine.
I noticed that the engine speed was spiking up to impossible levels but this appears only to happen when the car is idling very slowly (700 rpms's). As soon as I increased the idle the engine speed appeared to be solid (800 to 900 rpms).
Anyway, I'll continue with this tomorrow and hope to try it out on the road soon. At this point, the engine bay is practically untouched so if the MS fails, I can just swap back in the LJet in moments.

Thanks to everyone who has helped with advice, especially Steve for the excellent instructions at:
http://www.alfagtv6.com/MegaSquirt/Megasquirt.htm

One curiosity I had was with the function of the fuel pump. Mine now switches on with the ignition whereas with LJet, it only switched on when the engine was cranking (or running).

Cheers,

Michael,

The fuel pump runs with the ignition on with MS as the AFM used to control this. ie, the AFM had to open by cranking or running to start the fuel pump. This is one of the great reasons to get rid of the restrictrive AFM.
The fuel flow is still protected by the emergency cut off in the event of an sudden impact.

David

Michael,

The engine speed spikes are due to R10 being too small.

The 370ohm isn't big enough for our high-energy ignitions. Change it out to 2.2k ohms and install the wing diode (D5) and all should be well.

If you still get some occasional spikes increase R10 to 3.3kohm. Check the temperature of R10 while running. It shouldn't get too hot to touch with your finger.

Congrats on getting it running. I'll see your install in July! I got some good experience from the MS session in New Hampshire last year so I think I'll spend a LOT more time on tuning this time around. It might be helpful to you. If I don't bring my car, maybe we can do some tuning on yours for demonstration.

-Peter

Hi Peter:

Your post reminded me that I'd not completed this section - I had XG1 and XG2 jumpered for testing (with the Stim).
I do already have a larger resistor in R10 - not 2.2 K Ohm, but 1600 ohm (½ watt) as per Steve's directions. Is this okay?
I removed the XG1 --> XG2 jumper and installed a wire from XG1 to the case (which is itself well grounded). I then installed a .01 uF 500V capacitor on the back between Diode D5 and XG1.

By the way, what is the difference between Ceramic disk capacitors and Ceramic radial capacitors? I could only find the disk capacitors.

I'll test again tomorrow.
Thanks,

Congratulations on getting MS up and running Michael. I had the same problem with my R10 being too small. I changed it from 1.6K to 3.3K after consulting with Professor Webb.

I'm still having a problem with stalling though. I don't think its R10. The rpms read wrong only after the engine sputters - its not like a misread tach signal changes the injection pw and then the engine sputters. I'm using Steve Rosser's map and I have Greg Gordon Stage II injectors.

I think I need to redo the map at low rpms. I think its too rich, at least that's what my air-fuel gauge reads when the engine starts bucking during idle. Off to the Megamanual to read about how to tune via Megatune.

For now I've raised my idle to 1,000. That seems to avoid the problem for now.

Did a comparison G-tech dyno run. Below is L-jet (red) vs. Megasquirt (black, with AFM still in place). Once I tune, I need to get that AFM outta there.



Thanks everyone for the help,
Bill in MD

Looks like my dyno graph upload doesn't have very good resolution. Both runs basically overlap with rear wheel horsepower of 147HP at 5060 rpm. If you assume a 20% power train loss, that's 183HP at the flywheel. Big woop, doesn't sound like anything more than stock. I have S-cams, S-pistons too. Bummer. What power train loss should I be assuming?

If you want to see a slightly better quality version (not much) than the above version, go to http://bmacfarl.fastmail.fm/alfa/MS%20v ... 9-2005.jpg

Looking forward to your comments.
Bill

Bill,

Do another run with the G-Tech and send me a datalog of the idle, the idle problem and the power run.

Your last idle datalog you sent me showed a tach fluctuation up to 6400 RPM at idle. Make sure all your wiring is good. Clean the connectors on the ignitor/amplifier under the coil, and the coil connectors.

Email me a datalog and your MSQ file. I'll take a look at it.

-Peter

Bill, I have to make a couple comments. Although Peter would know better, I don't think Megasquirt will give a big power increase until you remove the AFM. After all, your L-Jet does a decent job of managing fuel on a nearly stock motor. Removal of the AFM is where most of the gains will come from, as well as flexibility on a motor with really hot cams etc.

Regarding the dyno run. Are those numbers corrected for atmospheric conditions? A lot of dyno operators just give you raw data which is semi meaningless since atmospheric conditions can vary enough to cause a huge variation in power output. I have a small discussion on this in a post I wrote today under my 2.5 supercharging results. If it's just raw data I can adjust if for you, just give me the where and when you did the dyno pull. A 20% driveline loss is about right.

Without seeing the datalog, it's hard to say whether those power numbers are the best you can do.

Greg is, in theory, correct about the L-Jet. This is based on 2 factors though. AFR is 14:1 and BSFC is .45.

We've previously assumed the Alfa V6 can get better than a .45 BSFC, and we know max power (insert Homer voice here) is achieved at somewhere around 12.5:1.

We also know that an engine is an air pump, and without more air, more fuel isn't going to do much good.

So, with the above qualifier re: the datalog, at 14:1 AFR I'd expect the MS to give similar power to the L-Jet.

The advantage to a programmable over the L-Jet is being able to adjust and monitor the AFR for max power. This is where the datalog is the best tool for tuning before the G-Tech. If you can tune the VE table (and req_fuel) to get a consistent 12.5:1 under acceleration and 14:1 under cruise you should be extracting the most out of your engine.

Now the biggest problem IMO is getting air into the engine. I think we all agree that removing the AFM is a giant step forward. The MS allows you 2 advantages. First, it's a speed/density system so no measuring device is needed in the airflow, and second that once you increase that airflow, you can adjust the fuel curve to keep the AFR in the max power range.

All this is under the restriction of the stock timing curve, which is again optimized for emissions and economy. Give the static timing a little bump along with the lower AFR. I'm timed at 25degBTDC static at idle which made a huge difference in power. I can feel it in my trouser seat from 22degBTDC to 25degBTDC. Too much of a good thing isn't going to help. I bumped to 33degBTDC static and added a lot of fuel to stop detonation (unburned fuel for slower burn and no ping) but the power really dropped off. I could get good power at that timing but too much pinging for safe running.

Welcome to the world of programmable engine management. I told you guys tuning was the hardest part!!

-Peter

Good points, Peter and Greg. Although I know I am far from peak performance (1. still have AFM, 2. haven't learned how to tune), I thought it couldn't hurt to post some results.

Regarding Greg's question on atmospheric conditions, I guess I have to say I don't know how this little G-tech gadget works. Its a thing that figures out torque & HP based on monitoring your ignition signal. There's nothing in the manual about compensating for atmospheric conditions. I'm here in Maryland though and we can't be too far above see level if that helps.

I was a little surprised with the HP/torque plot results. My Milano sure felt livelier with Megasquirt, but then that may have been the psychological effects of trying to build and install MS over 6 months. 147HP at the rear wheels is a little depressing, but the G-tech really only provides me a run-to-run comparison, not real HP/torque numbers.

To Peter's point about tuning, damn right its complicated. Might take me another 6 months to have this thing running right.

If Michael is out there, let me know how your tuning is going. We seem to be at a similar stage of the process.


Thanks for the help ,
Bill in MD

Yep, tuning is difficult. Each engine is different.

There's no way the G-Tech can tell power from ignition. However, the MS can gauge HP from fuel based on BSFC.

BSFC = fuel_flow/HP or HP = fuel_flow * BSFC

See this thread to add the Torque and HP gauges to MegaTune:

http://www.msefi.com/viewtopic.php?t=17 ... +hp+gauges

Obviously, AFR is a factor. Dumping fuel isn't going to increase HP. The gauges take AFR into account but also assumes a .45 BSFC that can be adjusted in the custom.ini file.

Copy this and save it as custom.ini:

;###############################################################################
;## DO NOT edit this file in the MegaTune release directory, edit the copy ##
;## in your car-specific directory. That would be ##
;## ##
;## C:/Program Files/MegaSquirt/car1/mtCfg/custom.ini ##
;## ##
;## in the standard installation. ##
;###############################################################################

[MegaTune]
MTversion = 2.25

; This "defaultVersion" is only used for off-line editing, once you
; connect to MS, this is discarded.
defaultVersion = 2.00 ; Controller code

[Defaults]
engineDisplacement = 350 ; Cubic inches
injectorFlow = 42 ; lb/hr

[Units]
TPS = "%" ; Either "%" or "raw" are valid.
#if CELSIUS
#log "Celsius enabled."
temperature = "C" ; Either F (Fahrenheit) or C (Celsius) are valid.
#else
#log "Fahrenheit enabled."
temperature = "F"
#endif

;-------------------------------------------------------------------------------
; Add your customizations here, they will be read at the end of the
; standard megatune.ini processing and override any settings there.
;-------------------------------------------------------------------------------

#unset AUTOMARK_LOGGING_FEATURES
#if AUTOMARK_LOGGING_FEATURES
[OutputChannels]
; Make up some variable names...
highLoad = { rpm >= 3500 && throttle >= 50 }
enable = { rpm >= 2000 }

[Datalog]
enableWrite = enable ; Log entries are only written when the variable
; "enable" is true, see it's definition above.

markOnTrue = highLoad ; A datalog marker is written when "highLoad"
; transitions from false to true. The marker
; looks like "MARK 001 highLoad" to distinguish
; it from the manual markers.
#endif

;-------------------------------------------------------------------------------
; Uncomment this section if you wish to change the defaults. The commPort
; used for burst mode is normally the same as used for MS, but if you have
; your Palm hooked to a different port, then change it here.

#unset I_WANT_ALTERNATE_BURST_MODE_SETTINGS
#if I_WANT_ALTERNATE_BURST_MODE_SETTINGS
[BurstMode]
commPort = 1
commRate = 115200
getCommand = "A"
#endif

[OutputChannels]
totalFuel = { 206.5 }
BSFC = { 0.45 }

#include "bhptorqueGauge.ini"

; [FrontPage]
; gauge8 = bhpGauge
; gauge4 = torqueGauge

;-------------------------------------------------------------------------------

Copy this and save it as bhptorqueGauge.ini

[MegaTune]
MTversion = 2.25

[OutputChannels]
; At max power, BSFC is usually between 0.42 and 0.58 lbs
; (Supercharged:0.55, an efficient N/A:0.45)

maxpower = {afr > 13.0 ? 0.0 : afr < 12.0 ? 0.0 : accDecEnrich > 100 ? 0.0:accDecEnrich < 100 ? 0.0: 1.0 }

; So at max power + no accenrich, BHP roughly=(dutyCycle1 / 100.0) *
; (lbs/hr of all your injectors / BSFC)

bhp = {maxpower < 1.0 ? 0.0 : (dutyCycle1 / 100.0) * ( totalFuel / BSFC ) }
torque = {rpm < 1.0 ? 0.0 : maxpower < 1.0 ? 0.0 : 5252 * bhp / rpm}

[GaugeConfigurations]

; Name Var Title Units Lo Hi LoD LoW HiW HiD vd ld

bhpGauge = bhp, "Engine Power", "BHP", 0, 400, -1, -1, 350, 400, 1, 1
torqueGauge = torque, "Engine Torque", "ft/lbs" 0, 400, -1, -1, 350,400, 1, 1
;

Then add the following line to megatune.ini

#include "custom.ini"

That will give you the gauges in MT to measure Torque and BHP within MT based on fuel consumption. You might want to play with BSFC to say .40.

The calculation is simply injector rate, elasped time and PW.

Do a power run (which it won't datalog without some custom programming in the megatune config files). Watch the telltale in MT for the max setting. Use the tuning screen (preferably with a ride-along assistant) to adjust the VE table to max power and AFR.

HTH

-Peter

Even at a give altitude power can vary by 20% due to atmosperic conditions.
Peter, it's also worth pointing out that removing the AFM will improve BSFC, so the power gain is not just from more air. In fact it might be interesting to see if there is more air which you can do if you have a way of very accuratly measuring manifold pressure. I would not be suprised if most of the gain is from the improvment it BSFC. Thats a big factor, every horsepower you save turning the motor is a horsepower increase at the flywheel.

Greg,

That's what I was trying to say IRT BSFC. The assumption under stock configuration is .45. It can/will improve.

The MAP sensor will very accurately measure plenum vacuum under certain conditions. A datalog with and without the AFM will show something but without a pot-type TPS, it's hard to say whether there's more air without knowing the airflow at a particular TPS ADC count. If Bill gets the 164 throttle and 122 001 TPS installed before the AFM is removed, we can prove that very easily.

We can also measure BSFC on a chassis dyno by using the BHP and Torque gauges in MT and RWHP on the dyno. By removing the AFM, the BHP should remain constant in MT with no changes but the RWHP should go up.

-Peter