LPG injection up and running: my tuning story

[kikkegek]

I changed the name of the topic to "...my tuning story" and will be sharing my tuning story as I move forward on my project in this topic. Hopefully it will help others and it will give me a place to upload and store info for future use.

Some info on my setup:
GMC C15 Suburban, see projectpage in my signature
engine: 454 Chevy Big Block low compression about 8:1
single fuel: only LPG injection
ECU: Megasquirt MS3X (see tune in attachments)
Reducer: Tomasetto AT13XP set at 150 kPa gas pressure (rail pressure - MAP)
(replaced this reducer: Femitec FR400)
Injectors: Valtek 30 BFC, 3 Ohm
(replaced these injectors: Femitec GIAC-01 (2 Ohm)
Peak&Hold boards (4A/1A) from SECU-3.org
I have extra sensors for LPG connected in one itegrated unit:
- gas pressure (mpxh6400a)
- gas temperature (10 kOhm NTC)
- external MAP sensor (mpxh6400a)

Ignition is:
FAST dual sync distributor for crank and cam signal
8x LS D585 coils at 3.5ms dwell

Transmission: 4L80E
TCU: Microsquirt connected by CAN to the MS3X

YES! with the help of a lot of people I talked to and some of there here on the forumI got my LPG injection system up and running.

I even went for a testdrive tonight. Got my Suburban up to over 70 kph even and felt really good driving it!

I blocked the idle fields in the Autotune, cause somebody told me Autotune can be a little rough in these fields, and then went on a drive.

I drove great. First a little through the neighbourhood at low speeds and then took the open road. Hit the throttle and got it up to about 70kph. Pulled into a parkinglot to see check out my laptop and the and what autotune was doing. At first it idled fine and then suddenly for no reaso I can think of so far it started surging.

I didnt want to get stuck in the parkinglot so put it in drive and limped home. It ended with a big LEAN bang in front of the door. I dont think the log file gor the LEanb bang though. But the surginng is clearly in there in the 2nd part of the log file

Could you guys give my MSQ and log file a look and see if you can give my any pointers as to what might cause this surging?

I have added the MSQ file to this topic, but the LOG file was over 1MB so I have shared it from my Google drive. Hope that works:
https://drive.google.com/file/d/17bGoUF ... sp=sharing

thanks...Im so excited.
here is a video of my first cold startup I made: https://youtu.be/xuskv7_qB6Q

and this is my Autotune settings:

first testdrive surging 2021-01-07_19.59.12.jpg (511.82 KiB) Viewed 4675 times

[shawge]

Nice job! For the surging, I'd suggest flattening/smoothing the VE table around the areas where that is happening. Maybe flatten the area highlighted and smooth the transitions beyond that. Especially around the 65-60kPa to 55kPa at 800rpm where it jumps from 45 to 60.

I think the motor will be sensitive to large transitions of VE with the RPM range in the table being close together.

[kikkegek]

I edited the first message to make clear I'll be using this topic to share my tuning story as I move forward and get to know my setup better.

Lot has happened since this first post.

my starting problems were caused by a leaking injector. Flooding the engine after each first start and making a second crank allmost imnpossible. We found it making a FLIR video:
https://www.youtube.com/watch?v=_nwfUuwFrO8

#2 injector was clearly causing the problem. I opened it up and it has a small metal piece on the rubber pad causing it not to close. Took it out and it was fixed. I did order new injectors, just to be sure, but for now the system holds pressure. Even after a couple of hours. So its safe.

I wasnt able to start my car at first, because I had far too much fuel going in on cold cranking. Through my friend that has a caddy500 V8 I learned that he used only about 20-25% cranking pulsewidth when really cold. I lowered mine too and now it even cranks at -1C recently.

From a couple of guys that have shared their tunes with me I found that you need to have a opposite compensation for fuel temperature compared to gasoline.

gas_temp_correction.png (26.08 KiB) Viewed 4589 times

I went for my third testdrive today. I even got to take it around the blok for a small spin, without it stalling at the traffic light. I swapped my Thrush Welded 2.5" mufflers for some 2.5" Turbo Mufflers:
https://youtu.be/8AxK4C8EZ2s

video of just the turbo mufflers:
https://youtu.be/oJ2b13jK6mY

I have attached my latest MSQ file and uploaded the big LOG file to my Google Drive. Maybe somebody can check to see if its dfownloadable this way:
https://drive.google.com/file/d/1BGeIK3 ... sp=sharing

3rd test and new exhaust 2021-01-17_14.50.35.msq

(285.47 KiB) Downloaded 121 times

About half way through the testdrive I have turned ofd EGO control. As soon as the fans come on it will cause surging and for some reason I dont understand yet I was unable to increase rpm aboout 1200rpm without it going lean and Autotune would not adjust the VE to fix it.

As soon as I turned EGO control to "no correction" it was fine and autotune adjusted the VE table just fine. What might cause this?

For now I have a pretty safe (little rich) AFR table. Stoich for LPG is 15.5 but for some reason it will not idle strong at 15.5 . It idles best between 13 and 14 AFR. Any ideas what might cause that?

Together with a buddy Im trying to find out how to adjust for LPG gas temperature. Im using the "" fuel temperature option" for that. LPG is a funny fuel as it is richer when cold and leaner when warm. Just still finding out how much.

I do notice that when the fans come on, my voltage drops to 13.7V and I notice it starts surging a little. Its probably visible in the logs too. What might cause this?
My 2 efans are controlled by PWM though the Generic PMWA and B options.

[kikkegek]

I recently read here on the forum somewhere how to change your scale for reading in Megalogviewer. So I made a short video and thought I'd share it also over here for future reference.
https://youtu.be/QC42nThb6iM

[kikkegek]

So I just went on my 4th test run and took my car to the gastation to fill it up. I was pretty nervous, cause I had never been further then around the block, where I have spots all over the place I can stop in case of tuning problems.
But since I got the idle under control and been able to drive it when gas temperatures are 50-60C (so around or close to opperating temperature for the reducer) I felt it was time to take the next step.

the drive was very nice. Felt really good to drive my car with a working LPG injection system. I stopped a couple times along the road to check if the fuel temperature compensation was doing OK. And its actually already pretty well tuned. Cause the VE cells in the idle spot where I have only had a chance to tune a little didnt alter a lot. Stayed within 48-50.

The only thing that I havent gotten to work is the EGO control. As soon as CLT gets above 70C and it activates the AFR starts bouncing high and low AFRs like crazy. Its visible in the log file from a CLT temp of 70C and up. I stopped the car and just turned it off. And let the autotune on to do its job.

Maybe somebody can take a look at my tune and the big log file and maybe give me a pointer to why it fluctuates so bad?

My MSQ file is attached to this post.

This is the log file in my Google Drive for the drive to the gas station:
https://drive.google.com/file/d/1YoWZDz ... sp=sharing

This is s short log file when I was allmost home back from the gaststation:
https://drive.google.com/file/d/1xaobkx ... sp=sharing

[kikkegek]

went on another run and made another log file.

its also in my Google Drive because of its size
https://drive.google.com/file/d/1fVmRaw ... sp=sharing

Also added my latest MSQ file to this post. I feel like I'm getting pretty close to the temperature correction for the gas temperature of the LPG (using fuel temperature correction option).

Something I'm batteling is that when its warm and I go from Neutral or Park to Drive that the rpms drop so much, that the engine will stall if I dont catch it with the throttle. I think it happens because the stepper motor for IAC closes fast to compensate for the rise in rpm. When I zoom in on the log moment I see that
the rpms rise first,
the stepper starts closing,
then rpms start dropping,
then log seems to register the gear change (why so late?)
then rpms drop too fast for the IAC to step up
and the 2 degree raise in advance of the timing arent enough to catch the falling rpms....hmmmmm

any thoughts? anybody?

stall_when_put_in_gear.jpg (481.89 KiB) Viewed 4535 times

[kikkegek]

went on my 6th testrun tonight and took it to the highway. It performed really nice.

Video during my testrun
https://youtu.be/80pgXrzS7vU

I have set the Autotune to:
Coolant > 70C
and
Fuel temperature > 40C

The reducer takes a lot longer to get up to operating temperature then the engine. The highway drive gave me a nice opportunity to see where the LPG gas temperatures stabalizes during a cruise at 100 kph. It seems to stabalise at about 40C during this weather while cruising the highway. Somebody that has been driving LPG injection for 15 years now to find this number and choose that as the point for no fuel temperature correction and correct the fuel above and below this temperature in my Fuel temperature correction table.

6th testrun highway drive.jpg (498.15 KiB) Viewed 4503 times

I have made two nice big logs and uploaded them again in my Google Drive for other to view.:
Highway drive forth
https://drive.google.com/file/d/1ou1ne9 ... sp=sharing

Highway drive back
https://drive.google.com/file/d/1ZwN_aK ... sp=sharing

I also added my latest MSQ tune file to this post.

Whenever my EGO would come on it would correct the AFR like crazy. making it bounce all over the place. Somebody pointed me to the manual again. Where most videos tell you to tune the P first, the manual says to tune the I first. So I did that a little while idleing. And I noticed anything over 15% would make the AFR oscilate. So for now I have set the I to 10% and P and D to 0%. And that seems to work pretty good for constant conditions. Next thing I need to figure out is how I need to adjust it to make it respond to quick changes, without over reacting.

Something I also noticed is that my engine seems to run cooler now. On the highway it would normally have an operating temperature of 85-87C. This was with a IMPCO 300 LPG mixer mounted on the throttlebody. And now it was at 79-80C on the road out and back.
Maybe thats from all the cilinders getting fueled correctly. And not rich center cilinders and starving outer cilinders?

[kikkegek]

well...the ride keeps improving day by day.

I have set EGO control to onlye 10% for the I. Andy higher and its start iscolating. And with 10% it really nicely corrects so far. Havent really looked into flooring it. But under staedy conditions like cruise and mild driving it does a really nice job.

This are two big log files from today.
https://drive.google.com/file/d/1TmGqcL ... sp=sharing
https://drive.google.com/file/d/1gYFW4P ... sp=sharing

and added my latest tune as attachment to this post.

I have slowly been adjusting my fuel temperature adjustment and cranking. And I can now crank the car in the morning without touching the throttle and I seem to be getting close to the correction that the gas temperature needs (fuel temperature correction).

Only thing Im fighting is the engine allmost stalls when I go from N to D or from P to R. I already adjusted the ignition around my idle area to raise max 2 degrees when it drops in RPM and I have a IAC. But they both have a hard time catching it.

Maybe somebody can take a look at my MSQ and give me some pointers?

this is my timing. I idle at about 700rpm @ 50-55kPa MAP

ignitiontable.jpg (303.63 KiB) Viewed 4481 times

[kikkegek]

this is a part from my log where you can really clearly see what happens.

I put the trans from D (gear 1) into R (gear 255). It first rises in rpm from going through Neutral. Then gets into Reverse and then allmost stalls.

log_trans_P_to_R_or_D_engine_allmost_Stalls.jpg (494.52 KiB) Viewed 4479 times

[kikkegek]

I have started a seperate topic for my problem that the engine stalls when I put it in R or D from N or P at idle. before I had the LPG injection it had no problem catching the extra load from gear engaugement of my 4L80E. But now it will stall or allmost stall.

viewtopic.php?f=131&t=74550

during my testrun today I made a couple full throttle pulls. I noticed that full throttle isnt 100%. Typical for these cars that the throttle body has more range then the throttle pedal. I need to look into why my pedal wont go 100%. But...from the data I also saw that my injectors were allmost maxing out. They are at 91% duty cycle at 4400rpm with 91% TPS. I believe I read somwhere that you should keep that under 80% at peak RPM? which is 5,000rpm for my engine, although I make my 4L80E shift at max 4,500rpm.

I have a replacement set of Valtek 3 Ohm injectorrails on the way. My current injectorsnozzles are drilled at 2.6mm but my buddy with his caddy500 engine has these Valtek injectors drilled at 3mm. Since my 454 is pretty close I probably need to open them up too this size too.

90p_injector_dutycycle_at_4400rpm_and_91p_TPS.jpeg (207.8 KiB) Viewed 4452 times

[BigBlockMopar]

Nice progress Jochen!

[kikkegek]

Nice progress Jochen!

He thanks BigBlockMpar!

I have done some more driving today and connected my Cold air intake hose. it doesnt look nice, but I logged a noticeable difference in Incoming air temperature (MAT). Its connected to the original intake opening behind the grill and you can really see the difference. Its allmost 15 degrees cooler and you can see it drop as you start driving.

Cold_air_intake_hose_connected.jpg (207.8 KiB) Viewed 4236 times

I didnt relly change enything in my tune today, so no updates there. Still chasing the stalling when putting it in gear. which went fine when I had the LPG through a mixer on top of the carb.

I did notice a small leak at the header where the O2 sensor is also in the exhaust. I have new donuts so I'll replace those when the exhaust has cooled down. Exhaust leak is never good of course when tuning of a wideband

this the log file from the morning drive:
https://drive.google.com/file/d/1OKGQf_ ... sp=sharing

this is the log file from after I connected the Cold Air Intake hose:
https://drive.google.com/file/d/18VjiKy ... sp=sharing

I also learned something cool yesterday from a video that Andy Whittle shared:
https://www.youtube.com/watch?v=Rie98I1y-xo

In my log files the Reverse gear was logged as 255, while all forward gears from Neutral/Park are logged as positive numbers 0 to 4.
I didnt like looking at the log file "autoscaling" this. I wanted Reverse to be viewed as -1.
From Andys video I learned how to change it.
There is a complete list on Efi Analytics website of options:
https://www.efianalytics.com/MegaLogVie ... mulas.html

this is what my log now looks like:

Improved_log_viewing_for_gears.png (7.41 KiB) Viewed 4236 times

and this is the simple formula that I used for changing that view

Changed_R_255_value_to_-1_value.jpg (36.2 KiB) Viewed 4236 times

[Marek]

The throttle doesn't need to go to 100% to be wide open, except at very high rpm. This is because the engine will get all of the air it can use at low revs with a medium/low throttle opening. (You can confirm that by plotting throttle opening percentage numbers on to a load versus rpm graph.)

Your gas injectors appear to be too small. Either increase the gas pressure or increase the injector size to get the duty cycle down.

Also consider whether your reducer flows enough gas volume. The line for gas pressure appears to move about quite a bit. I would expect it to be flat except for a very sudden increase in acceleration, when it might briefly dip for an instant.

kind regards
Marek

[kikkegek]

The throttle doesn't need to go to 100% to be wide open, except at very high rpm. This is because the engine will get all of the air it can use at low revs with a medium/low throttle opening. (You can confirm that by plotting throttle opening percentage numbers on to a load versus rpm graph.)

Your gas injectors appear to be too small. Either increase the gas pressure or increase the injector size to get the duty cycle down.

Also consider whether your reducer flows enough gas volume. The line for gas pressure appears to move about quite a bit. I would expect it to be flat except for a very sudden increase in acceleration, when it might briefly dip for an instant.

kind regards
Marek

Thanks Marek,
could you maybe share one of your log files with me? I'd like to see what your gas presure looks like in you logs.

Yes I have new injectors on the way, cause one of my current injectors was damaged. Im going to open up the nozzles from 2.6mm to 3.0mm, because I dont want to raise the reducer pressure any further.

[kikkegek]

I seem to have fixed my stalling engine going into gear from Neutral of Park. IInfo is in this topic, including my latest MSQ tune

viewtopic.php?f=131&t=74550&p=567222#p567222

[kikkegek]

test

[kikkegek]

took my car to a quiet road tonight and did couple WOT runs to see if the 0-100kph time improved with the LPG injection versus the traditional IMPCO300 LPG system.

well...Ladys and gentlemen, the results are in.
And we have a winner
LPG injection 1 - IMPCO300 0

took 0.8 seconds of a 0 to 100kph sprint

and as you can see from the curve it seems to accelerate quicker. Road was cold and had wheelspin on take off. And was rich mixture all through the pulls...but it clearly feels stronger...too bad I dont have any dyno info from the IMPCO300 system

Dragracer_LPGinjection_vs_IMPCO300_small.jpg (107.26 KiB) Viewed 4197 times

This is the log file I took:
https://drive.google.com/file/d/19mw-oE ... sp=sharing

I did find in the log file that indeed my injectors are maxing out. At little over 4400rpm they went up to 99,6% duty cycle.

WOT_run_99.6p_dutycycle.png (142.13 KiB) Viewed 4197 times

I had some new Valtek 30 injectors on the way and got them in the mail today. By pure luck I seem to have ordered BFC (Improved flow injectors for big cube engines). Drilled at the correct nozzle diameter that should fix my dutycycle problem.

Valtek30_BFC_02.jpg (96.86 KiB) Viewed 4197 times

Valtek30_BFC_01.jpg (97.26 KiB) Viewed 4197 times

[kikkegek]

OK another small update

I made a dashboard for my smartphone for MS Shadow Dash, so I dont need to have the laptop open en look sideways to see whats going on. The tune is good enough to drive around whatever the LPG gas temperature is now. I got that pretty good tuned for now by driving and adjusting on the go, without modifying the VE table and just seeing how much more or less fuel it needs.
https://youtu.be/HHf5FFMgHck

Conclusion for this dashboard is:
1. I really like it.
2. My smartphone (Huawei Mate9) is too small or I need a 7 or 10" tablet for dashboard hahaha

Something I got worked out is my cold engine start. I had figured out my LPG gas temperature correction through the fuel temperature correction. But now at cold start I'd be pulling up to 7% out of the fuel. And when I started the engine it would search and be lean like 17-18AFR. So I'd up the correction to pull only 2 or 3% and then would start driving and it be too rich hahaha...so...what I did is I left the fuel temperature correction alone and used the WUE to add 10-15% fuel when the coolant is cold. Because I noticed as the coolant warmed up, the engine ran fine while gas temperatures were still low (gas temperature warms up really slow) and pulling up to 7% fuel.

Also not always the engine would stay running after cranking. I fixed this by using the after cranking enrichment and made a taper of how fast after cranking that will taper of.

And now it cranks withouth touching the throtttle just by the IAC and calms down to a really nice idle.

I dont start EGO untill 40C coolant. Untill that temperatue I run the engine on the rich side through WUE at about 14-14.5 AFR. I have noticed it needs that when cold.

driving it around now not using autotune at the moment and EGO seems to be able to correct fine. Just not sure if my EGO is fast enough when driving conditions change. Like take off from a stand or when increasing throttle. I have tested a little in the parking lot with dfifferent I and P settings, but settled with I at 10% and P at only 1%. Any more of both causes more oscilation and doesnt get a quicker matched or close AFR to target. Im just not sure how fast is OK. Maybe somebody can take a look at the last part of my log filewhere I go from a stand still and do some driving untill I get home and tell me what you think about AFR vs target AFR and EGO correction.

thanks and I'll keep sharign what I find here.

log file from tonight:
https://drive.google.com/file/d/1Wa1n3H ... sp=sharing

and I added my latest tune

[kikkegek]

Well...that didnt go as planned. Installed the BFC Valtek 30 injectors and drilled them to same size for first startup. So 2.6mm.

Car was very hard to start and didnt have any power beyond idle. Sounded like it had a very bad cold. So I went a little research, because I thought I was overseeing something. I had to up the VE from 47 to 70 just to get an idle. Something wasnt right, because these were supposed to be bigger.

Nozzle ID Femitec GIAC-01 vs Valtek 30 BFC.jpg (157.61 KiB) Viewed 4137 times

Nozzle intake ID Femitec GIAC-01 vs Valtek 30 BFC.jpg (150.97 KiB) Viewed 4137 times

Valtek 30 BFC valve seat ID.jpg (171.72 KiB) Viewed 4137 times

Femitec GIAC-01 valve seat ID.jpg (134.63 KiB) Viewed 4137 times

I checked all the spark plugs, thinking I had a bad sparkplug. But they were all fine.

Then decided to check the injector pulses in test mode. Here is the video of the Valtek 30 BFC injector test on the car. You can clearly hear they all sound different and InjD isnt even responding in sequential mode, only in individual mode and after a couple pulses. So proabably sticking. I bet they all dont open 5mm as they should from the specs
https://youtu.be/9cSbfZ-D6Ps

So I ended up putting the Femitec injectors back and Im glad I saved and stored tunes with clear names. So it started back up right away. I'll stick with these untill I get these Valtek injectors figured out.

[LAV1000]

Hi there, kikkegek.

Don't know how you run the current injectors.
But if you use resistors or Peak and Hold drivers then you need to adapt the values to the specs of your new injectors.
And after you figured the above out then you also need to check out deadtime and voltage compensation of your new injectors.