Josh's turbo 1098cc carburetion sucks lets try EFI
Turbo Mini Forums

so TS whont connect think its a usb problem with the computer

little progress used a different computer to get it going and its running but crap as to be expected with no real tune and not at all setup for the 5 port

a little improvement

spark was ragged so took 4 deg out from everywhere after 1200rpm it had to much ramp i think anyway


its hot today about 34-35c hence IAT 45c

Attachments:

43.86kb

46.93kb

Edited by Turbo This.. on 3rd Dec, 2017.

Good work. What injection timing are you running? I usually find 460-470 degrees works well at idle with the single pulse method.

EDIT: Added 360 to the numbers.

Edited by Paul S on 3rd Dec, 2017.

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."

I really like your persistance, engineering and trying new stuff.

I was wondering about the position of the injectors; won't they suffer a lot from manifold heat?

Anyways, keep it up!

Dazed and Confused....

paul that log was with out any timing that's with seq off and 5 in the trim table on two injectors and added or took from the ve table i found spark had quite a lot of influence

i did go in and turn on seq and set it for both 105 and 465 i assume the added 360 is co the cam sig is out ? and both hardly run the idle hunts real bad so i double the injector size for fun as it seemed the balance was about right at 180 but it was going lean rich wildly the dubbing of injector helped but its still sounds off on the exhaust note like a lean burn back fire but not quite

more correctly when the revs come up it goes lean and rich by a lot

i did re check the base spark locked it and matched it in ts with the light also checked the plug order both ok

bit annoyed as i cant make heads


yo han thanks mate and the injectors well we will see if its a problem i was going for a flatter aim at the center line in the port but yes they may be unhappy with the heat but they are largely in alu the mani and rail so should sink most of the heat its the body thats getting the radiant heat
the ex is quite a way tho more than it looks anyway

Edited by Turbo This.. on 4th Dec, 2017.

so it helps if i use the correct injector wires.. had H not A like a dope so its now sounding correct on the exhaust note..

but still hunts on rpm a lot like nearly stall to 3k
it swings full scale on the o2 gauges however they both go the same way and both together

just to rule out the spark made the whole table the same at 7 deg thinking the table was making it hunt

not sure what else to try?

sounding a lot more promising then,

first thing i would do is make the advance table around 15 deg at idle, 7 deg is way low, or put the original table back (ndepending what it looked like... but basic rule of thumb, around 15deg at around idle, high 20's to 30 at full bore NA or at 100kpa, tapering back to 20 deg at 1 bar, over run at high revs can be up around 40 deg. that should be a good safe map to get it running on.

once youbve done that and established the sparks are in the right place compared to whats on the screen, swing the req fuel, you might need to piss around with the throttle opening ( idle speed) a bit, I always get the engine runing without any form of electronic idle control valve first,

also bear in mind once you get somwhere near a stable idle you need to get the motor up to temp, sort the idle, and then go back and do the warm up mixture once its cold again. that will take a few times, but you might find the cold idle needs richening a lot,

Sounds like you need to gain an understanding of the issues of port injection on a 5-porter.

If you do not time short pulses through the open inlet valves, then it will not run right.

What is your injection strategy?

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."

ok ill get the spark back looking like a real table and ill grab a spare plug and test each output in the test window to insure the ecu is firing the cyl it thinks it is

i had bumped the ve table around a lot from about 25 to like 200 most of the time it didn't make a lot of difference as long as it didn't stall just pig rich or dead lean but not the same on both O2s

i have the throttle open a fair amount quite a lot more than it ever will be when all sorted but this is just so im not hitting the starter every 5 seconds

the IAC stepper is un plugged and the valve is normally closed

my limited under standing is that i can use one injector per runner and fire it once with its window of time or spread of fuel spray start as the first valve in the runner is closing and finish as the second is opening swinging the balanced by the injector timing table over the two ports? adding or taking fuel from one to the other by adding or taking timing


what i not sure on is how the time is shown in the ecu and how it relates to the cam card from kent and if at idle the injector has time to do two separate pulses only when the respective valve is off the seat?

is the timing table 0 as in TDC firing and with intake timing of 12/56 248 dur is that 124 in the table going to put the middle of squirt into the open valve? assuming the event is set at mid not start or end

i have injector wires
A and D joined and go to 1 and 2 cyl runner
B and C joined and go to 3 and 4 cyl runner

i think the ecu is sending the injectors to fire in the same order as the spark but at the timeing table degrese but is that the same as spark baised on TDC for cyl 1?
so mid valve open time would be 124 in the table take a bit to allow air speed and injector open time? so say that 100 or 110 is window for start of spray?

but then how dose this apply to the others 2 3 and 4 cyls with one getting the air speed going the other getting the higher VE anc cyl filling needing more fuel can each cyl be trimmed timeing wise or is that in the indival cyl trims? ie just add say 10 in 4 and 1 to get it back richer

at the time of the above two logs i had 180 deg set for timeing but at the time i didnt know i was useing injector wire H not A haha oops the wire labels have faded a lot and hard to read

is there timing tables for each cyl? and is the table in crank deg or cam?

injectors ar 950cc each one per runner and i have a base fuel press at 40psi as low as my reg and pump will go it is how ever an adjustible rate reg not fixed at 1:1 its base is 1:1 but it can add say 3psi fp for 1psi boost to make the injector flow like a bigger one but idle like a small one i think this will be inportant as i want to once its all set and ive got my head straight go back in E85 on the same injectors to use the flex fuel and dual tables

what should i have in the general settings 2 squirts per engine cycle 4 injectors Evan tho im using 2 i need it to think theirs 4? engine size 1150cc and 950cc injector size required fuel being 2ms at 100%VE?

Right a lot of questions, I'll try and answer them.

Firstly to get it to work, you will need to set it as Sequential.

Edited by Paul S on 6th Dec, 2017.

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."

well im a dope..

the crank angle sensor is mounted to the steel bracket for the rad and engine mount the fixed part that has the two long horizontal bolts that pinch in to the alu casing on the box now apparently there is a bit of slop in the holes and the bolts not 100% tight as the engine has dropped a fraction relevant to the bracket and pushed the hall sensor into the trigger wheel =( so backed it out to an air gap and no sync seems its gotten to hot from the rubbing and now dead

not sure why i didnt check the trigger logger thing or notice the sync loss flashing down the bottom of TS

frustrated for that simple thing causing me to go round on this trail of messing with stuff

also seems the return line has some kind of resonance or something as the line sets up a pulse wave thing and makes the pressure bounce
not sure why as its an 8mm hard line same as feed and minimal rubber hose with no real restriction all fittings are full id and tank vent is ok to

might try a restriction in the feed line pre reg to reduce to flow rate so the down stream can keep up this will make the pump do more work but i think it may be worth a try to see and then perhaps run a larger ret line

ball ache having to wait like a week for the hall sender but its an opportunity to get a flesh loom as well for when its all worked out i can do it nice and neat

really wanted to get it sorted before the years out

all right im back on the mini

crank angle sensor swapped and working as it should

messed about a bit and gotten it sounding semi happy altho still a hunting idle

log

Attachments:

36.87kb

file if it works?

nope says ignoring msg file

well this hunting rpm is driving me insane not sure what to look for it seems so close yet so far

it sounds good and revs well if i can only get a stable idle/ low rpm i should be right

ive been through everything in ts to make sure its off like idle control settings ect however in not sure if ive gotten it all or if theres some setting out of whack??

is there a basic file available or a file from a successful 1098 project with one injector per runner at the gasket like i have?

beginning to think its got a vacuum leak or something stupid

Attachments:

57.36kb

The forum only allows a limited number of file types as attachments, they are listed when you click on "add attachment" and don't include the megasquirt proprietary stuff but does include "zip" files so just compress it.

The logs in jpeg form show OK for me, and yes, you do have a significant AFR imbalance on the first log and lots of hunting on the second one, what did you change ?

EDIT - typo

Edited by Rod S on 22nd Jul, 2018.

Schrödinger's cat - so which one am I ???

ahh i see

think this should work?

im not really sure what numbers where before hand i was just adding or takeing from the whole table both timeing and ve or what i see as average fuel volume for rpm/manifold pressure i beleve the spread or inner vs outer ratio is sorted in the timeing table for the single ingector per runner methord as ive chosen

please excuse the poor spelling its been a long day

i feel like i should prove the contraption of an intake for air leaks that means i need to remove it make plates for the throttle body and cylender head flanges and put say 15 psi in it with a ball valve and see if it holds it for a perid of time

also as mentioned above i should un bold the idle air valve and place a block off plate there or duct tape it a few times and re bolt it to insure no flow

im just not sure if its a computer or maniacal problem

perhaps there is some weird resonant pulse thing in the intake ?

A zip file should work but I don't see one attached.

Looking again at the plots just as jpegs it seems the cycling (hunting) is around 2Secs which could easily be the IACV not setup correctly in the software rather than leaking air.

But your plots don't include the IACV steps (or PWM, without the msq I can't even be sure what the IACV is).

Also, you field maximum an minimums aren't consistent, particularly the AFRs which one one plot is 10 point something to 18 point something (they are not even both the same) and on the other plot, 9 point 8 to 14 point something.

It makes looking at data very difficult. Ideally you should set MLV for the minimum and maximum values that your widebands display (usually 10 - 20 or 9 - 19, Innovates are a slightly greater range but still specified) and keep it like that for all your plots.

Also, timing appears to be a fixed 15 degrees so it can't respond to any instability ?

Schrödinger's cat - so which one am I ???

i will have to try again on the zip file

i went and turned off anything i could find thet may be worm up or correction related to try get something stable along with setting the whole ignition table to 15 deg and the area where its hunting in the ve table to the same value with the thought that its chasing its self around the tables
work is crazy at the moment 13+ hours days and building my house on that makes it hard how ever i dont want to trailer the mini over the the new house and also dont want to risk another melted piston

perhaps i should start a fresh tune file and see where that goes

im not sure what your saying about the max min values in the log files? are these not just the top and bottom end of the recorded values in order to get the best resolution?

could i send the files in a PM? or email?

For the min/max values, yes, the default setting is it looks for the highest/lowest recorded data and sets the scales to those values.
But if you go to the "calculated fields" menu, then "Field Min/Max value editor" you can set them manually there.

With the paid version you can save the settings so every file you open will have the same range but, IIRC, with the "Lite" version you have to reset them every time you open the programme.

You should have my email address from a couple of years ago if you want to send the msq direct (if you don't have it any more, PM me), what I'd really like to know is what is your IACV and how it's set in the msq.

Schrödinger's cat - so which one am I ???

so the current setup is

one 950cc injector mounted in the bottom of each runner with the nozzle inline with the gasket and about 5-10 degrees from flat with the edge of the nozzle basically on the bottom of the port

coil per plug remotely mounted LS style coils

fuel pressure set at 4 bar 58 psi with a boost reference line before the throttle body that is only a boost reference this reg can be set to be 1:1 or grater as it has an air bleed just like if you want more boost on a stock actuator you bleed air off to fake the signal line pressure this feature will become useful once on ethanol fuel that i plan to use later on when the boost goes up and the injector becomes to small

Walbro high pressure F90000274 in tank pump rated for upto 120psi @3.5L per min 75psi @5.5L per min so loads of fuel available

fuel lines feed and return are both 5/16 or 8mm hard line in the car with compression fittings no rubber line in car what for no fuel fumes in cab
flex fuel sensor in the return line this can also see fuel temperature this should be good as the high flow and pressure rate with the rail near the manifold may show some temp problems if there are any

pistons are Russell Engineering high silicon content flat tops with holes not slots behind the oil rings
2.602" or 66.104mm bore (60 thou over std 1098)

the crank is lightened via knife edging and wedging the throws as well as turning down the over all diameter to about 20 thou from the journals
it was then hand blended and finished with Nitrite hardening and final grind 10 under size retains the stock stroke @ 3.3" or 83.72mm

displacement is 1150cc compression is 10.75:1 cranks 195psi +/- 2-3psi total combustion chamber volume 29.48cc plain to run 7-10psi on unleaded switching to E85-E100 for more knock resistance and more boost

the intake is my own franken creation build in charge cooler and a 2" i think? 1987 ford escort throttle body with three wire TPS and two wire idle air control valve seems like a good bit of kit nice progressive linkage and its cable operated for a proper connected feel not like today's fly by wire junk im yet to try the idle air valve lets get the car running correctly first

turbo is a brand new Garrett 2052 part no 727264-5001S rated for 140-230 hp my goal is 160-180hp so perhaps a little big that's unAustralian to say aint no turbo to big in the performance world as many say how ever when i go to ethanol fuel the turbine housing needs to be bigger so it should be close

manifold is also my own creation with a short dump pipe at the front wheel its plenty loud and sounds great with that turbine whistle i will be adding a v band and creating an easy on/off "friendly neighbor exhaust pipe" for worming it up and for playing around in the garage

sample chamber technique for the wide bands might also add one for over all mixture post turbo i have the bung welded in already so its easy if i want to

36:1 crank wheel and single pulse cam just a cap head screw in the cam sprocket (i put one in at 180 out on the other side to keep it in balance in case it matters its only half crank rpm anyway) with the hall sender in the steel backing plate

charge cooler radiator is a huge 24x12x1 inch right up front behind the grill it should have no problem getting the coolant down darn close to ambient air temp to give the charge cooler a good chance to get good intake temps pump is a Bosch unit 3/4" lines

ive gone back to the stock rad and position for the coolant just for space i will soon find out if its up to the job and deal with that when it comes around

using a cast water pump and std plastic 12 blade fan this may change but it should get the job done for now

i stuffed up the gearbox and its stuck in reverse so it will be getting pulled and broken down when im in my new house around September at witch time i will sort anything else that i find needs attention im pushing on tho as i want to do asmuch as i can now like getting the base idle and no load stuff sorted worm up and idle control etc

im trying to use one long spray of fuel spread over the two cylinders chasing an erratic idle at the moment probably just some setting in tuner studio

Thanks for the description.

(I asked Josh for it after he emailed me his msq....)

A two wire idle valve could be either a simple open/close solenoid valve, a totally weird thing that I know GM (so maybe others) used which had a bi-metal strip which was heated once the ignition was turned on so slowly closed, or hopefully a PWM controlled valve. As you say, worry about it later but it would probably be best to start by blanking it off for the time being as the solenoid or PWM type might be floating freely between open and closed with the electrics disconnected.

Having been back through the photos, even after moving the injectors, your setup can only really be described as port injection, with very long runners and a plenum (incorporting a cooler) fed through a single throttle body.

In which case you should really be setting up the msq (project and tune file) as a speed/density algorithm (multiply by MAP as they say in MegaSquirt language), not Alpha-n which is what the msq you emailed me is.
Alpha-n is really aimed at throttle body injection, esp. multiple throttle bodies where there is no plenum to get a good average MAP reading.
Presumably your plenum has a MAP tapping point and an MAT/IAT sensor ?

I wouldn't be surprised if the hunting you are experiencing is because there is no feedback on actual air pressure and you have a large volume of it between your throttle body and the inlet valves.

The single pulse spread over two cylinders is proven, I even currently use it with MS2 above 2K RPM (MS2 always had the option of two pulses, one long one, or a combination of both) but I found the injection timing is a lot more sensitive with the single pulse than two discrete ones.

The MLV plot in one of the earlier photos suggests you need to move it quite a bit. The mis-match in AFRs is probably making the hunting worse as well.

But first, I'd suggest changing from Alpha-n to speed/density.

Schrödinger's cat - so which one am I ???

Just picking these few bit out:

Edited by Graham T on 29th Jul, 2018.

’77 Clubman build thread
http://www.turbominis.co.uk/forums/index.php?p=vt&tid=618189

Siamese 5 port EFI testing
http://www.turbominis.co.uk/forums/index.php?p=vt&tid=611675