im building a tt383, and have a few questions about which FI system to use... first i want to say i want it to be a tpi. but SD, or MAF? i have read about the limitations of each but arent sure how they will affect a turbo motor, and the tuning of a turbo motor. and the motor is using 2 varible vane turbos.... if that will change anything

#1
MAF... the over reading of air as the throttle is snaped open and the plenum fills... i know this will make it add more fuel then normal which is better then not i guess, but how bad will this be under a positive pressure? Its normaly caused by the vacum lessinging in the plenum as it fills with more air... i know the motor wouldn't be under boost here, but is there anything i need to be concerd with?

#2
MAF... in respects to tuning.. i am gona max it out real fast.... what are my options?? i have assces to a wide band 02, but i dont like the idea of flying blind on a turbo motor....


#3
MAF... the plain and simple restriction of it in the intake.
Is there a comon swap to a bigger one.. maby a larger bar sensor?? with a larger diamiter??

#4
SD.... im not sure i understand how reliable this will be under boost... as the temp changes and so on i dont see how it can keep up, without leaning it out too much...



#5
MAF, and SD... the plain tables are frikin small. what 8bits??? what are my options here, i could wright a 16bit table but holy crap does that take time....lots of time...


any help would be very awsome... i posted it here because it is going to under a turbo app, and my questios are directly regauarding the 2 types injection aplied agenst a boosted app.
thxs in advance
i searched and came up with this artical
https://www.thirdgen.org/techbb2/sho...ight=2+bar+maf

 

When mine goes turbo, I'm running a 749.

Easiest most simple way to go and have adequate engine management.

20 yr old stockers can still do everything newer $2000 systems can for a lot cheaper.

Plus the code has been hacked and you can do as you will with it if u have the knowledge.

Works for 3 bar too :-)

later
Jeremy

 

the concern i have will be present with eather the old system (tpi) or new systems, they both measure the air the same way... the only difrence is the newer systems have a 16bit tables, and i tihnk theres smome with even more like 32bit.

 

My main concern with our MAF systems is the limit, as you said it's maxed real quick. There are possible ways around it, but they are all band-aids, and that always means there is something better to be had.

If you convert to a '749 (sy/ty), or repin the '730 (thirdgen sd) to use $58 code (sy/ty) such as Jeremy said, there isn't anything holding you back from getting a perfect, proper tune, short of your tuning skills. If you'll be running more than 15 psi, the 3-bar patch works wonders from what I've read.

That being said, -overall- I like MAF better than SD, but only the newer MAF systems, as they don't have the limits (In fact, I have an LT1 Camaro MAF with a signal converter on my Talon). SD requires more tuning input, but once dialed in there's no reason it can't work perfect, judging by the results of some of the member's boosted thirdgens here.

Now, tell us more about the variable vane turbo setup you'll be using. Very interested.

 

the varible vane(VGT) turbo is out of a late modle 6L ford f350. its simular in displacment to my moter, but turns half the rpms.. so it works out to be very close when using 2 turbos and twice the rpm... farilmy normal for a j-yard turbo system... but what sets this apart is the use of the varible vane. Its controled by the computer, but is actuated by oil pressure on stock ford apps, so its not to hard to find another way to control the hydro pressure entering the turbo. of all the diofrent turbos to use there is some better then others. finding one out of an 05 or later f350 is imporant because they have large bearing, a bigger thrust face... and a few other goodies... heres a link to a site with some good info for you. http://www.backglass.org/duncan/ps60...l/ps60_036.jpg

 

Oh I know what a variable vane turbo is, I was just curious as to how you were going to set it up.

Has anyone been succesful in using one of those Powerstroke VVT's on a performance app? I haven't heard of any.

 

1) Airflow is airflow
2) A TPI MAF will max out. Use a larger one.
3) Yes, it is a restriction
4) reliable?? Use the correct IAT (DIY PROM board)
5) How many bits do you need? Have you done any calcs to check what you need for resolution with the dynamic range?
Some TPI SD and MAF calcs are done using 16bit. If you write code then you might want to look at the megasquirt SD ECM.

Post this in the DIY PROM board and you will probably get more responses (and a lot of "do a search").
This has been discussed over and over at DIY PROM.

 

Huh, the oil pressure controls the vanes? How are you going to control it real-time? Using a linear regulator or pulse-width-modulated(PWM)? I am guessing that it would not like PWM. Is there any info out there describing the pressure vs. vane opening?

It seems like a good project. Are you building the computer/controller yourself? The control loop seems like it will take some tweaking without the specs on the vanes.

 

i apoligise ahead for my spelling....im a mechanic, not a english teacher.
the vgt are limited position, not infinatly varible.
the way the vanes are operated in the ford trucks is like so... the computer reads rpm vs. load. there is a magnetioc field generated by this signal which moves this shaft in the control valve. (pic) this movment meters the oil threw the valve to eather side of the piston. whe one side is presured the other is vented. depending on which hold in the shaft oil is aloud therw the vanes open or close. for the vanes to be in the full open position, oil is directed to the open side of the shaft in the valve. to close the vane the oil is directed threw the other hole in the shaft, and the vane is forced closed. when low oil presure is present, the shaft is in a neutral position and the vanes are open. as for the copmuter part of the ford, the first signal from the pcm\ecm to all sensors is a flash +5.0V to establish a referance. All the sensors will change the +5V in some way. except for those that make there own voltage.. anywho, the vgt volve is more or less a actuator then a sensor, but has both charitoristics. (SP???) All it does is take a signal and turn it into a mag field to operate a vlalve, which in turn would control oil pressure etc, . there is also a constant +12v source lead to the actuator. ground is threw threw the chasis. along this signal wire to the vgt actuator, there si a constant 5v signal, the the computer changes this signal to eathier a lower to +2V or higher +8V voltage resulting in the magnetic field being stronger and puching the shaft in the valve farther to reveal a spicific hole in the shaft. the signal as far as i know is pulsed, but the pulse is so fast that the actuator cant change that fast so it dosent matter. this sounds way confusing as im writing this.

so my plan to operate this turbo is to make just make a constant +12V source to the actuator.. easy enough. then make a control board that reads engine rpm, and for anythign under lets say 0 to 2000 rpm gets +2V and anything 2001 to 3500 get +5V, and anything 3500 plus get +8Vs. Im not sure which voltage yeild what position on the turbo yet.. i will figure that out when i get that far... or get the motavation to mess with the new ambulance at work and test it. then just add oil! lol we will see how this plan changes when it starts going togther.

as for fueling, the table im going to need im gona have to make obviously, but what size i have no clue.. probly at least 32bit cause im **** retentive.

 

Most of us around here can't spell. Spelling doesn't really matter. It is more about the way one writes and explains things. You do both of those very well. The post that are written in a thirteen year old letter to a friend kind of way with grammar like "ur so kewl, wat ya doin?" are the ones that drive people crazy.

It is not confusing. You explained it very well. That is what would be called a pulse-width-modulated (PWM) valve. The ECM sends out voltage pulses to the actuator. The pulses get smoothed/averaged and hold the actuator in a constant position.
Thank you for describing it. I wondered how the VGT worked. I stock TPI SD ECM (730 ECM) has a few extra PWM drivers available that may be able to handle the electrical current that the Ford actuator requires. There is a current top 10 topic at DIY PROM talking about the SD ECM PWM drivers. It would be worth a look.

You are on the right track. You might be able to use a factory ECM to control it. It would take a lot of learning of the ECM to figure it out. I would follow the path that you are planning to take and go with a custom controller. A small microcontoller like ATMEL AVR or Microchip's PIC would work well. I would use the AVR because it has a built in analog-to-digital (A/D) converter that can be used to read the throttle position sensor (TPS), MAP sensor, and tach signal (RPM). You will need some extra circuits for filtering those input signals and a driver (FET) to control the actuator.
As a first test, I would be tempted to use just use a 555 Timer with a variable resistor that controls the pulsewidth stuck on the dash and play with varying the vane to see how it responds before going the full ECM control method. It would be a lot faster to build and give you a feel of how you want the ECM to control it. You could make notes about RPM, TPS, MAP, and how it responded.

Using too many bits is not a bad thing. It does create more work. As an example, at 6000rpm on a batch fire (TPI MAF and TPI SD) the fuel injectors have 20ms to squirt into the engine. The ECM controls the injectors in the 10s of microsecond range. This is plenty of resolution. At idle, resolution is still adequate. If you really want more resolution then I would use a different ECM all together. Not for the resolution, but for the fueling method. SEFI instead of TPI batch.


I like this project. It relates to a thread where a few people were just talking about sequential turbo setups. Using VGT is todays tech. of replacing sequential turbo configurations.
A little more detail about your 5 questions:

1) Both MAF and MAP use more than just that sensor(s). When you open the thottle to fill the plenum the ECM sees the TPS opening so it gives it extra fuel (accel. enrichment). It does not just use the MAF or MAP signal. The accel. enrichment (TPS opening) is what makes for a non-existant lean condition.

2) Flying blind is why most people convert to MAP. All kinds of automotive MAP and non-automotive MAP sensors are available up to 3BAR and beyond. MAF sensors that do high CFM are expensive so not many people use them. With proper tuning a MAP sensor works just as well as a MAF sensor.

3) If you use a high CFM MAF then it will not really be a restriction because of the large diameter pipe.

4) It is reliable. Look (or search) for intake air temp. (IAT) sensor. Most people use the LT1 IAT (non-threaded) or 1990 (and other years) V6 Camaro IAT sensor (threaded 3/8 NPT). They respond much quicker to temperature changes than a V8 MAT/IAT. The sensor thermistor is open instead of having the brass cap on the end so it can change temp. quickly, but is weaker mechanically. If you use a stock TPI MAT/IAT then you are flying blind to intake air temps. I have looked at ALDL logs using both and the stock v8 TPI MAT/IAT just doesn't respond and causes a lean condition.

5) Not sure what to say. It just isn't worth it for the time involved. If you really want 16bits or more then I would look at www.megasquirt.com. Lots of stock ECM calculations are 16bit on an 8bit microprocessor. The LT1 ones were all 8bit microprocessors (two of them). It would take a lot of proof to convince me to use a 16bit or better processor. The only calcs. I would like to have as 16bits or better would be using a 3BAR or bigger MAP sensor. The thing is that you need a 16bit A/D in order to do this. None of the ECMs have that so going to more bits isn't worth it. Although, doing the math at a higher bit count is worth it. The GM ECMs using 16bits when doing 8bit * 8bit multiplies because the result is 16bits. It also does 8bit * 16bit multiplies and keeps what it thinks is the most important 16bits. Just because the base values in a table are 8bits doesn't mean everything is done with 8bits.

Good project. I hope it works well for you. I like JY turbo setups. Using two of those Ford units should make some good 1000+ HP from the info I have seen about them and spool very well. With the proper control it would be one of the best street/race configurations that I know of.

 

I followed a post with pics at Turbomustangs and a guy put one on a 5.0.He just JB welded it up and said it worked great in fact so good he lost head gaskets.
Does Ford use variable vane to control boost or make for faster spool?

 

faster spool, and so it dosent over spin the turbo

 

I meant one that was used with operational variable vanes.

Yeah, faster spool. (In the most basic form) it's essentially making a variable 'size' turbo. 'Smaller' at first when you need spool, transitioning to 'larger' to get top end power.



Good info above.