5.8 Speed Density(SD) with VGT Turbo

wwhite

New member
Going to turbo my engine, keeping SD.
Bought a 72.5mm VGT turbo, 5 years old, 100,000km, price was right.

Looking for information with how others have made SD work with boost,
What parameters had to change in the tune.
changes to the MAP sensor.
One way check valve on MAP sensor vacuum line.
etc...
What worked for you?


If your going to say "SD won't work" or "convert to MAF", please do not reply.
I'm not converting to MAF, and I will make SD work.
 

Raymond_B

Administrator
I Support the NLOC
Going to turbo my engine, keeping SD.
Bought a 72.5mm VGT turbo, 5 years old, 100,000km, price was right.

Looking for information with how others have made SD work with boost,
What parameters had to change in the tune.
changes to the MAP sensor.
One way check valve on MAP sensor vacuum line.
etc...
What worked for you?


If your going to say "SD won't work" or "convert to MAF", please do not reply.
I'm not converting to MAF, and I will make SD work.
I used to say the same thing many years ago when I started Performance Specialties and tuning :)

Anyway, if the engine has a stock or SD friendly cam you can make it work. Most supercharged folks just work off the WOT fuel multiplier and injector size. However with a turbo you really won't make boost in that linear of a fashion so you'll need to watch that.

What you'll also run in to is the stock MAP pegs under almost any sort of boost, and to make matters worse it is a frequency based MAP sensor so you cannot just grab an easy replacement. A loooong while back some of the diesel guys messed around with analog to digital converters to try, in essence, to make a MAFia setup where you could alter the signal going in to the EEC. You could also try to source a Ford turbo MAP sensor, but without knowing the output of KPa vs frequency you won't get much use. I remember digging through the code (even with Jerry W.'s help) hoping I'd find some sort of easy transfer function similar to a MAF, but never did as I believe the MAP input goes in to a series of calculations based on MAP, engine size, air temp, etc., etc. so it is not very straightforward to tune.

So yeah I'm not saying go MAF, that's only an incremental step and it is sort of a fun intellectual exercise to try to hammer the square SD system in to a round performance hole it get's old after a while. Aftermarket EFI is where I have evolved to now some 27 years later :)
 

wwhite

New member
I used to say the same thing many years ago when I started Performance Specialties and tuning :)

Anyway, if the engine has a stock or SD friendly cam you can make it work. Most supercharged folks just work off the WOT fuel multiplier and injector size. However with a turbo you really won't make boost in that linear of a fashion so you'll need to watch that.
Cam is on the edge of not being SD friendly, but with a tune it is great. Idle needs more work.

What you'll also run in to is the stock MAP pegs under almost any sort of boost, and to make matters worse it is a frequency based MAP sensor so you cannot just grab an easy replacement. A loooong while back some of the diesel guys messed around with analog to digital converters to try, in essence, to make a MAFia setup where you could alter the signal going in to the EEC. You could also try to source a Ford turbo MAP sensor, but without knowing the output of KPa vs frequency you won't get much use. I remember digging through the code (even with Jerry W.'s help) hoping I'd find some sort of easy transfer function similar to a MAF, but never did as I believe the MAP input goes in to a series of calculations based on MAP, engine size, air temp, etc., etc. so it is not very straightforward to tune.
The only function built in is if the MAP sensor fails, then uses a transfer function based off Throttle Position(TP). If TP has also failed, a fixed value constant is set to 29.875. Stock sensor is in the 89 to 162 Hz range. I'll do some more research on this, kinda remember something about using a GM map sensor.

So yeah I'm not saying go MAF, that's only an incremental step and it is sort of a fun intellectual exercise to try to hammer the square SD system in to a round performance hole it get's old after a while. Aftermarket EFI is where I have evolved to now some 27 years later :)
I've taken this route because I want to understand everything about EFI and tuning.
My square SD now has rounded corners, and is more of a performance ellipse rather than circle, getting closer.

A few more years of playing with the stock stuff for me, then I'll go aftermarket.

Totally appreciate your response, THANK YOU!
 

wwhite

New member
Run open loop tune and get a good logging wideband. Then you can slowly and painstakingly alter the VE table to get where you need.
I have a AEM (30-4110) UEGO Air/Fuel Ratio Gauge.
Basically continue tuning as normal.

Keep it in open loop, never go back to closed loop?
Or after open loop tuned, start tuning in closed loop?
 

FMOS Racing

PSI=:-D
GM MAP sensors are voltage. There's, sadly, nothing compatible with the Ford freq-based MAPs.

Like Ray said, turbos are a different animal than a centrifugal supercharger. With a turbo, you can make full boost at part throttle and lower RPMs. This makes tuning tricky when the same conditions may not make boost at all under different circumstances.

If you're not dead set on optimizing everything, an FMU and a boost retard will get you something drivable and survivable into the 500-600 hp range.
 

slow94

Active member
I have a AEM (30-4110) UEGO Air/Fuel Ratio Gauge.
Basically continue tuning as normal.

Keep it in open loop, never go back to closed loop?
Or after open loop tuned, start tuning in closed loop?
Is this a logging wideband? You will need to log A/F and rpm at least. But I suggest locking it in open Loop. That way you can add fuel in low cell areas where turbo will make boost. Also will help with the cam. Aside from that the wot fuel multiplier will be your friend. If you do keep it closed loop you will need to set the tp for open loop very low so you don't go lean with the turbo.

Sent from my moto z3 using Tapatalk
 

wwhite

New member
GM MAP sensors are voltage. There's, sadly, nothing compatible with the Ford freq-based MAPs.

If you're not dead set on optimizing everything, an FMU and a boost retard will get you something drivable and survivable into the 500-600 hp range.
Since the GM 3 bar MAP sensor are voltage, I can disable EGR, and wire the voltage MAP to the EGR input. The EGR input is a 0-5V analog input. I'm actually working on adding two(2) more analog inputs to the EEC-IV(off topic).
I'm not too worried about the little nitty gritty right now, turbo is not installed yet.

With a turbo, you can make full boost at part throttle and lower RPMs. This makes tuning tricky when the same conditions may not make boost at all under different circumstances.
Never thought about that, great information, thank you.
 

wwhite

New member
Is this a logging wideband? You will need to log A/F and rpm at least. But I suggest locking it in open Loop. That way you can add fuel in low cell areas where turbo will make boost. Also will help with the cam. Aside from that the wot fuel multiplier will be your friend. If you do keep it closed loop you will need to set the tp for open loop very low so you don't go lean with the turbo.
Yes, Wideband, currently have it logging via EGR, with EGR disabled, logging what I want with QuarterHorse. Soon to be adding another analog input so I can keep EGR and log boost.
Going to research different MAP sensors, 1995

OK, think I have a plan, to at least get it running.
- data logging (done)
- locked in open loop (done)
- WOT multiplier(research)
- GM 3 bar map (research), 1995 Ford Escort Cosworth Turbo SD frequency MAP research
- enable boost logging(research)

Finally, or Firstly, get the turbo installed!

Thanks for your help!
 

Raymond_B

Administrator
I Support the NLOC
Since the GM 3 bar MAP sensor are voltage, I can disable EGR, and wire the voltage MAP to the EGR input. The EGR input is a 0-5V analog input. I'm actually working on adding two(2) more analog inputs to the EEC-IV(off topic).

That might work for datalogging, but doesn't get around the fact that the computer has no idea what to do with those extra analog inputs. Like I mentioned, some folks tried to make analog (0-5v) to digital (frequency) converters. Since you know the frequency range the software will accept you could (in theory) use a GM MAP sensor and then try to convert that analog signal to a frequency output in a range the computer will accept. **BUT** it comes with issues mostly you will be decreasing the resolution of the sensor because you cannot change what the computer expects.
 

wwhite

New member
because you cannot change what the computer expects.
I can and have changed what the computer expects. Heavily hacked and modified SD48 hardware, removed and added chips. As well as added and modified the software using hex/binary modifications. Little off topic, as an example, I have taken a Thunderbird TripMinder, and wired it to pin 34. The traces on the SD48 board for pin #34 basically went no where, go to a couple of blank spots where a diode and a chip sit. Copying the MA46G hardware, figured out that chip 71001FB takes a digital signal and converts it to pulse width modulation(PWM) 0-12v that the TripMinder uses. Now my truck has a clock that includes mpg averages, etc. data coming directly from EEC-IV.
Adding another input shouldn't be too bad.
The EEC-IV can have up to 12 A-D converters, bone stock mine has 10 that are used. That leaves me with 2 more A-D inputs for me to do what I want. Currently, one of those is for Wideband logging, and by the looks of things, I'll probably find a MAP that outputs 0-5v rather than freq, start with logging that data, analyze and research with how to use that data to be part of the controller.
 

Raymond_B

Administrator
I Support the NLOC
I can and have changed what the computer expects. Heavily hacked and modified SD48 hardware, removed and added chips. As well as added and modified the software using hex/binary modifications. Little off topic, as an example, I have taken a Thunderbird TripMinder, and wired it to pin 34. The traces on the SD48 board for pin #34 basically went no where, go to a couple of blank spots where a diode and a chip sit. Copying the MA46G hardware, figured out that chip 71001FB takes a digital signal and converts it to pulse width modulation(PWM) 0-12v that the TripMinder uses. Now my truck has a clock that includes mpg averages, etc. data coming directly from EEC-IV.
Adding another input shouldn't be too bad.
The EEC-IV can have up to 12 A-D converters, bone stock mine has 10 that are used. That leaves me with 2 more A-D inputs for me to do what I want. Currently, one of those is for Wideband logging, and by the looks of things, I'll probably find a MAP that outputs 0-5v rather than freq, start with logging that data, analyze and research with how to use that data to be part of the controller.

That's awesome, we needed you 20 years ago!
 

wwhite

New member
That's awesome, we needed you 20 years ago!
Couldn't afford a new(er) truck back then, didn't think SD was tunable, heck, didn't even know what SD or MAF stood for. Also don't think I had any resources like internet to find and research this stuff.
Only had(still have) a carbureted 1964 Mercury short box.
 

slow94

Active member
Sounds like you certainly have the software know how. Would be cool to see a complete new way to tune these computers. I bet you get along great with Decipha over on efidynotuning.

Sent from my moto z3 using Tapatalk
 

wwhite

New member
Sounds like you certainly have the software know how. Would be cool to see a complete new way to tune these computers. I bet you get along great with Decipha over on efidynotuning.
Yes, I respect Decipha's knowledge and opinions very much, that doesn't mean I always agree with everything either. If I had less knowledge and was using Decipha's tunes and/or custom tune, I'd probably be asking him all my questions. He was the one that I asked if this turbo would work, and he said yes. He would probably recommend I switch to EEC-V for tuning, but if I'm spending that much time switch computers, I'd probably go aftermarket like what @Raymond_B had already suggested.
 

slow94

Active member
Yes, I respect Decipha's knowledge and opinions very much, that doesn't mean I always agree with everything either. If I had less knowledge and was using Decipha's tunes and/or custom tune, I'd probably be asking him all my questions. He was the one that I asked if this turbo would work, and he said yes. He would probably recommend I switch to EEC-V for tuning, but if I'm spending that much time switch computers, I'd probably go aftermarket like what @Raymond_B had already suggested.
Yeah if you start making good power you will reach the processing limits of the ECU pretty quickly. In my opinion it's around the 600hp mark. I know when I started making real power on stock SD I had more problems getting the truck to shift on time. It simply couldn't process the info fast enough to pull off consistent shifts. Also I would have erratic A/F readings on same tune and boost.

Sent from my moto z3 using Tapatalk
 

Raymond_B

Administrator
I Support the NLOC
It’s funny I vividly remember Jerry (GJLS) calling me up in the early 2000’s saying he was having issues with his EEC-IV keeping up with the HP he was making. He ended up trailering down the black truck and just free reving it the A/F would go crazy and the EEC would reset. That’s when I started to realize beating old code to death was not where it was at. Jerry ended up going FAST(I think) and proceeded to run 8’s :)
 

FMOS Racing

PSI=:-D
That's pretty much my thoughts at well, I ran pretty decently into ~600 hp range. After that, bad things started happening, and that was with a MAF.

Low 12s/high 11s type of range, you can make an EEC-IV live well. Beyond that (boost only), things begin to get a little iffy.

Our engine computers were designed in the same era as the 80386 computer. A modern CPU is several thousand times faster without even thinking about the cache and improved instruction sets, just on core count and frequency alone.
 

slow94

Active member
I was running like 7.0's in the 1/8th mile which is dang near close to 10 second 1/4 range but it was so inconsistent. It was all just basically guess work. Now in the 400 hp range all was good.

Sent from my moto z3 using Tapatalk
 
Top