91 T3 SBEC1 Tuning Questions

[jckrieger]

The day has come to finally begin tuning my 91 R/T. I'm replacing the head gasket and the car already has a nice front mount intercooler. I have a few questions that I figure you guys can answer. For background, I used to flash eproms for my 87 CSX and SMEC cars, but I have never touched a SBEC car or used a flashable module.

1. I have (2) 91 T1 SBEC modules. One is a MP module and the other is stock. Will these computers run a T3 or are they missing parts?
2. Is there a good guide on building a flashable SBEC1?
3. What parts are needed to build a flashable SBEC1? Assume I have nothing other than a Windows based laptop with no serial port.
4. Is the current Turbonator T3 SBEC1 cal streetable? I'm looking to run +40's with a 3 bar map at 14psi of boost. Yes, seriously, 14psi. I'll add alcohol, a larger turbo, and more boost when it's time to replace the clutch (at 90K miles now).

[ShelGame]

The day has come to finally begin tuning my 91 R/T. I'm replacing the head gasket and the car already has a nice front mount intercooler. I have a few questions that I figure you guys can answer. For background, I used to flash eproms for my 87 CSX and SMEC cars, but I have never touched a SBEC car or used a flashable module.

1. I have (2) 91 T1 SBEC modules. One is a MP module and the other is stock. Will these computers run a T3 or are they missing parts?

Nope. The T1 computers only have 1 coil driver. The TIII SBEC's have 2 for the duel coil pack setup.

2. Is there a good guide on building a flashable SBEC1?

I have instructions for installing mt module, but thtere's no complete DIY method that I'm aware of.

3. What parts are needed to build a flashable SBEC1? Assume I have nothing other than a Windows based laptop with no serial port.

To use my module, you just need to socket the SBEC (with a 32-pin socket); and get one of my cables. The SBEC needs a protection circuit for bootstrapping becuase the boot signal is MUX'd on the Rx line. So, you need to protect the USB side from the 12v boot signal. You can DIY it if you're handy, the circuit isn't complicated.

4. Is the current Turbonator T3 SBEC1 cal streetable? I'm looking to run +40's with a 3 bar map at 14psi of boost. Yes, seriously, 14psi. I'll add alcohol, a larger turbo, and more boost when it's time to replace the clutch (at 90K miles now).

I would say it is. Ask Eski, he's run it more than anyone else. He's running the SBECII version, but cal-wise the 2 are identical. I do need to update the posted SBECI cal to the latest specs. It's old and probably not as good as Eski's cal.

[jckrieger]

It sounds like I need to find a good 91 R/T SBEC before I can get started. I really don't want to risk damaging my only 91 R/T SBEC. After that I have no problem purchasing your flashable module and cable interface. I look forward to running the Turbonator cal.

[ShelGame]

It sounds like I need to find a good 91 R/T SBEC before I can get started. I really don't want to risk damaging my only 91 R/T SBEC. After that I have no problem purchasing your flashable module and cable interface. I look forward to running the Turbonator cal.

I actually have a 91 R/T computer core...

[jckrieger]

You just moved to the top of my holiday list!

I actually have a 91 R/T computer core...

[shackwrrr]

I daily drove my spirit for about a year on robs code. He was still working on bugs here and there while I was driving it. I just updated my cal every time he fixed something. I think everything is worked out by now. I used an ostrich on mine but I wouldn't recommend running one on a SBEC. you need to provide usb power all the time and if you don't sometimes it would corrupt the cal. Also it would kill the battery after a couple days. When I get my shadow on the road (with the old spirits TIII drivetrain in it) I will probably go with one of Robs flashable setups after all the problems with the ostrich. Even on my SMEC now the ostrich causes resets and hiccups now and then.

You can't beat the price of his setup over the ostrich.

[jckrieger]

I'm going to install the flashable T3 computer I received from Rob this weekend. The plan is to keep the stock MAP + injectors, then log some data and determine where I need to add/subract fuel/timing. I've noticed I get some wicked knock on 89 octane around 3,000 rpm at 1/2-3/4 throttle. I just looked at the 3D map for part throttle timing and I see there's a huge timing spike at that condition. Having an editable cal will allow me to make a safe low octane calibration for winter driving.

The plan is to have an 89/93 octane cal. Low boost for 89 octane, high boost for 93. It'd be nice to have low and high boost timing maps, but honestly I'm not sure if I would ever fully take advantage of it.

Now I just need to learn how to use MP Tune. Once I learn how to scale the tables for injectors and the map sensor I'll be in good shape. Of course, I could just start with the Turbonator cal that is already scaled...

[ShelGame]

I'm going to install the flashable T3 computer I received from Rob this weekend. The plan is to keep the stock MAP + injectors, then log some data and determine where I need to add/subract fuel/timing. I've noticed I get some wicked knock on 89 octane around 3,000 rpm at 1/2-3/4 throttle. I just looked at the 3D map for part throttle timing and I see there's a huge timing spike at that condition. Having an editable cal will allow me to make a safe low octane calibration for winter driving.

The plan is to have an 89/93 octane cal. Low boost for 89 octane, high boost for 93. It'd be nice to have low and high boost timing maps, but honestly I'm not sure if I would ever fully take advantage of it.

Now I just need to learn how to use MP Tune. Once I learn how to scale the tables for injectors and the map sensor I'll be in good shape. Of course, I could just start with the Turbonator cal that is already scaled...

The cal I shipped it with is my 3-bar +40 cal that has been tuned quite a bit.

Scaling the 3D cals is not simple in MP Tune. There are still a ton of manual adjustments to make just to get it to run right as a 3-bar cal. And, MP Tune won't do them for you. I had to do it in a text editor. I don't recommend doing it yourself. When you get ready to go 3-bar, start with my cal.

[jckrieger]

The cal I shipped it with is my 3-bar +40 cal that has been tuned quite a bit.

Scaling the 3D cals is not simple in MP Tune. There are still a ton of manual adjustments to make just to get it to run right as a 3-bar cal. And, MP Tune won't do them for you. I had to do it in a text editor. I don't recommend doing it yourself. When you get ready to go 3-bar, start with my cal.

That answers my questions! I know I need larger injectors if I am going to go much over 11psi, so I'll just install the map and injectors at the same time and run with your 3 bar cal. I'll have to tweak the boost tables and setup the switchable boost, as I'm not ready to break 14psi just yet. I really appreciate the work you have done for the T3 cars. This really makes me want to put a T3 motor in the GLHS!

[jckrieger]

Ok, I'll update this thread with some new questions now that I have the Boostbutton flashable SBEC and SCI cable up and running in my R/T.

First, in the fuel and/or timing 3D tables, what value does the computer use when you are outside of the map range (RPM or MAP)? Does the computer try to extrapolate linearly, or does it simply use the latest value in the table before it went out of range?

Second, why in the world do the injectors go to 100% duty at the very highest RPM and engine speed? Is there a physical limit to how quickly the injector can close and re-open, so anything above 70% might as well be 100% since the injector can't respond that quickly? I'm planning to compare the 2 and 3 bar maps to see how everything was scaled.

Third, what is going on with the timing map? It looks like the T3 cal was designed to limit the torque at the highest RPM and highest boost points. Why in the world would you pull timing after 4600 rpm at 11psi but not at 6psi? It seems like this was to nerf the WOT torque at full boost. I ran the car around today with the stock timing/fuel at 13psi and there was absolutely no knock under any WOT condition. I plan to implement "SMEC Style" timing maps in order to help improve high speed engine torque.

I took a course on internal combustion physics and on paper, you can run more timing at higher engine speeds because there is less time for the "activation energy" to trigger a detonation event. This is why my friend can run 87 octane on his Hayabusa motorcycle with 12:1 compression... he's not WOT until he's over 7,000 rpm! I find it very strange that the T3 cal adds 2 degrees of timing from 3,000 to 4,000 rpm, but then subtracts 2 degrees of timing from 4,000 rpm to 5,000 rpm. I really feel this timing map is in no way setup for "mean best torque", but rather for some sort of artificial torque management or emissions control. What have you guys discovered? This is an area where I will be paying close attention, as the factory timing map can be quite... sub-par when raw performance is concerned.

Fourth, what the heck is up with the massive timing spike at 5.5 psi? This is the point where I get some audible knock with 89 octane, but could I possibly add over 10 degrees of timing to this cal without knock on 93 octane at the higher boost levels? It seems like the timing spike at the lower boost level helps illustrate just how much cylinder pressure margin there is in the cal...

[ShelGame]

Ok, I'll update this thread with some new questions now that I have the Boostbutton flashable SBEC and SCI cable up and running in my R/T.

First, in the fuel and/or timing 3D tables, what value does the computer use when you are outside of the map range (RPM or MAP)? Does the computer try to extrapolate linearly, or does it simply use the latest value in the table before it went out of range?

Uses the last value in the table.

Second, why in the world do the injectors go to 100% duty at the very highest RPM and engine speed? Is there a physical limit to how quickly the injector can close and re-open, so anything above 70% might as well be 100% since the injector can't respond that quickly? I'm planning to compare the 2 and 3 bar maps to see how everything was scaled.

They did it to save the headgasket. Basically using fuel to cool the cylinder and keep power and cylinder pressures down.

Third, what is going on with the timing map? It looks like the T3 cal was designed to limit the torque at the highest RPM and highest boost points. Why in the world would you pull timing after 4600 rpm at 11psi but not at 6psi? It seems like this was to nerf the WOT torque at full boost. I ran the car around today with the stock timing/fuel at 13psi and there was absolutely no knock under any WOT condition. I plan to implement "SMEC Style" timing maps in order to help improve high speed engine torque.

That's exactly right - see above comments. When I tuned the cal for eski, I actually took a little bit of timing out (just a little) at higher RPM and boost, and it he still got knock. I actually had to take out more than I thought I would to keep it from knocking. It will be interesting to see what you find.

Fourth, what the heck is up with the massive timing spike at 5.5 psi? This is the point where I get some audible knock with 89 octane, but could I possibly add over 10 degrees of timing to this cal without knock on 93 octane at the higher boost levels? It seems like the timing spike at the lower boost level helps illustrate just how much cylinder pressure margin there is in the cal...

No idea what that spike is all about.

[jckrieger]

Rob,

Did you ever get an answer to how much boost could be ran with the stock injectors? I can spend a little extra time with the stock injectors if you'd be interested to see what can be done at 13-14psi.

[ShelGame]

Rob,

Did you ever get an answer to how much boost could be ran with the stock injectors? I can spend a little extra time with the stock injectors if you'd be interested to see what can be done at 13-14psi.

Nope, I haven't done a 2-bar cal yet ofr the TIII. Only the 3-bar +40. I think there's 4-5 guys running a version of it now...

[jckrieger]

I added some timing to make the WOT timing map more logical compared to other SMEC calibrations in the past. The car still has peak timing at 5psi of boost, but it no longer drops to the floor at higher boost levels. I'm also adding a small amount of advance with increasing engine speed instead of pulling timing with increased engine speed. I will run this timing at 13 psi and check for knock. If I get knock, then I will switch to low boost (11 psi) and try again.

---------- Post added at 10:47 PM ---------- Previous post was at 10:43 PM ----------

One note I'd like to make is it appears the .lst file may not have exported the correct ram locations for MPScan to use. When I record data, all of the gages display out of range. The software will log samples when connected to the ECU, and provides a connection error when the cable is not connected to the ECU. MPScan is also setup to use the correct COM port and SBEC configuration. The ram locations look like they're importing correctly to MPScan, but MPScan just isn't pulling in the data. I will try to investigate this more tomorrow.

To be more clear about what MPScan is logging, I exported the data to a .csv file. It appears the program is reading 255 for each of the one byte ram locations. All of the values stored are maxed out. Any ideas?

[ShelGame]

I added some timing to make the WOT timing map more logical compared to other SMEC calibrations in the past. The car still has peak timing at 5psi of boost, but it no longer drops to the floor at higher boost levels. I'm also adding a small amount of advance with increasing engine speed instead of pulling timing with increased engine speed. I will run this timing at 13 psi and check for knock. If I get knock, then I will switch to low boost (11 psi) and try again.

---------- Post added at 10:47 PM ---------- Previous post was at 10:43 PM ----------

One note I'd like to make is it appears the .lst file may not have exported the correct ram locations for MPScan to use. When I record data, all of the gages display out of range. The software will log samples when connected to the ECU, and provides a connection error when the cable is not connected to the ECU. MPScan is also setup to use the correct COM port and SBEC configuration. The ram locations look like they're importing correctly to MPScan, but MPScan just isn't pulling in the data. I will try to investigate this more tomorrow.

To be more clear about what MPScan is logging, I exported the data to a .csv file. It appears the program is reading 255 for each of the one byte ram locations. All of the values stored are maxed out. Any ideas?

Don't use the SBEC configuration in MP Scan - use the SMEC T1 configuration. They should be the same, but the SBEC COnfig in MP Scan doesn't work right. The T1 SMEC and SBEC use the same baud rates and protocol, the SBEC just has some 'extra' tests that the SMEC doesn't have.

[wowzer]

OH! i have the sbec 1 setup differently then the smec t1! don't know where i got that info but its been like that for years!! i'll make the sbec 1 the same as the smec t1 then, correct rob?

[ShelGame]

OH! i have the sbec 1 setup differently then the smec t1! don't know where i got that info but its been like that for years!! i'll make the sbec 1 the same as the smec t1 then, correct rob?

Yes, they use the same baud rates, and overall protocol. The SBEC has an 'extra' logging mode that the SMEC doesn't have, but we haven't used it yet. It could be useful, but I think we can ignore it for now.

[jckrieger]

Thanks for clearing this up! I need to extend the SCI connector into the passenger cabin, then I can start logging on my drives to work.

I ran my aggressive timing cal today at 13 psi and 11 psi with some interesting results. At 11 psi I added 7 degrees of timing above 5,000 rpm and had no knock at all. At 13psi I added the same amount of timing and sometimes had some knock that started at 5,000 rpm on 3rd gear pulls and no knock on other pulls. I'm going to pull timing back a couple degrees and can post my map if there is any interest. I'm pretty sure this bottom end can take reasonable timing at 13 psi if my old 2.5L bottom end held 30 psi without any trouble.

Once I settle on a map, I will do some back-to-back 3rd gear pulls with my G-tech GPS based performance meter and report any gains. If the performance difference from the timing isn't measurable, then I'll probably put it back where it was.

[jckrieger]

Ok, I'll provide an update here. Rob, your suggestion to use the SMEC T1 configuration fixed my logging problems. You are the man!

I do have a Turbonator question regarding the anti-lag feature and switchable boost. I have high boost set at 13 psi with low boost at 11 psi. I also have the anti-lag setup as default with 4 degrees of retard. Is there any chance that the switchable boost messes with the anti-lag and causes the low boost setting to always pull timing? There is a very noticeable difference in torque between 11 psi and 13 psi!

I am going to reset my logging channels and run around with the car tonight. This will allow me to answer my own question. What are the limits for logging with the SBEC? I think I read about 10 ms is about what the ECU is capable of. If that's the case, then could I have 10 channels with a 100 ms polling frequency? What are the risks with adding more channels? Will this cause the ECU to become unstable or will the data rates just decrease?

-Clark

[wowzer]

mpscan will try to log as fast as it can no matter what you set the polling frequency to. the screen and sample frequencies only determine how often to update the screen display and how many samples per second to log/save. after you have gotten a log if you post the log i can get an idea of how fast it is really running and you can then adjust your sample rate to elimnate duplicate samples.

---------- Post added at 05:31 PM ---------- Previous post was at 05:27 PM ----------

also, i believe the last version allows you to set the priority/frequency of each gauge, so if you have certain gauges that should be updated more often then you can set its specific sampling frequency to a higher (closer to 1 on a scale of 1 to 10) sampling rate.