But, I wouldn't be so quick to do it. My axle seal is leaking at that particular tripod. It might require a bearing/seal assembly, or perhaps I just plain have a leaky seal. Although it works just fine, this issue did come up. See a few posts above for that.
Great news today! I bypassed all conditionals for Quasi-Asynchronous mode, affectively forcing it to run in that mode. Heres a little clip of what the difference sounds like:
you can't hear any change in the quality of the idle, which is great. It was running pig rich in this video, otherwise it would have shut down when transitioning to normal firing mode.
Then, I just went ahead and pulled back the fuel, and added in a little bit of timing.
Now, I just need to figure out which conditional is failing to allow quasi asynchronous pulse, and fix that. I think it's the road speed to disallow QAP, which is set to 255 mph (at L86C2). That may be set to disable QAP entirely, where it APPEARS like it is set to be active all the time.
Either way, It looks like I can probably restore fuel pressure back to stock, as this tuning roadblock is lifted. Now, It will even idle in closed loop, but the corrections are still far to aggressive, I need to figure out which "Idle Proportional gain count" or whatever it is that I can dial that in with. It's making the AFR's go all over!
It's been a little while since an update. Tuning, tuning, tuning.... That's what my life revolves around nowadays.
However, having done a TON of work on the program, I am pleased to say that I've got the idle sorted out. It wasn't easy, and I burnt a half a tank, JUST IDLING! The bottom line is this: Today, the state of Nevada recognized that a Turbocharged 3.4 DOHC with 60lb injectors easily passed emissions without any hiccups.
Our test is three parts: A 90 second high speed emissions check at 2500 RPM's in neutral An idle emissions test. Visual inspection of all emissions components.
Since I have all the factory emissions equipment installed and functional. The visual was an easy pass, but they did want to see the cars emissions label, which I could not provide,"Oops, sorry, no the label was lost a long time ago" Even though it was simply taped over. They did not know or figure out that it is an engine swap.
This all makes me feel proud of what I have done. I'll get to the chronicle of tuning, and where I stand so far later on.
Like anything that gets tuned, you start with a base file. Mine is from a 90 Grand Prix McLaren Turbo, It is an $8F mask, ARZC code running in a 1227730 ECM, which is the in-car version of the 1227727 that was used in the 89-90 Turbo Grand Prix's. This ECM has everything I need to do this whole project. Conveniently, it is a very common, and cheap ECM.
I also know that there are three other tunes that contain valuable data for guiding the tuning of this engine.
$DF, BCFA which is the 91-93 Grand prix's with the 3.4 DOHC, and a manual transmission, uses a 16196396 ECM, which is similar to the 1227727, and 1227730, but it has some extra ram. $6D, ATWX which is for a 89 Cavalier 2.8 with a manual transmission, Used in the 1227730. $6D, ARYH which is also for a 89 Cavalier 2.8 with an Automatic transmission, Also used in the 1227730.
The two $6D codes were especially handy for helping show me what GM changed in an otherwise identical car, to make them compatible with a manual transmission. $6D is nearly the same code as $8F, minus the boost stuff, so this little treasure of information was very helpful!
I have a hack for $8F and created most of one for $6D, so it wasn't too difficult to compare the differences to figure out what might be changed for a manual transmission. I have only implemented a couple of the changes to my current code.
The very first thing I did to the $8F AZRC code was change the table called "Base Pulse Constant vs Desired EGR", which were all set to 105, down to 42, which reflects the change in injector size from 22# to 60#, as well as the change in displacement from the 3.1L Turbo Grand Prix, to the 3.4 DOHC, This allowed the engine to actually run. Otherwise, it would have been FAR too rich. It still injected FAR too much fuel to start, so I could only start it with the throttle pressed to the floor, which shuts off all fuel during cranking. (clear flood mode)
I decided to decrease the cranking fuel to 40% of what it was, since that is roughly what I did to the "Base Pulse Constant vs Desired EGR". The table that I used for this was "Crank fuel PW vs Coolant Temperature" I just grabbed all the values and multiplied them by 0.4. Now the engine starts right up.
The next change I found I needed to make was to disable closed loop. It would surge and stall when Closed loop would start to make fueling corrections. This was simple. I changed the "Temperature threshold for C/L determination" from 80.15F to 304.25F, effectively disabling it. This made it run better, but not good enough to drive yet. I also decided that it would be a good time to disable Charcoal Canister Purge "Disable CCP if coolant < this" set to 304.25, and the EGR "If coolant <= this Disable EGR" also to 304.25. I'll deal with those later.
I was having issues where I would hit the throttle and it would stall every time, with the WBo2 reading pig rich when I'd open the throttle. I went "Async Factor Vs Coolant" (under Fuel AE) and multiply them by 0.4. This allowed the throttle to be opened without any bogging. At this point, the car suddenly became drivable. And so I did some breaking in, and had some fun with it.
I have a bit of a dislike for the way That $8F (and $6D, $DF) calculates it's Volumetric Efficiency. It has 3 tables. One is a "Base VE table", which just looks up an RPM and coughs out a simple VE value. Then, it goes to either the "Idle VE table" or the [Open throttle] "VE Table" where it looks up a second number based on MAP and RPM, then adds it to the Base VE table value for the final VE number.
Seems silly to me.
Why not just add the base VE table numbers to all the values on the Idle VE table AND the [open throttle] VE table, and skip having a Base VE table altogether? So that's what I did. Except, I made a new VE table that goes up to 6400 RPM, stuck it in an open spot on the chip, and re-directed the call to it. As well as opened up the RPM lookup cap from 4000 RPM to 6375 RPM so that it could see the whole table. This makes tuning much more straight forward now.
With that sorted, I Imported the main VE table, and the Idle VE table from $DF BCFA, and made the necessary changes so that it would fit the new tables. Then I made a new "Main Spark Table" which was based on BCFA's Main Spark Table in the Naturally aspirated section, and blended in the boosted timing table from $8F AZRC.
This gave me a good basic tune to start playing with.
This seems like a lot, but it is where the tuning STARTS.
My car's inspection was due Dec 7. So that set the deadline to get it running well.
The first thing I tried to do, (and failed at) was to get the engine to idle in open loop, at 14.7:1 on the wideband. All I had to do was decrease the Idle VE table till it would hit 14.7:1 Only, it would die anytime I got near 13.5:1 because the injectors couldn't inject a small enough pulse before it wouldn't inject anything at all. In the end, the adjustable fuel pressure regulator wasn't even enough to get the fuel trimmed back. What I found, was that the engine wasn't running in Quasi Asynchronous Pulse mode, so it was firing the injectors once per engine revolution. I modified the program a little to ignore the MPH disable parameter for Quasi Async mode. This change caused the injectors to fire every other revolution, effectively allowing me to trim the fuel back further. (see video above) Suddenly I was able to adjust the Idle VE table down till I was idling at 14.7:1
With that issue settled, I tried enabling closed loop. "Temperature threshold for C/L determination" from 304.25F to 80.15F. The engine started surging with the closed loop corrections. But I noticed that the INTegrator was also fluctuating heavily with the surges. So clearly the INT was changing too fast, causing the corrections to be too large. I went to this table: "Integrater Update Delay Vs CLFLOW" and multiplied all the values by 3.0. This slowed the INT corrections down enough to keep the magnitude of the corrections reasonable.
One thing that I decided to do to help idle stability, was to import the idle RPM tables from $DF BCFA. This made a nice difference.
I realized that this engine REALLY doesn't like when the cooling fan kicks on. It would stall almost immediately. This one had me scratching my head for a little while. I combed over the logs carefully and noticed that the IACV jumps open a few counts just before the fan is actually commanded to turn on. I found a couple of parameters: "Fan 1 Turn on delay" (0.3 second) AND "Num of IAC steps for Fan 1" (which was 10 steps). So I increased the fan delay from .3 seconds to 6 seconds, to help me see how the IAC steps affected the idle. So I saw the IAC bump open 10 steps. which caused a slight idle surge, then the fan kicked on and the engine died anyways. I bumped the IAC steps up from 10 to 15 and tested it again. It had a slightly larger idle surge, and I set the Fan 1 turn on delay to expire exactly when the idle surge peaked, which was 1.6 seconds. This solved my stalling issue.
This is where I was this morning. It ran well enough to pass emissions. I'm far from done, but getting a car to idle correctly certainly makes me happy, and saves a lot of gas.
One other thing I noticed today that may have a surprise effect on tuning, I've been using a Memcal from a 3.1L N/A aluminum head "vin T" engine to do my tuning. I had a factory Turbo Grand Prix Memcal hanging around that I would re-burn when doing my final tune. I knew It's resistor chips are identical between the 3.1L N/A and Turbo. But it turns out, that the ESC filter on the Turbo memcal is different then the N/A. This is VERY important. This means that Knock was probably being incorrectly reported to the ECM. I was seeing unreasonable knock out of the engine constantly under low boost. I have a feeling that putting in the Turbo Memcal is going to make a big difference in how I tune timing.
Been having a ton of fun with the car lately. Lots of tuning to do still, I've been working on the main VE table a bit.
I had a couple of problems pop up. I had a fuel leak near the fuel filter. It turns out that the barb that I was using to go from the nylon tubing to the fitting for the fuel filter wasn't aggressive enough, and the fuel was leaking around it. The barb was from a "Nylon Fuel hose repair kit" that you can get from Autozone. So, I'd recommend against those crappy barbs. I replaced it with a barb that I cut off of a metal quick disconnect end, and welded it to the fuel filter fitting. Problem solved.
The other issue was the intercooler pump. I was using a cheap generic pump that came with my intercooler. It was making some funky noises, and I suspect it was possibly leaking too. So, I bought a Bosch "Cobra" pump and installed that instead. Slight difference in size and quality...
On the upside, I did a bit of driving, a bit of WOT fun, and let me say, It's pretty fast at 10 PSI. Leaves me giggling maniacally when I look at the speedometer and it just blows through 40-80 in just a few seconds, then 80-100 in just a couple more. I've only driven a few cars that accelerate this fast. It's quickly getting to the point where it may be the fastest car I've ever driven. But it's not quite there yet...
You are a true craftsman, I must pay you for your wares!
Also, that bosch pump can be tricky to prime. I run the same pump for my w2a system. The only thing I dont like its the old school GM fuel injector style retaining clip. Otherwise, its a great bit of kit.
[This message has been edited by BV MotorSports (edited 12-27-2013).]
It's not uncommon for people to have trouble with those clips. Many try to remove the retaining clip with a screwdriver. If that's your problem, you are not alone. You need to press the center of the retaining clip in and it will release the connector...
The dipstick is a gift. It wasn't a big deal to make, I just needed to get off my butt and do it.
Priming that pump wasn't too bad. But I had a lot of issues getting the system filled though. I found that the Anti-freeze was too thick to pour into the 1/2" fill neck's hose and effectivly have the air come back up through the fill neck. It just wouldnt go! But when I was putting the distilled water in, it worked just fine. But I figured that out the hard way.
Some testing on a fairly deserted desert road. It's quite a bit of fun!
Sort of. My speedometer needs calibration badly. It's way off. Second gear to 80 on a F40? More like 60-65 probably. But I was taking my sweet time shifting...
I am having one strange issue with tuning. Any time I'm coming off the throttle fairly quick, I'm getting a heafty burst of knock. But under 10 PSI of boost, there is no knock and the timing is ramping back in. So when I shift, I'm starting back in to boost with a large amount of retard that tapers off over time. If I'm in second gear for a while and I just whack the throttle, I'll have no retard issues and it pulls so hard!
Maybe I still have boost in the intake between the throttle blade and the valves when decel enleanment kicks in creating an instant lean condition for a couple of engine rotations? It's really strange! Not sure just yet what's causing that.
Sort of. My speedometer needs calibration badly. It's way off. Second gear to 80 on a F40? More like 60-65 probably. But I was taking my sweet time shifting...
I am having one strange issue with tuning. Any time I'm coming off the throttle fairly quick, I'm getting a heafty burst of knock. But under 10 PSI of boost, there is no knock and the timing is ramping back in. So when I shift, I'm starting back in to boost with a large amount of retard that tapers off over time. If I'm in second gear for a while and I just whack the throttle, I'll have no retard issues and it pulls so hard!
Maybe I still have boost in the intake between the throttle blade and the valves when decel enleanment kicks in creating an instant lean condition for a couple of engine rotations? It's really strange! Not sure just yet what's causing that.
I can hit 70 mph in 2nd gear at 7K with my F40, so it didn't seem too far fetched... all comes down to tire diameter at that point.
You might try turning off decel enleanment to see what happens. Also its quite possible you just have too much timing in the high to moderate vacuum and high RPM areas of the timing table. I know I had to tweak mine some on the LS4 since they never expected shifts to happen at 7K and see those areas of the timing map at with the throttle closing.
I have a $DF mask I modified for boost kicking around here somehwere if you want it. Lifted the spark/fuel control off of $8F but never got around to the wastegate. Was a daily driver for 2 years running 36# injectors from a 3800SC and if I remember right there were very little tuning changes below 100kPa that I had to make.
[This message has been edited by brian89gp (edited 01-13-2014).]
I'd love to look at your $DF code. Curious though, why go through the troubles of adapting DF? I know it's an improvement over the existing 8F, but is it worth the effort? 8F is ok, I haven't really had much trouble trying to get it to do what I want. DF does have the advantage of being manual, and is compatible with my EGR...
I'd love to look at your $DF code. Curious though, why go through the troubles of adapting DF? I know it's an improvement over the existing 8F, but is it worth the effort? 8F is ok, I haven't really had much trouble trying to get it to do what I want. DF does have the advantage of being manual, and is compatible with my EGR...
I'll dig it up this week. One other thing that I just remembered, I ran two MAP sensors, one 1 BAR as a dedicated barometer that was open to the atmosphere and one 2 BAR for manifold pressure. The barometer calculation in 8F I considered inadequate.
Which is more difficult, retuning all of 8F for a different engine? Or modifying DF but leaving the tuning for stock 0-100kPa untouched?
1. It is already tuned for the DOHC, and a turbo doesn't change most of the tuning. You get an instantly drivable car that you only have to tune for boost and tweak in a few other spots 2. The DOHC with a manual trans can be interesting to tune to be hospitible when done from scratch 3. Stock the RPM range on the tables is higher
[This message has been edited by brian89gp (edited 01-22-2014).]
Sounds good! I'd definitely give it a shot, just so long as the code doesn't require the extra SRAM of the 9396. I'll take a peak at BCFA and look at it's memory map.
Looking into using $DF, BCFA based... I found one possible issue in the memory map. L0200. I don't believe the 7730 has access to that address, it is written to but disabled from reading by calibration (manual transmissions only). It's towards the end of the code not sure what it does, all I know is that it appears to only be able to write $00 to it. Since it is always zero, not having any storage hardware at that address may not be an issue at all.
In the meantime I managed to calibrate my Speedometer.
My old Getrag was a 30 tooth reluctor, so I worked around that number: 225/45/17 tires are 25" diameter, so 78.5" circumference. That's 807.13 rotations per mile, times 30 reluctor teeth is 24214 pulses per mile. So 24K PPM. Got it.
So, the first thing I did was set my IP Pulse divisor to 128 ($80)
My notes on pulse divisors... A B C 0 = 0 = 0 0 0 Divide by 1 (4000 ppm VSS) 4 = 128 = 1 0 0 Divide by 6 (24000 ppm VSS) 2 = 64 = 0 1 0 Divide by 7 (28000 ppm VSS) 6 = 192 = 1 1 0 Divide by 8 (32000 ppm VSS) 1 = 32 = 0 0 1 Divide by 9 (36000 ppm VSS) 5 = 160 = 1 0 1 Divide by 10 (40000 ppm VSS) 3 = 96 = 0 1 1 Divide by 11 (44000 ppm VSS) 8 = 224 = 1 1 1 Divisor disabled, no output
Wiring and configuration of the Dakota Digital SGI-5C module
Power and Ground are straight forward. Sensor ground: I wired to the Purple wire from the VSS Input: I wired to the yellow wire from the VSS Out 1: goes straight to Pin B10 on the ECM I deliberately left pin B9 on the ECM OPEN. It is already grounded inside the ECM. ECM pin B11 goes to the pull-up circuit, here:
The calibration switches are all turned OFF.
First I started with the calibration I first thought would be accurate. 6 Coarse, 9 Fine. Which is a calibration factor of .383. By dividing the 30 teeth of the getrag's reluctor, by the 78 teeth of the F40's ring gear/reluctor. I got .384; .383 was the closest setting.
I verified accuracy on a police speed display, which seemingly to be perfectly accurate according to our other vehicles.
So now my speedometer is accurate. Worked like a charm. Now I just need to fix my road speed constant so that the ECM MPH matches the speedometer. That should be pretty straight forward.
I'm still tweaking values here and there in the code. Fueling is pretty much bang on. I'm curbing the issue I'm having with knock while shifting by pulling my timing in the 4800+ RPM column so that it is 17 at boost, and 23 at full vacuum. There was a 35 degree peak in between, but now its just a straight line. I think that 35 degree peak while coming off throttle was causing a knock burst. We'll see!
Thought that $df requires '9396 but can't say I ever researched it in depth.
Can't figure out how to attach files to this forum. If you PM me your email I'll get them sent over. What I have: 1. HTML files with the assembly changes commented and documented 2. the BIN files I used (125 of them in total) with an Excel spreadsheet documenting the tuning change (and/or assembly change) for each one
[This message has been edited by brian89gp (edited 02-19-2014).]
I decided to take the a/c pressure transducer off and use the wiring for a separate a/d channel for an additional MAP sensor. Every place the stock a/d variable for the a/c sensor was called I loaded a dummy variable so I gained the ram address for use elsewhere. Also to do this you must turn off the a/c pressure transducer present option flag.
Stock
code:
C720 @2 ldaA #$00 C722 call LF04E C725 staA L014C C728 ret D067 ldaA L014C D06A cmpA #$D0 E74B ldaB L014C E74E cmpB 226, X F0BB @2 ldaB L014C FOBE cmpB 0, X
Since I didn't have a cutoff in the barometer calculation the first time I hit boost the baro would be pegged at 103kpa and it would stay there until the car was shut off. Since I had the MAP running off of a single 2bar sensor and had an extra channel open I decided to use an independent baro sensor.
The baro is set up on the regular MAP wiring, the MAP sensor is on the a/c pressure transducer wiring. It just worked out that way (I suck at planning)
Stock:
code:
F1E1 ..............bunch of stuff F27C
Modified:
code:
F1E1 ldaA #$30 ;a/d channel F1E3 call LFD5D ;a/d read routine F1E6 staA BARO F1E8 clra F1E9 staA L01F4 F1EC ret F1ED-F27C nop
I originally had 2 map sensors but eventually figured out how to use a single 2bar MAP. Turns out it was on the a/c pressure transducer wiring so that is where it is today. (#00 is a/c pressure transducer, #30 is MAP sensor)
Stock:
code:
9907 ldaA #$30 ;load a/d channel 9909 call LF04E ;a/d read routine 990C staA MAP ;store result 990E staA L0054 ;store result the others are similar, just a different a/d routine is called
Modified:
code:
9907 call FE2E ;jump to routine 990A nop 990B nop 990C staA MAP ;store result 990E staA L0054 ;store result 9CC9 call FE2E 9CCC nop 9CCD nop 9CCE staA MAP C51E call FE3C C521 nop C522 nop C523 staA MAP
FE2E ldaA #$00 ;a/d channel FE30 call LF04E FE33 staA L014C FE36 ldaA L014C ; load 2bar FE39 cmpA #$7E ; see if it is 1 or 2 bar FE3B bcs @+3 ; if <7A then jump (1 bar) FE3D ldaA #$FD ; else load max 1 bar value FE3F ret FE40 ldaA L014C ; if 1 bar FE43 lslA ; multiply by 2 FE44 ret FE45 ldaA #$00 FE47 call LFD5D FE4A staA L014C FE4D ldaA L014C ; load 2bar FE50 cmpA #$7E ; see if it is 1 or 2 bar FE52 bcs @+3 ; if it is 2 bar FE54 ldaA #$FD ; load max 1 bar value FE56 ret FE57 ldaA L014C ; if 1 bar FE5A lslA ; multiply by 2 FE5B ret
The spark table needs to be changed to run on a 2bar scale. The kpa scale is roughly this now: 10,21,33,45,57,69,81,93,105,116,128,140,152,164,176,188,200
There you go. I did all of this back in 2003 so I can't really answer as to why I did things the way I did (switched around the MAP and BARO sensor....) but I know it worked. Drove it daily for a year or two and the drivability was the same pre and post modifications.
The gist is there is a 1 BAR MAP on the standard MAP sensor pinout that is open to atmosphere and acts as an independent barometer and feeds the RAM value directly instead of calculations and tables off of the MAP sensor. The 2 BAR MAP is on the old AC pressure transducer pinout and replaces the stock MAP in the code, the scaled 1 BAR is still fed into the same RAM address and a 2 BAR RAM address is added that is referenced by the BPW table and the main spark table. There is a BPW multiplier table for fuel and the spark table is rescaled to use a 2 BAR MAP. Not much else was really required to get it working.
The car had A/C and I noticed no difference before/after removing the pressure transducer. The BIN's I have are tuned for the use of stock L67 injectors.
[This message has been edited by brian89gp (edited 02-20-2014).]
Cool! Seems much simpler then I would have thought!
couple of notes:
You could liven up some extra analog input pins with a simple change to the accumulator load before the A/D routine. OR using the MUX read routine. The 7730 and similar has a ton of spare analog and digital inputs. I've livened up a slew of them for this project, mostly extra IAT/MAT compatible inputs.
I've never had much luck with $01 NOP's They always seemed to cause the uP to crash, probably the COP watchdog (?) so I usually address my way out of open address slots.
That being said; Thanks so much for your conversion, I will try it out when I have the time.
I had to pack up shop and move across town. So needless to say, life is busy again.
What I did to get my '7730 + F40 to use the stock Fiero speedo is the same digital to analog circuit above but I used the 2000PPM output instead of 4000PPM. Reason being the F40 outputs a 60,000PPM signal and the stock getrag uses 4000. The 7730 by default expects a 24000 signal so a divisor of 6 works great for the getrag. For the F40 you need a divisor of 15 for the Fiero speedo and the 7730 only supports none and 6-11. So using the 2000PPM output is like an extra division of 2 so a divisor of 7 or 8 gets you within 7% error but if your tires are over or undersized you can reduce that error. Ideal divisor is 7.5 using the 2000PPM output but that's not supported. Depending on whether you use 7 or 8 is your margin of error.
Cool! Seems much simpler then I would have thought!
Suprised me too. Taking the easy way out with the independent baro sensor made it a lot easier
quote
Originally posted by Fierobsessed:
couple of notes:
You could liven up some extra analog input pins with a simple change to the accumulator load before the A/D routine. OR using the MUX read routine. The 7730 and similar has a ton of spare analog and digital inputs. I've livened up a slew of them for this project, mostly extra IAT/MAT compatible inputs.
I was thinking that few input pins had the same internal electronics as the a/c pressor sensor and MAP sensor. I did purposfully take the easiest way out though..
quote
Originally posted by Fierobsessed:
I've never had much luck with $01 NOP's They always seemed to cause the uP to crash, probably the COP watchdog (?) so I usually address my way out of open address slots.
That being said; Thanks so much for your conversion, I will try it out when I have the time.
I had to pack up shop and move across town. So needless to say, life is busy again.
I think its time to do something I've never seen done on this forum. A full disclosure of cost. It's something I put together during the entire build. I kept tabs on what I was spending on items that went into the build. Items that wound up not going into the build were left out, and plenty of misc. hardware, hoses and paint are definitely not on the list, though they should be.
So here it goes!
Turbo and accessories: GT3582R Turbo with anti-surge housing $1200 Tial stainless turbine housing, 0.82 A/R $400 Tial MV-S wastegate $263 Tial 50mm "Q" bov $260 Oil feed line 36" Long 4-AN, 90 on one end $32 Oil feed fitting 1/4NPT to 4-AN $5 Oil Feed Oriface restrictor $17 Oil drain tube (24" Gas hose) $25 Oil drain flange kit 1/2"NPT $29 Weld on pan bung 8-AN $8 Water connection kit $47 Wastegate water feed 4-AN 90 degree $8 Wastegate water feed hoses 10" 4-AN $31 Turbo blanket $43 half couplings (1/2"NPT Bungs) for drain $17 Total, $2385
Suspension: 2" Rod Delrin $12 4X LH end links HAL-HML8-10SZ (20.97) $84 4X RH end links HAL_HMR8-10SZ (20.97) $84 4X LH jam nuts HAL-JNL10S (2.99) $12 4X RH jam nuts HAL-JNR10S (2.99) $12 2X 10" link bars KYS-100-T010 (10.99) $22 2X 9" link bars KYS-100-T009 (10.99) $22 3X 6 pack 1/2" Rod end seals (11.97) $36 2X Timken Wheel bearings $155 Total, $439
Grand Total, $8508. Which is a bit conservative. Its probably closer to $9000
My reasoning behind this disclosure is to expose what the real costs were to do this type of build. Keeping in mind that I did almost all the work myself, except for the flywheel and head surfacing and the powder coating. I Probably put a couple hundred hours of work done in about 2-3 hours per day. And, this build is not over. I intend to slowly push the boost up till the MAP sensor says It can't do any more. I intend to Dyno it then, and in the meantime I REALLY want to do the body work and get this car painted. It is a really rough looking car at the moment, but mechanically, it's a beautiful beast, and a whole lot of fun to drive.
Today I decided to tackle the axle leak. It was getting pretty bad.
So, I had to come up with a plan. The most obvious choice was to get one of these and make it fit.
But, I've seen several complaints about these, and that the stub material is too soft on the axle cup and tends to spall a bit. So I started looking for just a standard bearing that would fit around the seal collar that I pressed onto the axle stub. There weren't any that were that size. So my next idea was to find one that would fit the smaller diameter of the stub itself, which just so happened to be a hair under 30mm. And that was no problem at all! I also needed a bearing that's OD was smaller than the OD of the seal. I wound up getting a 30x47x9 bearing.
And so I very carefully cut off the pressed on seal collar in the lathe.
Then I sanded off the paint and polished off the sealing surface
I did a bit more AutoCAD, to layout my piece, then began machining a chunk of aluminum to match.
I deliberately oversized the piece to be a hard press fit at .003" over. But, I had a trick up my sleeve to make it easy to press... That's right, an air duster. When held upside down, it produces an extremely cold liquid which I used to freeze the piece to a temperature too cold to safely touch, It still needed to be somewhat lightly hammered in with a dead blow, just not as hard as it would have been otherwise.
Then, I took the two seals off the bearing, then washed out the grease with a solvent bath. These bearings will be sitting in transmission fluid, so they didn't need the seals, nor the grease.
Then install:
tap in the new seal
Test fit the axle cup
Seems to be going great, so I re-greased and clamped the CV assembly back together
At first I was going to re-fill the transmission via the fill port, but it would have been nearly impossible to get to the nut. So, instead I removed the VSS and filled it through the hole. Worked great.
And 40 miles later... No leaks!! Just like it should have been.
While I had the rear wheel off, I got a hold of some hub centric rings to help center the wheels. The car was a bit shaky on lug centric. It seems a lot better now.
I am still working on code a little here and there. The car feels incredibly fast. It now tends to get a little squirrely in the back in second gear. Which, for a Fiero with good tires, means you're making a butt-load of horsepower.
My focus over time with this car is shifting though, I'm starting to really think long and hard about doing the body work, get the interior all squared away and actually have a complete car that I am proud of.
Looks good. 400 thow with a .003 press fit It should not walk out. Just keep an eye on it. You have two tapped holes on the engine side if you need to make a hold in bracket
I've put the project away for a bit while we've been getting settled in our new house.
But recently, I've taken it out for a few more tuning and shake down runs
I wanted to tune the car to handle the onset of Air Conditioning load, but it turns out that my compressor is no longer displacing Freon. So I'm betting that the control valve on the V5 compressor is busted. Common problem, but aggravating. So I abandoned that effort.
Instead, I've been desiring to push the boost up slowly. I've done some reading, and feel like I know how to approach stepping up the boost via code changes. But first, I had to install the original fuel pressure regulator, which I did today.
Shortly, I will find some web space and release my code, along with the research that went into making it happen.
Today has been a fun day. It started with an oil change. I'm now switched over to Mobil 1 synthetic 10W-30.
Once I had the engine warmed up, it was time to start incrementing the boost. Fortunately, I have one of the best boost controllers available. Stock GM Equipment
Since 8F has native wastegate control, I am using the turbo Grand prix's boost control solenoid (AC DELCO 214-474)
My Tial wastegate has a "Half Bar" spring in it. so 7.25 lbs of boost theoretically. In reality, its right around 8.2 peak sustained boost. Between manifold pressure and the wastegate control is the boost control solenoid. The BCS is a normally open valve that can "throttle" the pressure that makes it to the wastegate. When no power is applied, the pressure from the intake manifold controls the wastegate directly. The valve itself is controlled by pulse width (Duty Cycle) And the more duty cycle is applied, the more boost gets restricted from opening the wastegate. Which increases boost. So the goal for today was to slowly increase the duty cycle on the BCS, and document how that impacted the sustained boost level.
First thing I did to the code, was to increase the boost fuel cutoff limit to 14.5 PSI, restore at 7 PSI, the I filled out the desired boost pressure base table. 14 PSI in the 100% throttle column.
Next, I turned off the ability for the computer to adjust the boost upward by turning off the boost step-up update interval to zero. And, "Positive step adjustment for wastegate DC to zero". So now the computer can only put out the maximum DC that is specified in "Initial Value of WG DC vs RPM" (in all RPM columns)
I did a few runs starting at zero DC and incrementally stepped up the value: in return I got this data: %DC PSI 0 8.2 10 8.4 20 8.6 30 9.2 40 10.0 45 10.7 50 11.5 55 12.3 60 13.0 65 13.7
At about the 55% DC mark, I started to really lose traction in second gear. But I pressed on, working my way to 65% DC, mostly in 3rd gear and 13.7 PSI of boost. I never actually had to adjust the fueling. It was rock stable at 10.8:1. Some of the early "Just Guessing" changes seemed to have been pretty much on the mark. Just a touch richer than I'd like, but perfectly good for that kind of boost.
How does it drive? A bit scary. The boost comes on and the back end gets squirrely between 35 and 55 mph, traction comes back in at right around 55 mph again. I can say now, that the power the engine makes is pretty dialed in. However, I've lost some drivability with the fuel pressure bumped back up with the stock regulator. I need to work on getting the day-to-day driving bit nailed down. I had it downright reasonable before. Now it stalls, and throttle response is boggy off idle.
I'm also finding that I am having temperature management issues that I need to work on. The engine temperature is much higher than I'd like. Its a 195 degree thermostat, fan comes on at 213, and shuts off at 204, but my CTS regularly registers 220 or so. Perhaps the crossover and turbo is cooking the CTS? I thought I shielded it plenty, but I suspect it still is being heated. The fan is running all the time practically. I'll have to get a thermal gun and do some investigating. I really should have installed a 180 degree unit, It seems logical with a turbo engine. It's a complete PITA to change out now, but I think I will make the effort sometime soon.
I'm also going to have to add heat protection for the decklid. The turbo wasn't really running hot at 8.5 ish PSI. But today when I pushed it up to 13.7 The heat coming off the turbo was impressive. So much so that I sat with the car as it was cooling off to make sure nothing started burning, because I could smell cooking paint once again, not necessarily the decklid.
The worst part. Intake air temperatures are frighteningly hot. Today, I've seen the highest temperatures that I have ever seen. Going into the turbo, which is my temperature sensor just inside the air filter. It hit 188 deg f. At the same time the turbo outlet temperature skyrocketed to 290 f. Which is very close to the highest temperature that the sensor can register (304 f.). That aside, the intercooler did it's job. Dropping the turbo outlet temperature a whopping 110 f. to 180 f. Slightly cooler than the air going into the inlet. So... I really need to devise a cold air intake, with little to no room to do it. That sounds like the rest of this project. Difficult.
All that aside, I think a dyno session really sounds like a good idea now that I'm at the boost level I built this engine to operate at. Not sure when I'm going to do that. But I NEED to know how much power am I making! And I know you all want to know too
[This message has been edited by Fierobsessed (edited 08-29-2014).]
I drove all over the valley this morning hitting every dyno shop in town. Came away empty handed unfortunately all were either booked solid, or closed up shop. I'll have to schedule one someday in the near future.
In the meantime, I've had my first casualty of boost. My EGR.
It's got some carbon soot buildup all over the intake manifold around the EGR. I'm guessing the pressure on the crossover tap is too great on boost. It didn't have any issues till I turned up the boost. The discoloration from the heat started to show up on the EGR tube near the EGR when I turned it up, which is a not so subtle way of telling me that the exhaust has been leaking through the EGR, and the soot told me exactly where. I do still intend on having a working EGR when this is all said and done.
I chose to tap the EGR feed from the crossover because the Turbo Grand Prix had it that way. But I guess I didn't consider that perhaps the back pressure would be too much for the EGR. Having boost push on it too from the other direction probably isn't helping matters. This EGR wasn't ever used in a turbocharged application either. Guess I better think about a workaround for this issue too.
However, tuning is going well, I've really committed some time to nail the idle down. Since the fuel pressure change it became problematic again. It turned out that much of the idle surging was caused by the fueling map in idle having too steep of a slope on the X axis. Flattening out the 600 and 1000 rows really cured the idle surging.
Stalling was another issue, that seemed to be caused by a lack of spark timing in the area on the chart that becomes active during a stall. Originally I spliced the 3.4 DOHC's ECMs timing table into the Turbo Grand Prix's naturally aspirated section. This caused the 80-100 kPa section at low rpms to have Zero timing. So I back filled that part of the table with more of the original Turbo Grand Prix's original timing table, cleaned up. This made a pretty dramatic change to throttle response as well as helped curb the stalling issue. It's becoming more drivable all the time Blended table before: Blended table After:
One thing that I am finding the more I dial this thing in, is that any changes that I've made to the code that I wasn't completely aware of its effects, needed to be removed. Every change made to the code has to be done with purpose. Guessing at some settings is ok, just as long as the effect is fully understood.
So far, I've started the code changes over 3 times from the original donor bin. Each time only starting with the changes that I knew were effective in achieving the goals.
[This message has been edited by Fierobsessed (edited 09-09-2014).]
VRI likes to deal with Fords. If I went there, their dyno will explode. It would be the first time it's ever seen any real horsepower Poor Dougs dyno doesn't stand a chance! It hasn't had to measure any torque from those Eclipses yet. It would be in for a big surprise.
Hah! well, on that note, I wouldn't mind ganging up on a dyno day type event. It would probably be wise to wait for a crisp day, or get an ice tank for my intercooler.
I did some preliminary HP estimates based on injector utilization. I've hit 65% on 60 lb injectors, so It looks like I've seen right around 470 ish crank horses. But that could be a bit off. Wouldn't mind tuning that calculation to match the dyno results. My guess is I'll probably see 430 or so to the wheels give or take some.
I'm now working on re-coding 8F for 3-Bar operation. I'm hitting the limit of what 8F allows for with boost. My boost set point is too close to the limit of what the code/sensor combo can handle, so sometimes I over-boost just a little and the ECM has no idea what to do. about it.
all were either booked solid, or closed up shop. I'll have to schedule one someday in the near future.
