Tuning a high CR engine

[SappySE107]

Its been safely done to 12:1 already, no knock with stock spark timing...and even more power with custom tuning. There are MAF metered 60V60s but the MAP tuning is fine as well for all intents and purposes NA.

Tuned port was just equal length intake runners, which the Gen 3 engines have.

Coolant flow, I am not aware of any hot spots that would need looking into, but that doesn't mean none exist.

[RacerX11]

The optispark is a piece of shit...I would't want it on our engines. It is one of the most failed parts on an LT-1. It isn't anything more than a re-worked distributor with an optical trigger mechanism. They went to the individual coils for the LS-1. And they went away from reverse-flow cooling on the LS-1 as well. All these things that were supposed to work so good in theory were introduced on the LT-1, but later quietly dropped.

Marty

[quaddriver]

Its been safely done to 12:1 already, no knock with stock spark timing...and even more power with custom tuning. There are MAF metered 60V60s but the MAP tuning is fine as well for all intents and purposes NA.

Tuned port was just equal length intake runners, which the Gen 3 engines have.

Coolant flow, I am not aware of any hot spots that would need looking into, but that doesn't mean none exist.

btw - there is a problem with the new board, whenever I first select the 'quote' button, it throws me out of the thread back to the header page and I have to reenter, then I can hit quote the second time. I have not seen this before.

anyways, Id have to see these 12:1 motors - MAF has a chance (albeit slight), MAP - no way. You cannot alter even basic engine parameters on a speed density system and even have it be close. if anyone has boosted compression on a MAP system, Im betting if they look, its always in negate mode once off idle (you can check this by running a bi-scan tool while driving, but not one that puts it in field test mode (which negates the EST effect anyways)) The entire fuel/spark curve needs reprogrammed and so few have even the basest equipment to do this - its something you cant guess at, you have to have the engine operating on a brake and the programmer has to program each stepped cell one by one (way different from the 16 cell BLM - this has over 1000 cells) while adjusting the timing from an independant BTM modified to use no boost gauge, adjusting injector bandwidth and measuing output gasses. (it is illegal to replace a chip with one that has no cert - it is illegal to sell a chip for onroad with no cert - if anyone tries to sell you differently, it means they guessed with a simple power programmer an no telling what it will do.

but tuned port is a lot more than just equal length runners, the runners also spin the airflow (as mentined in another thread) and they have the peak of the vortex timed (or tuned if you will) to be right before the actual injector, so it grabs the fuel and pulls it along, futhering atomization. tuned port is also tuned to do this in a relatively narrow operating band. you cannot movethe hp peak effectively without going shorter/fatter runners and keepingthe tuning effect - its modifications not for the faint of heart, thats why I am doing the research I am in another thread - remember my reasonable goal: 2hp/inch on a NA race motor.

for the coolant flow when running elevated cyl head temps (which he will be) the process/practice is simple...

the pump pushes cool water in the 'front' of the block down both sides, where it goes past the liners and up into the head, back to the front and out in a longitudinal motor. the transverse 660s exit the coolant at the 'back' so the problem of 'short loop' coolant (coolant that goes in and out quickly without getting to the back) is alleviated, but the base problem still exists - by the time any coolant reaches cyls 5/6 it is heated twice. What you have to do on a performance 660 is drill the heads to allow *some* coolant out at 2/3 and tap thru the intake into this passage using coper tubing and route coolant back into the 'rear' crossover where the thermo is. once this is done, you restrict the front most coolant holes in the block face with plugs with holes in them, so more cooler water is forced backwards. then you restrict the 2/3 holes a little less. It would not hurt to block the 'front' crossover (if in fact you manifold does not have it done already, I recall seeing some that do)

this way, the coolest water that hits 1/2 is limited and you force more cooler water to the rear - this will still only work perhaps with all other changes to 11:1 - that is about the most that chevy has ever reliably got out of any motor without using major cooling system changes, block filler, reverse cooling etc.

without all the above changes, you will fry it. no way around that.

[quaddriver]

The optispark is a piece of shit...I would't want it on our engines. It is one of the most failed parts on an LT-1. It isn't anything more than a re-worked distributor with an optical trigger mechanism. They went to the individual coils for the LS-1. And they went away from reverse-flow cooling on the LS-1 as well. All these things that were supposed to work so good in theory were introduced on the LT-1, but later quietly dropped.

Marty

the mechanicals of HOW optispark worked are unimportant as to WHAT it did. the system allowed hyper accurate cylinder by cylinder tuning changes independant of each other - this was done in the spark module and the dist was direct cam driven to eliminate the lash problems so it would work (.001 of lash at 5000rpms is a significant amount of time) the individual coil packs are much more able to accomplish this now, using as you will note hyper accurate cam sensing.

as for the cooling, the LS1 block and heads were entirely recast to provide priority cooling to the heads without using a complex and failure prone mixing pump.

the name of the game is simply to reject cyl head heat quickly and uniformly and react quickly to detonation. there exists no parts or programming for the 660 to accomplish this. hence, a lot of modifications are needed.

[SappySE107]

anyways, Id have to see these 12:1 motors - MAF has a chance (albeit slight), MAP - no way. You cannot alter even basic engine parameters on a speed density system and even have it be close. if anyone has boosted compression on a MAP system, Im betting if they look, its always in negate mode once off idle (you can check this by running a bi-scan tool while driving, but not one that puts it in field test mode (which negates the EST effect anyways))

I have tuned 1 with 12:1 and a wild cam, and the cam was the only real issue with the tuning because the idle vacuum was low. Compression alone is rather simple to tune for. Even on a stock computer, the compression wasn't doing more than a few degrees KR on a 3.1. MAP has 0 problems with increased compression so I don't know what you are talking about here. As a note, I have tuned 2 others with less wild cams that had no issues with tuning at all.

The entire fuel/spark curve needs reprogrammed and so few have even the basest equipment to do this - its something you cant guess at, you have to have the engine operating on a brake and the programmer has to program each stepped cell one by one (way different from the 16 cell BLM - this has over 1000 cells) while adjusting the timing from an independant BTM modified to use no boost gauge, adjusting injector bandwidth and measuing output gasses. (it is illegal to replace a chip with one that has no cert - it is illegal to sell a chip for onroad with no cert - if anyone tries to sell you differently, it means they guessed with a simple power programmer an no telling what it will do.

1000 cells? Try 3 for the 727/730 computers. There are 16 for the F bodies but for the FWD OBD1, there are only 3, 1 for idle, 1 for cruising/wot, and 1 for decel. OBD2 may be much different but not with what I have played with. With increased compression, it would be best to reprogram the fuel and spark tables, but its no where close to the ordeal you are making it out to be. BTW, what is a simple power programmer? Are you talking about OBD2 and the hypertech programmer for them? I don't think his fiero would require or care for that unless he is swapping the computer for some reason.

As for the cooling, GM addressed the cooling problem at the number 5/6 cylinders with oil squirters for the bottom side of the pistons. Your changes for the coolant flow are pretty interesting. Id like to see some pictures for example of doing this, since I know I am more of a visual person in terms of modifications like this.[/quote]

[quaddriver]

aha, something is different, Im able to post 1 click, ok let me try and address ur concerns 1 by 1

I have tuned 1 with 12:1 and a wild cam, and the cam was the only real issue with the tuning because the idle vacuum was low. Compression alone is rather simple to tune for. Even on a stock computer, the compression wasn't doing more than a few degrees KR on a 3.1. MAP has 0 problems with increased compression so I don't know what you are talking about here. As a note, I have tuned 2 others with less wild cams that had no issues with tuning at all.

ok, when you change compression AND/OR porting, valve size etc in any way, you affect the viewed manifold vacumm at any throttle setting, any rpm when you compare the base motor vs the modified motor. The problem is now, the 'puter uses the MAP, ACT and TPS guess on the incoming air speed and density (hence the name speed density). Assuming the air is the same temp and you are on the gas the same (lets say WOT in a race) then the MAP readings are all wrong for the new motor. It will supply the injector duty cycle and spark timing as if it was the old motor - well kinda, with no puter changes you can see perhaps up to a 5% difference as the puter will interpret the first part of the MAP variance as simply being more on the gas[1]

1000 cells? Try 3 for the 727/730 computers. There are 16 for the F bodies but for the FWD OBD1, there are only 3, 1 for idle, 1 for cruising/wot, and 1 for decel. OBD2 may be much different but not with what I have played with. With increased compression, it would be best to reprogram the fuel and spark tables, but its no where close to the ordeal you are making it out to be. BTW, what is a simple power programmer? Are you talking about OBD2 and the hypertech programmer for them? I don't think his fiero would require or care for that unless he is swapping the computer for some reason.

Ok, I used the term cell because I know you were familiar with the 16 cells used in the BLM strategy. Actually, when over BLM cutoff (1500-1800rpm depending on motor) there is a 3-4 column assosicative memory lookup table that contain base settings that correspond to load (RPM vs TPS reading and the associative change), desired mixture (at idle or in BLM, a mix leaner or lessor than 14.7:1 is deisred, at WOT when very hot, 12.5 may be tolerated, but typically 12.7-12.8, and then of course 'overfueled' (13.5ish:1 required when maintaining RPM against a load, but not wishing to accelerate - i.e. up a slight hill) Misfire counts, KS counts and other factors all modify this. The first column is associated with injector bandwidth, the next spark timing, the third is a modifier based on elevated temps, the 4th if used is a modifier based on excessive isfire/knock. Please note, a valid modifier at 2500rpm is gonna differ from one at 3500 rpm. Hence the elaborate (simplified in my example) engine test bed. the engineers and anyone who reprograms have to test the motor under the different factors but using rpm slices like 10rpms.

As for the cooling, GM addressed the cooling problem at the number 5/6 cylinders with oil squirters for the bottom side of the pistons. Your changes for the coolant flow are pretty interesting. Id like to see some pictures for example of doing this, since I know I am more of a visual person in terms of modifications like this.

I might be able to lay my hands on a busted head from a scrap heap and either show the mods via digicam or simply mail it, I wont be around till later in the week and would need a few hours to make the mods (its a bone stock head, incl the crack )

Lastly, hypertech has been around (and their ilk) since the early 80's using OBD1. nowadays the ROMs are PROMS and unremovable, but programmable, in OBD1 days they were removable, but for the most part unreprogramble so you would get a simple EPROM chip and use the hypertech et al to set up what you wanted.

Notes:

1: In all the articles and shootouts I have ever seen, a chip addition has never done anything when the engine was actually tested other than raise emissions, cyl head temps and fake out the TPS and ECT(CTS) reading. Professional tuners who do actual engine changes in all cases compliment the changes with chip work, but they use a setup like I described and in most cases carry a CARB EO number

[SappySE107]

Damn chat room...it locked up and when I killed it I lost the post I was typing! argh!

First, an example dyno test of some of my work.

http://www.fastfwdperformance.com/3.1%20chip%20dyno.jpg

I am aware of the many variables for the end result the engine sees as far as pulse width and timing advance. Its no problem for me to change the coolant vs X tables, or the base tables, or any of the other constants. I dont know what the "guess" is after its tuned, since all the data tables are modifiable. Some mods don't require much change for normal driving, like the milder grind performance cams from crane. WOT is just about the only area worth changing.

I don't know what "over BLM cutoff (1500-1800rpm depending on motor)" is. I know what the BLM and INT ranges are, but nothing in terms of RPM and BLM. I do my tuning by driving the car under different conditions, but not all at once to get the fuel tables within a close spec to optimal BLM. After idle, its cruise, and then decel, and then near WOT with the PE mode parameters changed to keep it out of PE to get a wider range of constants with the O2 adjustments in working order. Anyway, this isn't a computer tuning discussion, so ill cut that short:P I just wanted to point out that MAP does in fact work with engine changes until the MAP range is too small vs TPS and RPM. If it gets too touchy, then MAF would make tuning much easier. Its a lot of fun manipulating the IAC to deal with drivability.

For the head pics, we have some on the site of a 3.4 DOHC and a 3.1 Gen 2 head cut up. I dunno if the 3.1 pics will help but maybe?

http://www.60degreev6.com/index.php?p=pages&pid=72

I would hope franz still has those pieces, but he is in the middle of moving so he won't be on here for a while. I can email him though and see if he has them, it would save you from cutting a set up.

[brian89gp]

Ok, I used the term cell because I know you were familiar with the 16 cells used in the BLM strategy. Actually, when over BLM cutoff (1500-1800rpm depending on motor) there is a 3-4 column assosicative memory lookup table that contain base settings that correspond to load (RPM vs TPS reading and the associative change), desired mixture (at idle or in BLM, a mix leaner or lessor than 14.7:1 is deisred, at WOT when very hot, 12.5 may be tolerated, but typically 12.7-12.8, and then of course 'overfueled' (13.5ish:1 required when maintaining RPM against a load, but not wishing to accelerate - i.e. up a slight hill) Misfire counts, KS counts and other factors all modify this. The first column is associated with injector bandwidth, the next spark timing, the third is a modifier based on elevated temps, the 4th if used is a modifier based on excessive isfire/knock. Please note, a valid modifier at 2500rpm is gonna differ from one at 3500 rpm. Hence the elaborate (simplified in my example) engine test bed. the engineers and anyone who reprograms have to test the motor under the different factors but using rpm slices like 10rpms.

Want to expand on this a little? I know in the $DF mask that there isn't a BLM cutoff till at least over 4800 and thats only because it requires a MAP/TPS reading that throws it into PE to get it spinning that fast. OBD2 might be able to detect misfires but I am pretty sure that the 727/730 on these engines could care less.

[quaddriver]

Want to expand on this a little? I know in the $DF mask that there isn't a BLM cutoff till at least over 4800 and thats only because it requires a MAP/TPS reading that throws it into PE to get it spinning that fast. OBD2 might be able to detect misfires but I am pretty sure that the 727/730 on these engines could care less.

Im not sure what ur asking me...here is a *representative* sample of a breakdown of the data area in the ECM: (I hope to god this formats, else its gonna look like major shit)

Address Range Purpose 
$0000-0001 Checksum  
$0002-$0003 Scan ID 
$0004 EPROM Template ID 
$0032-0106 Main Spark table, 3d  
$015E-0165 WOT Spark adder 
$017F-0187 Knock sensor retard at WOT 
$0188-018C Knock sensor retard non WOT 
$030E-0312 MAP Power Enrichment 
$0314-031D TPS Power Enrichment 
$037F-03C6 Main Fuel table, 3d 
$03C7-03D7 Main fuel table Adder 
$044A-0458 Crank Air Fuel Ratio Vs Coolant 
$0459-0468 Air Fuel Ratio Vs Coolant 
$047D-0485 Percent of TPS to enable PE 
$050D-050F Error Code Masks 
$05F5-0605 Air Fuel Ratio Vs Coolant 
$060E-0616 Idle Speed Temperature Corrections

This is for OBD1, TBI in fact. GM has well over 100 variants so your mileage will differ. Note that x616 is 1558 bytes of data, some of these tables are 2 or 3 dimensional and are summed (net value) to produce the desired number, for example, the spark number sent to the DIS might be the proper value in the main spark table, the WOT spark adder (if WOT detected) and max spark retard based on RPM or MAP readings (2 other tables.) The main spark table prolly does not go above 3600rpm (max timing is still used) and the retard will set prolly 4000 as its highest value. Spark I think is in 200rpm steps.

As per previous convo, for example the main fuel table adder (aka VE table) - if you change the hardware in the motor such that you have altered this, the VE for that motor is no longer a function of the same RPM!!!!!

as for the BLM, there are only 4 RPM 'columns' in that table, the 4th column is 1500 orf 1800 rpm+, meaning the BLM does not play much of a role once off that rpm limit

[SappySE107]

Thats TBI, which appears to be massively inferior to the port injection code masks that brian and I have used. 4 RPM columns! yikes! Ill link to a few pics for example of a stock 3.1 from a 91 grand prix.


Spark table (just one of them, but shows the detail)


Main VE 3D Table


Power Enrichment Table


The Base VE table is broken up 0 - 6400 RPM in 400 RPM increments. The spark table has a breakout spark RPM setting associated with a timing advance, so that over 4800 RPM you can increase spark timing a set amount of degrees every 1000 RPM if you want. There are also many spark adv/ret correction values based on many sensor inputs. Same goes for fueling needs. It appears that the MPFI is a ton more accurate for tuning. For TBI, i dunno, i guess even a small change is able to totally screw it up, but I don't have that issue.

[v6h.o.]

i'd the fiero ECM from the car, and then get a custom chip programed for it.
your only loosing out on the SFI feature with this setup. which the SFI engines only use untill about 3000 or 3500 RPM and then they revert back to batch fire for the higher rpm.

thats what us Jbody guys have been doing for all the 3400 swaps, which work fine.

all i know is that i ran over 11:1 compression for most of this summer. no issues with cooling or overheating. no KS sensor activity or knock. all that on 92 Pump.

[2M6]

^^^ Thanks for the info. I will keep that in mind.

[SappySE107]

I know there are people on fiero.nl that could tune any engine with the stock ECM. You COULD use the 94/95 3.4 DOHC ECM to keep the SFI but you have to use the DOHC MAF and retune it. Brian is working on it now, and im kinda just tagging along to learn the maf so I can tune those computers as well. Either way you go, I would NOT use the stock 3400 ECM. OBD1 is cheaper/easier to tune and if you want to, you could learn how to tune it yourself.

The stock Fiero ECM is different than the FWD cars, and so are the code masks. I haven't looked at it so I don't know its specifics. Someone can help you though im sure.

[quaddriver]

Ben, (Im not gonna quote your post and cause the pics to repost)

thats an interesting software tool you have, windoze compliant it appears (mine is in pascal and runs in a DOS window for OBD1)

but you see what I mean - many many cells that you have to reprogram.

question, from your spark table it seems you have a base timing of 80-some degrees? how in the heck do you avoid bending rods?

ps - I had put up the map for tbi because that was what I had locally, Im working right now on my K motor using an H ECM and trying to figure why it wont idle smooth when cold.

[SappySE107]

Thats the hex value, not the actual spark advance. I don't refer to them as cells, but yes, there are quite a few. Using a scan tool and recording the engine data while I drive, its pretty easy to go through and modify the values based on KR and BLM/INT.

That tool is winbin, there are others but I had that one handy for posting some example pics for the tables. The range you see for the numbers is just a part of the whole thing which is shown as the chart. Most of the arrays are smaller increments til 1200, for tuning idle.