new tuner completely lost please help

[JR92RS]

The line is drawn wherever you want it to be drawn. Some people want perfection. Some want close to perfection as they can get. Some just want to make to work 5 days a week and not break down on the freeway in a burning heap of liquid metal.

i think the last one should have been first....lmao, i'd settle with close as i can get and no burning heap of metal

[SiriusC1024]

While you are at it compare the injector data for a 02 Fbody and a 02 Vette that both use the exact same injector and figure out why they are characterized so different by GM calibration engineers. Or find an injector manufacturer that doesn't spec their injector flowed with n-heptane.

Yeah that's what I'm going to do. Gotta see if it's consistent across the tunes.

[SiriusC1024]

just got a tip of the day type email from hp tuners about crank relearn....and realized that hasn't been done yet, could that be the miss feel i was getting yesterday?
and in looking at their process with applying the parking brake then the foot brake.....i don't know if my ecm is getting the parking brake signal or not. i think i'm good on the footbrake as the torque converter would need that signal for lockup and disengaging, but if i don't have the parking brake signal could that make the process crash or cause any undesirable affects???

Of course you should do a crank relearn. It's done with a new ecu, cps, or engine/crank.

Also, you should do a tps relearn.

[JR92RS]

Of course you should do a crank relearn. It's done with a new ecu, cps, or engine/crank.

Also, you should do a tps relearn.

i will look into this tomorrow, thanks

[JR92RS]

Of course you should do a crank relearn. It's done with a new ecu, cps, or engine/crank.

Also, you should do a tps relearn.

i'm assuming both are under the special functions icon?

[SiriusC1024]

Crank relearn is.

TPS:

Key off.
Unplug TPS and IAC.
Turn key on for 5 seconds.
Key off.
Plug in TPS and IAC.
Turn key on. Closed throttle should be 0%.

https://sites.google.com/site/sloppy...ls/iac-relearn

[JR92RS]

Crank relearn is.

TPS:

Key off.
Unplug TPS and IAC.
Turn key on for 5 seconds.
Key off.
Plug in TPS and IAC.
Turn key on. Closed throttle should be 0%.

https://sites.google.com/site/sloppy...ls/iac-relearn

thank you for that

[JR92RS]

Crank relearn is.

TPS:

Key off.
Unplug TPS and IAC.
Turn key on for 5 seconds.
Key off.
Plug in TPS and IAC.
Turn key on. Closed throttle should be 0%.

https://sites.google.com/site/sloppy...ls/iac-relearn

ok i think i did a successful crank and tps relearn attached is your tune file - the crank fueling (not used no starting issues) and a log file of your tune (crank relearn in this log)seriousC.hplbase tune + step5 sirious.hpt

[eficalibrator]

Can you do gen V's or DI's and what do you charge for the different styles? I see it the most on gen V's where customers have to remove them and send them in for cleaning and flow testing - just curious as this would deliver Exact Data for that given set...

Testing DI injectors for characterization requires DI pressures (100Bar+) My machine is only built for PFI pressures. Anyone who says they can derive calibration data for DI injectors at PFI pressures is selling BS. Cleaning and flowing them (relative to each other under similar, but not representative conditions) is easier, but it's not going to get real calibration data.

My PFI characterization testing is pretty thorough across pulse widths and voltages. I provide the service for $389.00 here in house. I understand that the casual DIY tuner probably won't justify the cost, but I am an engineer and don't shortcut the process.

From the logic above you're saying that even a set of factory injectors with known data need to be characterized due to variation. Save that for the race teams.

Production injectors are allowed to be +/-6% from the mean, but are usually within about 3% unless they are damaged or clogged from what I have seen. Companies like Bosch and Conti sell injectors to the aftermarket with an even wider range of tolerances, more like +/-8%. Luckily, many injector resellers will do a quick relative flow test and group them in sets that are within a few percent of each other before shipping them out. These sets may be right on average for that part number, or you may get 8 that are all ~5% off the mean together. Many amateur tuners won't notice this.

Your method estimates, not measures, the non-linearity region. The CSI model still has up to 10% error. So where do we draw the line on what's acceptable error? Sometimes it's a case of diminishing returns.

I can assure you that I am taking thousands of raw measurements in the nonlinear flow region with my machine. There is always some averaging in the data analysis because injectors that big can have shot-to-shot variability too. Keeping the model that close all the way down to zero fuel delivery is pretty good. It would be much tighter if we were looking at lower flow injectors like the 36# units the OP has. Either way, my data is ~4X more accurate than the other model in this case. I will agree that there can be diminishing returns at some point.

What sense would it make for OP to pay even more for injector characterization when the manufacturer gives him the data and guarantees 1% variation? Further, those Accel's are rapid-opening. The non-linear region is much smaller than factory Bosch's.

Unfortunately, it is RARE that injector manufacturers/sellers have data that guarantees 1% variation. ID comes close because they have rigorous quality control checks and are willing to scrap outliers, which raises the cost of the remaining injectors. Companies like Accel are not even in the same ballpark there.

What is "rapid-opening"? All solenoid injectors open pretty rapidly, but a difference of 0.1ms is very measurable with the right equipment. It also changes with voltage and pressure. Even with modern injector design, the nonlinearity is measurable as seen in this set of Bosch 36# injectors that I tested a while ago. If we take enough data below 3.0ms, the mismatch becomes obvious. As a tuner, you will have to decide if this error is acceptable for the quality of work you plan to do.
36lb.JPG

SAEj1832 tells us to measure with N-Heptane because it has similar viscosity to pump fuel. We can easily make the gravimetric correction to E0 or E10 if needed. Some companies will test with virgin gasoline, but it can only be flowed through the injectors once before it loses some of the light ends and begins behaving differently. They also have to be aware of the obvious fire and explosion hazard compared to heptane. I correct my data back to E10 numbers since that is what most users will have in their tanks.

[SiriusC1024]

Testing DI injectors for characterization requires DI pressures (100Bar+) My machine is only built for PFI pressures. Anyone who says they can derive calibration data for DI injectors at PFI pressures is selling BS. Cleaning and flowing them (relative to each other under similar, but not representative conditions) is easier, but it's not going to get real calibration data.

My PFI characterization testing is pretty thorough across pulse widths and voltages. I provide the service for $389.00 here in house. I understand that the casual DIY tuner probably won't justify the cost, but I am an engineer and don't shortcut the process.


Production injectors are allowed to be +/-6% from the mean, but are usually within about 3% unless they are damaged or clogged from what I have seen. Companies like Bosch and Conti sell injectors to the aftermarket with an even wider range of tolerances, more like +/-8%. Luckily, many injector resellers will do a quick relative flow test and group them in sets that are within a few percent of each other before shipping them out. These sets may be right on average for that part number, or you may get 8 that are all ~5% off the mean together. Many amateur tuners won't notice this.


I can assure you that I am taking thousands of raw measurements in the nonlinear flow region with my machine. There is always some averaging in the data analysis because injectors that big can have shot-to-shot variability too. Keeping the model that close all the way down to zero fuel delivery is pretty good. It would be much tighter if we were looking at lower flow injectors like the 36# units the OP has. Either way, my data is ~4X more accurate than the other model in this case. I will agree that there can be diminishing returns at some point.

Unfortunately, it is RARE that injector manufacturers/sellers have data that guarantees 1% variation. ID comes close because they have rigorous quality control checks and are willing to scrap outliers, which raises the cost of the remaining injectors. Companies like Accel are not even in the same ballpark there.

What is "rapid-opening"? All solenoid injectors open pretty rapidly, but a difference of 0.1ms is very measurable with the right equipment. It also changes with voltage and pressure. Even with modern injector design, the nonlinearity is measurable as seen in this set of Bosch 36# injectors that I tested a while ago. If we take enough data below 3.0ms, the mismatch becomes obvious. As a tuner, you will have to decide if this error is acceptable for the quality of work you plan to do.
36lb.JPG

SAEj1832 tells us to measure with N-Heptane because it has similar viscosity to pump fuel. We can easily make the gravimetric correction to E0 or E10 if needed. Some companies will test with virgin gasoline, but it can only be flowed through the injectors once before it loses some of the light ends and begins behaving differently. They also have to be aware of the obvious fire and explosion hazard compared to heptane. I correct my data back to E10 numbers since that is what most users will have in their tanks.

Rapid-opening is Accel's terminology for those injectors. They claim low-mass disc results in rapid opening. This shows in the datasheet where the injector is full open and flowing linearly at 1.38ms instead of somewhere around 3ms.

No doubt your method is the best way of approaching injector characterization. OP just wanted good enough, and there is enough injector data from the manufacturer to get it there. The only thing missing is short pulse adder, so assumptions have to be made there.

Thank you for your time. I'm not arguing with your method, just the necessity in OP's case.

[JR92RS]

in a speed density tune how does the ecm know weather to use the high or low octane tables in the spark tab?

[SiriusC1024]

in a speed density tune how does the ecm know weather to use the high or low octane tables in the spark tab?

You should make a new forum post about each subject you're trying to learn.

[JR92RS]

You should make a new forum post about each subject you're trying to learn.

ok will do, but before i do....the tuner that i was talking too had an opening yesterday (because a car was having mechanical problems and couldn't be tuned) and after he sent me a base file, and 1 revision after i sent him back an idling log, i then drove across town and had it tuned. going to post it below too get feedback as i am not even closely qualified to make judgements on his work. car seems to run fine. also going to put a list of the statements he made....so you can see the why behind his tune.

1. the knock sensors were being overly sensitive and pulling timing during the dyno runs(he wanted to know the brand of sensor and the torque applied to them on install / both are new ac Delco, new factory harness, 15 flbs torque, and the caps sealed on the valley cover), he attempted to desensitize them, then turned them off altogether before telling me about the issue and leaving it up to me.....i decided to have him turn them back on in their now desensitized state for safety.

2. he said that since the car was running on regular gas when dyno tuned i would have to keep running regular gas, and (i believe he said, don't quote me....we talked about a lot) the tune won't adjust for me switching to 93 octane down the road. hence me trying to learn how the ecm decides which octane table it tries to use and why. (my question after looking at his tune) cant this problem be overcome with two separate octane tables that don't appear to be identical?

3. during a phone call prior he had stated this is just a run of the mill cammed 5.3 (which in my mind said he would not give it the attention it deserved, but i think he did end up spending about 4 hours on it which i believe to be normal tuning session)

4. toward the end he said he was getting abnormal airflow readings (more intake than he believed the ls1 intake was capable of) so i believe he scaled back the injectors.

i can't really remember anything else that stood out, but i'm sure the following comments will spark something. i have my thoughts on this tune but as i said i am nowhere near qualified to judge his work. i will say that i did compare his injector work to seriusC1024's work.....and i still reserve judgement. i will get a log of this tune in action later today

so off we go, pick it apart tuning addicts

[gtstorey]

I think you would typically tune on the higher octane fuel and then you can make some some assumptions for the lower octane. Hard/dangerous to tune on low octane and push the tune up for higher octane. It also uses the knock sensors to know when to bump down to the low octane table so if they are desensitized or off, you might have problems changing from a higher octane to lower.

[JR92RS]

I think you would typically tune on the higher octane fuel and then you can make some some assumptions for the lower octane. Hard/dangerous to tune on low octane and push the tune up for higher octane. It also uses the knock sensors to know when to bump down to the low octane table so if they are desensitized or off, you might have problems changing from a higher octane to lower.

good feedback thank you, in this case it would be switching from the low octane(that it was tuned on) to the high table.....i was hoping assumptions could be made, or i could find a stock 5.3 tune high octane table and re-install it to my tune....but the desensitized knock sensors is a valid point

[SiriusC1024]

Short pulse adder is zero. Guess he just made up his own offset table; however, offset is scaled for vacuum, which I forgot to do.

Kind of a sloppy tune overall. Look at that hard transition on the VE table. Hard to get with 6200rpm redline maybe, but still that cam is online there. Spark is a little odd looking. Any time you see power enrichment flatlined you know they didn't take their time maximizing power.

Knock sensors are a known failure point on gen 3. Get good sensors, new harness, and a torque wrench.

What were your dyno numbers?

[JR92RS]

Short pulse adder is zero. Guess he just made up his own offset table; however, offset is scaled for vacuum, which I forgot to do.

Kind of a sloppy tune overall. Look at that hard transition on the VE table. Hard to get with 6200rpm redline maybe, but still that cam is online there. Spark is a little odd looking. Any time you see power enrichment flatlined you know they didn't take their time maximizing power.

Knock sensors are a known failure point on gen 3. Get good sensors, new harness, and a torque wrench.

What were your dyno numbers?

why doesn't it show tune file was viewed?
the knock sensors are brand new ac delco oe, so is the harness, and i torqued them to 15 flbs


Dynojet Graph.pdf

[GHuggins]

Did I miss something here? Was there another timeline change or something This is a gen3 right? OK - just to clarify some things - when you fail the MAF and run a SD tune ON A GEN 3 it AUTOMATICALLY defaults to the low octane spark table. Just the way GM set them up "for safety". Your tuner tuned this on the low octane table for low octane gas. IF it is left as a SD tune it WILL NEVER try to go back to the high octane table as it will stay defaulted to the low octane table.... You want to put it back on MAF and tune the MAF in, THEN it can use both. Timing isn't perfect per say. Actually looks like it ramps up then back down. Plus there's an area in the lower load where it's lower. This does tell me it's not optimized, but not horrible. Usually an engine will run above 4800rpms the same amount of spark or a touch more up to redline. He has it dropping off. Why? I don't know... This is also shown in the dyno where things start to drop off. His VE table is smooth and does have a hump - even factory tables do. This hump area is probably running too much timing in the timing table and cylinder pressure is peaking. Not a "bad thing", just the way it turned out. It does drop off after it - should maybe be smooth there, but anyway...

[JR92RS]

Did I miss something here? Was there another timeline change or something This is a gen3 right? OK - just to clarify some things - when you fail the MAF and run a SD tune ON A GEN 3 it AUTOMATICALLY defaults to the low octane spark table. Just the way GM set them up "for safety". Your tuner tuned this on the low octane table for low octane gas. IF it is left as a SD tune it WILL NEVER try to go back to the high octane table as it will stay defaulted to the low octane table.... You want to put it back on MAF and tune the MAF in, THEN it can use both. ALSO the spark table IS NOT flatlined out. Actually looks like it ramps up then back down. Plus there's an area in the lower load where it's lower. This does tell me it's not optimized, but not horrible as it was hinted. Usually an engine will run above 4800rpms the same amount of spark or a touch more up to redline. He has it dropping off. Why? I don't know... This is also shown in the dyno where things start to drop off. His VE table is smooth and does have a hump. This hump area is probably running too much timing in the timing table and cylinder pressure is peaking. Not a "bad thing", just the way it turned out. It does drop off after it - should maybe be smooth there, but anyway...

i wasn't saying, insinuating, or stating it was a bad tune.....because i have no idea, i just had questions. thank you for your response....i didn't know in a sd tune it defaults to the low octane table (i guess that is why they are both the same) and why i asked.

[GHuggins]

It was never meant to come off bad In fact I even misread one comment and have since corrected my post. Just some - for what it's worth info...

But yes, on the gen 3's when anything "act's up" or "fails" it will automatically default to the low octane table. It was GM's earlier way of thinking to protect and "de-power" things...