dude that car lights the wick!!!!!!Wow! I want a ride.
dude that car lights the wick!!!!!!Wow! I want a ride.
Have a look under Engine Diagnostics table 39020 Airflow Max/RPM & 14064 Peak Airflow WOT.
See what baby airflow numbers you are working with/against.
You need to log more parameters like I had in log I posted earlier so you can see what is happening.
Thanks I will give that a look. I have adjusted the throttle air flow tables and reduced oscillation significantly. I have worked on adding some of the parameters you have listed. I am not having luck pulling those PIDs, and I am not sure why. I sent an email listing the parameters I need, and I used on my Ram. Not sure if it is a Pentastar thing or unlocked PCM issue.
Thanks for the log. That is very illuminating, but you can log so much more stuff than me I'm not sure how applicable it is to my old NGC4 PCM. Demanded torque and desired airflow would make things much easier but I don't have those parameters with the latest beta.Originally Posted by Hemituna
Unlike my PCM, your actual and expected torque differ. Outside of idle mine are always the same. Yours also seem to be way too low relative to what the Aircharge to Torque and Friction Torque tables indicate.
I'm surprised it wants that much timing: 18 degrees at 4,000 RPM and an atmosphere of boost. I only give my naturally aspirated 449 17 degrees at 4,000, though I am knock limited on 93 octane. Makes me want to see how much more power is there with E85 and more advance.
Unfortunately I don't have that table either.
I have figured one thing out: the torque to airflow equation is definitely as presented by HPTuners: airflow = (torque - offset) / factor
If I apply the above equation to the Aircharge to Torque table (times 8, because 8 cyilders) I get almost exactly the right amount of airflow (as implied by the axis, aircharge and rpm). If I do torque / factor - offset, the result is not even close to correct.
QUOTE=Grant;638229]Thanks for the log. That is very illuminating, but you can log so much more stuff than me I'm not sure how applicable it is to my old NGC4 PCM. Demanded torque and desired airflow would make things much easier but I don't have those parameters with the [latest beta.
Welcome to 6-10yrs ago on these...we only had diablo, logging was nearly non-existant, very few parameters available so we were pretty much tuning blind.
Just had to try stuff and see what worked. When HPt came along things got way better.
But knowing what we do now from logging later OS's some of early stuff we did made sense and some of the things that got messed up with those changes could now be explained.
Unlike my PCM, your actual and expected torque differ. Outside of idle mine are always the same. Yours also seem to be way too low relative to what the Aircharge to Torque and Friction Torque tables indicate.
Your tq numbers stay similar as you dont have boost. so airflow/tq is as per tables and numbers stay good....add boost and actual tq goes up...get too aggressive with the blower relative to pedal and it will shoot past and then probably get clipped.
This is the surge, let off, surge, let off that someone complained about earlier in this thread.
Now we all hate low numbers right...cant be making any power right....but try not telling the PCM about your huge engine, giant airflows and big tq and it will let the party continue, it has no idea what is really getting to the tyres if you dont inflate all those numbers.
Your OS is missing a table that usually is under the Aircharge to torque table, it is an inverse table Torque to aircharge...the 2 need to match so changing just the one you can see probably wont work so well.
The torque model is a little complex, hence why I say it is easier to get a result by lowering airflow numbers but just try it yourself, its like fitting a bigger TB (gives more airflow but PCM sees same values).
On that log I posted, yes the aircharge is high but tq is low. This is because there is a heap more tq to come and I dont want to run out of scale.
STD Hellcat uses a roughly 0-900nm tq scale. When you add boost with a smaller pulley, the max tq number wont actually go up much as it is at the top end of the scale and already nearly maxed out std.
Most everyone just leaves it be and maybe adjusts the trans WOT shifts to handle the extra tq. However, take it a few steps further and put in a 1300+hp (nearly double stock 707hp) setup and you will be full scale waaaay before WOT.
It can be made to work from an engine tune perspective, but the trans is horrible as it doesn't really know what tq it is getting.
So what i do is to stretch the tq scale a little to 1000nm (which the trans is ok with) and then squeeze the nearly 2000nm engine into the 0-1000nm scale.
Idle, off idle numbers will need to stay same-ish but from there the actual tq is nearly halved for same aircharge (so tq stays same but aircharge needed to achieve that tq is much larger) and i see a max of 950-980 actual tq.
Airflows need to controlled from getting excessive and worked in with airflow/factor numbers to make it all play together. Pedal response is very linear and easy to drive.
Power is even and smooth and 1/2 pedal gets you 700hp and full pedal gets all 1300+.
I use Tq management in the first few gears and that now really works well too as tq is linear and scalable.
Rescale the trans to suit the new bs tq scale and it will shift flawlessly.
Smooth at PT and firm at WOT and nice everywhere in between as there is good tq info from PCM.
Even though aircharge is way up (weight of air trapped in cylinder), airflows and tq and only slightly increased which means no limits being reached (you cannot raise them all and even then you are still off the end of scale).
Hellcats especially like to open the bypass valve and limit boost if you overstep the mark at all.
I'm surprised it wants that much timing: 18 degrees at 4,000 RPM and an atmosphere of boost. I only give my naturally aspirated 449 17 degrees at 4,000, though I am knock limited on 93 octane. Makes me want to see how much more power is there with E85 and more advance.
Pretty standard timing with 9.5:1 comp, the right cam and setup and of course E85. Runs 1.7bar boost at WOT so obviously runs less timing with the added boost.
If yours is hi comp na and currently knock limited, E85 may show some good gains.
Unfortunately I don't have that table either.
If your OS doesn't show it, it is hidden in there and will be slowing you up if you dont stay under it.
Look at other similar OS's and if they have it, assume yours does and keep the numbers under that or at similar levels.
Sure, all this above doesn't matter much at WOT and you can get around most things and make a big power number,
but to get it to drive properly and be usable is even more challenging/rewarding.
[/QUOTE]
Last edited by Hemituna; 01-18-2021 at 06:43 PM.
[QUOTE=Grant;638270]I have figured one thing out: the torque to airflow equation is definitely as presented by HPTuners: airflow = (torque - offset) / factor
If I apply the above equation to the Aircharge to Torque table (times 8, because 8 cyilders) I get almost exactly the right amount of airflow (as implied by the axis, aircharge and rpm). If I do torque / factor - offset, the result is not even close to correct.[/QUOTE]
Now imagine if you can reduce the amount of airflow it thinks it is getting.....it will allow you more TB opening (actual airflow) to get to the same tq number
Remember its all just numbers, so unless you are stock, the stock numbers will now be wrong.
Unfortunately this realization doesn't seem to have helped me figure out how anything else is working:
Actual torque -> airflow is still far short (67-82%) of logged Total Airflow.
Actual torque + Friction -> airflow isn't quite as short, at 84-91%.
Aircharge -> torque ends up higher than logged Actual Torque. If I subtract friction from the answer it's still 110-120% of Actual Torque.
Using speed density calcs to do Aircharge -> airflow gets me really close (93-99% of) to logged Total Airflow, so I think that's what where Total Airflow comes from.
Actual / Expected Torque (they're always equal for me, outside of idling) must then be what the pedal is requesting? But Actual torque + Friction -> Airflow -> Throttle Voltage (using the phi tables and current PRatio) yields a number that is 0.11-0.20 V too low.
Maybe Actual / Expected Torque isn't scaled right, or isn't what it claims to be.
In any case, at the part-throttle-WOT limit, gross torque (Actual Torque + Friction) is flat for me, as is Aircharge. Since reducing calculated torque values doesn't seem to get around the problem, my guess is it's Aircharge / Total Airflow based as hemituner says.
Whoops, didn't notice hemituna's responses:
Yup, I fixed it last night by reverting all TB settings to stock. Thanks! This caused cruised-control problems, and the lightning bolt icon to flash on the dash (meaning limp mode).
I fixed cruise control by adjusting friction losses. Since the entire curve is stock, I don't get sudden unintended acceleration like I was when I messed with portions of the curve. I'm going to try to use the friction table to prevent tip-in stall and leave the curve stock.
I fixed the lightning bolt by adjusting engine diagnostic parameters (which weren't available in my old NGC3, which is why I've never tried this before). Calculated MAP is way off from actual MAP, which is what throws the TB / lightning bolt code.
The problem doesn't seem to be entirely airflow based. One of the first things I did was reduce engine displacement to stock and increased stoich ratio to eliminate the speed density error . This drove fine, reduced airflow and torque, but did not resolve the problem. I could do this again to get calculated MAP in-line with actual MAP, if only to get the throttle diagnostic protocol to work properly (having experienced a powerful car with a stuck throttle, this is definitely something I want the computer to catch).
I also isolated what about my 1/2 torque experiment (described above) made the problem worse. Turns out it was the Aircharge to Torque table. Decreasing it decreases Actual Torque and PT voltage, while increasing it increased Actual Torque but not PT voltage. I ended up leaving it stock, but kept the Y axis scaled up for more aircharge. Oddly adjusting the Airflow to Torque Factor tables oddly didn't seem to do anything.
It's more like there's a limit to the amount of sonic airflow the PCM will request, but this limit varies with RPM. That doesn't really make any sense, but I don't know how else to interpret my findings.
With all the time I've sunk in to learning and experimenting with this PCM, I feel like I could have just decoded the car's CAN bus protocol and used a standalone... I'd already learned a bit of it when trying to get the car to run with a manual transmission.
Last edited by Grant; 01-18-2021 at 10:56 AM.
(Quote Grant) It's more like there's a limit to the amount of sonic airflow the PCM will request, but this limit varies with RPM. That doesn't really make any sense, but I don't know how else to interpret my findings.
Here is a table you dont have defined...but it is in there.
Its from a 6.4 but your numbers (6.1) will be very similar.
Looking at your log (airflow), it tracks these numbers very closely.
You will have to go to a bigger TB or make it read less airflow to improve it.
If you are getting the lightning bolt and limp...
either make the airflow adapt limits larger or go to sensed map enabled so it no longer uses calculated map.
Last edited by Hemituna; 01-20-2021 at 07:18 PM.
Turns out I do have 39020 Airflow Max/RPM. Those figures are similar to my limits at lower RPM, but are way too high at higher RPMs. It is definitely not strictly an airflow limit. I halved my displacement and stoic ratio. The car drove exactly as it had before I did this, despite airflow getting cut in half.
I'm able to fix the problem by setting the Small and Large TB airflow tables to stock, but keeping my scaled TB Airflow Model because:
I wish I'd known this sooner! I had no idea my PCM was using alpha-n instead of its perfectly good MAP sensor, until some fiddling with the TB tables resulted in a huge WOT lean condition last night. I guess I'll experiment with both modes and see how well it drives. As far as I can tell, the sole purpose of the TB Airflow Model table is calculating sensed MAP.
39020 Airflow Max/RPM & 14064 Peak Airflow WOT are two different tables.
39020 is the max limit (WOT), you can can do whatever you want with this table but it will still follow the 14064 tables numbers during PT.
If you have the 14064 table available and raise it, PT still does not change just the total numbers go up at WOT, then needing an even higher limit in 39020.
Its as if the PT airflow numbers are embedded in the firmware related to engine type and capacity.
Anyways, fight it if you like but I have found it best to sneak under it and you can get it to do exactly as you need.
Don't worry about seeing the TB fully open, look at the MAP and if low down in rpm you can get to 100kpa, then that is enough TB opening at that rpm.
Tweak the airflows enough (smaller) to get it in the range you need (only above .5-.75V) and you shouldnt mess up too much other stuff lol.
Last edited by Hemituna; 01-21-2021 at 01:52 AM.
(Quote Grant) As far as I can tell, the sole purpose of the TB Airflow Model table is calculating sensed MAP.
There is a whole other table available in later OS's for MAP estimation that is used for calculated map.
Obviously when you mod the engine you mess this up hence why we sometimes go to sensed map.
Why do they use calculated map?? Switching between the two, calculated map is softer (filtered), probably better for emissions/economy.
@hemituna Thank you for your detailed information sharing on this thread. I've read it at least 10 times and have seen good results as I continue to experiment with the airflow and demand tables. One part that I still haven't been able to connect yet is the relation to Driver Demand Torque and Expected Torque. In one of your example logs (002) they track nearly perfectly. In my log they don't aligned until there is a WOT pull, and then they sync up pretty close.
Where should I be looking and experimenting with to align the Driver Demand TQ with Expected TQ? Overall my car feels much better, but there are still some spots where it feels like the TQ is not aligned to what the car will allow.
Log below.
Thanks in advance.
Maybe I can shed some light on this. From your first post there...it isn't that the PCM has a limit, but rather physics has a limit. For any pressure ratio across an orifice/throttle/hole below 0.528, the air flowing through it reaches the speed of sound. And then it doesn't get any faster. Even if you slammed the throttle closed at 150mph in a hard decel and brought your MAP down to like 15kpa with a 100kpa baro...the air would still move at the speed of sound through the throttle. Because this ends up being a super reliable bit of physics, it's used as a way to calculate MAP in Chrysler software, or in GM's case, airflow.
MAP is calculated as well as measured with the sensor for a few reasons. One, if you characterize your throttle body and have a speed density model of the airflow that the engine is consuming, then you are able to calculate MAP. Comparing it to sensed MAP will tell you if the throttle flows slightly more or less than it should, as in manufacturing variation. It can also pick up on a throttle that is getting old and dirty and doesn't flow as much air anymore. Then that info can be used to update throttle flow adapts, which are needed to keep the baro calculation accurate. There are actually separate ways of calculating barometric pressure too, one obviously being at sonic throttle flow where the engine airflow is low. The other is during near-WOT throttle positions where the throttle restriction is minimal. And if there really needed to be anything else to calculate...those two different baro values are compared, and if the baro calculated during sonic throttle flow is higher than WOT baro, then the intake air filter is getting dirty.
Sort of seems wild how circular all of it is, but it all somehow works lol. This redundancy is used very often for rationality diagnostics.
I will post a log up in a bit. I have a 2020 Magnuson supercharged Gladiator with the 3.6L. i have experimented with drastcally lowering and shapping the airflow tables. It works, but it brings with it a host of other changes that make the calbration a total package. If you are looking to stop throttle osscilation and get good power this works. The fall out, not necessarily because of this but revieled by it:
1. You will need to rework injector mass and acceleration enrichment as now you can easily transition into boost part throttle
2. Idle control. The low air flow numbers seem to throw of min torque and idle tables. So the closed loop on drive idle and pn idle needs adjusted pretty extensively along with min driver demand torque in neutral
3. Variable camshaft tuning. Now that boost is part of you part throttle range. There are areas in the camshaft map the shift the camshafts quite a bit as you near 100kpa. I assumr NA this is not noticeable as this is happening while nearing max power with lots of airflow and max torque. In a boosted state you are at part throttle and this cam movement is super noticeable just cruising around. It can be tuned up but it is a task.
4. Throttle gain and control in the boosted zone. This will help the throttle from doong what I call a fly away when you shift gears.
5. Torque demand table. This seems to drastically help with part throttle tunning and power delivery. I cannot seem to get to greedy or the throttle gets chunky. In higher RPM once squarely in the boost camp it seems fine to set it at a reasonable spot that fueling can support
6. Pedal demand. This will smooth it all out. Ton of work and a bit of a black art.
7. Transmission, enuff said
My biggest issue right now, as the jeep is coming up to temp and then sometimes, what seems random at the moment. It will pull all the timing for a VERY quick bit like 10 milliseconds. No shift or anything is happening. Just a quick annoying tap you can feel and all is back to normal. Anyone seen this or have any ideas?
Well I answered my own question.
Now it takes way less throttle to accelerate as such the shift maps are a little out of wack. The pedal is in a mich lower spot for the output shaft speed, so the trans is trying to execute shifts early or at the wrong time, and it cuts timing to do so. I went and did some mapping and I can get rid of these dips.
I have installed a 105mm TB and having some idle issues and what I call recovery when coming to a stop. I have adjusted the TB air flow table and the small and large range. Also adjusted the air charge to torque and the torque to air charge. CS is a little rough but the TB response and drive ability is better. I had my car tuned and 20 minutes after leaving the shop started getting the lightning bolt symbol as some call it. Tuner said bad TB so I went from a BBK 90MM to a VMP 105 since I had to buy a new one and it didn't take 3 minutes with the new TB to get the lightning bolt......frustrated...did not go back to tuner.....trying to figure out myself..........was told by a guy air isn't right...made several changes, i guess I haven't hit the right area yet
Depending on how much the airflow has changed, you will need to look at the min torque tables, in the demanded torques section. The values are to high the idle will stumble quite a bit. If you cannot decrease the torque down enough, then the idle control values will need tweaked, this can get pretty daunting. I started with idle speeds for the different driving situations and once that was close, I ended tweaking the PID values a bit, but that is where it gets to be an interative process.
While I have been overhauling and modifying my Sprintex I decided to see how the stock tune and TB model/opening worked while being N/A.
I made a comparison at a low rpm (1750) of both cases with the factory TB airflow table values. This clearly shows the issues with the stock TB model once the manifold pressures rises above baro, eg the TB opening gets clipped very early and is the main reason why the TB airflow tables need to be tweaked significantly to "trick" the system to allow better PT drive-ability and transition to WOT when boosted. Currently this is the only way to improve it.
TBComparison.JPG
At 1750rpm NA the throttle opens up to a max of 2.03V which is significantly more than the 1.03V once boosted. Yes the accelerator pedal voltages are quite different, but in either case its the maximum allowable TB opening under PT conditions until the pedal voltage hits the WOT threshold.
Again this was tested with the same TB airflow tables. On page one, I showed that even after significant reductions of the TB airflow model I was only able to reach was ~1.7V TB at 1850rpm and that was met with other issues such as the "cruise control effect".
Just thought I'd share. Would be great to know whats actually missing in the TB model that requires tweaking for it to respond correctly under boost.
Reply With Quote