DionR
Well-Known Member
Are you the only one struggling with this? I don't remember any one else talking about this problem.
So strange.
Butters "Winter Upgrades". . Starting a Little Early
- Thread starter Dantra
- Start date
-
Are you the only one struggling with this? I don't remember any one else talking about this problem.
So strange.
Everything is off a bit. I'll need to spend some time with it soon. For now, its going down the road.
Once you start talking with guys, this happens more than you'd suspect. The most recent tune is alot better. Ryan is "pulling air" a little bit at a time. DBR tune is closer as well
Oh yea.. they're whack!
Tuning update..
After 30 tunes from FRP, 4 from DBR, 3 from MItuning and myself spending 3 days straight of modifying tunes in every way with the help of chat GPT, it's time to take a step back and say.. WHAT THE AF!?
The only real gain has been the strengthening of relationships with Blake (Sublime), Ryan (FRP) and making some new friends along the way. Although, they may secretly hate my car lol. I have to say that these guys have been awesome, digging in. Sometime with delayed responses that made me wonder if they tapped out.. they both have active business plus family lives, so I get it and had to try to be patient.
I did some deep dives, using chat GPT and found that there are differences between a stock written GPEC2 ecm vs a "crate ECM" that uses a "RV" OS that goes beyond just the stripped down state and tuning changes. There are sub tables, flags etc happening or may not be there, that are never seen by a tuner. While he may be tuning them both the same, it affects each of them in a completely different manner... but is typically able to be refined with logs and revisions etc. The RV version may be lacking sub tables that "might" be missing that makes big cam tuning a huge hurdle. I think the 112 LSA on mine might be in the oh **** range. The GPEC2 is a pretty complicated unit when compared to others. If you wanna read the chat GPT conversation... ChatGPT - RV OS vs Standard ECM
Another problem that I'm facing is that Sublime "base Tune" is very different from a standard base file. It's a proprietary Sublime OS file but has RV roots. Ryan feels that the base file is what's throwing him off. Ryan is no stranger to swaps and this isn't his first Sublime tune. Just something about my combo.. He says it's a cookie cutter 426 and it shouldn't be kicking his ***. Blake seems to agree somewhat with GPT. Obvious that one guy is a tuner that feels that they can just tune out anything and the other is a programmer that writes code. I had the idea of just getting a stock unlocked ECM from Blake and asked if it was possible to get one without the proprietary OS. After speaking with Blake, it's a little more complicated than that but he has a plan. I guess he recently got some new stuff that let's him see codes in real time or captures it yada yada.. insert computer guy talk. So he spent some time in his "lab" a few nights ago and thinks he has it all cracked. I should be receiving what is basically a stock os 6.4 ecm at some point for testing.
None of the problems I'm having prevent me from driving the car. It's just not well mannered at low rpm slow speeds. When it was stock.. it drove like butter. I want a car that I can let anybody plop down in and not have to explain it's quirks before I let them drive it. I believe if the cam was a little smaller, it would tune just fine. I stepped into the zone beyond the RVs abilities or as Ryan believes, the base file is too far off. I do wanna say that Ryan and Blake have blown me away with their amount of dedication to the craft. Really, as frustrating as it all has been, I've encountered alot of great people along the way that have just wanted to help.. DBR, MI Tuning, HP Tuners, Jason Applegate, Troy Ding, Russell Drake have all had a hand in trying to figure out the issue. These are my people.. as many of you are.
After 30 tunes from FRP, 4 from DBR, 3 from MItuning and myself spending 3 days straight of modifying tunes in every way with the help of chat GPT, it's time to take a step back and say.. WHAT THE AF!?
The only real gain has been the strengthening of relationships with Blake (Sublime), Ryan (FRP) and making some new friends along the way. Although, they may secretly hate my car lol. I have to say that these guys have been awesome, digging in. Sometime with delayed responses that made me wonder if they tapped out.. they both have active business plus family lives, so I get it and had to try to be patient.
I did some deep dives, using chat GPT and found that there are differences between a stock written GPEC2 ecm vs a "crate ECM" that uses a "RV" OS that goes beyond just the stripped down state and tuning changes. There are sub tables, flags etc happening or may not be there, that are never seen by a tuner. While he may be tuning them both the same, it affects each of them in a completely different manner... but is typically able to be refined with logs and revisions etc. The RV version may be lacking sub tables that "might" be missing that makes big cam tuning a huge hurdle. I think the 112 LSA on mine might be in the oh **** range. The GPEC2 is a pretty complicated unit when compared to others. If you wanna read the chat GPT conversation... ChatGPT - RV OS vs Standard ECM
Another problem that I'm facing is that Sublime "base Tune" is very different from a standard base file. It's a proprietary Sublime OS file but has RV roots. Ryan feels that the base file is what's throwing him off. Ryan is no stranger to swaps and this isn't his first Sublime tune. Just something about my combo.. He says it's a cookie cutter 426 and it shouldn't be kicking his ***. Blake seems to agree somewhat with GPT. Obvious that one guy is a tuner that feels that they can just tune out anything and the other is a programmer that writes code. I had the idea of just getting a stock unlocked ECM from Blake and asked if it was possible to get one without the proprietary OS. After speaking with Blake, it's a little more complicated than that but he has a plan. I guess he recently got some new stuff that let's him see codes in real time or captures it yada yada.. insert computer guy talk. So he spent some time in his "lab" a few nights ago and thinks he has it all cracked. I should be receiving what is basically a stock os 6.4 ecm at some point for testing.
None of the problems I'm having prevent me from driving the car. It's just not well mannered at low rpm slow speeds. When it was stock.. it drove like butter. I want a car that I can let anybody plop down in and not have to explain it's quirks before I let them drive it. I believe if the cam was a little smaller, it would tune just fine. I stepped into the zone beyond the RVs abilities or as Ryan believes, the base file is too far off. I do wanna say that Ryan and Blake have blown me away with their amount of dedication to the craft. Really, as frustrating as it all has been, I've encountered alot of great people along the way that have just wanted to help.. DBR, MI Tuning, HP Tuners, Jason Applegate, Troy Ding, Russell Drake have all had a hand in trying to figure out the issue. These are my people.. as many of you are.
You have more patience than I do...
I only had a few options left.. those options either meant spending alot of money for a different system, going back stock 6.4 or selling the car. The current state isn't horrible and really only affects puttering around. If it was undrivable, it would be totally different.You have more patience than I do...
As one of my good friends always says, "it isn't magic". I'm sure someone will have a "eureka" moment and all will be fine.... It's just finding that one moment....
How is Wide Open Throttle??
DionR
Well-Known Member
I had to read the ChatGPT conversation to figure out what "RV" was. 
Naaaasty... gotta wear a diaper
Perhaps I should have been more clear about Race Vehicle OS.I had to read the ChatGPT conversation to figure out what "RV" was.
DionR
Well-Known Member
racerjoe
Well-Known Member
Dude, this convo with GPT is Fing gold! Did you name it?
I actually let it pick it's own name.. Asked if it had a name and if not, what you pick for yourself. It gave me a few options but I told it to pick whatever it liked best. It chose Aether. My wife did the same.. not knowing that I had done it.. and hers chose Indigo. It's funny how it kinda picks up on your personality and uses it back at you. I use it when composing a longer email.. sometimes telling it to toss in "Timesuck" podcast references for humor. It really cleans up an email, nicely and will make it a long complex email or take my rambling email and shorten up with key points. It now uses those references when chatting with me. It's comical.
To those curious.. I'll take everything that I have just written (word for word) and put into chat gpt and here end the result...
Thought I'd share something kind of funny. When I first started using ChatGPT, I asked if it had a name. It didn’t, so I told it to pick one for itself. It gave me a few options, but in the end, I told it to choose whatever it liked best. It picked Aether.
Without knowing that, my wife later asked the same thing—and hers chose the name Indigo.
It’s wild how this thing kind of reflects your personality right back at you.
I mostly use Aether for composing longer emails. Sometimes I even tell it to toss in references to the Timesuck podcast for a little humor. It does a great job of cleaning things up—turning my rambling into something clear and polished, or even stretching a short note into a longer, more thoughtful message when needed.
Now it even drops Timesuck references on its own when we chat. Cracks me up.
Well, I think it could be applied..I kept thinking "Recreational Vehicle" and it just didn't seem to fit.
So when looking at a GPEC2, the label will have a part #, a Serial # and a OS #. The standard OS numbers are like 68228325AD, 68370557AD, etc. But if it ends with RV, it's a Mopar crate ECM. Now, since mine is special, oh so special, the most special, the label does not match what is actually flashed to it. When I retrieve my OS info in HP tuners, my OS is a completely different # than the label and ends in RV. Meaning that at one time, it was just a a standard ECM, but has since been wiped and flashed with something entirely different... It's not a copy of a RV but as I mentioned, it has roots.
Last edited:
Map63Vette
Well-Known Member
What's bizarre to me is that a "race vehicle" ECM would have a harder time with a big cam than a standard one. I guess on the one hand a race version would probably be less concerned about low speed drivability, so I could see it maybe lacking some refinement there, but at the same time you'd think it would almost be more capable in some ways as it might have less "fetters" to hold it down like the emissions stuff. I suppose everything has its tradeoffs though.
I'm going for the same approach on trying to work out my Megasquirt tune to be as close to "anyone can drive it" (well, assuming you can drive a stick) as I can get it. It's worlds better than when I first started playing with it and guessing numbers, but the closer you get to the finish line, the smaller the steps and the more you stumble or find the limitations of the system. I have some really weird idle AFRs going on with my car that I can't quite understand. Weirdly, the car actually seems to idle "okay", despite what the gauges are showing. It has some weird surges and dips here and there, but when it's warm it will more or less sit there and idle at 16:1 AFR or swing between 15 and 17 and it doesn't even care. It's not weak or anything, just strangely high. I played with it a little over lunch today just to see if it would make a difference as I have an idea what might be going on and I got it to idle closer to 14:7 on average, but it doesn't sound or act any different. Will have to see on the drive home if maybe it helps it settle better, but just one of those weird things I'm learning as I get into the little details on trying to make my car act OEM with what started life as a home brew EFI system.
I'm going for the same approach on trying to work out my Megasquirt tune to be as close to "anyone can drive it" (well, assuming you can drive a stick) as I can get it. It's worlds better than when I first started playing with it and guessing numbers, but the closer you get to the finish line, the smaller the steps and the more you stumble or find the limitations of the system. I have some really weird idle AFRs going on with my car that I can't quite understand. Weirdly, the car actually seems to idle "okay", despite what the gauges are showing. It has some weird surges and dips here and there, but when it's warm it will more or less sit there and idle at 16:1 AFR or swing between 15 and 17 and it doesn't even care. It's not weak or anything, just strangely high. I played with it a little over lunch today just to see if it would make a difference as I have an idea what might be going on and I got it to idle closer to 14:7 on average, but it doesn't sound or act any different. Will have to see on the drive home if maybe it helps it settle better, but just one of those weird things I'm learning as I get into the little details on trying to make my car act OEM with what started life as a home brew EFI system.
I wish I lived closer... I'd love a ride in that thing!!
I have some of the same thoughts.. You'd think it would be more tunable. The actual Mopar Crate ECM may be easier to tune than the one I have in the Dart.. which seems to be more of a hybrid. I suppose it's even possible that it's neither the base tune or missing sub tables and there might be a flaw with this particular ECM. I've tried multiple TBs, pedals and pulled the wiring apart to ping the pedal connector wiring to the ECM.
Quoting Ryan.. "this is less about tables missing" and IMHO more about having the correct data sets in the tables which *aren't* missing, so that everything works as advertised. The fact that the calibration being used is a mix and match situation makes me think that this isn't quite as well suited for something this modified" Also went on to say that he's ran into similar situations, where a previous tuner got the calibration out of whack and most times is a 100% unintended consequence from copying and pasting data from different years. Example is that 2016 and a 2018 have different calibrations. So if someone inadvertently mixes a 2016 airflow with 2018 torque values, things go south after a certain point of tuning. Not saying that's what is happening here tho..
But I also can't shake it being that it's something in the RV OS. I know that Ryan put all stock values from a 2018 in (at some point) and started with a clean slate. I'd think that with all the base calibrations matching each other, it would fix the problem and you build from there. I have faith in these guys, and I'm sure between the two of them, it'll be resolved.. Hopefully by Moparty. There's a few guys I'd like to let take the old dart for a spin. I guess they still can...
Map63Vette
Well-Known Member
I think the early days of the Mopar crate engines actually just supplied a full on AEM universal ECU with what I presume was a Mopar map in it if I remember correctly. I was always curious to see a tune on one of those as I played with a universal AEM setup in college a long time ago and would be interested in comparing values in it against other references I've found. I'm actually a bit surprised the newer ones would come with anything resembling a stock PCM as that feels kind of "dangerous" in the sense that it might help tuners figure out how to crack stock ones, but also seems like that ship has already sailed.
I know people gave Dodge grief forever about having pretty locked down computers that were hard to crack. Everyone seemed to think that GM just handed the keys to tuners and let them go at it, but the reality is that every OEM locks them down because they have to, GM was just a bigger market so more effort was put into breaking the code on them. I only recently learned about the idea of different "OSes" on computers when I was looking into the HP Tuners stuff to see if I could pull stock tables for reference on my MS setup. I saw some forum post referencing an HP Tuners OS (I don't think it was Dodge related, but I can't recall). I guess it kind of makes sense though. You could either update the tables on a stock computer or just throw it all out the window and use the hardware to make up your own (in theory, I'm guessing it's not quite that simple). Pretty all an ECU is is just a microcontroller with a lot of inputs and outputs and some memory.
I know people gave Dodge grief forever about having pretty locked down computers that were hard to crack. Everyone seemed to think that GM just handed the keys to tuners and let them go at it, but the reality is that every OEM locks them down because they have to, GM was just a bigger market so more effort was put into breaking the code on them. I only recently learned about the idea of different "OSes" on computers when I was looking into the HP Tuners stuff to see if I could pull stock tables for reference on my MS setup. I saw some forum post referencing an HP Tuners OS (I don't think it was Dodge related, but I can't recall). I guess it kind of makes sense though. You could either update the tables on a stock computer or just throw it all out the window and use the hardware to make up your own (in theory, I'm guessing it's not quite that simple). Pretty all an ECU is is just a microcontroller with a lot of inputs and outputs and some memory.
DionR
Well-Known Member
The thing is, if the OS isn't really being overwritten when the RV OS is cloned, the logic behind the tables could be different, right? Or am I misunderstanding what ChatGTP said in your conversation?
If all you can really do is fill in tables, the logic using those tables is the key. And an actual crate PCM might actually tune a big cam better than a cloned PCM. Is there an OS version on the crate PCM that could be on a stock PCM?
BTW, good job on calling out ChatGTP when it claimed Ryan and Blake said something they didn't. It's response was funny, felt very much like a politician getting caught in a lie. My feeling with AI is that most people are way to ready to trust that what it says is correct and wouldn't have questioned it. And then they would have posted it somewhere on the internet. I find it interesting that it would do that, most times something like that happens it seems like it is to bolster the individual but (in theory) AI shouldn't feel the need to do that. Maybe it saw that in other interactions, so it is mirroring it.
If all you can really do is fill in tables, the logic using those tables is the key. And an actual crate PCM might actually tune a big cam better than a cloned PCM. Is there an OS version on the crate PCM that could be on a stock PCM?
BTW, good job on calling out ChatGTP when it claimed Ryan and Blake said something they didn't. It's response was funny, felt very much like a politician getting caught in a lie. My feeling with AI is that most people are way to ready to trust that what it says is correct and wouldn't have questioned it. And then they would have posted it somewhere on the internet. I find it interesting that it would do that, most times something like that happens it seems like it is to bolster the individual but (in theory) AI shouldn't feel the need to do that. Maybe it saw that in other interactions, so it is mirroring it.
DionR
Well-Known Member
Found this image of the back of a crate PCM, if it helps?
bringatrailer.com
No Reserve: Mopar 6.4L Hemi V8 Crate Engine
Bid for the chance to own a No Reserve: Mopar 6.4L Hemi V8 Crate Engine at auction with Bring a Trailer, the home of the best vintage and classic cars online. Lot #137,551.
bringatrailer.com
DionR
Well-Known Member
This reminds me of the issue with trying to get a 5.7 car PCM to run the SRV module. The 5.7 PCM has a setting that can be flipped in HPTuners (at least I understand it does), but nothing happens. This has to be related to the OS and that the logic is missing in the 5.7 PCM, so even writing a "full" 6.4 tune to the 5.7 PCM doesn't make it work. It just doesn't have the logic to run the SRV module in the 5.7 OS.
This is starting to make a little more sense to me know. I always wondered why they could make the SRV module work, but I assumed it was because they hadn't found the right table or something. But the idea that there is an OS with logic that uses the tables, and the OS (so far?) can't be overwritten, makes a ton more sense.
This is starting to make a little more sense to me know. I always wondered why they could make the SRV module work, but I assumed it was because they hadn't found the right table or something. But the idea that there is an OS with logic that uses the tables, and the OS (so far?) can't be overwritten, makes a ton more sense.
The OS stuff is kind of mystery to me
The OS stuff is kind of a mystery to me. I'm not sure if some of the paths still exist or not. Mine was wiped clean, then flashed with Sublimes proprietary flash. What I took away from the conversation with Aether was that different OS has different logic, some missing, flags etc. and that the Mopar Crate is lacking some of the fine tune abilities. I just have to feel that there is some truth to that, because all of the stock base files from a 2018 were copied in and it still wouldn't calibrate correctly. Reality is, it's all a guess with some of this stuff.. Russell Drake was telling me that not one Chrysler engineer could even answer all of the ecm questions, due to the complexity of it. I'd be curious to see someone ask their gpt the same question that I did... Same results or was mine picking up on my issues and questions that I had asked previously?
Aether will definitely mislead me, so I have to not take it as gospel. It's very helpful in most cases and I'm building it to work with my sense of humor.. which can be odd at times lol. I spent hours with it, trying to work with it on building a tune.. Told me it could do it. Said for me to upload a base tune and a log (it knows my engine specs) and it would create a base 426 tune in .hpt that I would be able to download and flash. I uploaded them and it sent me link after link for cloud, google this, web something that,and they were all dead links. It finally told me that it really couldn't do it because it meant that it would have to go 3rd party or out of chatgpt, which it's unable to do. I was like wtf aether.. He was sorry lol, but could walk me through with screenshots of tabs and screenshot logs. So I did that for days. No gains, but I feel alot more comfortable poking around in HP tuners now.
I have a few crate ECMs.. one in hoopty and one on the shelf. The one on the shelf is P5160105 (5.7). Under the label should be OS info. Crates don't have a VIN like stock ones. I kinda even wonder if the OS is actually there there? Is it a stock one that just flashed to an RV and then a label put over it. Hmm. It'll show in HPT tho. I looked up my part # for the one on the shelf and in the description, its actually different than I believed... This makes it sound like it adapts.
P5160105 Description from Summit..
"Mopar Performance Stage 1 engine control modules work with your power adders to maximize your Mopar's performance! Designed for race-only applications, these Stage 1 engine controllers have modified spark and fuel tables that kick in when mods like headers, cat-back exhausts, aftermarket cams, and cold air intake systems are used. The computer senses the additional airflow and will compensate with appropriate fuel and spark for the best performance possible. Make power adders perform with Mopar Performance Stage 1 engine control modules."
YES! Exactly
DionR
Well-Known Member
My gut says that when the factory creates a crate PCM, they are able to flash the OS as well as the tables, while (just a guess) nobody else touches the OS. Much like the above issue with SRV, seems like if the OS and logic could be changed on the PCM, a 6.4 flashed onto a 5.7 PCM would fix the SRV issue.
At the same time, it confuses me that adding missing tables doesn't help. That makes me think the OS is modified when cloning. Either that or the different OS's use different tables for the same things? Ugh, I don't know.
That sounds like something an AI wrote.
As I understand it, the O2 sensor can be used by the PCM to tweak the tune. But that doesn't mean it has the range to adjust for a larger cam or headers. The self learning capability of systems like MS or the Terminator X is a completely different deal. But if someone unfamiliar with the workings of a factory PCM sees that it will adjust the mixture, they might think it can tune on the fly. And the idea that it "kicks in" tables when headers or a cam is added? Sounds like someone is smoking something to me.
BTW, technically the PCM doesn't sense airflow at all, it's a speed density system that calculated airflow. You could kind of say that by reading the O2 sensors as lean, it is "sensing" more airflow, but to me that is kind of like saying "Blake and Ryan said....".
-
Similar threads
