Creating a Custom Throttle Cracker Calculated PID

[ArKay99]

I have my car running very well. Now that's it running well, I just HAVE to make it run better, right? One of the 'small' problems I have is what me and my buddy call, 'the hurky jerky'. It's a hesitation that occurs under very slight load between 25 and 35 mph,'cruising speed' and 1200 to 1500 rpm's. Sometimes it's very slight, sometimes it's quite annoying, like the engine just quits running for a revolution then snaps back on. When it's bad it bangs the rear end, when it's slight you almost can't notice it.
---
We've tried logging every table while driving and inducing this. It's not any spark, VE, MAF, yada, yada. The other day it was occuring and when it did I punched my clutch in and the engine went from 1300 rpm's to 1600, then quickly settled to 850. I thought it must be the Throttle Cracker. I pulled some air out of the table and went and found it was much less, but it happened in other places, like at 35 instead of 30 and 1400 rpm's instead of 1300. Also, while it was happening in another area when I punched the clutch in the rpm's would drop, or rise...
---
It definitely seems that throttle cracker is contributing to this, but just multiplying the whole table, or changing values by row or column isn't giving me the granularity I need and the table provides, so... maybe a custom PID?
---
I'm now guessing what PIDS contribute to the final airflow sum, or if that is the one I need to base it off of. I'm wondering if the PID should be based on IAC_SUM_DMA, or MAF. I'm thinking I could make a sort of 'BEN factor' map by using the measured airflow and the final calculated airflow, and subtract the contributing PIDS to come up with a 'BEN factor' to apply to the throttle cracker table. I am doing one now with IAC_SUM_DMA + (IAC_TF_DMA - IACDES_B). Not sure, only guessing at this point. I can't see the tables in the .prn file to know how the DMA's are being combined to form the calc'd pids. The only ones I have in sae_generic.txt are the few hp, kw, VE, and AFR customs.
---
I'll definitely report what I find, but I know there are some experts with this stuff here that may be intrigued by this and be willing to give me an educated guess as to where to start first or stop wasing my time and thought process. Also, if this technique works, there are a number of other tables that would benefit from this also, Throttle Follower being another possibly.
---
I just read the the link above The sticky above. (http://forum.efilive.com/showthread.php?t=4709) and it looks like GM.CYLAIR_DMA * 8 is what the map should be based on. Going on from here...
Ok, after more research here is what I came up with....
({GM.CYLAIR_DMA.cylair}*8)-(({GM.DYNAIR.gps}*8)-{GM.IAC_TF_DMA.gps}) , just thinking out loud...

[redhardsupra]

that's a damn good question, no wonder you got no replies :/
The bigger question here is which table is used to determine the airflow at idle and near idle:
1. desired airflow?
2. airflow parked?
3. MAF?
4. airflow calculated from SD calcs?

what role, and what other factors alter the idle airflow numbers? LTIT? STIT? Cracker/Follower stuff?

until we know exactly which value gets used when, we're shooting in the dark.


from my own experiements i know that in a regular MAF/SD hybrid mode, you gotta alter both, and they better both be yielding the same results. RAF is obvieously in use, because without it, the idle is rough. the suprising one is that in what we normally think of as the 'normal steady condition', at idle, VE is used in addition to MAF. if you dont believe me, jack up your VE and run it in MAF mode, it will either die promptly or surge like mad. so my guess is that the airflow at idle is a rather complex set of 'reality checks' and consistency comparisons, otherwise we'd be able to tune idle with just one of these tables.

any other observations out there?

[joecar]

I'm still reading and absorbing at the moment...

[SSpdDmon]

I have my car running very well. Now that's it running well, I just HAVE to make it run better, right? One of the 'small' problems I have is what me and my buddy call, 'the hurky jerky'. It's a hesitation that occurs under very slight load between 25 and 35 mph,'cruising speed' and 1200 to 1500 rpm's. Sometimes it's very slight, sometimes it's quite annoying, like the engine just quits running for a revolution then snaps back on. When it's bad it bangs the rear end, when it's slight you almost can't notice it.
---
We've tried logging every table while driving and inducing this. It's not any spark, VE, MAF, yada, yada. The other day it was occuring and when it did I punched my clutch in and the engine went from 1300 rpm's to 1600, then quickly settled to 850. I thought it must be the Throttle Cracker. I pulled some air out of the table and went and found it was much less, but it happened in other places, like at 35 instead of 30 and 1400 rpm's instead of 1300. Also, while it was happening in another area when I punched the clutch in the rpm's would drop, or rise...
---
It definitely seems that throttle cracker is contributing to this, but just multiplying the whole table, or changing values by row or column isn't giving me the granularity I need and the table provides, so... maybe a custom PID?
---
I'm now guessing what PIDS contribute to the final airflow sum, or if that is the one I need to base it off of. I'm wondering if the PID should be based on IAC_SUM_DMA, or MAF. I'm thinking I could make a sort of 'BEN factor' map by using the measured airflow and the final calculated airflow, and subtract the contributing PIDS to come up with a 'BEN factor' to apply to the throttle cracker table. I am doing one now with IAC_SUM_DMA + (IAC_TF_DMA - IACDES_B). Not sure, only guessing at this point. I can't see the tables in the .prn file to know how the DMA's are being combined to form the calc'd pids. The only ones I have in sae_generic.txt are the few hp, kw, VE, and AFR customs.
---
I'll definitely report what I find, but I know there are some experts with this stuff here that may be intrigued by this and be willing to give me an educated guess as to where to start first or stop wasing my time and thought process. Also, if this technique works, there are a number of other tables that would benefit from this also, Throttle Follower being another possibly.
---
I just read the the link above The sticky above. (http://forum.efilive.com/showthread.php?t=4709) and it looks like GM.CYLAIR_DMA * 8 is what the map should be based on. Going on from here...
Ok, after more research here is what I came up with....
({GM.CYLAIR_DMA.cylair}*8)-(({GM.DYNAIR.gps}*8)-{GM.IAC_TF_DMA.gps}) , just thinking out loud...

I think you'll find that equation will calculate negative numbers. Personally, I started out by killing the throttle cracker in the 1000 and 1600 RPM rows. Then, I tackled the surging with the throttle follower. Remember, cracker is mainly for off-throttle (i.e. decel). Throttle follower is where you want to tackle on-throttle problems.

Check to make sure timing isn't making huge drops in your base timing tables, which should be active for off-throttle application if you have an M6. From 1600 to 1200 RPM, timing shouldn't have a delta greater than 1* if you can get away with it. For example, my low airflow cells for 1600 are set to 24* and 23* for 1200. That means, timing is only going to change 2x in (0.5 degree increments) in that 400 RPM span. Limit the change and it'll help limit the herky jerky. :) Light on-throttle application should be close to this as well. Send me your tune (saj79 at aol dot com) and I'll give you my spark maps so you can see what I'm talking about.

Too much timing will add to the problem as well. If you're running an after market cam and have more than mid~20's timing commanded, it's most likely too much. I idle at 23* with a 232/238 cam and it's pretty smooth. The rest is in your Desired Airflow and idle learning tables.

[ArKay99]

Sorry, I've been away from this for the last week. I accidentally dropped one of the very thick black steel washers that come with the FAST manifold into the #7 hole and didn't notice. When I put it back together and cranked it, it lit right up but sounded like I ran over a pile of wrenches with my lawn mower. I shut it down immediately, but by then it was too late. I gouged up the piston, cracked the ringland, gouged the cylinder wall, and put a 2" crack in the cylinder wall. The thing that is driving me crazy is I'm so anal about this stuff.Everything has to be perfect or I won't move on. I've had the manifold on and off so many times I just got lazy and sloppy. Here is a link in case you are interested in looking further. How to ruin $3000 of machinery with a 10 cent washer. (http://forums.corvetteforum.com/showthread.php?t=1795340).
A look on the bright side. The motor is already out and I'm purchasing a Lingenfelter 403 along with some other goodies, big cam, at least for me 242/242 144LSA 112ICL, Tony Mamo is fixing and porting the heads for the 4" bore, etc. When the smoke clears from the install, I'm going to need some hand holding to get this up and running in a good way. I'll post up for advice when I'm ready.

[joecar]

Ouch... :bawl:...what are the odds of that washer standing upright like that...:bash:

[ArKay99]

Ouch... :bawl:...what are the odds of that washer standing upright like that...:bash:
I've been laying awake at night thinking about that. Kind of like throwing your change on the table and a nickel landing on it's edge and staying there. Also, it is a fairly large washer, so the intake valve on #7 had to be wide open or it probably would have jammed in the bowl. A series of unfortunate events...

[Chalky]

Sorry to hear about it. Is does make me feel a little better when I lose things when working on my car or generally screw up.

I feel lucky when I drop a tool under the car and I can reach it without a broom.

[2002_z28_six_speed]

Can't be of much consolation but here goes.

We all learn from our errors and they give us an excuse to buy something even better. :)