When you click on links to various merchants on this site and make a purchase, this can result in this site earning a commission. Affiliate programs and affiliations include, but are not limited to, the eBay Partner Network.
ahh I didn't realize port mod was something only he did.
that is a shame.
It "looks" simple. It is just the square circuit board in the middle left area of picture above that matches green of EBL board. It has some leads that get soldered onto the board in a few spots.
After 11 years of running open loop with wideband for reasons explained in my alternative intake build thread, I have installed a heated NB and activated closed loop. Initial drive last night went very well. BLMs didn't seem too far off but that isn't a huge surprise with my VE tables having been tuned with my WB previously. Learns were making progress especially at idle with the last learn showing 0 adjustment. Idle is very smooth, seems smoother compared to when running OL even while using base bin PRP and other CL settings. Seemed to be pulling fuel for the most part, and my wideband appears to have been reading a little off. In closed loop I am seeing consistent mid to high 15 afrs on the wideband while NB and BLMs show on the rich side. Likely due to my leaky Y-Pipe issues I mentioned in my build thread.
With it being so long since I have messed with closed loop I am a bit foggy on some of the details. Have been reading up to jog my memory but wanted to ask a few specifics.
I noticed I was coming out of closed loop on decel. Every time this happened my INT and BLM would reset to 128. Is this proper function or did I miss a parameter when reverting back to CL? I know about the open loop decel checkbox and have unchecked that to give that a shot on my next drive. But in case I have surging issues under decel, I am curious on the reseting INT/BLM being a correct function when coming out of CL. With it being a stick shift that means it was in and out of CL frequently therefore resetting to 128 frequently.
PE mode. My WOT was being tuned with the WB previously (obv) and was pretty close with AFRs in the mid 12s. Should I put PE mode inactive or with the VE table in that area being "close" will PE mode work correctly? My spark advance is all from the main tables with no adders btw so keeping PE active should only affect fuel.
With a heated o2 sensor, I should be safe to reduce some of the parameters to get into CL sooner correct? I lowered the CTS threshold from 50*C to 40*C and the warm timer from 30sec to 20sec. Haven't touched the cold timer which is at 120 sec.
Edit: One more question and probably a dumb one. E10 vs Non-Ethanol Stoich. I've never really concerned myself with this as I was running open loop and targeted lower 14s for my OL AFR for a majority of the cells. Does the narrow band sensor swing on "stoich" for whatever fuel is in the tank or do we need to change the parameters to target 14.3 vs 14.7 via the mV parameters?
Tonight I unchecked open loop decel. Was pretty impressed with how smooth it ran on decel in closed loop. I see no reason to keep it in open loop if it continues to deliver the results I got tonight.
At this stage my next steps are getting my exhaust leak on my passenger header/ypipe joint and other portion of the ypipe repaired so that my wide band is more trust worthy. My drivers header is sealed nicely and I have more faith in the narrow band located there. I have an appointment scheduled for Tuesday at a exhaust shop to fix the ypipe. From there I will continue to dial in my VE tables. It has been taking a decent bit of fuel out so far and continues to run smooth. I have noticed on cold start up my idle hunts for longer than it used to, I made sure and confirmed my VE table in that area was smoothed out after the learn I did tonight. Hopefully that is all it was, if not I will move on to messing with choke settings. I am thinking with the fuel being pulled, I have leaned out cold starts beyond where it likes to be. Warm idle is very solid at 800 rpm and 45kpa.
I think I sell the second unit as is. It runs great besides the lockup, and the quad chip needs a replacement. Where is a good place to sell it? I think 200$ is a reasonable price.
P/s. I have instructions from BobR RIP what and how to fix the unit.
yariv, I’d be interested in buying this from you. Reach out to me.
kristofferwickstrom@gmail
So I have a question about the VE learns.... one day I'll hop on the closed loop train but for now it's open loop season lol. My question is is when do you feel like the VE learns start having diminishing returns? Sometimes it feels like I'm better off just datalogging and making adjustments then. Will reducing the smoothing factor from 5 to 1 make smaller adjustments for fine learning?
So as someone who just recently moved to closed loop from open loop after ~10 years of sporadically tuning the car I will give you my thoughts...
Yes, it pretty much always felt like I was chasing my tail doing OL VE learns. That is the nature of the beast. Do a tune in the morning when temps are lower and then do a tune in the afternoon when temps are higher and you will basically erase the mornings changes. I installed a heated o2 and instantly noticed improvements. I was actually seeing 0s on learns for the first time in forever. I even activated closed loop decel expecting surge issues...nope it decels great even engine braking with the 5 speed! I can't believe I put up with open loop tuning for this long. I have zero intentions of ever looking back. It will only continue to dial in too now that I feel like I can confidently head in a forward direction.
So as someone who just recently moved to closed loop from open loop after ~10 years of sporadically tuning the car I will give you my thoughts...
Yes, it pretty much always felt like I was chasing my tail doing OL VE learns. That is the nature of the beast. Do a tune in the morning when temps are lower and then do a tune in the afternoon when temps are higher and you will basically erase the mornings changes. I installed a heated o2 and instantly noticed improvements. I was actually seeing 0s on learns for the first time in forever. I even activated closed loop decel expecting surge issues...nope it decels great even engine braking with the 5 speed! I can't believe I put up with open loop tuning for this long. I have zero intentions of ever looking back. It will only continue to dial in too now that I feel like I can confidently head in a forward direction.
I'm sure this has been covered but how do i go from open loop to closed loop? and I know exactly what you mean. i had a decent tune set up yesterday morning and did another learn and its like Im starting from scratch. Also, when adding AE, do i go for positive or negative numbers?
There are several ways to lock into open loop. Best way to see what has been done is compare your tune in tunerpro to say 3006 base tune. Then look for any parameters that are related to closed loop. Most likely you have the option flags for open loop idle/decel/cruise checked. Then you might also have the CTS threshold for closed loop maxed as well. You will need a narrow band sensor. Heated is recommended. I used the AFS74 sensor from Delco.
For AE, the pulse width tables a larger number is more fuel. Then the filter tables affect the duration and magnitude of AE. And finally the multiplier tables do just that, act as a multiplier to the final calculated AE (TPS and MAP AE combined).
There are several ways to lock into open loop. Best way to see what has been done is compare your tune in tunerpro to say 3006 base tune. Then look for any parameters that are related to closed loop. Most likely you have the option flags for open loop idle/decel/cruise checked. Then you might also have the CTS threshold for closed loop maxed as well. You will need a narrow band sensor. Heated is recommended. I used the AFS74 sensor from Delco.
For AE, the pulse width tables a larger number is more fuel. Then the filter tables affect the duration and magnitude of AE. And finally the multiplier tables do just that, act as a multiplier to the final calculated AE (TPS and MAP AE combined).
I use the same heated sensor myself, so that's covered. i played more with my VE tables last night and was able to richen things up. What I'm tuning is a TPI ZZ4 heads and cam swapped 92 k1500. The goal isn't full on power but mileage and reliability. It my winter rig so I'm trying to hurry and get it ready lol
You should easily be able to run closed loop with that setup. That cam is just slightly smaller than the XFI 280 cam I run. Closed loop will gain you mileage and reliability. Reliability will be gained through consistency in the AFR/tune. Can even give highway mode a try once you get VE dialed.
You should easily be able to run closed loop with that setup. That cam is just slightly smaller than the XFI 280 cam I run. Closed loop will gain you mileage and reliability. Reliability will be gained through consistency in the AFR/tune. Can even give highway mode a try once you get VE dialed.
i actually run the 280 cam in my LT383 lol. I'll definitely look into getting closed loop down. So i should dial it in first in open loop then try for closed loop operation?
Probably could do it either way. Get it close on open loop then jump into closed. Or just start closed. Work your way into high load areas of the VE tables slowly because mine ramps up very fast unlike the base files.
It really breaths some air at high load. I would imagine the massive throttle body on the FIRST intake probably plays a role in this table shape too. At 3000 rpm I can hit 80 kPa easily at 50% TPS. Worth noting that I used to have a hump around 1500 rpm in this table before going closed loop. I could manually smooth it out and it would always returned. Running closed loop that has since been learned out without me touching it. This table has basically smoothed itself to what you see the last handful of tunes. Here is the last open loop ve table I had before switching to closed loop.
Since last night I'm getting a notification that says "OVERFLOW ON VE 1700 rpm 70 MAP. what does that mean? Also, when I let off the throttle it seems to run super fat for a touch in the 20-25 map areas. If I lower the VE tables in those areas even one point it goes super LEAN in those areas. What tables do i need to modify for this so I'm not going back and forth?
Last edited by robertfrank; Oct 29, 2024 at 01:32 PM.
Are your VE tables hitting 100 in any cells? That would be about the only thing I can think with overflow, can't say I have seen that notification.
Regarding the off throttle areas, it is tricky. I have been actively looking at this area right now as well now that my VE tables are coming together finally. Being a stick shift car, there is a lot more throttle movement compared to an auto. Typically on a casual shift I see a rich drop on throttle off and then a further drop on throttle application. My next test run will incorporate more deceleration enleanment and less acceleration enrichment for low delta tps/map areas to see how it does. You might be experiencing the opposite. The VE table doesn't necessarily handle the transitions, but more so the steady state. Granted you still want a smooth VE table, because a rough table can make transition tuning a nightmare.
Are your VE tables hitting 100 in any cells? That would be about the only thing I can think with overflow, can't say I have seen that notification.
Regarding the off throttle areas, it is tricky. I have been actively looking at this area right now as well now that my VE tables are coming together finally. Being a stick shift car, there is a lot more throttle movement compared to an auto. Typically on a casual shift I see a rich drop on throttle off and then a further drop on throttle application. My next test run will incorporate more deceleration enleanment and less acceleration enrichment for low delta tps/map areas to see how it does. You might be experiencing the opposite. The VE table doesn't necessarily handle the transitions, but more so the steady state. Granted you still want a smooth VE table, because a rough table can make transition tuning a nightmare.
the area in question (70 MAP 1700RPM) is at 76.86 so i don't have it at 100. it also gave me a VE OVERFLOW at 65 MAP 1800rpms with the graph saying 76.08
On the AE, you don't want it to interfere with the closed loop fueling. That's another way to start chasing your tail.
In 8D there's a parameter that says "Reset Integrator when AE in progress". I have that box checked (the DA2/DA3 LT1's do as well). For the transient conditions during an AE event, you don't want the ECM confusing that for steady state fueling rich errors and start yanking fuel back out.
With my other thread about playing around with the O2 filter coefficients and stuff, I'm wondering if that helped with the overall AE as well... The INT is the short term fueling correction, but the O2 sensor will react before the INT does, so that may have helped with what feels like a much sharper/crisper feel to the AE.
just a thought that I read somewhere, something about duty cycle of injector vs rpm being a limit factor. is it possible it's warning you about duty cycle being high even thou ve isn't at 100? Just a thought, seems like I remember you can log duty cycle and tbi injectors being picky about it. if you're tbi, I only guess because I saw it's in a truck.
on ae and de and this is way to far back to remember details but I was working on a old accel dfi 6, super easy to use ae and de on, live 2 d graph, showed you where it was live and you moved it up and down to tune. but I thought I needed a flat line when I lifted off the gas and when I hit the gas. someone I believe on this board said to go check a log from even a stock stick shift car, the af is all over the map every gear shift. I had a buddy with a stock t5 305tpi car and we logged it. it wasn't flat at all. so at that point I took their advice and it basically close your eyes and tune the ae to what feels the best driving. it's going to spike, but usually richer feels a better. so you end up using less throttle movement each shift. even my newer obd2 car you can see every shift. seems like a waste of fuel in theory but I guess it's not enough for the oems to tune perfectly either.
the EBL itself is telling me that there's VE Overflow
Is it possible to screenshot this? I've never seen that notification before.
Originally Posted by ???
just a thought that I read somewhere, something about duty cycle of injector vs rpm being a limit factor. is it possible it's warning you about duty cycle being high even thou ve isn't at 100? Just a thought, seems like I remember you can log duty cycle and tbi injectors being picky about it. if you're tbi, I only guess because I saw it's in a truck.
on ae and de and this is way to far back to remember details but I was working on a old accel dfi 6, super easy to use ae and de on, live 2 d graph, showed you where it was live and you moved it up and down to tune. but I thought I needed a flat line when I lifted off the gas and when I hit the gas. someone I believe on this board said to go check a log from even a stock stick shift car, the af is all over the map every gear shift. I had a buddy with a stock t5 305tpi car and we logged it. it wasn't flat at all. so at that point I took their advice and it basically close your eyes and tune the ae to what feels the best driving. it's going to spike, but usually richer feels a better. so you end up using less throttle movement each shift. even my newer obd2 car you can see every shift. seems like a waste of fuel in theory but I guess it's not enough for the oems to tune perfectly either.
Good points, I am sure many people get into the mindset of a perfect solid AFR no matter what. And with a stick shift being so transient with the throttle, I just don't think that is really something that is necessary or even achievable for that matter. "Give the motor what it wants" comes to mind.
You are definitely running out of room in your VE table. Like Byron mentioned, fuel pressure could be a issue. Or you could just need to adjust your BPC. Please attach the bin that was created from that learn.
This is the actual BIN I run. What I've been doing is running learns but instead of loading them in the EBL, I technically use the learns as a datalogger, go back to my "main" BIN and look at what the Learn changed and make even smaller changes to fine tune the VE tables manually. at this point I pretty much ignore anything -3 and below and modify -4 to plus or minus 12. What is weird for me is that looking at the VE table in TUNERPRO RT shows different numbers than the EBL does.
one other thing I've noticed is that it seems like the truck kinda runs hard and starts hard. i think that may be in my timing. Luckily my knock counts are pretty dialed in now. I'm getting a barely getting a few knocks now. I'm running 47psi with an acdelco ep241 pump and a Holley TPI regulator with Bosch 3 22lb injectors.I may upgrade to an EP381 pump soon
Last edited by robertfrank; Nov 7, 2024 at 04:11 PM.
Reason: added spk pic
Your learn is attempting to max out the VE table. Hence the warning of overflow. It is warning you that it cannot raise anymore.
I would try to adjust BPC to give your VE table breathing room then let the learn function try to do its thing. No need to make it that complicated. If you want to reduce the learn affect, lower the smoothing factor in WUD preferences. I personally think you should move to closed loop learns.
So I think i found the issue. I checked my BPC vs VAC and it was set to 45psi and 186.00. I'm running 47psi. I used the EBL utility and reset the table to 182.00. I'll see if that makes a difference.
Keep in mind you now also need to multiply the VE table by the difference in BPC. In your situation, don't always need to follow the utility perfectly. Because with you lowering the BPC, you would need to increase your VE tables which would only make the maxing out worse. Below is a quote from RBob discussing BPC when VE table needs more head room. Found in this thread https://www.thirdgen.org/forums/diy-...ot-tuning.html
Originally Posted by RBob
With TPI (multi-port) the "BPC-BPC vs VAC" table needs to have the same value throughout. As you increased this BPC table, the VE table needs to be reduced by the same percentage.
This gets the low and mid range back in line and provides more head room in the VE table.
Think of the BPC table as more of a centering device for the VE table(s).
I would go 10% higher BPC and reduce VE tables by 10% (multiply whole tables by 0.90). Then use the VE learn function to its fullest for a few drives to see what it is wanting to do with fueling. Will need to manually flatten low map low rpm area as idle comes together (see my VE table above).
Ok so I did what you had said and it worked out great. My next question is is when tuning in open loop, are the VE learns trying to force the ebl to tune in accordance to the "Open Loop AFR vs RPM & VAC tables? If so, what would a good table represent?
Correct. It reads the data from the wideband and tries to match the commanded AFR. Which in open loop is primarily coming from that table. With a handful of modifiers like the CTS multiplier, choke, crank, highway afr tables. At temp commanded should match the main table though unless in highway mode.
You set the table up to your needs. Most probably are around 14-15 during cruising/light load scenarios. Then idle at whatever the motor likes. And ramp down to 12.5-12.8ish from there for naturally aspirated motors at WOT.
I disabled PE mode when running a strictly open loop tune. I am not sure if EBL even activates PE mode when running open loop only, would have to do some reading to verify that one.
Fighting prop gains at both idle and off-idle. I've never ran a closed loop tune till recently so it's a new learning curve. VE tables are pretty tight. Were showing little changes during VE learns until started messing with PRP then that started playing with INT/BLMs. My only reasoning for even messing with prop gains is I wasn't seeing the typical "good" NB o2 swings that were commonly shared by Bob and others. Anyways, with increased PRP - Gain vs o2 Error I get good o2 swings but bad AFR swings that are felt while driving. Note, when I say I increased that table it is going from say 3 to 6 in the low o2 error areas and 4 to 7 in the higher error areas. That table is shown below too.
At idle I still lack in good o2 swings but my AFR is steadier and idle is smoother.
It is worth noting that my o2 constants at cruise are at a mean of 600mv with +- 50mv for rich/lean. My idle is 625mv with same rich/lean split. I am confused why my idle afr is closer to what I would expect with 625mv averaging 13.8ish. But at cruise my AFR is much higher in the 15s even though target mv are basically the same. Maybe it is just my PRP gains being off. Highway mode is definitely disabled.
Anyone have any thoughts or recommendations? Are the PRP gains just that sensitive? I can only adjust them +- a full digit. So the resolution isn't very high in that table. So If I try to add 10% it will just round back to 3 from the 3.3 value. If I go 20% it will round up to 4. Which is actually a 33% increase. With the actual gain equation being (gain vs o2 error)x(gain multiplier vs airflow) I am thinking maybe I can get the gain vs o2 error table slightly higher than needed then use the multiplier to dial it in in each airflow area. I assume the final value spit out from the equation isn't going to be rounded like the o2 error table?
I know in EBL we have the option to tune with BLM or WB.
What are the advantages of each?
I've been using BLM, but I feel like I'm going in circles tuning the VE. Should I be using WB?
WB tuning has the advantage of being able to tune where BLM tuning can't. Namely WOT. That being said, in the past I have felt like I was going in circles more so with WB tuning than with BLM tuning. I still have a WB hooked up but am now doing most of my VE adjustments through BLM and then manual changes to WOT area.
How are you going in circles exactly? Major changes back and forth in the learns? Like +6 one day -6 another day? Or +-1 to 2? As the VE table comes together you can adjust your smoothing factor in the WUD preferences to dial it back some. Typically start with higher factor than lower it to like 4 or so. Also keep in mind with the colder weather the engine will likely want to be adding fuel compared to two months ago. Or even just a week ago where I am located.
So I'm battling a few things that I'm sure you guys can help me with. I've noticed that I'm having too much residual fuel in my low map (20-25map) when i let off of the throttle. its causing super fat spikes then levels out. What tables should i adjust to dial that back a touch? if I take fuel from there its fine for the before said situation but if I'm trying to tune those areas with throttle I'm way too lean i.e. when going downhill with throttle. Also, I'm chasing knock like a ****. Attached are my current Main Spark table and the most recent learn I did showing what I'm seeing. I really don"t want to keep removing spark to not make the engine lazy. What are your thoughts?
I couldn't get your log to play, but looking at the knock you posted I would try to address that first. Is this a new build? Or miles on the engine that may have carbon buildup? Compression ratio? New knock sensor?
Check spark plugs for physical signs of detonation. You could have external noise causing those issues. I solved some exhaust leaks, exhaust hitting panhard bar, and other misc suspension noise and that cleaned up a lot of my false knock. I still have some in low map/low rpm areas though but keep an eye on my plugs for any indication of true knock.
As for the rich spike on decel. Look at your deceleration enleanment parameters. Note: that isn't decel fuel cut off (DFCO), it is a reduction in fueling at throttle decel to help handle that rich spike. If it spikes rich but then levels out then your VE tables are probably good but the DE isn't pulling enough initial fuel on decel.
I couldn't get your log to play, but looking at the knock you posted I would try to address that first. Is this a new build? Or miles on the engine that may have carbon buildup? Compression ratio? New knock sensor?
It's 170k former TBI 350 with a with ported ZZ4 heads and cam with a ported TPI intake with 22lb Bosch 3 injectors. I know my compression definitely went up. I'm thinking I'm around 10:19.1 according to the summit compression ratio calculator. So premium gas is all I can run. I also have Comp Roller tip 1.52 rockers which are a bit louder than stock.
Check spark plugs for physical signs of detonation. You could have external noise causing those issues. I solved some exhaust leaks, exhaust hitting panhard bar, and other misc suspension noise and that cleaned up a lot of my false knock. I still have some in low map/low rpm areas though but keep an eye on my plugs for any indication of true knock.
As for the rich spike on decel. Look at your deceleration enleanment parameters. Note: that isn't decel fuel cut off (DFCO), it is a reduction in fueling at throttle decel to help handle that rich spike. If it spikes rich but then levels out then your VE tables are probably good but the DE isn't pulling enough initial fuel on decel.
Is your current tune the active tune or the compare tune in those? The compare tune has both the DE TPS Threshold and DE Enleanment Factor changed, the active tune looks like base bin values. DFCO has nothing to do with your rich spike issue, so ignore those for now.
