Discuss all things tuning in this section. News, products, problems and results.
I was reading through the MoTeC PDF on tuning flex fuel with M1 ecus, they obviously want the VE map initially tuned on 93oct.
My question is when im doing the drivabilty and base boost tuning on 93oct what should I do if I develop knock while optimizing cam advance? This car sees E85 almost always but i still want to do my due diligence in setting up the flex fuel strategy correctly. Would it be wise to get a rough idea of what the cam advance optimization will be using E85 and then switch back to 93oct and simple adjust the timing accordingly?
2JZ vvti 3.4L 10:1 compression
In the flex fuel guide there are sections about setting the ethanol charge cooling gain and ignition timing compensation. Keep those in mind. Here is a way to do it that I prefer:
1. 93 octane VE table first with VVTi zeroed out and lower boost.
2. 93 octane startability and cold idle/warm up optimization.
3. 93 octane spark timing and boost optimization with 0 VVT.
This is your baseline calibration from which the rest of the calibration process is based (high temperature, higher ethanol concentration, VVTi active, etc).
Now you add features.
4. Boosted VVTi calibration on 93 octane including spark and VE changes. 93 octane calibration is complete now except high temperature or low octane conditions. Part throttle VVTi is time consuming and optional. Its a small fuel economy benefit.
5. With zero VVT switch to fuel E85 (not an intermediate blend). Calibrate charge cooling gain per the Motec guide.
6. Calibrate startability and warm up on E85. This could take a while.
7. Calibrate spark and boost, with zero VVTi.
8. Calibrate boosted VVTi including spark and VVTi changes. Again part throttle VVTi is optional.
9. Switch to E40 and take logs to check your compensations and startability.
If that sounds like a lot of work, well it is. This is basically the process used on OEM flex fuel programs. That's how you get a really thorough flex fuel tune with a boosted VVT engine. There are areas where you can make educated guesses in the maps and basically skip steps to reduce work load. It's easy to trip yourself up though if you don't do separate VVT and non VVT tunes, and separate tunes/checks for the different ethanol blends. Its ultimately your judgment call as to how much of the above plan to truncate.
While I am not a Motec certified tuner(yet), I am in several other systems. Not only 2j's being one of the most I have tuned, I do tons of VVTi and variable cam apps.
If I read your question correctly, you asked about what to do when/if you develop knock while tuning optimal cam advance?
If so, type of fuel in a proper system should not change your ve. Cam timing however does. Hence why its impossible to properly tune a variable cam with SD. If your constantly altering VE by moving cam timing then you'll chase your tail in your SD map. VE is not a fuel table, its a airflow model that you, the tuner plot. Timing on the other hand is absolutely affected by type of fuel. So if you have developed knock mapping your vvti tables, while on 93, then you simply are running too much Ignition Timing and need to pull some out or target a richer afr, des eq or lambda and possibly both. Which ever you are using.
You have simply shown yourself that you have not completed either tables for tuning. If and engine has become more efficient and requires less timing due to the changes in your cam timing, you simply need to adjust accordingly.
I understand all that. Let’s say I’m tuning base boost on 93 and I’m adding 10 degrees of cam advance after every pull to determine when the advance isn’t producing more torque. Will I be able to optimize the cam advance on base boost or it’s it possible that the I will hit the limits of the fuel before I do. And if I do hit the limit of the fuel then it’s impossible to build a proper VE map since on E85 I’ll be able to run more cam.
Garth, we met at Sound Performance once a couple years ago during a lay over you had.
What is your base boost/spring pressure?
So one scenario is, you have tuned the VE table with 0 cam advance (step 3 in my baove post). For step 4 you advance the cam at each rpm until you hit knock limit, and then stop advancing, leaving spark the same. So in that way, the max VVT advance is limited by the fuel. But there are a lot of combinations of VVT and spark and boost to see what the best torque is, which is why VVT is so complicated. Typically on a WOT curve for a VVT system of that era (intake only, maybe up to 35-40 degrees advance max) the VVT runs at max at very low rpm and then tapers down to zero around say 4000rpm for best VE. Then you repeat the process on E85, step 8 above.
The other way to do it, and this is how stock ECUs work, is you map out the VE for each cam position (doing WOT first as part throttle is low priority). So you make a VE table with 0 advance as mentioned in step 4 above. Then for step 4 you set the entire WOT area of VVT to 5 or 10 degrees, adjusting the VE table up or down as you go (expecting VVT to go down at medium to high rpm). You end up with a bunch of logs and a series of VE tables that you will have to track in Excel. With all the data you've collected you basically have to merge together all the optimum VVT and VE settings together. Then you repeat/check the VE mapping process on E85 (step 8) and check/confirm the interpolation of fueling on E40 (step 9).
In theory if you map the VE and VVT correctly on E0/E10 and aren't knock limited, you only have to tune the charge cooling gain for E85. But to your point, you might be knock limited, and you run yourself into a corner trying to trade off spark, VVT, and boost on E0/E10. However the ECU doesn't have a separate wall film model for E85, so that can add an extra complexity to it as you now have extra fuel sticking to the walls (mostly with a cold engine though).
Stock ECUs (think like a Honda K series engine for Civic Si, or any modern Ford) basically have separate spark and VVT tables for each combination of cam position, and then use an interpolation algorithm instead. The problem is that even a Motec hasn't caught up to a Honda ECU from 20 years ago, and you're stuck with 1 VE map despite all the possible cam positions.
I remember you for sure man. I didn't even see the screen name when I replied. I got hooked into doing the engine building course. Lord knows after all these years I needed a refresher. I keep finding my self in the tuning section reading unanswered questions.
There is zero complex about vvti at all. Its simple mechanics. How much extra time can you give the intake valve to draw additional air until it makes no positive difference. That is it! And if you go too far past 25 deg(mech limited around 30), or attempt to rather you will cross your crank and cam trigger throwing the ecu to a fit. I think if you use the rear cam sensor its not a possibility. Cant recall. But there is no benefit beyond its limit, I tried a few times.
To better answer your question now let me put it like this. Bear with me as this overall general rule of thumb and is the norm not the odd.
All the advance from a pedal stab and once you have reached peak torq the following factors will determine from a mechanical point of view how to proceed.
So... A pee shooter with say 264 cams on a vvti 2j with stock intake if peak torque occured at 4700, I can almost guarantee you'll need to be at 0 cam advance by 5200 for the best curve
On the other hand lets say S2 or better, 7675, vvti 2j will roll cam advance to 0 much slower.
Generally speaking the turbo is the biggest factor of how when you begin rolling the advance out and then at what slope to end up back at zero.
When you crack 1k ft/lbs at 5000 with a 2j and no spray, you sir, have won the battle.
Best advice I can offer is to trust your dyno with cam advance and be methodical. It will not change much amongst boost levels and fuels. The important part is it be optimized when you want all of it, not when your going to get groceries. If you do not have a nice linear ve and trq curve you have a issue. Whether it be tuning or mechanical something is up, find it and fix it. Unhappy motors don't tend to last.
Keep me posted Bob. I will certainly be looking at your progress. I am certain you'll be a lot of help to people here.
Are you referring to tables given by hondata to incorrectly tune a variable cam car in "SD" and comparing to a Motec while suggesting they aren't decades ahead of the epa designed ecu strategy found in a factory car from 20 years ago?
I think the best explanation for what I'm talking about is what Cobb has written about the factory Ford Speed density and their simplified speed density VE tables which are designed for intake and exhaust VVT. This explanation of the OEM system is here:
You can see that it uses a series of VE tables (well, tables and modifiers etc) depending on VVT position. In this case it's more complicated than a 2JZ because of the wider VVT angle limits on newer engines and the fact that these Ford engines have intake and exhaust VVT rather than just intake.
Tuning the Ford OEM speed density models explanation is here:
Cobb has a modified/simplified speed density system to bypass some of the complexity of the stock Ford control system. In that VE mode there are VE compensation tables based on VVT position.
Scroll down to the Cam Timing (VCT) Tuning section https://cobbtuning.atlassian.net/wiki/spaces/PRS/pages/96481299/Ford+EcoBoost+Speed+Density+Guide#FordEcoBoostSpeedDensityGuide-_Toc311462550__RefHeading__3610_178852463TuningSD–AdditionalTopics
I don't think Motec has implemented VVT compensation tables to the VE maps at this time, or allowed for separate VE maps based on VVT position. So the ECU can't automatically adjust VE as you move the VVT angles. Therefore you have to be really careful not to end up chasing your tail as you move the VVT around, as the relationship between throttle angle, MAP (either boost or manifold vacuum/pumping loss), VE, and torque changes every time you adjust the VVT.
Let's try and really simplify things here as this doesn't need to be too difficult.
When you're optimising the VVT cam mapping you're trying to optimise the cylinder filling or VE of the engine. With this in mind you're going to optimise the VVT map and then leave it alone - It will want to be the same on pump gas or E85 (the only caveat here is that if you're running maybe 15 psi on pump and 35-40 psi on E85 then it's possible that at higher boost levels that you can't reach on pump gas that the VVT may then want to be altered but essentially in all the areas of the map you'll run on both fuels the VVT will be identical).
If you strike detonation while tuning on pump gas (which is probably quite likely with 10:1 CR) then you need to either reduce the ignition timing, boost, or perhaps both.
Coping with fully varying cam timing for a given combination of load and rpm is a real PITA on a SD system as you'll need multiple VE tables. In my own opinion in the aftermarket this is totally unnecessary and just adds additional complexity to the tuning.