VE Lower Than Expected (and more)

Hey all, my apologies for the pretty basic (and yet convoluted) question but I've been working on sorting out my tune some more and the VE table in general just seems a bit low.

Idle is in the high 20s to low 30s, it peaks around 80 at 5k rpm, and then drops into the 60s towards redline. Presently this is the primary concern, although some other possibly related issues I'm having are a weird bump in 5k rpm region under mild vacuum, a high tendency to lean stall coming back to idle (temporary band aid is the fuel bump below the idle region, doesn't work great though), and occasional extremely rich (pegged rich on lambda, 0.44ish) on decel.

This car has given me quite the headaches and I'm getting close to out of ideas. Something seems off in general with the fueling system but I can't figure out what. Fuel pressure compensation seems fine, I'm using the MaxxECU provided settings for the ID1050Xs I'm running, etc. Just not sure where the issue is.

The engine in question is a boosted 12A on MaxxECU running a custom fuel injector adapter on top of the factory carby intake. It's weird, I don't like it, but it's what I've got for now. Transients are also a problem with this setup (likely thanks to the great distance fuel has to travel from the injectors) but that's not my present concern.

Any help appreciated. Thanks.

Ben,

If you add a fixed 20-25% to all those numbers they don't seem so odd, and IDs should be well characterized, so a few potentials come to mind.

Have you done a compression test? Maybe it's just hurt.

Are all the coils working well?

Any chance the injectors have debris in them, aren't flowing as they should any more? (lots of people don't install proper filtering)

Any chance the fuel properties aren't accounted for accurately at the moment? (stoich point etc.)

I haven't used that ECU, but do you perhaps enter engine displacement in a way that requires fudging it for a rotary?

I have other ideas, but I think that's a decent bunch to start with.

Hey Mike, thanks for the ideas! The numbers do seem reasonable if adjusted that 20-25% or so like you said, which is why something definitely seems amiss in either configuration or hardware. Regarding your questions in order:

I have not done a compression test in a while, although the engine is pretty fresh and still runs and starts great. Weak rotary engines are usually pretty tough to start when hot and this one has no issues. I'd like to run another test although I have misplaced my rotary engine tester for the time being.

Coils should be doing great. They're IGN1As and I've never had any significant issues with them. 4.5ms dwell, holds power great to redline, don't get hot, etc. Not sure how else to measure their health but they seem fine.

Injectors could be one potential culprit and have been on my list for some time. I am running a 10 micron filter before the injectors, although the injectors sat for about six months when the car was down and I never had them cleaned after that. It has been something I've been meaning to get done, I just haven't gotten to it quite yet. If they were clogged with debris though wouldn't they be more likely to flow too little fuel? Or could it be either way?

Fuel properties are another thing I had wondered about, although the difference between E0 and E10 should be minimal I would think. Maxx does have the option to change the fuel type (which I currently have set for E0) although I haven't touched that and I wasn't sure if I should. Pump gas is theoretically around E10, so perhaps it would be more accurate to change the setting? The difference in AFR is only 14.1 vs 14.7 so I don't think it could account for the full error present.

Maxx ECU has a rotary engine type option, so no number fudging should be required. I have the exact displacement for my engine entered (1146cc for the 12A) and so there shouldn't even be a rounding error present.

One other thought that I had was my fuel pressure mapping. The ECU is tracking fuel pressure instead of assuming it's 3 bar all the time. Fuel pressure doesn't seem to do anything odd in the logs, and I checked the calibration for my sensor and it should be fine, but it was another idea nonetheless. Sensor in use is a Honeywell MIPAN2XX100PSAAX.

Thanks again for the ideas. Hopefully one of my responses might click for some possible solution.

Little update, I tried changing the fuel type to E10 (compared to E0) and it made the car run even richer which is the opposite of my issue presently. If anything this means my VE is even further off than expected.

I also did some reading on the MaxxECU pages and found that they collect their injector data using ethanol rather than the more common heptane. Supposedly this is good for about a 5-10% difference in flow data, however I don't know which direction that would factor. It also wouldn't add up to the full error present.

Ben,

Honestly the stoich point change may seem a step in the wrong direction, but any time you can change a setting so it's more correct, it narrows down where improvement can be made, and then when everything is working as expected it will be more clear.

I haven't used that ECU yet, but are all your lambda targets set in lambda or AFR? One quirk of stoich point adjustment is if targets are set in AFR, you may have to adjust them after changing stoich point, if that changes how an AFR target table works. One of the perks of working in lambda is not having to worry about that.

Injectors do flow differently on gas vs. ethanol so I would say you're likely getting a bit more flow than expected IF the characterization was performed on ethanol. That said, I don't know that ECU so do they perhaps compensate flow as you change fuel settings i.e. stoich point? Do you enter specific gravity etc. in that ECU?

In terms of fuel pressure, is the fuel pressure target set to the differential pressure the vehicle operates at? If you're running say 3 bar differential pressure, but the target is 3.5 bar, then you may have a constant positive fuel correction occurring. You could disable active fuel pressure comp for a moment just to see what the impact is, confirm it's minimal, re-enable it, then continue your investigation.

Again I'm sorry I've not used MaxxECU, so are you absolutely sure that when you select rotary engine type, you can enter displacement that way? I ask because I've had ECUs where you can select something indicating it's a rotary, but you still enter double the displacement.

Hey Mike,

Regarding the stoic point, I was talking with a tuner friend of mine and he suggested leaving it alone for the time being. I know it's technically more correct have it set as E10, but to avoid adding more variables to the mix I think it's probably best. Most tunes for pump gas I've seen also tend to assume stoic is 14.7:1 anyway so it matches convention better this way. For this particular ECU all lambda targets are in lambda, you don't even have the option to work in AFR. That said, changing the stoic point still affects the fueling calcs as I noted.

As far as the injectors go I may have actually made a bit of a breakthrough, although we'll see how it pans out. I have four ID1050X injectors on this setup running in a staged configuration. Previously I was using the "ID1050" injector preset, although I noticed there's now an "ID1050XDS" preset. After chatting briefly with the folks at ID I did confirm that all ID1050 series injectors should use the same data, so what I can't explain is why switching to this new preset made VE fall better in line with my expectations. My theory is that Maxx's "ID1050" preset is actually for the ID1000 but I can't confirm that yet. I've actually been suspicious of the ID1050 injector preset for a long time but eventually gave up messing with it until I saw the new XDS preset.

Anyway, after changing to the new preset fueling requirements increased 10-20% across the board, putting VE closer to where I was expecting to see it. Idle VE is now around 36 and the table tops out in the upper 90s. It's perhaps not the full 20-25% difference you had mentioned, but it's certainly a lot closer. Lambda also seems perhaps a little more stable at low pulse widths but that could totally just be placebo as well.

For fuel pressure my target is pretty spot on to the actual differential pressure. Target is 300kpa and in the logs I generally see right around that, and at lowest it's dropping into the 280s on occasion. Any sort of pressure compensation should be little to none, although I can double check it.

And no worries regarding the lack of experience with MaxxECU, there's too many out there to know them all. I am 99% certain that rotaries are supported without needing to fudge the displacement. I've never heard of this being an issue on the platform and I know multiple folks who run them on rotary engines.

I'm glad this sounds like progress.

1050xds only flow about 10% more than ID1000 at 3 bar, so I wouldn't expect that to make a 20% difference in higher IPW areas where both injectors are linear and dead time is a small percentage of total on time. Then again if the characterization info isn't visible, their changes could be more drastic.

In terms of the stoich point, setting it correctly should always be the right move unless there's a software bug.

Perhaps now you're largely down to the flow difference between ethanol and gas.

Definitely progress. It also seems to be helping some of the exaggerated shape that was present in the VE table. In particular I'm noticing it wants a smoother transition and less fuel around the 60kpa row instead of the weird shelf that used to be present there at lower rpm.

Conveniently I wasn't seeing changes of 20% in higher IPW areas, that large change is mostly present in the low end. If we compare the VE cell at 7000rpm and 200kpa it used to be 71.7 before the preset change and now it's sitting at 82.7 (about a 13% change). This still notable difference at higher IPW could also be the result of the second injector stage being active at that time so the error would be compounded. Either way it seems within the realm of plausibility. I'm still dialing in the rest of the table as the shape has changed some as noted but overall things seem to be happier.

Figure I'll leave the thread open for a little while longer yet in case anything pops up, but so far this is seeming like the magic bullet. Hopefully the discrepancy is just down to the flow difference between ethanol and gas now as you said.

Awesome, sounds great!