Discuss all things tuning in this section. News, products, problems and results.
Trying to work thru a hole. Engine stumbles hard and dies with minimal load between 2K and 4K. Little bit about the setup.. 4 Rotor, Twin Garrett G42-1200 turbos, Haltech Elite 2500 with REM, 3 Stages of ID2000 injectors, 102MM throttle body, OEM CAS crank trigger, Turbo Rotors, Full Turbo Perifial port, Widebands at outlets of each turbo.
Free Revs nicely it only falls apart once the engine gets a load on it.
Fuel Axis is RPM and TPS , Lambda is RPM and MAP, Ignition is RPM and MAP, Boost is RPM and MAP
I'm not familiar with Haltech, but posting datalogs or even screenshots of logs showing the relevant channels might help people help you. It could be as simple as an acceleration fuel setting, some systems have the acceleration fuel as a multiplier and different values might be needed for different loads.
Try different values for injector time/phasing, this might be especially important if all your injectors are mounted far from the intake ports. With 3 stages of injectors, watch the actual pulsewidths for all the injectors in the datalog and see if the stumble is related to the second or third stage activating. At part-throttle and atmospheric pressure a rotary engine shouldn't need more than 2000cc worth of fuel per rotor. Watch fuel pressure, add a fuel pressure sensor for the ECU to log if you don't already have one.
On engines with a lot of overlap, the O2 sensor might be showing leaner than the actual combustion if unburned oxygen goes straight out the exhaust port (not through the engine) during overlap. If the spark plugs look like the mixture was very rich but the O2 sensor doesn't show that, the O2 readings might be skewed by overlap. People worry about detonation but you're not going to hurt a rotary from running it lean at 3000 RPM in vacuum. Don't try to run too much ignition timing, rotary engines don't need much spark advance in the first place. Search in the HP Academy webinar archives if you have access, there's at least one video about rotary engine ignition timing and trailing split.
I've tuned a PP 4 rotor turbo drag engine many years ago and they are a little tricky to tune due to the effective port overlap. This is more problematic I found at part throttle (which wasn't a big deal for us). I set the ECU up conventionally with the fuel table being MAP based and this worked pretty well.
I'd suspect that your issues may be related to injector staging. You need to log what's happening around the problem areas with regard to measured AFR/target, and the injector pulse width on each stage of injectors. As Scott has mentioned, often you can't entirely trust the measured AFR due to the overlap and in this case I would suggest trying a wholesale change to the VE table in this area to see how the engine responds.
Would you recommend a more traditional MAP based instead of the Alpha N? Have been working with a few tuners to get my hands around this and unfortunately EVERYONE has a different thought on it. I have worked a little bit with Haltech on this and they scrapped the Alpha N and reworked to total Map based. Still no drive-ability and fell to pieces once clutch let out. The gent I'm primarily working with believes Alpha N is the only way to get there. Unfortunately with the profile this build has many of the tuners available that have rotary experience may be less interested in protecting the engine and more about big numbers.
I will work at getting some log files pulled and posted. I know that will help with getting guidance. Have been working thru issues of a new one off build and fuel system has been giving me fits. Faulty parts can be the devil.
All 3 stages are located close to the plenum first two on top of the runner and 3rd is underside. AFRs at 99% of the no load rpms are very good and consistent. Its once load comes on it falls to pieces. To the point where i dont even want to go to the dyno with it, partly as its all i can do to load on the trailer let alone try and get some ramp runs to develop a good VE table.
I can say this from binge watching webinars posted here I have a COMPLETELY different view on tuning this.
You need to go to a load bearing dyno, that will help you sort this out quickly. OK, I see from your intro post you are in Texas. Consider taking your car to someone with a 4-wheel hub dyno (T1 Race or perhaps check with Dynotronics). You can start with holding the engine at a steady RPM, and figure out what it wants for fuel as the load changes.
Good Luck -- Sounds like an awesome project!
David I would love to get it on a dyno. Have tried to just get some simple straight line low MPH driving in and thats been a challenge. Im in South Texas so Cali is a bit of a haul to go to the Dyno.. :)
I agree with the suggestion to get on a dyno so you can load the engine and slowly sweep through different load cells of the fuel map without worrying about acceleration fuel. In some cases, spending a lot of time getting the quick free-rev behavior dialed in before the steady-state throttle behavior can be counter-productive. It's best to get steady-state tuned before touching the acceleration fuel settings, or your accel fuel settings might be overcompensating to mask bad steady-state numbers.
I prefer Alpha-N (with MAP compensation) for cars that need part-throttle manners, especially if they don't pull much vacuum at idle. MAP-based should work also, really it's best to check with people with Haltech experience to see if one mode works better than the other for the ECU you're using. If your drivability problems seem load-related, like 10-20% throttle feels fine driving in first gear but too lean at the same RPM in 3rd gear, it's possible the MAP sensor isn't doing a good job of detecting engine load. In that case, you may need to compensate by running a bit too rich at light loads to avoid running too lean at higher loads. It's also possible the fuel injector locations in your custom intake manifold just aren't doing a good job of mixing fuel with the air. Look at the inside of the runner and make sure the fuel injector openings aren't shrouded by the mounting bungs. Try to have your first stage injectors pointing toward the intake ports, angled so the fuel spray doesn't go straight into the opposite wall when airflow is low.
I can't imagine any rotary engine would have good manners using 2000cc/min injectors for primaries, especially when mounted far from the ports. Try temporarily replacing the injectors closest to the intake ports with something that flows in the 400-800 cc/min range, that should be much easier to get started with. You can swap back to the 2000's later if you need that much fuel.
I haven't been able to get the 4 rotor periph that I'm involved with on the dyno to tune yet however I have done my fair share of peripheral 2 and 3 rotors as well as every currently viable port configuration on a rotary (I think), HOWEVER, my boosted experience is limited to side ports so please take whatever I say as a suggestion rather than gospel.
That said, I've found that injector timing plays a much much larger roll in partial throttle drivability on large, and more specifically light weight rotational assembly, peripheral ports as well as the fact that they need far more fuel than you'd expect at very low load. By a lot. I also suspect that this is a situation that has been glossed over or secreted in the past by the rotary tuning community 'gurus' since configurations like this have historically been in all out race vehicles with not insignificant budgets rather than the current trend of crazy billet everything shoved into a street car (still big budget, but generally self built).
We have worked with the injector timing a little. With the injectors far away from the ports its a bit of experimentation as well as a large handicap. Space in that area is a premium. The inserts that are in the plates are solid so putting a primary in the side port is out. Runner lengths are all within 1/4" of each other and the intake plenum comes from a popular machine shop that has several cars over the 1500hp mark using it. Its got machined in velocity stacks at each port and a machined in ramping system for the MAP sensor port to help minimize turbulence around the sensor. I have to say this may just be my bad for wanting to run a PP in a engine thats not in an application where its at 50% throttle or more all of the time. I thought with good tuning this would be possible.
I have done some simple tests to see how much fuel hangs out in the runners. With the injectors turned off the engine will start and run for about 4 seconds. I think a majority of this is from fuel pooling in the bottom set of injectors as they point upwards and do have a small pocket around them. I have tested which stage of injectors are running at idle and low boost. I may need to experiment with this a little more.
That's a lot of fuel wetting out on the walls if you ask me, though with time I'm sure you'd be able to deal with it. I'd still try for Alpha-N as has been mentioned by Scott above on a peripheral port.
Still looking for the needle. I have attached the ECU file as well..
MAP correction table.. Seems Off.. Shouldn't there be little to no correction in the vacuum sections? Isn't this to help compensate for the manifold pressure reference that ISNT used in the fuel map? Mainly boost ?
Do you have any logs you could share to watch what the engine is doing vs what the ecu is doing vs what you potentially think its doing?
I will see what I can round up.. Car is in the trailer and have no plans to pull it out until I have space in the shop again. I'm in the process of replacing the torque tube as the first one I built isn't rated for the full RPM the engine can produce. I had plans to rework some other systems while the driveline is out as well so it wont be in startable condition for a few weeks.
I will pull some logs when I get it out of the trailer to put in shop.
I did notice yesterday that 3 and 4 rotor EGT was about 200deg cooler after the engine fully warmed up at idle. Might have a fouled trailing plug there. Car idles awesome, just sits there, never loads up.. its when you touch the throttle all hell breaks loose.
How much work has been done with the transient fuelling? One thing that stands out is how much fuel is being dumped in, since you mention its been through a few tuners i'd imagine several thing have been tried. Regardless when I was first getting my elite 1500 dialled in the transient fuel was horrible in the base map and it seems to be a common problem amongst rotors running Haltech, big misfires on tip in, and you've got additional fuel on top. I came across a thread from a bloke that happens to be running a 4 rotor PP that really helped get me close. HERE.
Yours is a totally different beast to mine but I ended up with a "load acceleration dead band" of 800% and pulling additional fuel out of the "enrich rate" table for fast throttle movements. Enabling disenrich was a must for me too. Old mate with the 4 rotor ended up with something like 2000% on the load acceleration dead band and adding synchronous fuel etc. Post #6 is where it gets interesting. No guarantees that's your issue but worth looking into