Sale ends todayGet 30% off any course (excluding packages)

Ends in --- --- ---

Engine goes lean if RPM drops on takeoff

General Tuning Discussion

Forum Posts



Tech Articles

Discuss all things tuning in this section. News, products, problems and results. 

= Resolved threads


Hi guys,

I have a highly modified classic VW Beetle fitted with a Haltech Elite 1500 ECU. I have the tuning pretty spot on except for one particular scenario. When I am attempting to takeoff, for example at 2000rpm, if the RPM drops at all, the engine loses most of its torque and attempts to stall if the clutch is not dipped for another attempt. During this event, the lambda goes lean to around 1.15. All of the cells that then engine transits through are tuned correctly and are on target in steady state operation, however if RPM is dropping through these cells, the engine goes lean. If I manage to keep the RPM constant while slipping the clutch, this does not happen and the engine torque is strong.

My theory is due to a larger camshaft, as RPM drops, the manifold pressure increases causing a similar scenario to a transient throttle event where more fuel is required. Has anyone experienced this, and any ideas on how to correct? My thoughts were creating a correction table where additional fuel is added with a negative RPM trend, or possibly switching the transient throttle trigger from TPS to MAP, however I'm not sure if that will create more problems than it will solve.


Sounds like you need more acceleration enrichment, so when you add more throttle (in response to RPM's going down), it has enough fuel to make more torque. Might just need the proper ignition timing (or more flywheel), so there is enough torque to not have the RPM drop.

Is the engine mapped with TPS as the load or MAP. If TPS, you might need to add a map compensation to deal with this.

As David said.

I assume, from your description, you're using an induction set-up with a single TB and a plenum feeding each cylinder - these are prone to the condition you describe, especially as the flat four will probably need longer runners and/or a larger plenum to reach the heads.

As I understand it, it's caused by the delay between the engine demand and the fuelling compensation. Using a suitably placed MAP 'should' help this.

If the engine is N/A, and you are using a single point TB, converting to "WEBER" style throttle bodies for each head should improve response and, maybe, power/torque.

Thanks David, ignition timing has been Dyno optimized to MBT throughout the range. The lean spike makes me think it's fuel related. The engine is tuned with MAP as load and Transient Throttle is triggered by TPS (the default option). Because of this, the transient throttle is not even being triggered when this happens, I am usually maintaining a constant throttle and using the clutch to modulate rpm. With a constant throttle position, the MAP increases as the RPM falls "off cam". Do people have any success running Transient Throttle triggered by MAP? I feel like it could fix this particular situation, as long as it also works well for all other transient conditions.

Yes you are correct Gord, it is a single 75mm DBW throttle body, followed by a series of Y-sections to split the intake to either side of the engine and then cylinders, so the runners are fairly long. The engine is Turbocharged so I am no longer running the Weber style throttle bodies.

My injectors used to be in a bad location about 8 inches up the runner and this condition was far worse, causing the car to stall almost every time. I have relocated the injectors to point at the back of the valves and it has made a great improvement. The event still occurs but the engine now receives the required fuel in time to prevent stalling.

I'm curious about your "suitably placed MAP" idea. I am using the Elite 1500's onboard MAP sensor, and reference is taken just after the throttle body. Can you explain how changing the location would cause a difference? I assumed that pressure throughout the runner would be balanced.

As a driver, I would request more torque if the engine slows. In other words, I balance the clutch release with throttle application to keep the RPM at my target RPM until the clutch is fully released. I also dip the clutch if grabs and slows the motor too much.

It might be interesting to explore improving the drivability by matching the torque so that it remains constant as it moves from 2000 RPM minimum throttle (no load), to say 1900 RPM with 5kPa more MAP (or whatever your map change was). Since you can't increase the torque at 1900 RPM, how about you reduce the torque at 2000 RPM by retarding the timing. Think of drawing a torque map showing lines of equal torque, now it might turn out that this forced you to open the throttle a bit more at 2000 RPM so there is less of a MAP change as the RPM decreases. This is just something I would play with on the dyno to see what happens.

I would also try using MAP acceleration enrichment, usually the problem is that MAP lags TPS, so it comes too late. In this case MAP changes are independent of TPS changes.

I have made some good progress on this issue.

I first tried MAP as the trigger for Transient Throttle. The issue was resolved however as David suggested, it was too slow above around 2500rpm, causing a lean misfire before enrichment. I went back to TPS based Transient throttle.

By decreasing the Transient Throttle - Load Rate Dead Band from the default 25% to 15%, the pumps now activate with smaller throttle movements. I then added some additional fuel in the Transient Throttle - Enrich Rate table under the 25%/sec and 50%/sec columns. I think this is appropriate as I am running a fairly large throttle on a 4 cylinder engine, so there are larger MAP changes for small TPS changes. The bogging issue was easier to manage, it still occurred but small movements of the throttle would activate the Transient throttle and the engine would regain torque.

The final thing I did was create a generic fuel correction table. I added a 10% additional fuel correction between 1000-3000rpm when the MAP-Derivative is 10psi/sec, tapering back to 0 correction at around 3psi/sec. I also put a condition on this correction to only activate when Transient Throttle Fuel Asynchronous Output is less than 1L to avoid it overlapping with a genuine transient throttle event. The result is with a constant TPS, as RPM drops and MAP increases, additional fuel is supplied. The car is now almost as easy to take off as most factory cars that I have driven, just with a more aggressive clutch.

Thanks to David and Gord for the input.

That is great news. Thanks for sharing the details of what you changed.

Although I had an idea of the issue, I have learned from your details of what worked. Thank you for sharing.