My AFR curve is spikey/jagged not linear - thoughts..

[24OZ]

I just got back from the dyno and wanted some thoughts on my AFR plot for my L28ET. I've been told that the tuner tuned the map via the O2 sensor and the readings from the graph where taken from the exhaust, hence in reality the numbers should be slightly richer.

 

The car feels great on the road but I've shown the graph to a few people and they've told me that there shouldn't be so many spikes in my AFR plot and it should be smooth.

 

Is this really the case and is it worth spending more dyno time to smooth out? The guy who tuned the car for me I guess doesnt think so or he would have addressed it?

[ryan95i4]

You could lower the resolution of the software and the result would appear to be a smoother a/f curve, or you could increase the resolution and have even more spikes. Unless other cars roll off that same dyno with flat a/f curves, I wouldnt stress about the small peaks and valleys too much.

 

 

The small spikes from 4000K rpm on wouldnt concern me much as the slight trend towards lean as redline approaches. If you have enough injector left, I would probably want to even that out.

 

You also have a lean spot from ~2500 to ~3500, but Im assuming this is during spool, which Ive always tuned lean during spool-up also as it decreases spool up time (unless theres something different about the L28, but thats how Ive tuned other cars).

Edited September 27, 2012 by ryan95i4

[24OZ]

Thanks for your input Ryan.

 

Here is a bit more details on my set up. I'm running 320cc injectors, stock turbo cam and twin turbos (old school units that came with the HKS twin turbo manifold I'm running). I am also running two innovate O2 sensor in each turbo down pipe, ie each turbo feeds of 3 cylinders.

 

I am attaching my other graphs, what's interesting is to see the boost with regards to the lean spot you mentioned between 2,500 to 3,500. These are lazy turbos and they don't wake up until quite late. I think I could benefit from a better cam as power falls off at 5,000 revs, not sure if this is due to the nature of the turbos, or the cam, or a combination of the two. Looking at other peoples graph it does appear that the turbo cam runs out of revs around 5K.

 

Anyway, appreciate all input.

Edited September 27, 2012 by 24OZ

[Tony D]

I would say the jaggies are because ther instrument it fast scanning. software should smooth that somewhat.

i think you are rich past 4000, i think you will broaden the torque peak by trimming fuel back above torque peak. on Jeffp's engine we were at almost 13.5/13.8 after torque peak and still making power with good EGT's and no detonation. the more we pulled, the more power we made.

[ryan95i4]

Thanks for your input Ryan.

 

Here is a bit more details on my set up. I'm running 320cc injectors, stock turbo cam and twin turbos (old school units that came with the HKS twin turbo manifold I'm running). I am also running two innovate O2 sensor in each turbo down pipe, ie each turbo feeds of 3 cylinders.

 

I am attaching my other graphs, what's interesting is to see the boost with regards to the lean spot you mentioned between 2,500 to 3,500. These are lazy turbos and they don't wake up until quite late. I think I could benefit from a better cam as power falls off at 5,000 revs, not sure if this is due to the nature of the turbos, or the cam, or a combination of the two. Looking at other peoples graph it does appear that the turbo cam runs out of revs around 5K.

 

Anyway, appreciate all input.

 

Yea lean during spool is typical, as it helps spool-up times. For example, on my Avenger (2.0L DOHC with a very large Holset HX40 turbo at 24psi on pump) I would tune very lean, 13.8-14.0 or so during spool, to help with lag. If your turbos arent quick to spool, your tuner was likely trying to help get things going -- running fat during spool would hurt even more.

 

 

 

I would say the jaggies are because ther instrument it fast scanning. software should smooth that somewhat.

i think you are rich past 4000, i think you will broaden the torque peak by trimming fuel back above torque peak. on Jeffp's engine we were at almost 13.5/13.8 after torque peak and still making power with good EGT's and no detonation. the more we pulled, the more power we made.

 

These are the type of intricacies of the L-Series that Im interested in and need to learn. Different motors respond differently, back to my Avenger, we tend to tune very fat (11.0-11.4) at full boost because the 420a tends to make more power that way. If the L-series is happier and stronger running leaner, then ignore my comment above about fattening things up near redline.

Edited September 28, 2012 by ryan95i4

[Tony D]

The issue is also the stock cam that power peaks at 5300! Get an Isky cam that pulls to 6800, port the hell out of the head, and lament you don't have forged slugs to ly o 8500, which really is where this would work better.

Modern wheel cuts and compressors could drop your spool as low as 2500 or 3000, and THAT would really make the car fun to drive!

Pair of eBay disco potatoes...with the small hot side A/R!

 

Playing with cam timing to move the power peak up may help but your bottom end will o away completely!

Edited September 28, 2012 by Tony D