Critique my AFR graph

toddu · 10-25-2014, 04:14 AM

I've been playing around with my new AFR controller after I've done a little tuning on the carbs. I had put smaller idle jets in and then went back through the lean-best method (still need to work on this) and did a run.

I'm a bit lean through the RPM range, but not sure if it's a concern or not. You can see below it's for about a second. Is this a problem? The graph below is zoomed into just one part of the graph. But, it's similar pattern through the gears. I'm not sure if that's the accelerator pump kicking in.

Todd

JohnJL · 10-25-2014, 05:09 AM

Can you zoom in on a full-throttle 4-th gear acceleration from 2000 to redline?

toddu · 05:20 AM

Don't have that in this run. So, 4th gear and WOT from 2k up?

Just to add a little detail, I was quite rich at cruise (10'ish). Put 55 idles in and adjusted mixture and now just over 12 at cruise.

Todd

Raceboy · Raceboy's Garage

12's at cruise is still too rich, 14-15 in cruise and 12.7-13.0 at WOT is good.

Steve@Rennsport · 10-25-2014, 11:06 AM

Quote:

Originally Posted by Raceboy

12's at cruise is still too rich, 14-15 in cruise and 12.7-13.0 at WOT is good.

JMHO, but thats too lean for air-cooled engines and this is especially important in hot weather.

13.0-13.2 is OK for light throttle cruise, but WOT, full-load should be 12.6-12.8 for the engine to survive.

toddu · 10-25-2014, 12:13 PM

Steve - WOT would be mains and air correctors? How about a quick spike through the RPM range, normal or something to correct?

btw - this is a modified STREET car.
Todd

Gordo2 · Gordo2's Garage

Quote:

Originally Posted by toddu

You can see below it's for about a second. Is this a problem? The graph below is zoomed into just one part of the graph. But, it's similar pattern through the gears. I'm not sure if that's the accelerator pump kicking in. Todd

Todd,

Does the software that you are using with the wideband provide a smoothing feature, and if so have your smoothed the AFR any in the posted graph?

The Innovate LM2 "Logworks" software that I used has an adjustable smoothing feature that you can apply to take out the high and low peaks, which helps to see what the AFR looks like over few seconds vice within a second or two. I would have been chasing my tail if I tried to adjust my AFR based on the un-smoothed data.

I ended up applying just enough smoothing to take out the 1-2 second AFR spikes. Not sure if this is acceptable though. With the smoothing on, my graphs turned out similar to most fuel injected AFR graphs that I found, without smoothing my AFR was all over the place

Unfortunately, there isn't a lot of Porsche carb AFR graphs around (believe me, I looked - I can post links to the one's I found if you are interested).

As such, I'm not sure what you should expect regarding momentary spikes (lean or rich, lasting 1-2 seconds). I spoke to Richard Parr about it (I was lean on acceleration) - although I didn't get a specific answer, I came away with the understanding that a 1-2 second spike was normal.

BTW - I did most of my runs 2k RPM up in 2nd or 3rd gear Not sure if this is right either, but I recall finding that this recommendation somewhere.

Bottom Line - My Take on Your AFR Graph: you appear to be lean during the time period 111.7s to 114.0s or from 2.3K RPM up. Smoothing would probably show this averages out to ~ 14-15 AFR for ~ 3 seconds. As such, I agree with Steve (which is always a safe bet...) - you appear to be lean.

If you could, I think it would be helpful if you could post a graph that shows your AFR over a longer time period/duration. Tough to do - I had to find some remote roads where I could stop and then wind up to 80 mph or so...

Good luck,

Gordo

Lapkritis · Lapkritis's Garage

How do the plugs look on all 6? Looks like you might have a cylinder blinking out. Should be more steady AFR.

Steve@Rennsport · 10-25-2014, 09:53 PM

Quote:

Originally Posted by toddu

Steve - WOT would be mains and air correctors? How about a quick spike through the RPM range, normal or something to correct?

Yessir, thats correct. The spike is something else and should be investigated.

Raceboy · Raceboy's Garage

Quote:

Originally Posted by Steve@Rennsport

JMHO, but thats too lean for air-cooled engines and this is especially important in hot weather.

13.0-13.2 is OK for light throttle cruise, but WOT, full-load should be 12.6-12.8 for the engine to survive.

I should have been more specific, 12.7-12.8 in max torque area and can lean up to 13.0 at higher rpms.
Regarding cruise, target afr is a combination of many things, but I forgot that he does not have programmable ignition, because with leaner afr one can advance ignition so the EGT's are cool and not transfer heat into heads from exhaust ports. I have tuned quite a few normally aspirated 911 engines like I described, so far they have been running for few years on the track and on the street (even long road trips to western europe). But we have only hot summers, not all year long.

Just my 2 cents.

toddu · 10-26-2014, 05:12 PM

I'll try and get some more runs recorded next weekend based on the feedback above.

Gordo - no, the software I'm using is WinLog and doesn't have a "smoothing" option. Not bad, but somewhat basic.

Thanks all!!

Todd

Eagledriver · 10-26-2014, 07:44 PM

Any time you are less than WOT, lean doesn't matter as long as it drives ok. For a street car with 9.5CR or less and timing of 30 degrees or less, lean not going to be a problem. I would run it as lean as you can and still have drivability. Racing is a little different.

-Andy

VFR750 · 10-28-2014, 08:55 AM

I think you will find it is impossible to lean it much more than 11.5:1 at cruise and still have a reasonably stumble free transition. It is just how carbs, especially weber/PMO/Zeniths work.

There is a reason they got such poor mileage and polluted so much. They were always pretty rich. Now we have AFR gages and discover how sloppy they really were.

Wot is different. 12.7 to 12.9 best.

There is also why CIS and motrinic offered better mileage. They can do a better job of managing transition. So they can run a lot leaner in cruise.

Of course this is my input. Having spent the better part of four years mapping my zeniths, with both zenith and weber IDA transition ports, I concluded that is what it is.

Rich cruise equals good transition. 11.5:1 is about right.

toddu · 10-31-2014, 12:04 PM

I've done a little more work with the wideband controller (iDash). I slightly leaned out the idle a little more, and I'm just a bit over 12 at steady cruise now. I have to go back through lean-best tuning as I sort of rushed through it last time.

Also, light throttle load (cruising in 4th, give it ~ half throttle) I see AFR go up a bit as in graph above. Since it's only part throttle, is this an issue?

Graph below is 3rd gear, went WOT about 2500 rpm. See a quick spike which is probably accelerator pump which hasn't touched. Graph below is same run but zoomed in on the spike. Pretty short lived. I need to verify that I'm actually getting full WOT (been playing with linkage). Bad weather this weekend, so maybe I'll get more runs next weekend (need a longer road too :-) )

Todd

Gordo2 · Gordo2's Garage

Todd,

I think your AFR looks good - appears to average out in the mid 12 range as you accelerate.

I also think the spike is of such short duration (above 13.5 AFR for ~ 0.3 sec) that it's an anomaly.

How does it feel - are you getting any kind of hesitation as it ramps up or transitions from idle circuit to mains? If not, I wouldn't worry about that brief AFR spike.

Gordo

simsalabim · simsalabim's Garage

This is an interesting thread as I may be heading down this road soon myself.

As a basic question, when one is doing his test runs, would the data be more accurate if the motor was under as much load as possible while monitoring its AFM for a particular rpm range ? Should one attempt to have as much weight in the vehicle as possible while doing the test ?

Thanks