Lotus Elan

Weber DCOE Jetting Criteria

PostPost by: type26owner » Wed Jun 29, 2005 7:55 pm

Due to two-phase flow behavior the amount of air allowed to mix in the emulsion is by far the dominant factor controlling the curve of the mixture strength versus rpms graph. The tendency is to install a much larger air corrector jet then is actually needed. This approach is reinforced by the jet sizing relationship formula stated in the Weber Tuning Manual. Some simple testing of jetting combinations will soon prove that that formula to be dead wrong. What you want to see in a graph of your engine is one with the rpms on the vertical column and mixture strength along the horizontal and your's charting a nearly straight vertical line over the desired mixture strength. This is quite possible and easily achieved it turns out with some swapping and sucking.

The approach to take is to step down the air corrector jet in size until the measured mixture stabilizes at some mixture strength at all rpms and loads. Deviations of one part air to fuel is about as good as the carburetor can be expected to deliver though. When the mixture does not change much regardless of the load and the rpms you've found the nominal air jet size to use. Now adjust the mixture strength by sizing the mainjet to obtain the desired value.

The same procedure applies for the idle jets. Start with a large diameter F-numbered jet and step down in size with just the F-numbers until the mixture stays steady from just off idle until about 2500 rpms. Once you find the right amount of air flow then stay with that F-number and change the jet fuel size to get the desired mixture strength. You can do this measurement by removing the emulsion tubes and driving the car.

I'll show how this is all done in the proper sequence in the video I plan on doing. I know, I know, it can't be this simple, right? Oh yes, yes it is. Easy peasy stuff! :lol:
type26owner
Coveted Fifth Gear
Coveted Fifth Gear
 
Posts: 1616
Joined: 18 Sep 2003

PostPost by: type26owner » Sun Jul 03, 2005 2:38 pm

Spent yesterday refining the tuneup on the DCOE 150s on my brother's Caterham with the intent of getting the Webers to pass the smog test without the need of a cat. The AFM signal was bouncing around so we first fixed the small exhaust system leaks but that did not solve the problem. So then I replaced the dizzy cap and rotor and that fixed it. It was caused by a partial misfire at idle, the unburnt oxygen would show up as a lean spike or the bouncing lambda.

Then I discovered that engine had a pcv valve with a manifold that introduced a huge airleak into each induction port. Really dumb design from Caterham like I'm surprised! By adding restricting orifices I got the system to still provide the required amount of scavenging of the crankcase fumes but reduced by one whole hell of a lot the airleak going into the induction system at idle. As it was the idle could not be reduced below 1100 rpms and the idle speed screw would have the throttle plates completely closed at that point. By adding the resistors that changed so I could reduce the idle speed by the speed screw down to 600 rpms. I know 600 is too low to have the throttle plated just downstream of the first progressive hole. So now we were ready to go and really do some fine tuning.

Which leads me into some interesting observations about the idle jets behavior. I was targeting the mixture to be 0.9 lambda (14:1). I could not get there. Could adjust it to be 0.7 (12:1) or 1.1 (16:1) but not in the middle. The F-numbers I was using were F9=1mm and F8=1.2mm with only a 200 microns difference in the air bleed hole size. That difference is too big and there isn't a jet selection in between that's available that I'm aware of! Looks like I have to make custom sized jets to really be able to dial this stuff in where it needs to be. Phooey! :x
type26owner
Coveted Fifth Gear
Coveted Fifth Gear
 
Posts: 1616
Joined: 18 Sep 2003

PostPost by: type26owner » Mon Jul 04, 2005 2:55 pm

My son and I spent yesterday trying to do some numerical analysis using the software provided by Innovate Motorsports. Caused more confusion by doing so and we're hitting the books to get a bit smarter on that subject (damn good thing my son has lots of experience analyzing raw binary data at the lab). Turns out the there is no way to do 3D mapping yet. The next release of Logworks is due out this month and it looks like it will have it. Determined that the datastream from a TPS is essential for that purpose. It's time to get the TPS that mounts on the DCOEs working.

One mystery we're trying to sort out on the Super7 is under acceleration at a steady throttle opening there are displayed a series of lean 1.5 lambda spikes at about one second intervals. The spikes could be real or maybe it's just a software glitch. The driver is unable to provide me with reliable feedback. Looks like I really need to invest in a handheld scope now to eliminate the ignition system from being the source of the problem. :( Could this be some type of two-phase flow which is causing this? This stuff is really getting exciting. Very bizarre!
type26owner
Coveted Fifth Gear
Coveted Fifth Gear
 
Posts: 1616
Joined: 18 Sep 2003

PostPost by: rgh0 » Tue Jul 05, 2005 1:00 pm

Keith

The only thing that could work on a 1 second cycle in the system is the float and valve system in the carb. If its working in an on off mode rather than maintaining a constant level you could get this behaviour. Have a play with fuel pressure to see if the frequency changes if it does then you know the culprit

Rohan
User avatar
rgh0
Coveted Fifth Gear
Coveted Fifth Gear
 
Posts: 7655
Joined: 22 Sep 2003
Location: Melbourne, Australia

PostPost by: type26owner » Tue Jul 05, 2005 1:52 pm

Hi Rohan,
Thanks, that's a good point. Makes perfect sense too. Hope it's really that simple afterall. He has some amplifier still attached onto the ignition leftover from the original Caterham reluctor ignition. Thought I might disable that unit since he's now running a Pertronix hall effect ignition. If there is a partial misfire that unburnt oxygen would also show up as lean spikes.

At the moment I'm going to go back and remap my Elan to use as a baseline. I've not been using the recording function of the AFM until this weekend so I lack the data files to make comparisons by overlays.

Gotta make those mistakes to learn anything. Boy, are we learning stuff now!
type26owner
Coveted Fifth Gear
Coveted Fifth Gear
 
Posts: 1616
Joined: 18 Sep 2003

Total Online:

Users browsing this forum: No registered users and 5 guests