Fuel flow and a lean condition a 3.5L 911 race engine

[stownsen914]

X-post from the AutoX and Track Racing forum ...

Hi all,

I am chasing a lean condition in my 3.5L vintage 911 race engine. It happens at full throttle / hi RPM conditions. I can see the lean condition clearly on the A/F meter. The car runs marvelously at all other loads and RPMs. If I keep the throttle to max 3/4, it runs without a hitch.

A few details:
- 3.5L engine making 300+ hp with 50 mm PMOs
- main jets 210, air correctors 170
- Fuel system is set up with a return line back to the tank. Pressure is set to 4-5 psi.
- One interesting driveability note is that after heavy braking, when I go to blip the throttle when downshifting, the engine coughs and spits. I suspect the float bowls are empty at that moment after the long straight. After the corner, it runs fine again (up to 3/4 throttle, that is).

I've tried a couple different fuel pumps, and have am suspicious that they are not supplying enough fuel. Both pump provide 60-70 gallons of fuel per hour (measured at the return line back to the tank with the engine off), which is what they are rated for. By my calculations, that volume might be marginal for this engine.

Note that since I've measured flow rate at the return line, this takes into account any restrictions that might reduce fuel to the carbs/engine, so I don't think restriction is the issue.

Is 60-70 gal/hour enough? I recall someone saying once that these engines like the Holley blue pump, which gives ~100 gal/hour. Is that what's required?

Should I be looking at different emulsion tubes in the carbs (not sure what I have honestly). Or could the needle valves be insufficient? I believe my carbs have the 1.75 and 2.5 needle valves.

I'm guessing there should be lots of experience out there with setups like mine. Hopefully someone can chime in on this

Thanks in advance!

[KTL]

Well the typical recommendation from PMO is to run a Mallory 4110 pump which supplies 110 GPH (no load). You've got the fuel pressure set well for a racing application, since it's typically recommended to increase the pressure from the street-friendly 3 to 3-1/2 psi.

You've already got quite large main jets. So those aren't a hindrance. Plus a main jet that's too small would generally rear its head throughout the rev range in terms of producing an AFR that is leaner than desired.

I have quite a few Holley Red pumps (noisy as hell but they work pretty good) that you're welcome to give a try. Free of charge no strings attached. I have so many because my car used to run two Reds in parallel and it also came with four spares- two good spares and two retired pumps I believe.

I think the emulsion tubes should be OK. PMO has those figured out pretty well and there's not much to fiddle with there for a race engine. He says the F16 tubes are pretty good all around. But if you're seeing top end leanness it might not be a bad idea to give the F2 a try

http://www.pmocarb.com/Images/Instructions-3.pdf

I think the needle valves are a non issue. PMO has figured these out and it makes sense that he puts a typical 1.75 valve in the bowl that serves one cylinder and a 2.50 valve in the bowl that serves two cylinders. Years ago when I had 46 Webers, I was futzing with the tuning and discovered I had 1.75 in the single bowl and 2.00 in the two cyl bowl. I asked for thoughts here on the forum and got the response that both should have the same size delivery valve. But then I got to thinking that it's probably no coincidence my carbs have never exhibited the common starvation issue during high-g cornering that prompts people to install the bowl baffles. Instead of baffling the bowl, one can ensure there's enough fuel in the dual cyl bowl by using a larger valve.

[stownsen914]

Thanks Kevin for the thoughts and for the offer. I actually have an older Holley blue pump kicking around, so it sounds like I need to try that to at least rule out the fuel pump capacity variable.

These PMOs were purchased around 2000 by the previous owner, so I wondered if maybe the thinking on emulsion tubes or needle valves had changed since then.

I've give Richard Parr a call to see what he thinks.

Scott

[stownsen914]

So I did a dyno test to get to the bottom of what was going on. It was very revealing and helpful ... We discovered two things:
1. I watched the float bowls run to nearly empty on runs longer than 8 sec on the dyno. We didn't do any runs long enough to actually starve it for fuel, but we did get a breakup at the end of a 7200 RPM run, though it didn't show as lean in the AFR graph. But the bowls definitely emptied as I'd speculated.
2. The engine was running rich (not lean) otherwise. Quite rich, in fact. The AFR graph showed 10-11 with the initial jetting throughout the RPM range.

So we made a couple changes during the session and retested. We went down from 210 mains to 190s (in a couple steps with reruns each time) before running out of time, and the AFR graph did improve. Still rich, but better. And I put in the Holley "black" pump. It's higher output than the blue and red pumps. After this change, the float bowls no longer emptied during a 10 sec heavy load run on the dyno.

It also confirmed that the AFR meter I have in the car is way off. Explains why I thought the jetting was just about right. It was probably off because the engine builder put the O2 sensors in the header primaries right next to the exhaust port, which is apparently too close to the heat source and can cause malfunction and bad readings.

I ran the car a few days later at Watkins Glen, and the car was tranformed compared to before. Finally! It ran very nicely overall. I went down to 180 mains and checked tailpipe color and sparkplugs to confirm it's not too lean. Also adjusted the ignition timing from 24 degrees to 26 at the top end, after asking around and finding that 24 was a bit retarded for a twin plug race engine running race fuel.

The bad ... the car was still coughing and spitting when blipping for downshift. I assume this is either too much or too little fuel from the injection pump, so I'm going to measure the injection volume, and depending on what I see, play with larger and/or smaller volumes. I've read that race engines like very small injection quantities, so that may be the answer - we'll see. I'm also going to fix the O2 sensor placement, and probably buy a wideband O2 setup with recording capabilities like the Innovate unit so I can tune more effectively. Any other suggestions welcome!

[KTL]

Well that's good to hear you sorted out a number of things. Also good that you have PMOs because changing the jets is a lot less messy (dumping the Weber fuel bowls for each main fuel jet change is annoying).

Regarding the accel. pump injection amount, be sure to consider switching the pump cam levers.

Gordo's PMO Carb Tuning

Wideband O2 sensor is a no-brainer nowadays IMO. They've become fairly economical and they're a really good DIY tuning tool. BTW Innovate makes a dual function gauge so you can monitor each bank of cylinders. Of course it requires two bungs and two WBO2. Or if you want only one for simplicity sake, that's certainly fine too.

I have an AEM digital wideband if you would like it. Brand new never used. I like the digital better than analog. I previously had an analog needle gauge because I wanted it to match the rest of my gauges. But it was a bit bothersome to watch the needle and have to make a quick read of what range the needle was in. I feel the digital numbers are easier to make a quick glance at, just to see what range you're in. This is of course assuming we're talking about quickly glancing at the gauge while driving. On the dyno, it's not a big deal to stare at the gauge for a few seconds to read it!

[stownsen914]

Thanks Kevin. I did get a pair of the new hatchet injection pump levers from PMO. I haven't measured my injection pump volume yet so I don't know what adjustment I want to make, but Richard Parr told me that that newer style levers give a quicker shot of fuel. I guess I'll see when I get into it ...

Sending you a PM about the AEM gauge.

Scott