Interesting CIS Issue - Clutch in stall

[sigchuck]

Hi guys, I have a relatively stock 1983 911SC that had a recent (< 2k miles) split case rebuild.

After some initial idle problems the engine builder seems to have addressed (or maybe it's been the 3 months of warm weather we've had??), the car has been running beautifully.

Driving to work yesterday it developed an issue in which holding the clutch in kills the engine. It presents as follows: I'll be in third gear approaching a stop sign/red light, clutch in, pop to neutral, and the RPMs will drop and the car quietly dies. No jerking, bucking, etc., just very quietly dies. It will immediately fire back up.

The car does have a relatively new clutch, so I've found if I barely engage it and quickly pop it out, the RPMs will still drop pretty low, but not to the point it dies. Additionally, if I brake down to a low speed and engage the clutch and go to neutral at that low speed, the RPM drop won't be as pronounced and it won't die.

Of note, the engine builder did disconnect and plug the small vacuum line to the decel valve. I examined it last night and the plug was still in the line, but there was no cap on the valve itself. I capped it, and while the issue is still there (the significant RPM drop with clutch in), it hasn't stalled. There also appears to be a large vacuum hose (approximately 1/2") near the decel valve that was simply hanging. I plugged this with a bolt, but I can't seem to find where it goes. It doesn't appear to be directly associated with the decel valve as the decel valve still has the two large hoses connected (only the small vacuum line was disconnected and plugged).

I've done some reading, and it appears that a working decel valve is intended to prevent an issue like I'm describing, but the only time my problem presents itself is when I clutch in; never a deceleration on brakes only and never when I take my foot off the throttle.

Any thoughts would be greatly appreciated!

[T77911S]

the mixture is wrong for some reason.
air leak or it is too rich.

check the throttle body and make sure there are NO uncapped ports.
make sure the 2 hoses from the DV are either plugged or capped AND where they connect or were connected to are capped.
the DV should have 2 large hoses going to it.

the DV is not needed. the DV holds the rpm hi when you let off the gas then it drops the RPM to idle. your engine should run fine without it.

you could also be a little too rich. if you let it idle for a while does it start to surge. if so it is definitely too rich.

the clutch has nothing to do with it.
rev the engine with clutch out and in neutral and it will probably do the same thing.

perhaps some pics of the DV and TB would help. the back side too

[sigchuck]

Thanks!

The decel valve definitely has the two large hoses hooked into it, though I can't very easily see where they are going. Any tricks to photographing/accessing the TB with the engine in the car? Any idea what the large vacuum hose I plugged should be connected to?

[DRACO A5OG]

+1 Vacuum leak and idle was richened to compensate making it go to rich when unloaded

[sigchuck]

One other item I forgot to mention - would an issue with the Lambda circuit have any impact on this? The 02 sensor has been disconnected since I acquired the car.

In trying to chase down another issue I unplugged the relay under the passenger seat which resulted in the car immediately running a bit rougher, so it appears at least some part of the system is still functioning.

[T77911S]

with the O2 not connected it goes into open loop which is a default.
as long as the way it ran got worse when the rly was removed that is a good sign.

unless it was done after the rebuild I would check fuel pressures, it needs to be done.

[sigchuck]

I have ordered a set of CIS gauges, and I did confirm that the problem was present in neutral with a quick rev and release of the throttle, just not as pronounced.

[tirwin]

Like others have said, decel valve is not necessary. It does sound like your mixture is off.

Take an engine that is functioning 100% correctly. Pull off the hose on the AAR on idling engine. The RPMs will shoot up from 950 to around ~3k rpm as the extra air makes it go lean.

Now imagine that you don’t have a big sucking vacuum leak, but a smaller one. You can tweak some things to make it seem like the vac leak isn’t there but it still is.

As T77911S said, the decel valve gradually slows rpms from when you’ve got your foot in it. With the decel valve disconnected, the throttle response is better but the change is fast. The change in mixture is happening when you deload the engine. As you noted, slow change doesn’t cause a stall. Your AFR is on the edge. When you de-load rapidly it’s pushed over the edge.

Your engine builder should’ve baselined the fuel pressures. At this point fuel pressures are unknown. You can try to find out from the builder. If they don’t have it, then you should check it. Maybe it’s fine. But right now it is unknown and everything else falls from that.

But it does sound like a vac leak though. There are a few recent threads on cheap homemade smoke machines. Unless you find a smoking gun (sorry for the pun) of something obviously left disconnected, that might be your best course of action.

With K-Jet Lambda the O2 sensor input is used by the lambda brain to adjust the frequency valve. But there is a limit to how much it can adjust. If the duty cycle is 100% and you’re still reading lean at the exhaust, there is nothing left it can do.

[sigchuck]

After looking at photos of the AAV (the disc shaped diaphragm, to be sure I'm not confusing it witht the AAR), it looks like the loose, open hose I plugged with a bolt is the right size for one of the openings.

Assuming it is a hose for the AAV, is it safe to assume it should be re-connected to the AAV? That is to say, would a half-connected AAV be a source of a vacuum leak?

Also, can someone point me in the direction of the AAV? I see the DV, and it looks like from internet photos that they're in the same neighborhood, but I'm having trouble seeing it.

[pmax]

There are only two flying saucers in your CIS system.

The one without the nipple is the AAV.

[sigchuck]

Thanks, I was just making reference to the fact the AARegulator, while not physically similar to the AAV, appears to be confused with it in a lot of older threads (people talk about applying power, etc., to the AAV).

Can you point me in the direction of the AAV?

[pmax]

You don't find this on your engine ? The upper right is the AAV.

[sigchuck]

I will take another look. The DV is definitely there as the small nipple needed to be capped (which did help the issue slightly).

I must just be overlooking the AAV.

[tirwin]

AAR has a 12v power connection. AAV and Decel Valve do not. AAR is on the same side of the engine as AAV and Decel Valve and is located between the intake runners.

Decel valve and AAV look similar (flying saucers). Decel valve has the nipple vac line connection in the middle. Both the hose and the metal nipple need to be plugged if the decel valve is not used.

[sigchuck]

Well the AAV is definitely there, I apparently just overlooked it. It definitely has one hose connected to it, but I can't get my hands in a place to feel for the other. I've attached a picture of the hose which was unplugged until yesterday when I plugged it along with the DV.


Look familiar to anyone? I don't see anything obviously disconnected on the TB, and new plastic vacuum hose was used for all of the small stuff, apparently. I do need to get an inspection mirror, but also wondering what would suddenly cause this to develop. Until now, it hasn't had anything close to an issue.

Also, in looking at the enrichment screw (not planning on messing with it yet), how exactly does one get a 3mm allen wrench in the space?

Thanks again for all the help!

[sigchuck]

Gauges and an a/f analyzer are ordered. Thoughts on hose? Brake booster?

[T77911S]

if you are talking about the hose in the pic you can remove it and plug it at the TB. just make sure it is not leaking.
brake booster and the hose going to it are possible.
you need a hand vacuum pump to test it. it is a great tool to have.
I have one that is just vac and I just git one that is vac and pressure.
pull a vac on the hose to the booster and see if it holds

you don't need the pressure but this is it.
got it cheaper somewhere else

http://www.amazon.com/exec/obidos/redirect?link_code=ur2&camp=1789&tag=wwwpelicanpar-20&creative=9325&path=https://www.amazon.com/Powerbuilt-648532-Vacuum-Pump-Set/dp/B003TYM3MM

[Jameel]

I've got a thread going on the same condition in my brother's '79. CIS Experts: Help evaluate a couple numbers

The one thing I have yet to do is check mixture. And I think I'll try and do that today. I will add here that the '79 only dies when rolling up to a stop with the clutch in maybe once every couple months (and he drives it daily). Hope this helps you find your issue.

[john walker's workshop]

3.5% co and 1000 rpm and see how that works. O2 unplugged and hose off decel valve. Clamp the booster hose tight and see if the idle changes. They can crack down by the torsion tube occasionally.

[sigchuck]

I wasn't able to find a single gas (CO) analyzer, so I purchased an LM-2 wideband AFM analyzer. I think the CO% compared to AFM is universal, and it looks like the 3% CO translates to an AFM of 13.38, BUT, I further understand 14.7:1 AFM is ideal.

Is it just a function of an older, relatively unsophisticated FI system that 14.7:1 shouldn't be expected?

Thanks for your help!