Non existent idle after FI install

Find/Fix/Report Back

Moderators: Sluggo, Amskeptic

User avatar
Ronin10
Getting Hooked!
Location: Columbia City, Seattle, WA
Status: Offline

Non existent idle after FI install

Post by Ronin10 » Thu Jan 16, 2014 2:44 pm

Greetings, Air Cooled Bretheren:

I will be well indebted to you kind folks if you can help me sort out this difficult idle issue I can't correctly diagnose and resolve.

I am struggling to get my bus back on the road after taking off the dual Webers and installing the fuel injection system. I've spent many, many hours reading the Bosch FI manual, my Bentley, this website, and the Samba too. Here's a timeline of what I've done:
  • >> Early 2013: rebuilt engine due to a failing camshaft. Installed dual carbs because my old FI system needed rehabbing.

    >> Late December: Installed FI system, no idle, backfiring, can run up RPMs okay, fixed leaking cold start valve, tested double relay, verified harness continuity, tested Aux Air Regulator, tested for vacuum leaks multiple times, changed out Temp Sensor II, checked injectors for spraying, tears, arguing with wife over the value of owning a bus, towed the bus to Perry's automotive in Auburn, WA, frustration ensues

    >> Early January: Perry's corrected my mismatched spark plug wires (careless, stupid mistake). He screwed the AFM mixture screw all the way in and was able to get it to idle while cold, but it wouldn't idle when it was warm, said the AFM was the last cause.

    >> Two weeks prior: drove around on the Perry's setup to get a feel for things and find some time to work on it: unable to go much faster than 45 mph or go up anything more than a moderate incline without downshifting to 3rd or 2nd if the hill is substantial enough. Significantly underpowered, even for a bus.

    >> This past weekend: Ordered a used AFM (an 018) from a seller off the Samba. Came from a running bus with too much body damage to save: Arrived on Wednesday. Appears to have had the seal broken at some point in it's past. Cleaned it and set it aside since my AFM hadn't been tampered with. I didn't want to go that route just yet.

    >> Last night: My AFM (an 018) has never had it's seal broken, but I did find the flapper door was sticky at it's extremes. A little cleaning and lube with WD40 and it now moves freely, reinstalled, no idle cold or warm, but can still run up the RPMs

    >> Today: I started with the assumption that my AFM was good since it had worked on my bus about a year ago (before the engine rebuild) and doesn't seem to be fussed with so presumably any issues would have been from mis-set mixture and idle air bypass screws, ran a sweep combining changing the air bypass screws (1/4 turn at a time) with changes in the mixture screw, found a few regions where the car would start, idle roughly for a bit then dies, minor tweaks to explore these regions revealed no change, tears, cursing, frustration, installed spare AFM, ran through a sweep of idle air bypass screw settings (full in to 10 turns out), no change. Didn't adjust the mixture screw because I want to keep this is a "control" variable since it came from a running bus, called Franklin's (just down the hill) about an hour ago, left a voicemail, but haven't heard back yet

    >> Now: waiting on Franklin's, divine intervention, or profound insight for the good minds here
There are probably some steps I accomplished, but missed mentioning. I think this covers the bulk of it. My feeling at this point is that the system is more or less in good order, but with some single item out of whack. Much thanks in advance.

Andrew

EDIT: One more thing that did occur to me: when looking down the s-boot side of both AFMs and pushing the flap open, the edge of the doors seem to be at a slight non-vertical angle. Is that correct? Seems like they'd be vertical.
Oscar: 1976 Sage Green Bus, Stock Motor, Solid Lifters, Manual Transaxle

User avatar
Amskeptic
IAC "Help Desk"
IAC "Help Desk"
Status: Offline

Re: Non existent idle after FI install

Post by Amskeptic » Thu Jan 16, 2014 3:05 pm

Ronin10 wrote:Greetings, Air Cooled Bretheren:

I will be well indebted to you kind folks if you can help me sort out this difficult idle issue I can't correctly diagnose and resolve.
Put the opened AFM on the engine. Remove plastic cap.
If you need an assistant to keep the engine running so you can back there to fool around, well, get an assistant, like your wife, get her invested in this.

STEP ONE

Ask The Engine What It Wants

You ASK by moving the little wiper inside the box barely a 1/4" in either direction with your fingers.This will temporarily provide more fuel (CCW), or less fuel (CW).

The engine REPLIES by either speeding up or slowing down. Speeding up means that the engine LIKES what you provided. Slowing down means that it does not care for what you gave it. Note that if you move the wiper too much in either direction the engine will slow down or stall, telling you nothing but that you are too ham-handed. Be subtle.

If you move the wiper CCW and the engine speeds up, then the current status of the mixture (i.e. when you let go) is LEAN.

If you move the wiper CW, and the engine speeds up, the current status of the mixture (when you let go) is RICH.

Let me know what you find. If it is lean, adjust the black plastic cog 6-10 teeth COUNTER-clockwise and check to see if it will idle correctly. Read the AFM procedure for how to set the mixture adjustment screw to the middle of its travel and adjust the black cog and the little wiper to compensate.
Colin
BobD - 78 Bus . . . 112,730 miles
Chloe - 70 bus . . . 217,593 miles
Naranja - 77 Westy . . . 142,970 miles
Pluck - 1973 Squareback . . . . . . 55,600 miles
Alexus - 91 Lexus LS400 . . . 96,675 miles

User avatar
Ronin10
Getting Hooked!
Location: Columbia City, Seattle, WA
Status: Offline

Re: Non existent idle after FI install

Post by Ronin10 » Thu Jan 16, 2014 3:11 pm

I was hoping we weren't quite at adjusting the AFM innards yet. I guess that's where I'm at then. You don't think the fact that two AFMs couldn't get it to idle is indicative of anything? I would have anticipated that one or the other could've made it idle.
Oscar: 1976 Sage Green Bus, Stock Motor, Solid Lifters, Manual Transaxle

User avatar
Amskeptic
IAC "Help Desk"
IAC "Help Desk"
Status: Offline

Re: Non existent idle after FI install

Post by Amskeptic » Thu Jan 16, 2014 3:21 pm

Ronin10 wrote:I was hoping we weren't quite at adjusting the AFM innards yet. I guess that's where I'm at then. You don't think the fact that two AFMs couldn't get it to idle is indicative of anything? I would have anticipated that one or the other could've made it idle.
We are AT adjusting the "mule" AFM. You want a running engine so you can diagnose, yes, but also it is easier on the engine to be running well as you research. You can then lean out until it is running slightly lean > then go find your vacuum leak with propane or carb spray, find your valve cover gasket leak, etc.

Do not be timid about committing this felony tampering adjustment. Always mark your starting point with a dab of red or white paint, and keep good records of your adjustments. Play. That is an order.
Colin
(and don't get all reverent on us just because your "good" AFM has not been opened. It is a time capsule of an era where these engines had factory mechanics and good valve guides. Now, they are frequently waaay too lean right out of the box)
BobD - 78 Bus . . . 112,730 miles
Chloe - 70 bus . . . 217,593 miles
Naranja - 77 Westy . . . 142,970 miles
Pluck - 1973 Squareback . . . . . . 55,600 miles
Alexus - 91 Lexus LS400 . . . 96,675 miles

User avatar
Ronin10
Getting Hooked!
Location: Columbia City, Seattle, WA
Status: Offline

Re: Non existent idle after FI install

Post by Ronin10 » Thu Jan 16, 2014 8:19 pm

Sorry for taking so long to get back. It was a few hours before I had a helper available.

Here's what I did and found:

>> 20 min warmup: stationary, RPMs in the 2000s (Weather here is in the low 40s and foggy)
>> Low Speed: CCW = RPMs up, CW = RPMs down
>> High Speed: CCW = RPMs up, CW = RPMs down (the rise in RPMs was less pronounced at higher RPMs)

So I'm lean/lean which means adjust the position of the wiper arm. You suggest 6-10 teeth above, but the guide says 1 tooth at a time. Same source so I'll override my conservative inclination to go 1 tooth at a time, but if you can clarify the discrepancy for me, that would ease my mind. Off to the pop it 6 teeth forward and see what happens. Will report back soon.

EDIT: I got out to the bus and kind of put together what you were saying...you were referring to the Dynamic adjustment (i.e. adjust the tension on the coil spring connected to the AFM flapper door, correct?
Oscar: 1976 Sage Green Bus, Stock Motor, Solid Lifters, Manual Transaxle

User avatar
Ronin10
Getting Hooked!
Location: Columbia City, Seattle, WA
Status: Offline

Re: Non existent idle after FI install

Post by Ronin10 » Thu Jan 16, 2014 8:44 pm

FYI, on the 2nd AFM (currently on the bus), it only looks like the lid was taken off, nothing else appears to have been touched...until now, that is. More to the point. 10 notches CCW gave a very low, rough idle. It died once or twice, but after long struggle to stabilize.

I'll sit tight for about 20 minutes or so and see if you happen to chime in. With no other guidance, I'll revert to the AFM adjustment guide and start from scratch: revert the cog to its original position, reseat the wiper arm and then check to see where I'm at.
Oscar: 1976 Sage Green Bus, Stock Motor, Solid Lifters, Manual Transaxle

User avatar
Ronin10
Getting Hooked!
Location: Columbia City, Seattle, WA
Status: Offline

Re: Non existent idle after FI install

Post by Ronin10 » Thu Jan 16, 2014 10:25 pm

Stopping for the night as it's cold, dark, need to walk the dog, and get to bed for work in the morning. I'll retackle this tomorrow afternoon after work.

Here's my current state:
>> Reset black cog to it's initial position and proceeded to follow the AFM adjustment guide.
>> Since I was lean/lean, I adjusted the wiper arm position CCW to enrichen the mixture. There was only about a tooth and a half of travel left. It only took about 1/2 a tooth to get it to idle. When it did, it came out way rich. Backing it off anything from it's current position and it won't idle at all.

Here's a video of the exhaust that is still coating the back of my throat:
http://www.youtube.com/watch?v=h0ri73uR ... ture=share

I tried adjusting the idle screw out to smooth out the idle, but to little affect, 2-3 turns only yielded about 50 rpm

>> Final state for the night: idling rough, belching smoke, mixture screw set to 4 1/2 turns out, idle screw to 5 turns out, black cog at original position.

My thinking at this point is to reinitialize the wiper arm, add the 10 cog notches back to get back into the nonsmoking section of VW life, and then re-ask the engine what it wants.
Oscar: 1976 Sage Green Bus, Stock Motor, Solid Lifters, Manual Transaxle

User avatar
Ronin10
Getting Hooked!
Location: Columbia City, Seattle, WA
Status: Offline

Re: Non existent idle after FI install

Post by Ronin10 » Fri Jan 17, 2014 9:01 pm

So I followed the AFM tuning guide and was able to get Greta to idle. I reset the wiper arm to where I felt the idle response from the wiper arm was as described in the guide. I then adjusted the black cog to 15 teeth CCW. The idle at that point was stable although low and rough. I tried adjusting the idle screw out to see if I could get the idle to stabilize a bit, but got very little response except in the fraction of a moment when adjusting the screw...

Screaming vacuum leak, idn't it? Been through this several times before, but gave it another go. I sprayed around with carb cleaner to see if I could see the rpms respond. All the hoses are new, but I shot them with carb cleaner anyway, did the valve covers, around the breather chimney, along the s-boot, around the throttle body...probably somewhere else that I can't recall.

At that point, I took it for a test drive, just to get a sense where I'm at. The route I took was roughly level, but with periodic, minor hills up and down. Speed limit was 50 mph. I then got on a freeway for about two miles and came back. The engine accelerated through 1st and 2nd just fine. I never really pressed it, but didn't notice anything out of sorts. 3rd and 4th gear were a different story. Initial acceleration after shifting was pretty good in 3rd but as I got into the heart of the gear, it started running rough. 4th gear was very weak. Any little hill would shave speed off and it was rough throughout the entire speed range unless I was cruising at very minor throttle when it was smooth.

I knew I wasn't there yet with the tuning, but I need to wrap my head around where the vacuum leak could be and what more I need to do on the tuning.
Oscar: 1976 Sage Green Bus, Stock Motor, Solid Lifters, Manual Transaxle

User avatar
Amskeptic
IAC "Help Desk"
IAC "Help Desk"
Status: Offline

Re: Non existent idle after FI install

Post by Amskeptic » Sat Jan 18, 2014 8:39 am

Ronin10 wrote:So I was able to get Greta to idle.
I need to wrap my head around where the vacuum leak could be and what more I need to do on the tuning.
Sorry I abandoned you to this wandering in the wilderness, it's good for ya.

Let's be be very careful with exploding variables. When the manual says "make damn sure everything else is perfecto" they sure aren't kidding.

You *can* inadvertently skip from too lean to too rich and have similar running symptoms, the only difference will be eye-watering acrid exhaust. That is why I suggest only the most irritating small increments. The wiper adjustment (static) is so powerful that one tooth is often waaay too much. 6-10 increments on the black cog is as far as I will go before having to declare "something else is going on".

Vacuum leaks are one tricky issue, timing advance is another. Fuel pressure under load is a possible issue IF it runs well at idle/low speed/initially, then it all peters out every time you try to get up to highway speed.

We need to be focused and be happy that *everything it isn't* is just as helpful as *whatever it is*.
Colin
BobD - 78 Bus . . . 112,730 miles
Chloe - 70 bus . . . 217,593 miles
Naranja - 77 Westy . . . 142,970 miles
Pluck - 1973 Squareback . . . . . . 55,600 miles
Alexus - 91 Lexus LS400 . . . 96,675 miles

User avatar
Ronin10
Getting Hooked!
Location: Columbia City, Seattle, WA
Status: Offline

Re: Non existent idle after FI install

Post by Ronin10 » Sat Jan 18, 2014 9:57 am

My adjustments last night before my test drive seemed to clean up the ugly exhaust so I feel like I'm back at the edge of lean. At least the best I can tell at this point and without an air-fuel meter.

I have a couple hours before I head back out to hunt down the vacuum leak, but it occurred to me that I haven't checked the push rod tubes. The tubes and seals are new there, but you never know I guess.

My guess at this point is that the vacuum leak is small and annoyingly hidden as its major impact is on idle and to maybe grow of my upper rpm tuning a bit, hence the roughness while driving in 3rd and 4th.

Please let me know if you see a flaw in that thinking.
Oscar: 1976 Sage Green Bus, Stock Motor, Solid Lifters, Manual Transaxle

User avatar
Amskeptic
IAC "Help Desk"
IAC "Help Desk"
Status: Offline

Re: Non existent idle after FI install

Post by Amskeptic » Sat Jan 18, 2014 10:05 am

Ronin10 wrote:My adjustments last night before my test drive seemed to clean up the ugly exhaust so I feel like I'm back at the edge of lean. At least the best I can tell at this point and without an air-fuel meter.

I have a couple hours before I head back out to hunt down the vacuum leak, but it occurred to me that I haven't checked the push rod tubes. The tubes and seals are new there, but you never know I guess.

My guess at this point is that the vacuum leak is small and annoyingly hidden as its major impact is on idle and to maybe grow of my upper rpm tuning a bit, hence the roughness while driving in 3rd and 4th.

Please let me know if you see a flaw in that thinking.
Push rod tubes are oil leakers before their contribution to air/fuel ratio is felt.

You have checked AAR rubber elbow? Are the breather and the decel valve hose nipples properly seated in the s-boot? The metal nipples have a ridge that some people think are "insertion stops" when they are actually supposed to go inside the holes to a groove.
Colin
BobD - 78 Bus . . . 112,730 miles
Chloe - 70 bus . . . 217,593 miles
Naranja - 77 Westy . . . 142,970 miles
Pluck - 1973 Squareback . . . . . . 55,600 miles
Alexus - 91 Lexus LS400 . . . 96,675 miles

User avatar
Ronin10
Getting Hooked!
Location: Columbia City, Seattle, WA
Status: Offline

Re: Non existent idle after FI install

Post by Ronin10 » Sat Jan 18, 2014 10:19 am

I do have a small oil leak that collects along the seam of the case and seems to be coming from the driver's side of the engine. My best guess is that it was coming from either the seals behind the oil cooler or the oil filter mount. I have to get my transaxle rebuilt later this year so I was planning to investigate then.

I have checked the AAR rubber elbow (new) and the Y shaped hose nipple was seated by warming the s-boot to enable the nipple to slide into the groove. Missed that one early on and fixed it already. When I head out again, I'm going to assume everything is poor and work from there. Since it'll now idle (and not suffocate me), it's easier to do that investigation.
Oscar: 1976 Sage Green Bus, Stock Motor, Solid Lifters, Manual Transaxle

User avatar
SlowLane
IAC Addict!
Location: Livermore, CA
Status: Offline

Re: Non existent idle after FI install

Post by SlowLane » Sat Jan 18, 2014 12:17 pm

Let's rewind to the beginning again for a bit.

You mentioned that the engine had been rebuilt due to a failing camshaft. Can we assume that the replacement camshaft is a stock profile, that the lifter and camshaft types are compatible (ie. solid lifters with solid lifter camshaft or matching hydraulic versions of same), and that the camshaft gear wasn't accidentally installed one tooth off the index?

Then can we assume that the valves have been adjusted correctly for the type of lifters in use, and that their variations in adjustment have settled down to be pretty much consistent (ie. the valvetrain isn't still "settling in" from the rebuild).

Can we assume that the distributor is providing consistently reliable timing, with the centrifugal advance smoothly returning to the same idle advance position each time the engine spins down to an idle?

Have you performed a compression test or leakdown test on your rebuilt engine? Are all cylinders contributing equally? Just what was replaced in the rebuild anyways? Just camshaft and lifters? Cylinder heads? Pistons and cylinders?

Is your exhaust leak-free? Even minor exhaust leaks can have a surprisingly deleterious effect.

The "Perry's setup" sounds suspicious, You shouldn't have to screw the idle mixture screw all the way to the bottom. That indicates that you're getting "false air" (ie. unmeasured by the AFM) from somewhere. Have you inspected your S-boot for cracks under a strong light (daylight, preferably, although that may be a rare commodity in western Washington)? You have said tha the AAR elbow is new. Be warned that the quality of the aftermarket AAR elbows is abysmal. I had one go bad on me in less than a month once, so don't assume that all is good there. Source a genuine VW one if you can.

There are a lot of variables here. Make sure you've covered all the basics before assuming it's the AFM.
'81 Canadian Westfalia (2.0L, manual), now Californiated

"They say a little knowledge is a dangerous thing, but it is not one half so bad as a lot of ignorance."
- Terry Pratchett

User avatar
Ronin10
Getting Hooked!
Location: Columbia City, Seattle, WA
Status: Offline

Re: Non existent idle after FI install

Post by Ronin10 » Sat Jan 18, 2014 12:43 pm

About to head out, but let me address some of the things you bring up.

Camshaft is a Webcam 107i the correct, corresponding Webcam hydraulic lifters. Yes, I know it's not stock for a 1976, but I'm confident it was running correctly since it was in good order on while running with carbs.

The new engine has nearly 3,000 miles on it since the rebuild. I haven't noticed any "settling" affects as of yet. New AMC cylinder heads, OEM muffler, gaskets, and annealed copper washers at the heads, and no discernable exhaust leaks. Distributor is consistent and reliable. Have performed a compression test with and without oil; 150 psi across all cylinders and adding oil saw maybe 5 psi or so, tops.

I'm on my 5th round of vacuum checks with little success including having Perry's review it for leaks as well, but this will be the 1st time I've been able to have the engine idle while hunting. Always before I had to have a helper keep the rpms up which can mask exhaust leaks.

All vacuum hoses are new as well as the tubing. All components were checked for correct functionality prior to install and the relevant ones for idle afterwards. S-boot sealed, immersed in water, and given mild pressure to check for leaks.

With all due respect, I'm not assuming it's the AFM that's at fault, I'm concluding it based on the evidence I've been able to compile. Yes, there's pretty solid evidence for a vacuum leak and I'll continue to hunt for that, but with tapping the AFM, I would be making progress. I've put considerably effort forward previously to avoid going into the AFM and now that I have, I understand my vehicle considerably better which helps in the long run.

I realize that last paragraph sounds a little pissy, but I assure you I don't mean it that way. I just want to be clear that I'm making every effort to be detailed and progress about this in the correct manner.

Off to test now.
Oscar: 1976 Sage Green Bus, Stock Motor, Solid Lifters, Manual Transaxle

User avatar
SlowLane
IAC Addict!
Location: Livermore, CA
Status: Offline

Re: Non existent idle after FI install

Post by SlowLane » Sat Jan 18, 2014 2:24 pm

Okay, sorry I may have rubbed you the wrong way, but understand that your original post didn't have any of that background, and one can't make assumptions. It does indeed sound as if you have covered the basics.

I do note, however, that your .sig indicates you have oversized valves. Don't know if that could be having an effect on the idle performance of your L-Jet system. Colin may have better insight into that.
'81 Canadian Westfalia (2.0L, manual), now Californiated

"They say a little knowledge is a dangerous thing, but it is not one half so bad as a lot of ignorance."
- Terry Pratchett

Post Reply