Fuel Injection Issue!

VW based Porsche. In a league of its own.

Moderator: Piledriver

C

Fuel Injection Issue!

Post by C » Mon Apr 05, 2004 1:00 pm

I have a 74 2.0 and love it. I took it in to a shop for an oil change and to re-thread a bolt and I when I got it back I have had fuel injection issues. I am very upset with the shop because it is a 914 specialty place. Anyway, they didn't give me much support. It runs well for 100 miles and then starts idling and firing very badly. I checked the plugs and they are very black with powder. All 4 plugs are like that. The fuel economy is very poor. So it seems to be runny very rich. Any thoughts? (temp sensors? pressure sensors? distributor? timing? ECU..??)

User avatar
raygreenwood
Posts: 11813
Joined: Wed Jan 22, 2003 3:01 am

Post by raygreenwood » Mon Apr 05, 2004 11:31 pm

First...what bolt did they re-thread? It sounds like a vacuum leak, or they jostled some connections while they were working on it. If it was something like anintake manifold bolt, they very well could have disturbed the gasket. Ray

C

fuel injection issue

Post by C » Tue Apr 06, 2004 10:38 am

They fixed one of the two bolts that hold the plate on the bottom of the engine where the oil temperature sensor is (and oil strainer). The plate was not sealing due to a loose bolt. It turns out the there were no threads for the bolt to fasten too. I don't think it the bolt. But I agree maybe something was knocked loose (I am hoping). I've checked for vacuum leaks and there doesn't seem to be any. I will go over it again. I haven't check the resistance on the #2 temperature sensor yet. I will do that. I recently adjusted the timing and put in a new distributor cap and rotor. The car ran great for about 100 miles. During that time the car idled fine. Could the mechanic have adjusted the CO2 knob ont he ECU and caused this?

C

fuel injection issue

Post by C » Tue Apr 06, 2004 10:54 am

They fixed one of the two bolts that hold the plate on the bottom of the engine where the oil temperature sensor is (and oil strainer). The plate was not sealing due to a loose bolt. It turns out the there were no threads for the bolt to fasten too. I don't think it the bolt. But I agree maybe something was knocked loose (I am hoping). I've checked for vacuum leaks and there doesn't seem to be any. I will go over it again. I haven't check the resistance on the #2 temperature sensor yet. I will do that. I recently adjusted the timing and put in a new distributor cap and rotor. The car ran great for about 100 miles. During that time the car idled fine. Could the mechanic have adjusted the CO2 knob ont he ECU and caused this?

pbanders
Posts: 103
Joined: Wed Jul 11, 2001 2:01 am

Post by pbanders » Tue Apr 06, 2004 1:34 pm

I suggest that the work your shop did and the fact that you had problems 100 miles later may be unrelated. Go through your FI system and check it out, using resources from my web page (http://members.rennlist.com/pbanders). For very rich running, I would be looking at the MPS (cracked diaphragm, won't hold a vacuum) and the head temperature sensor for starters.

User avatar
lmcchesney
Posts: 135
Joined: Thu Nov 20, 2003 11:11 am

Post by lmcchesney » Tue Apr 06, 2004 3:33 pm

Brad,
If they disconnected the CHT, would that not cause a lean mixture?
"C", you need to let us know how much you know about the FI system.
To check the MPS, remove it from the mount and connect the vacum nipple to a pice of tubing. Place suction on the MPS and clamp the tubing. Then attach a vacuum gague and see of you lose vacuum. I believe the rate is 5-10mmHg/min. Is that correct, Brad, Ray, Dave?
L. McChesney

pbanders
Posts: 103
Joined: Wed Jul 11, 2001 2:01 am

Post by pbanders » Tue Apr 06, 2004 4:56 pm

lmcchesney wrote:Brad,
If they disconnected the CHT, would that not cause a lean mixture?
"C", you need to let us know how much you know about the FI system.
To check the MPS, remove it from the mount and connect the vacum nipple to a pice of tubing. Place suction on the MPS and clamp the tubing. Then attach a vacuum gague and see of you lose vacuum. I believe the rate is 5-10mmHg/min. Is that correct, Brad, Ray, Dave?
L. McChesney
Disconnecting the CHT makes the mixture very rich. The CHT is a NTC (negative temperature coefficient) resistor - high when cold, low when hot. Going open-circuit is like telling the ECU that the motor is about 100 deg. F below zero.

You don't need to remove the MPS to test the vacuum retention, just pull the hose off the intake manifold (assuming the hose isn't rotten) and use a Mityvac to pump it to about 15 in Hg. A good MPS can hold this level indefinitely. An MPS bad enough to cause a rich mixture will lose most of its vacuum in less than 30 seconds.

User avatar
lmcchesney
Posts: 135
Joined: Thu Nov 20, 2003 11:11 am

Post by lmcchesney » Wed Apr 07, 2004 9:00 am

Thanks Brad,
I would appreciate your input on a concept of D-jet.
The CHT works as a neg. temp. resist. coeff.
Thus, as temp. increases, resis. decreases to compensate for cold condensation of A/F mixture.
A problem with increased displacement and D-jet is development of lean spots.
An excessively lean mixture results in excessive cylinder heat
If you replaced the CHT with a pos. temp. resit. coeff. would it be possible to mislead the ECU to increase mixture as a lean mixtures causes excessive cylinder heat?
Granted, you would have to use an inline potentiometer to get the correct resistance range and identify a resistor with a similar linear resistance.
Would not the T1 be adaquate for cold compensation in the altered system?
I would appreciate everyones thoughts.
L. McChesney

User avatar
Dave_Darling
Posts: 2534
Joined: Thu Apr 27, 2000 2:01 am

Post by Dave_Darling » Wed Apr 07, 2004 12:45 pm

lmcchesney wrote:Thanks Brad,
If you replaced the CHT with a pos. temp. resit. coeff. would it be possible to mislead the ECU to increase mixture as a lean mixtures causes excessive cylinder heat?
I suppose you could do something like that... But you'd run into some problems, I think.

First, the mixture has to be rich when the engine is cold. It simply won't run if the mixture is not "too rich"--BTDT!! So you can't get rid of the NTC thermistor because you have to have a rich mixture when cold and a leaner one when the engine is warmed up.

Second, that amount of "cold" richness is more than enough to kill the motor when it is warmed up. BTDT, again. Most especially when the engine is idling--a very fat mixture (appropriate for cold-engine operation) will bog the idle down to the point where the motor dies. So putting a lot more fuel in while the engine is running will be a Bad Thing. So we're talking a little bit of extra richness to try to help cooling.

Third, from reading through Brad's website I get the impression that the CHT circuit is somewhat non-linear in how it affects the mixture--so if you "bias" the resistance somehow you may put it into a range where small variations in operating temperature have drastic effects on the engine's driveability, causing relatively-large swings in mixture.

Fourth, the stock CHT location warms up and cools down relatively slowly. So by the time your sensor "sees" that you're too hot, you've been too hot for a while.


...So, you may be able to run a small PTC (Positive Temperature Coefficient) thermistor in series with the stock NTC one, but I don't think it will work very well and I am pretty sure it won't be worth the hassle.

--DD

pbanders
Posts: 103
Joined: Wed Jul 11, 2001 2:01 am

Post by pbanders » Wed Apr 07, 2004 2:09 pm

DD covered the main points in his reply. Yes, the ECU circuit that manages mixture enrichment from the CHT is non-linear, so that when the engine is warmer than the "fully warmed-up temperature" (something around 210 deg. F head temp), the mixture is stabilized and does not continue to lean out as head temp increases or changes with different running conditions. Because of this non-linearity, using resistors in the circuit to bias the overall mixture is not advisable. Adding a modest amount of resistance (50 to 100 ohms) can richen the mixture during warm-up, which is useful if you want to get smoother engine performance and eliminate light backfiring during overrun while warming up.

Best way to handle larger displacement is adjust the main mixture screw of the MPS (see my web page) while the car is on a chassis dyno and hooked to a shop-quality gas analyzer. Several people on Rennlist and PP's List have done this and reported good results. If you've modified the intake, exhaust, or changed the cam/valves/heads signficantly, then you've changed the volumetric efficiency (VE) curve of the motor. As the VE compensation is hardwired into the ECU for each specific D-Jet application, this cannot be compensated for by simple mixture adjustments, and requires modification of the waveshaping circuits - something I've never attempted to do.

Guest

Post by Guest » Wed Apr 07, 2004 8:50 pm

Thanks Brad and DD,
Brad, since I'm not an EE, it takes me several times reading your site, and each time I come away with something new.
DD, if you place a PTC in series with identical/similar resistance coeff. would it not result in a node? or delta r = 0.
If the response of the CHT to temp is so slow, does it not mean that you are seeing a just as prolonged enrichment as you would see a prolonged excessive lean mixture?
Brad, since T1 provides a 11% enrichment at open resistance, and T1 provides direct input to the PL, would a subsituting a PTC measuring CHT or oil temp. for T1 input provide warm engine excessive lean mixture?
From the block diagram of the ECU, it would seem alterations of T1 input along with MPS would have the greatest changes in mixture. Could you explain?
DD ......BTDT???? I,m not up with the neogolisms.
Thanks,
L. McChesney

Guest

Post by Guest » Wed Apr 07, 2004 9:45 pm

Mucking with the air temp sensor may give you some marginal improvements, but I still suggest that adjusting the MPS is the best way to affect overall mixture. Substituting a PTC for the NTC would likely result in very odd drivability problems, as variations in air temp would result in the opposite effect needed on for proper mixture.

BTDT = Been there, done that

User avatar
raygreenwood
Posts: 11813
Joined: Wed Jan 22, 2003 3:01 am

Post by raygreenwood » Wed Apr 07, 2004 10:02 pm

Brad and Dave hit on something very important. Since the CHT starts at 300+ ohms (rich) and drops to as low as 70 ohms (lean), you can have a problem. Its actually the worst in very moderate weather (spring, fall)...when its kinda warm, enough to give some extra heat. 200-250 degrees I have also seen to be the rangewherein, all of the change in the CHT is all in. That can be too low of resistance, running too lean at 70 ohms. Until about two years ago, I had been running a 0-1000 ohm ceramic variable resistor, to tune in an extra 125 ohms in the summer. I pulled it out in the winter. It can still drop too low in the winter.
The CHT, unlike a thermocouple for a gauge, does not qucikly and constantly change. THe sensor itself and the head bank heat to much. It generally helps you start up, gets you into the running range that the engine temperature needs to stay in. Unfortunately, as Brad noted, adding in extra resistance, may keep you from starting on the top end (cold) end of the scale when warming up.
I have been playing with a small relay board I made that uses 3 variable resistors, 4 subminiature reed relays and, and a bimetal thermostat to complete ground. On normal weather days, it stays straight through normal circuit...factory. When it gets warm enough, the thermostat kicks in the first relay and resistor and kicks out the stock line, adding resistance to the stack...after the car has already started warming up. I havn't decided to use a timer circuit yest, or a dash switch to tell it when to kick in. This way, it has normal high resistance to start normally always, but adds more as the resistance starts dropping. Yes, the best way to work with larger displacement is with the MPS. But...if the vacuum curve from the engine already works well with the settings of the MPS...and you are running the same volume of lean across the range of rpm....its better to raise fuel pressure to give across the board higher fuel colume. Ray

Guest

Post by Guest » Thu Apr 08, 2004 12:04 am

Ray, you might want to take a look at the section in my ECU page about cylinder temperature compensation. Once the value of the CHT sensor drops below about 100 ohms, there is no further effect on the mixture. Running at 70 ohms is the same as running at 50 ohms (which is the typical temperature of a fully-warmed up CHT). The CHT reaches 100 ohms at about 210 deg. F, so for any typically warmed-up motor, you're already in the region where mixture is unaffected.

Adding ballast resistance (to a point) to the CHT only affects the warm-up mixture. Too much resistance causes the ECU to intepret the car as never reaching warm-up, and varying the mixture (unnecessarially) when the head temp varies with normal running variation.

Substituting a thermocouple for an NTC resistor won't affect the responsiveness of the system, either - it's the thermal mass of the head that dominates the response.

pbanders
Posts: 103
Joined: Wed Jul 11, 2001 2:01 am

Post by pbanders » Thu Apr 08, 2004 12:10 am

Whoops, that was me in the posting above, wasn't logged in on this PC so I showed up as a "guest".

Jiggering with the CHT and air temp sensors to affect mixture, except for the purposes that they were designed to serve (CHT => warm up mixture compensation, Air Temp => variation of air density), is a band-aid that will cause more problems than it will solve. My first step in tuning other people's D-Jet systems is to remove large ballast resistance from the CHT's that I often find, and reconnect the air temp sensor, which is often disconnected to richen the mixture. If you want to change the mixture for normal running conditions, adjust the MPS.

Post Reply