Fuel Distributers

[William Miller]

I saw a post on PP where someone was modifying a 81-83 FD to replicate the early Carerra FD which seems to be a jewel.
Charlie, do you know something about this?
My car is still running lean in mid range and load. I checked the pressures last night. I lowered the system pressure a little by taking out some washers. This did lower the control pressure about 1psi.
It had some positive effects but not enough.
I saw a post there about being able to adjust the warm control pressure with a little screw inside the Control pressure regulator as well.
Is this much different than adjusting the mixture screw?

I really need to change the curve of the fuel mixture from rich at idle and then lean at midrange and rich at high rpm's. It needs to be richer in the middle. I'm afraid if I lower the control pressure it will make it richer across the board. This will be way too rich at idle.

Has anyone been here before?

[Charlie Stylianos]

The early Carrera 3.0 FD are a jewel but very $$ to buy new (~$2,800) and very hard to find used.

The best way to richen the mix across the board it to lower the control pressure by modifying the WUR. You can also richen the mix by increasing the system pressure, but lowering the control pressure is more effective as you have more adjustability. By lowering the control pressure, you can dial back the idle CO a bit and still get your richer mix you need elsewhere. BTW: Increasing/decreasing the idle a/f mix does not change the A/F mix throughout the RPM range in a linear fashion....it will continue to lean out as RPM's increase.

Here is a good thread on modifying the WUR to dial in a specific control pressure:

http://forums.pelicanparts.com/showthread.php?s=&threadid=89694&highlight=wur

[William Miller]

That was one of the posts I was refereing to.
My mixture control unit actually has 3 different angles in the cone.
One area is for idle which is steep, one is the mid range which is flatter and a third which is steeper. I can see the A/F getting richer as I hit this 3rd steep part of the cone. It happens somewhere above 4000 rpm. It really depends on the air flow.

I think the 80-later mixture control units don't have these steps or different shapes in the cone because the mixture is increased or reduced with the computer and frequency valve.

Mine does not have these so they accomoplished the open loop "Mapping" with the shape of the cone.

It looks like I am going to have to open up and adjust the control pressure regulator.

Has anyone explained what is so special about the early 3.0 FD.
What makes it different? The plunger shape, the slit sizes?

[Charlie Stylianos]

I don't know exactly what the differences are with the Carrera 3.0 FD. I'm guessing its slit sizes are different and fuel mappings are a bit different also. Another thing that people don't take into considerations is that the FD's are matched/mapped to their respective mixture control unit. So, if you swap an FD from a different year, you will also need its matched mixture control unit.

Steve Weiner actually recommended it for my project, but there was no way I'd spend that $$.

[William Miller]

If you changed it out you would lose the advantages of a closed loop system. I'm assuming that there are some drivability advantages and others besides emissions.

I recall you mentioning that it was on your list at one time. I was recently thinking that there would be a lot of things to match up. You would really need to almost replace (Backdate) the whole CIS system.
If you are going that far you might as well spend the money on carburators.

I think mine will eventually get sorted out. It runs very good when the A/F meter is telling me that I'm between 13-16:1. The problem is that that is not all the time like it should be.

I figure that somebody knows how to get these parts to match up to their optimimum. If I keep asking I'm bound to find someone who does.
I bet there is just one or 2 parts inside the FD that need to be swopped to get the correct curve. Possibly it's just the plunger.

I bought the 79 FD and control pressure regulator last winter. I was swopping the 83US distributer and WUR. I couldn't find any info at the time about an 83Euro Distributer so it looked like the 79 was the closest because the system works the same way they do however have different part numbers. My little spec book shows that the control pressure on the Euro Control pressure regulator is just 1-2psi (lower I think) on the warm control pressure. I've never seen any info on FD's.

[Charlie Stylianos]

My car is still running lean in mid range and load.

William,

I was talking with Steve Weiner this afternoon and he mentioned the exact same thing. At WOT, CIS will tend to lean out in the mid range, richen up a bit as the RPM's increase and lean out again around 6000 RPM's. Looks like this is the nature of the beast.

Increasing system pressure or decreasing control pressure will richen the mix throughout the whole range. There is really nothing you can do by modifying fuel pressures to compensate for a lean mix at a specific RPM w/o affecting the whole curve. The a/f curve (a/f mix at a specific RPM) is determined by both the fuel distrib and its corresponding mixture control unit.

Tuning tips:

13.2 is the magic ratio for max power output.
Anything above 14.0 is too lean. Anything below 12.8 is too rich.
This is all at WOT.

[William Miller]

I spent most of the day yesterday working on this.
I adjusted the warm control pressure to 46 psi based on the recommendation of 1980SC on the PP board.
Woth the vacuum hose off it drops to 40psi
System pressure is 75psi.
I spent a lot of time looking for vacuum leaks that have the potential to introduce un-metered air. I couldn't find anything sugnificant.
I ran the car at idle fot 2-3 hours and sprayed carb cleaner everywhere.
I also hooked up the vacuum gauge while doing this to see if there were any indications of vacuum change as I pushed, pulled and prodded every conection. (At least all the ones I could get to.) I also tightened all the hose clamps for good measure.)

The vacuum gauge stayed pretty constant 17-18 in vacuum. This was effected a little by leaning and richening the mixture. (Not much.)

I left the vacuum gauge connected with a long hose so I could watch the gauge as I drove. (The readings were as expected, and it's an education watching how the throttle effects the vacuum.)

Prior to doing this I had the control pressure way low for experiment purposes. (Something like 42-43psi) It's better at the current setting because the mixture screw is in a more normal range for adjustment.

I'll drive the car for a few days, but I don't think it's going to cut it.
The next option is to send the FD out to get it rebuilt to the 930/10
specifications like it should be. I have been trying to avoid that cost.
I recently found someone who can do this and I think it's about $250-350.

I also have a realitivly new FD that was on the car which was the 930/16
like you have on your car. This is what was on the car when I bought it.
I might put this back on and test it to see what A/F readings I get.
I also would be able to drive the car with this one while I send the other one out. (I hope it doesn't take forever.

[William Miller]

Guy's I think I solved the mystery!
After all the time I spent this weekend the car ran like crap on the way home last night. There was a period of time where the mixture went lean and stayed lean 17:1 plus. Normally If I shut it down and restart it this symptom would go away. Finally about halfway home and about the 3-4 times of shutting it down it fixed itself again. This is the kind of intermittant symptom that has been driving me crazy for 2 years.
As things usually happen the symptom has to get bad enough to identify the cause. Like a small noise that you can't identify untill it get's loud enough.

Anyway the lightbulb went off!
After all that checking and tightening and testing with carb cleaner what could let air in and seal itself back up when you shut the car off?.............................................. .................................................. .............


THE POP OFF VALVE!
This came to mind late last night.
This morning I had enough time to take a look.
I found the hinge pin for the pop off valve sticking out of the right side of the hinge about 1/2 inch. There are 2 arms on the hinge with the spring in the middle. The pin was out far enough that one of the arms was not being held in place. So, the valve flap was not centered over the hole and the o-ring.

I went over this several times this weekend. While the car was idling I tried to pull the flap open to see if it was sealed. It was! At idle the vacuum was 17-18 psi. I couldn't budge the flap without possible dammage. (I never noticed the pin sticking out.)

Here's what I think may have happened. The seal is stuck tight at idle. At WOT there is little or no vacuum, not much more on light cruising. So the only thing that holds the seal is the little spring on the flapper. If the mixture got too lean it probably popped the valve open. This is because there is little vacuum at that time. I think on an intermittant basis the flapper did not close in the correct alignment and caused the leak. When I shut it down There was enough time with no vacuum that the flapper could re-align itself! Ergo intermittant vacuum leak and lean condition!
The lean condition may continue to create more backfires up the intake making it pop open again. Like a programming loop. It wouldn't stop until I shut it down.

Anyway, I didn't drive it today. With the forecasted rain, the old seals on the top leak like a sive. I don't mind driving it, but if it sit's in the rain all day it begins to smell.

Has anyone had any experience with these valves? I pulled the o-ring to take a look at it and it seems OK, But I'm sure they ger dried out over time. Is there something your's supposed to lubricate them with to help them swell up so they seal better?

I think I can figure out some way to keep the hinge pin in place. Actually it's not moch bigger than a wire so maybe I'll find some wire that is the same size and bend the ends like a cotter pin.

I didn't have much time to look, but it didn't look like anything was broken or missing. What's supposed to keep the pin in.
Can someone look at thiers and let me know!
Thanks!

[Charlie Stylianos]

THE POP OFF VALVE!

Get out of town!!

Arn't these things supposed to help prevent problems :wink:

I think the pin is just a press fit into/through the small arms of the cap. One of the arms could have split at the seams or cracked causing a loose fit and the pin slipping out.

As cheap as POV's are (and you can confirm that its the cause of your lean running condition) I'd just replace it rather than rigging a fix.

[William Miller]

Good thought on replacing it except it's glued in, supposedly with epoxy.
It probably can be done but not as easy as installing the first one.
I'll have more time to look at it tonight. I'm going to search PP board to see if anyone has had the problem before. I wonder if you can just get the o-ring for it or do you have to buy the whole thing?

I can't wait to get home and drive the car and see if that solved the problem. It wouldn't be the first time I was wrong.

My frustration is that when I took on this car there were so many things wrong with it. It's not like it was running good and something broke. If something is not working I have to question the parts compatibility issues first. Then trouble shoot all the posibilities. There were definately several symptoms going on at the same time which effected each other. Because they were intermitant added another dimension. I would have never straightened this out without the knowledge that the A/F gauge gave me.

On that issue: I was reading your post about the meter with the wide band sensor. A few weeks ago I was reading some other posts on that subject. I was trying to clarify what people were commenting on that the narrow band sensor I have is not accurate. I began to doubt that statement because it is too general.

I found an emissions chart which shows the 4 exhaust gasses in relation to the A/F ratios. On that graph the O2 line is bassically flat below 13:1 I think at about 1/2%. If there is any variation below 13:1 it isn't much. Beyong 13:1 the % of O2 begins to climb in an arc and looks linear at about a 45deg angle beyond 17:1 (That's where the chart stops.) I think the Oxigen sensors were designed to work in that range because there are significant variations in O2 arround 14.7:1. I don't recall the control range of the K-lambda system, but the initial mixture control setting is 0.4-0.8% CO which is just a little richer than Lambda. (This is with the lambda control disabled, Right) To the lean side of lambda the CO line is flat at about 3-4%.

My meter's range is from 12:1-17:1. From looking at these tables I think it would have to be exponentially more precise to read below 13:1 because the O2 values don't change much there. It's probably ok from 13:1 up. 13:1 is max power as you stated.

The O2 gas was picked for a control input because it is a good indication of fuel mixture in the range that they want to measure it for emissions equipt cars.

I think the general statement made could be refined to say that if you are tuning at ratios below 13:1 then you will need a wide band sensor.
For lambda control cars the mixture should never get much below that.

FYI: The 78-79 k-basic mixture setting is 1.5-3.5% CO. 3.5% CO happens to be equivalent to the magic 13:1 max power ratio. On these cars the cone in the mixture unit is stepped. It's rich at idle, (This is the 1.5-3.5% setting) It get's leaner during crusing and light aceleration and at high engine speeds or high air flow it get's richer.
The 83 Euro like mine is tuned a little leaner at idle.
I've seen my A/F meter follow this general plan. Rich at idle (12-13:1 leaning out and then richening up at about 4000rpm.)

Since I have possibly fixed my air leak the ratio shouldn't exceed the aprox 16:1 ratio designed into the cone for light loads.

Do we really need a wide band sensor? I'm mainly intrested in the logging function! Where are we on the group buy anyway?

[Charlie Stylianos]

Narrow Band (0-1v output)

Anywhere .2 to .7 volts, your A/F meter will read around 14.7. Anything richer than 14.0 or leaner than 15.0 is just a guess.

http://forums.pelicanparts.com/uploads2/O2%20sensor%20voltage1065748483.gif

Wide Band (0-5v output)
A much more of a linear progression of Vout vs Lambda.
http://forums.pelicanparts.com/uploads2/wblambda1065749767.gif

I'm not quite sure about the Wide Band purchase yet. If I do, I'm going to wait until they release the inductive RPM pickup. Should be avail in a few weeks.

[William Miller]

OK, good graphs, I've never seen those.
It looks like they designed in a dead band into the O2 sensor so to speak.
How does the closed loop control even keep it close to Lambda with those wide swings in input from the sensor?
Maybe it doesn't and that was good enough to meet the regulations for the day. (Early 80's) I guess you will see the fluctuations on your new meter with the data logging!

I wonder how much better the montronic is at keeping the A/F ratios at optimum.

I know the meter in my car has some value, based on this info it's value is deminishing. It was a useful diagnostics tool.
I'll have to rel-earn how to interpret the meter.
When did they change to the wide band?

[Charlie Stylianos]

Wide Band sensors have been used in highly efficient vehicles (Honda Civic HX's for example) and can be used as an instrument for more precise a/f metering and a/f indication. Just recently they have been adapted for use for the DIY tuner.

Narrow band sensors are just used as an on/off switch for an enritioning mechanism (i.e. CIS frequency valve).

Just a guess as to how the lambda computer would use the narrow band output in a useful manner.

Too rich (voltage higher than .7v) tune down or turn off frequency valve operation until .2v-.7v. If near stioch. (.2v-.7v) continue current frequency valve operation, if too lean (lower than .02v) increase frequency valve operation until .2v-.7v.

[William Miller]

That's were I was going with it, but it seems too simple.
Are you sure the line on the graph is perfectly verticle?
If it's not then a slight change in A/F would cause a big change in voltage making it very precise arrount 14:7.

[Charlie Stylianos]

If it's not then a slight change in A/F would cause a big change in voltage making it very precise arrount 14:7.

True, but you won't know what's really going on at the other 2 extremes.

From what I've read is that that some narrow band guages are heavily damped to try and give an accurate a/f indication at either end of the spectrum.

Here's another:

http://westers_garage.eidnet.org/O2graph.jpg

[William Miller]

So I guess I really need the graph for the Oxigen sensor that they sold me with the meter.

Are you confirming my comment with your second graph? I should be precise in the readings within the white range near 14.7.

I will also ask them about the gauge.
I will need you to interpret their response if I get one.
I know just enough about electronics to be dangerous.

I read the pop off valve installation article on PP yesterday.
Their instructions mentions to put a dab of epoxy on each end of the hinge pin shaft so that it won't work it's way out like mine. I posted this falure on PP and got some good responses. I did the epoxy trick last night. I think it should hold! The intermmitant air leak problem seems to be solved. I actually enjoyed driving the car into work this morning.
I still have some tuning to attempt with the warm control pressure and the mixture screw to try and make the mismatched FD work well in my car.
Now I might have a good chance at doing this with out the air leak problem screwing up my results. I think now the readings will give me a true inducation of my tuning.
The car runs good now and if I didn't have the A/F meter I probably wouldn't know it's running lean in mid load conditions. With this knowledge I know I can make it better. The precision and accuracy of the A/F meter is more important now. It worked well for problem solving nos I need it for tuning.

I'll let you know about what info I get about my sensor and meter.

[Charlie Stylianos]

The a/f meter is just a damped volt meter (0-1v). You can also hook up a volt meter in parallel with a/f meter so you can see the sensor output voltage for a specific a/f ratio.

[William Miller]

Roger!

[Charlie Stylianos]

Here is some more good stuff.

From here: http://www.students.tut.fi/~hezekiel/bosch.htm

Lambda sensor:

Especially when you have K-jet installed to an engine, which doesn't conform to factory specs anymore, it's essential to keep track of fuel mixture in all driving conditions. Of course, the system could be tuned way on the safe side, but it's not very wise unless gas mileage is not of any concern.
The easiest way to monitor fuel mixture is to install a lambda sensor in the exhaust pipe somewhere within 50cm from the engine.
There are three main types of lambdas, some have only one wire, from where the voltage can be measured. Other types include 3 or 4 wires. Two of them are for heating the sensor for quicker operation after cold start (normally you'd have to drive 1 km or so to have accurate readings). On the 4 wire model the voltage is measured between two wires.

What the voltage tells:

The range is typically from 0mV to 900mV. The voltage rises as mixture gets richer. The "funny" thing is the voltage leaps from, say, 250mV to 650mV very rapidly at air/fuel ratio 1. It's not the best for fuel economy nor performance, but it's the range where cat-equipped cars try to keep the mixture.
So, you'll most likely to operate just "rich" or "lean". Of course 900mV means loads of black smoke from the exhaust and 750mV is better reading at WOT for an every-day car. On the other hand, the system can be tuned to indicate 0-200mV at moderate cruise and coast.
Monitoring the voltage is most easily done with a common electronics voltmeter, but it's hard to look at during driving especially in the night time and refresh rates might be inadequate. Everybody I know use a led bar, whose instructions can be found here.
BTW: Please don't take the voltage readings as absolute truths. Lambdas are slightly different from each other but the scale they use is the same. I tweak, therefore I am :)

[William Miller]

Thanks, it seems the writer confirmed that the line on the graph is not verticle so the narrow band sensor should be precise near lambda.
Each sensor probably has it's own charcteristics. I wonder how many diferent sensors there are out there.

I will try and get info from C&B performance who sold it to me.
You mentioned that the meter is probably dampened. This would explain it's seemingly slow response time. In my car K-basic the control pressure should drop immediately as you open the throttle. (As the vacuum drops)
I watched the vacuum gage and this drop is happens in milliseconds.
I think that's why vacuum is a great input source for many engine controls. I recall watching the control pressure drop pretty quickly when I disconnected the vacuum line. (Although I wasn't consentrating on that)
Now I wish I had a long hose for the fuel pressure gauge so I could watch it, the vacuum gauge and the A/F meter while driving. (This is the only way to see what is really happening.)

I would think the O2 content in the exhaust gas and the the ability of the O@ sensor to read it is much quicker than I am seeing the needle move on the gauge, and as you confirmed it's not verry accurate outside of the narrow range. (I think the needle is actually overshooting the true A/F ratio on both ends rich and lean.) If this is true my system is doing it's job pretty well now. One thing I have notices while taking the WUR apart is that the rubber diaphram on the bottom section (The richening control for WOT) is not tight like you would think a new one would be.

This is a rebuilt WUR that I bought from Vertex last fall. I wouldn't think it would be streached out that quickly. It always beggs the question "What does rebuild mean?"

It's a relativly simple device. I don't know why they wouldn't sell rebuild kits for it. They obviously sell them to the people that rebuild them.
It may be just a little more complicated than a carb rebuild.

I think the same is true for fuel distributers, except I think you need calibration equiptment to adjust fuel flows so they are even to each injectors.

A club or group such as ours could probably afford such equiptment.
I'm sure we could figure out how to do it.

Why won't Bosch sell the parts? For the most part I suspect K-jetronic went out in the 80's. Porsche may have been leading the pack with new technology, but with tighter emissions regulations the rest couldn't have been that far behind. Since most are over 20 years old they can't have that much business in rebuilding stuff. It seems they don't even support some of thes things anymore.

Does this make sence?
How could we lobby Bosch to sell these kits?
I would think PCA would be big enough group to push it.
Maybe I'm dreaming and I don't have a clue how the relationships between the Bosch and their distributers work.

I just wish I could get my hands on a rebuild kit!