Type Three Tuning Page -- Fuel Injection

Please select the area of interest:

Instructional/training slide show for the Type III fuel injection system, 24Mb. realPlayer can view it.
Hose diagram - FI parts list - Electrical diagram of late FI
From the Elfrink guide, which was published in 1972. Information on 1.8L and 2.0L components was not in the manual since those systems came out later.
Fuel Injection Troubleshooting:
Page 1  Page 2

Bosch D-Jetronic Trouble-shooting document
Operating Instructions for VW 1218 Electronic Fuel Injections System Tester
Download as a MS Word document (1.6Mb)

Pictures of the tester at work on my 1973 Porsche 914 with an unknown year 2.0L rebuilt engine...
Interface exploded fit  Interface together fit  Ready for testing  Passed the first test!
> I've had this car for close to eight years now. It came to me with the 028 
> brain (that's Bosch part # 0280 000 028; VW part # 311 906 021E), matching "E" 
> pressure sensor (Bosch # 0280 100 116; VW # 311 906 051E) AND throttle valve 
> switch (Bosch # 0280 120 024; VW # 311 906 111E). It also came with the 311 905 
> 205L distributor. All of it had worked well together until this happened so I 
> thought it best to keep the part numbers the same.

The "E" PS is good for both '72 and '73, but the 028 brain was for '72 only, 
and originally came with the expensive vac adv/ret '72-only distributor. I 
suspect that the differences in the brain were minor and had to do with finding 
a way to meet emission requirements without the complicated timing proceedure 
that the expensive '73 dist required. The 048 E brain superceeded the 028 E 
brain, but the reverse is not true. It's unfortunate that VW marked both as "E" 
but they probably did it when the 028 was discontinued from their stock and 
simply replaced by the 048.

The 311 L dist is from '68-9 and is probably the best dist out there for our 
cars, but you should time it like a '68-9, since it has more advance (~30 deg) 
than the OE dist. If you time it like the OE '73 engine (5 deg BTDC) that will 
give you more advance (~35 deg) than is healthy for this engine, so the best 
thing to do is to time it to TDC, as was the spec for 68-9.

In general, the rule of thumb is that you should time your engine according to 
the spec of the distributor year.

Jim Adney

...and along those lines...

>I had a thought....when a different brain goes in the car, does it need some
>run time to adjust to the motor?

NO, these brains are open loop, they don't learn like the modern ones.
028 E brain is for the dual advance/retard distributor [and the] 048 is the 1973
singe advance distributor, they must be matched.

Keith Park

According to an Allen Institute Bosch fuel injection training manual, the following cars used D-Jet injection: 
Some 75 thru 80 Cadillac, Mercedes w 3.5L to 70, 4.5L all, Saab 99E, Volvo 1800E, 1800ES, 142, 144 & 164E. Possibly some donors for injectors, etc.
914rrr, member in the www.914club.com forums

Deceleration Valve

The decel valve is (in my opinion) necessary. It's also easy to adjust if you have the manual one. It looks like a fuel pressure regulator with a locknut on the vacuum nipple. It has an air inlet and an air outlet...like 10mm hose?

The Decel valve is not supposed to be open until around 3000 rpm or so. Try this, you will notice that one side of the 10mm hose goes straight to the plenum, vacuum side. The other goes to the aircleaner or somehwere where there is no vacuum. Pop off the side that is feeding (not connected to the plenum or vacuum). If at idle you feel suction it is either broken or needs adjustment.

There is a locknut around the 3mm vacuum hose on the dome. Loosen that and turn the nipple clockwise (inward) with a 7mm wrench. Turn it until vacuum ceases on the supply side 10mm inlet hose nipple. Turn it a half turn more. Now rev the engine up slowly through 2500 rpm. If at any time you feel vacuum at the 10mm supply side nipple then tighten the 3mm nipple inward some more. You should feel no vacuum at any time while revving the engine (throttle open at any setting).

Now when you let go of the throttle at any rpm over about 2700-3000 you should then feel suction on the 10mm supply nipple.

The valve should only open when rpm is 2700 or above and the throttle is closed. If the decel valve does not adjust it is possible that you have the 3mm vacuum hose connected in the wrong place or the valve is broken.

What is this thing for? When you snap the throttle closed, vacuum becomes immediately very high and for just a seccond or two the engine can be very rich. The metering flap can actually be pulled forward by that much vacuum and airflow combined (the snapped shut throttle and high rpm). This also has a habit of activating the vacuum advance again. What the decel valve does is to bleed a little air past to lean the mixture for a second or two and bring the vacuum level down slightly. It's good for not overheating the exhaust in stop and go traffic, saves a little fuel, and is better for emmissions. It loses no performance at all because it only operates at throttle closed when adusted properly.

Ray Greenwood, from the Shop Talk Forums. Edited for readability by me.
There were 3 different generations of trigger points. 

Early: No washer staked on top of the 2 pivot pins, so these could  
be pulled off their shafts.

Middle: Washer staked on the top of each pivot pin, so they can no 
longer be slid off. These came out around 1970.

Late: An extra bit of metal was added inboard of each moving point. 
It's a kind of "shield" that keeps anything from being thrown from 
the cam onto the points. I don't know when this started.

--- Jim Adney from the Type III email list

Increasing discplacement (thus power) with stock FI:

> I have read and have been told that if all FI is working right on a VW
> 1600 Type III engine, and you add 200cc by putting 90.5
> cylinder/piston sets and nothing else has been modified, all one needs
> to do is increase the gas pressure by 3 pounds. The ECU still sends
> the same information to the injectors as it did at 1600cc. The engine
> makes no other demands except it needs more fuel for the added
> displacement to not run lean. The idea as I understand it is like a
> water hose. A water hose with a fixed nozzle has 28 pounds of pressure
> behind it which gives a certain fixed amount of spray out the nozzle.
> Now if the pressure is raised to lets say to 32 pounds, there will be
> a larger amount of water being sprayed because of the added pressure.
> Of course there is a limit to which this fixed nozzle and pressure can
> safely live together. BRET Instruments said that the stock 1600
> injector flow increases properly up to 35 pounds. 26 pounds is the
> minimum.
> Now what my question is, do you think it is this simple when all you
> are changing is the pressure/displacement? The ECU has no idea what
> pressure is it is delivering at, just that it needs to give a certain
> set time to the Injectors. Of course the sensors assist in what the
> ECU does, but that should not have any effect if all is running at
> proper temp.

...My thanks to Pat for a stimulating question.

Credit to Keith also, because he has espoused different injectors for years. His thinking about this was apparently more clear than mine.

As I got into this I realized that I had been looking at this question incorrectly. The problem, at least for me, was that I was assuming that the intake venturi produced a particular pressure drop for a particular air flow rate. It only takes a moment, however, to realize that the fact that the throttle valve is there, always in an unknown position, makes this untrue. I had to take a different approach and eventually found one that I could carry thru to a conclusion.

For Pat's question, upping the displacement by going to 90.5mm pistons gives an increase in displacement of (90.5/85.5) squared, which is a factor of 1.1204, or about a 12% increase. A 12% increase in fuel pressure should take care of this, so consider going to about 32-33psi, just as BRET said.


Engine modifications and the Bosch D-Jetronic Fuel Injection

The D-Jetronic FI system attempts to measure the amount of air entering the engine by measuring the absolute pressure in the intake air distributor. The absolute pressure is the atmospheric pressure less the pressure drop that the incoming air experiences as it comes thru the throttle valve and venturi. We usually think of the air inside the intake as being under some vacuum, but it is more useful to understand that the word vacuum just means some pressure that is less than atmospheric. On an absolute scale, zero pressure means NO air at all, or what we would normally think of as a perfect vacuum.

For the sake of simplicity, consider a stock D-Jetronic FI 1600 cc engine running under some static condition, like cruising, or any non- changing situation you care to think of. Now the engine is running at some RPM, some load, and some throttle position. Under these conditions, there will be some absolute pressure, P, in the intake air distributor and let's assume that the the D-Jetronic brain is working properly and delivering the correct amount of gas to the engine under these conditions.

Delivering the right amount of fuel to each cylinder is dependent on knowing how much air went into that cylinder, but the amount (mass) of air is proportional to the absolute air pressure, P, times the volume of air, V. In an internal combustion engine V is constant: ~400cc for a stock engine. Since the D-Jetronic system is a constant pressure system, the amount of fuel going into a cylinder is determined by how long the injector valve is open each intake stroke. The internal gains of the ECU have been designed to open the injector valves the right amount of time for each air pressure reading.

Now consider what happens if we increase the displacement of the engine to V' without changing anything in the FI system.

Suppose we now take the car out and try various constant loads until we find a situation where the absolute intake pressure is the SAME as before, so P is unchanged. If we drive like this, the ECU will see the same inputs as before so it should put out the same injector pulse widths as before. The intake air mass, however, has changed from P x V to P x V', an increase, by a ratio of V'/V. This means that if we change nothing in the FI system, the engine will now be running lean.

What can we do to improve this?

The simplest thing would be to just increase the fuel pressure, and in fact this should work just fine. Note that for a given intake pressure the air mass has increased by a factor of V'/V, so all we have to do is calculate V'/V and increase the fuel pressure by that same ratio. This will increase the amount of fuel delivered by the same ratio as the increase in air mass.

If you just go to larger pistons and keep the original stroke, the easy and accurate way to calculate this ratio is: D squared/85.5 squared, where D is the diameter of the new pistons.

A couple of follow-up notes:

It's still unclear to me what effect RPM has on the ECU output. The reasoning above assumes that there is only a minor effect. The same is true of minor (post warm-up) changes in head temp and intake air temp.

The pressure measured by the pressure sensor is somewhat larger than that actually achieved in the cylinder during the intake stroke. This is due to the pressure drop that occurs between the intake air distributor and the cylinder. The ECUs are built to take this difference into account, but they only do so for the stock intake geometry. If you port your heads, then the intake pressure drop will be less than the what the ECU is "expecting" so the engine should again run lean as more air is getting into the cylinder than in a stock engine under the same conditions. I don't know how much of an effect this might be, but a 1 psi increase might be reasonable.

Another way of increasing the FI fuel delivery for larger displacement engines would be to switch to different injectors which have higher flow rates. To do this, you should take the ratio of the flow rates: F'/F and change the fuel pressure such that V'/V = F'/F x Pf'/Pf. Pf' and Pf are the new and old fuel pressures, respectively.

There is a table of flow rates on the Web somewhere. Someone posted it recently, but I don't have it here at home. If someone would please send me the URL I will add the information to this write-up.

We should all stay aware of the fact that the D-Jetronic system is unique in that it measures the air mass indirectly by way of the intake air pressure. The later Bosch systems, L and K work on different principles and the conclusions above do NOT apply to them.

Jim Adney, jadney@vwtype3.org
Madison, Wisconsin, USA

By Dave Andrews Other great info can be found on his site. Thanks Dave!
Basic engine management systems and conversions: Great info for after-market FI systems. This is a must-read!