Audi 4000 Turbo quattro Conversion

The Intro

The Project








Total Bill

Engine Specs

Special Tools
With the motor disassembled I had some decisions to make.  I had always planned to make some changes while I had the motor apart.  I also wanted to keep my overall project costs as low as is reasonable.  Here are some things I considered in no particular order.

The 5 cylinder uses the smaller VW intake valve sizes used on non-GTI cars up through 1988.  I had hoped to change this head over to the larger 40 mm intake valves vs. the standard 38 mm.  I have solid lifter valves coming out of my ears, but had forgotten that the Digifant head (hydraulic lifters) that has been collecting dust for 2 years in my garage has 40 mm valves.  That means I would only have to buy one more ($22) instead of 5.  It turns out that the standard valve seat is too small in diameter to work with the larger valve by about 0.5 mm.  I guess you could have the stock seats removed and install larger ones, but this is outside the scope of my budget.

I had also looked at doing away with the sodium filled valves after comparing them to the solid lifter exhaust valves I had on my shelf.  After further inspection the hydraulic lifter valves out of the Digifant head are very similarly shaped and a few grams heavier so I will be sticking with the stock 5-clinder exhaust valves.  It is strange because the older solid lifter exhaust valves are shaped much nicer than the later ones.  I guess short of getting custom valves made you would have to cut down a set of solid lifter valves.  Maybe for some future project....

If you are willing to mix and match parts you can find later normally aspirated 10v heads with 40 mm intake valves.  If you are interested in this search in the archives of the quattro list on  Make sure it will fit on your block however.  For example an MC head (later turbos) will not fit on a KH block.  Been there tried that.

Most people won't care about the weight of valves, but I thought since I had them out I would measure them.  They were all still a little dirty with carbon so the weights might be off slightly overall.

10v 2.14L 5-cylinder turbo
70g (38 mm)
65g (33 mm)

Solid lifter 1.8L GTI (83-84) *
77g (40 mm)
64g (33 mm)

8v hydraulic lifter - Digifant (89-92?)
74g (40 mm)
71g (33 mm)

*Solid lifter valves have slightly longer stems, but the exhaust valve has a much smaller diameter stem at the back of the valve head (see picture on project intro page).

This one I was back and forth on.  Ultimate power in a turbo motor comes from low compression and high boost, but low compression gives poor off boost torque and in turn poor day to day drivability.  Since I am looking for a fun car, but not a rocket I kept the compression close to stock.  If I was planning to use programmable fuel injection like on my Race Rabbit I would go higher with the compression.  With the factory system I am worried about detonation since I don't have control over the spark timing.

When I started this project I had not realized that the KH engine had a fairly high compression ratio.  I had planned to deck the head some to get into the low 8:1 range.  Since the engine is higher than this to start I just decked as necessary to get a flat surface.  In the end I think I took about 0.009" off the head, but in reshaping the combustion chamber probably ended up with slightly lower compression overall.

Turbocharged Engines
My technical knowledge of turbo motors before this project was very limited.  I have reviewed some of my old textbooks and read many opinionated postings on the Internet.  Correct or otherwise I have come to the following conclusions:

I am looking for a motor with good torque, reasonable gas mileage, minimal turbo lag and good power.  First the gas mileage.  Obviously on boost I will use far more gas than I would in a NA 4000 motor.  The efficiency of a spark ignition engine is driven by its compression ratio.  My current motor has 8.5:1 which is very close to the KH compression.  For minimal turbo lag I want to minimize the volume of air between the intake of the turbo (compressor side) and the intake of the turbo turbine (exhaust side).   To get the turbo to spin up quickly you need a strong temperature and pressure delta across the turbo's turbine.  This means keeping the heat in the exhaust stream before the turbo and using large piping after the turbo's turbine.

Since most of my experience is with normally aspirated motors some of this has been hard to get used to.  For a turbo street motor equal length exhaust runners is of no help and could likely hurt things as this means more exhaust volume and more lag.  The purpose of an equal length header is for exhaust scavenging by creating low pressure zones in the header.  Well the last thing we want in a turbo exhaust manifold is lower pressure zones.  This made my life easier as I had been planning to make a new header.

Like I said this section is my opinion and many others abound.  Every hand made exhaust header you will see for Audi I-5 motors searching the internet will be a long runner equal length header.  While this is where ultimate power is found I don't agree that it is the solution for a daily driver.  I hope to add a section on alternatives shortly.  

Exhaust Ports and Manifold
I have some exhaust gaskets from some VW application that have a generous port hole.  I don't know what they are from, but they are bigger than the ones I took off the turbo motor.  Below on the left is a picture of this gasket on one of the exhaust manifold ports.  You can see that the hole in the casting is smaller and offset from the gasket.  This is also the case on the head side, but they are generally better centered.  The picture on the left is one of the other ports that I have already gasket matched with a die grinder.  I blended it back as far as I could reach into the manifold.  On several of the manifold ports I have found that the opening is actually the smallest part and it opens up slightly.


The collector side of the manifold also needs help in its stock form.  The picture on the left is how it came from Audi.  I am sorry for the poor quality of the picture, but I had already started porting before I realized this picture was not so good.  What is strange is the size of the openings verses what is feeding them.  The larger opening in the lower right hand is fed only by cylinder #1.  The other two manage the remaining 4 cylinders.  My main goal here was to open up the top opening.  I also blended the other two openings.  The opening in the bottom left corner has a sharp corner to turn just before exiting and there is ton of material available for removal here on the inside of the turn.  I cleaned some of it up, but the rest was hard to reach with the cylindrical burr I was using in the die grinder.


Additionally I have opened up the exhaust ports on the cylinder head quite a bit.  On the valve end of the port I found that the machining to install the valve seat left very sharp edges and also blocked flow out of the combustion chamber.  In order to fit in the head the exhaust port shifts to one side as it moves from the valve to the manifold.   I rounded edges and blended transitions.  The exhaust ports required a lot more work than the intake.

The valve guide blocks a great deal of the flow as seen in the picture below.  I had a little trouble getting a good picture, I may post a different one later.  Sadly the valve guide is already very short so I did not feel comfortable cutting it down (plus it would be a lot of work).  I did however change their shape from convex to concave giving a slightly larger cross section to the port.

You can see like the exhaust manifold earlier that the exhaust ports on the head are much smaller and offset from the gasket.  That has been fixed.


Intake Ports
On both the head and the intake manifold the gasket matches almost perfectly with the head.  The intake port size is already quite a bit larger than a ported GTI head I was comparing it to.  The ports are also slightly higher on the head maybe to minimize the turn the flow has to make.

I found that at least with my head the quality of the casting was not as good in the intake ports as in the exhaust.  It was not as smooth and the mold separation seams were more pronounced.  As with the exhaust side I had to clean up a little right under the valve seat with some minor blending.  I also tried to clean out some of the casting artifacts.

As with the exhaust side the valve guide protrudes quite a ways into the port with a casting bulge around this protrusion.   I changed the shape of these slightly as with the exhaust.  There was a little more room to work on the intake side, but I had to reach a little farther as the intake port is longer.

Combustion Chamber
As best I can tell the combustion camber is the same as other VW/Audi motors of the era.  I did not measure the volume of the head, but compared it to some I have by eye.  It is slightly different in shape than the Digifant 10:1 head.  Surprisingly the Digifant looks to be slightly larger in volume.  Below on the left is a picture of the untouched combustion chamber.  Next to it is a close up of the exhaust valve seat.  I found that the exhaust valve seat was sunken more than the intake.  This means the first few millimeters of lift would have very poor air flow.  The far right picture shows where I carefully removed some material to bevel into the valve seats.  It took some time since I don't want to buy new valve seats so I had to be very careful not to nick them.

These pictures are from my first head before I found it was badly warped.  I did similar work on the second head, but continued by removing all sharp corners and then polishing the entire combustion chamber after ceramic coating it.  Edges and corners get hottest first and lead to detonation, so I wanted the combustion chamber as smooth as possible.  I was careful to remove only what was necessary since this will slightly reduce my compression ratio.  The head has been decked slightly to compensate for some reshaping I did in the combustion chamber and to get a flat mounting surface.

Intake Manifold and Throttle Body
The KH motor turns out to have a desirable intake manifold.  The single round throttle body makes upgrades to electronic fuel injection easier as it allows room for a fuel rail where the spread bore throttle bodies make it difficult.

The first thing I noticed about the intake and TB where the features (or bugs) designed to make the single throttle plate feel like a progressive spread bore.  First the throttle plate has a huge ramp on the back side to minimize the open cross section at low throttle openings.  Second the manifold has two huge lumps to further block air flow at small throttle openings.  Needless to say these are all gone now.  Below are some pictures of before and after.  You can see even the choice of fasteners for the ramp cause a large flow restriction.

The throttle is a little twitchy now as you would expect with a single large plate.  I think a conversion to electronic fuel injection would help this.


Also you can see the after shot of the intake where I removed A LOT of aluminum to do away with the  humps.  I also used a sanding wheel with a long arbor as far up the intake ports as I could reach.  On most I was able to get at least 3 inches up.  As I recall the total runner length is about 12 inches.


Feb 12th Update.
Well a lot has happened since I last updated here.  It turns out my first head was unusable.  I had noticed a crack between valve seats when porting but this had not concerned me all that much.  I didn't really look the head over more than that.  I got a call from the machine shop and they said the deck was about 0.010" out of flat on the gasket side and about double that on the cam side.  When the cam was placed in the head it rocked like a see-saw.  The warping on the deck side would not have been a problem as they could have just machined it, but repairing the cam side would involve line boring and inserting cam bearings.  I am told there is no one in our area that does that and I just saw dollar signs.

So I ordered another one from a local salvage yard.  It turned out to be the wrong part (my fault as I thought the motor was from an 86 car until this week).  The MC head from 86 to 88 motors will NOT fit on an 85 block.  Hopefully I will have the correct head (and head gasket) within a week or so.

March 30th Update
I have had the new to me '85 head now for almost a month.  Extensive overtime at work has put this project on hold.  My confusion on the year of the motor also meant I had the incorrect rings as '85 motors have slightly lower displacement (bore).  I have had a chance to work on the bottom end more since the last update.

With help from Blackdog Motorsports my pistons now have ceramic coated tops and moly coated skirts.  Below is a picture of the now high tech pistons back in the block.  I have some other pictures of the pistons themselves that I will get up sometime.  The bottom end is most of the way together.

Below  is the engine 95% of the way together.  The coated exhaust manifold is the only external sign of all the coating work that went into the engine.  I still have a few little things to bolt on, but some of those will wait until it is in the car.