TDI install retrospective: Vacuum System

Continuing the process of back-documenting what I did to put a TDI engine into an old (1972) air-cooled VW bus. Today's post is all about the vacuum system. It is super short, because there really wasn't much to it. Before I start, Hapy ThanksGiving and White Friday to my US readers.

Bus Needs
The TDI has a few things that require vacuum, and the bus has one: the brake booster. The original engine supplied vacuum from the manifold through a pipe that runs along the driver side to the brake booster underneath and behind the driver seat (under the bus). Fortunately, 30 years of engineering marvels later, the brake booster in the modern car is supported by a very similar vacuum. It is so similar, in fact, I just needed to plumb the brake booster outlet on the vacuum pump to that pipe and that need was solved. Power brakes? Check.

TDI Needs

from TDIClub

The TDI engine has it's own needs. The turbo needs a signal to dump boost. That actuator is vacuum powered. The EGR is vacuum signaled as well. Each of these have a different valve triggered by a signal from the computer (the N75 and N18 respectively). Since I pulled my EGR equipment, I simply blocked the signal line leaving the N18 and left everything else in tact. I mounted the control valves to the floor support that runs just to the inside of the spare tire well. The system needs the vacuum ball, the pump, of course, a bunch of T fittings and a source of clean air. I source the clean air from the main air intake. I will get to the air system later, but simply put, I took an old air-flow metering (MAF) housing and replaced the sensor with a simple vacuum nipple. This air flow metering housing was the same size as the rest of the fresh air intake, so it was just one of the components that were strung together after the air filter.

I spent a great deal of time drawing and diagramming before I did this. Ultimately, that was unnecessary; I was working from fear. So many systems work the other way: air pressure, water pressure.. the direction of the media is the same as the force, so when you look at a vacuum diagram, its all upside down. Air goes one way, the force of the air used to open and close things goes the opposite direction. Basic vacuum versus pressure, like the switch on your MityVac. It is the media (air) leaving that is the force, not it arriving.

With that perspective, I could made a plan. Ultimately, all I was doing was changing the lengths of the hoses, so this turned out to be pretty easy. I set the pieces out and started replacing lines with new vacuum hose. In many cases, I cut a different length than what was originally in-place because my placement required it. I retained all pieces of the system with the sole exception of the EGR (which I may re-introduce). The hose that would have gone to the EGR was simply terminated with a cap.

Prior related posts:
Preparation
Fuel System
Physical Mounting

Next related posts:
Air, Intercooler and Exhaust
Primary Electrical
Cooling
Secondary Electrical
ECU, dashpod and Sensors

Nemo Head Install

Recall my prior post about Nemo, T's 1997 Audi A4 (See Sadist Engineering), I had exhausted all alternatives and had taken the plunge to remove the head to replace the head gasket. Today's post covers the re-install.

Head Shop

No, not that kind of shop, though I am in Oregon.... This head shop does work on the heads and block of your car. I looked around and found a small one-man operation 15 blocks from my house: Wilson Cylinder Heads and Machine. I figured that if he did good work (like according to the reviews), it would be really great to have a machinist so close. Keeping the money in the neighborhood is very important to me too, as you've probably noticed. Anyway, I dropped off the head and we talked a little bit about it. Since the car had not been misfiring nor throwing codes, he was not angling towards a rebuild. We turned it over and he could see that the center 2 cylinders had suffered coolant leaks, and figured that was caused by a head gasket failure. Remembering that the head bolts were none-too-tight, I was starting to think that the head was redone at some point by the PO and it was not torqued down properly. Recall, the PO did upgrade the suspension and the turbo, so maybe the head was redone at the same time? Anyway, Luke (the owner-machinist) went through the head, looking at the cam shafts and valves etc while he did the pressure tests and tests for cracks and other failures. The head came back clean. So, with a complete gasket kit in hand, but only a small subset really needed, I picked up the head and set to re-installing it.

Exhaust to Head

After the troubled removal, I decided that I would install the exhaust manifold (with a new gasket from that kit) to the head before putting the head into the car. While the head was at the shop, I scrubbed the manifold with sandpaper, freeing it of rust and then shot it with flat black header paint. It looks much better now. Before mating the manifold to the head, I put the turbo bolt which sits closest to the head through the manifold. It won't go through after it is attached. I installed all of the fasteners (random so they may be) and torqued them to spec. Once attached, the head is heavier, which is saying something because a fully loaded head is already pretty heavy already. I verified that the head was still in TDC position by setting the valve cover on, and checking the dimple on the gear aligned with the dimple on the cover.

Prep Engine
None of the online advice I have seen specifies the use of any liquid sealant for the replacement of a head gasket. It appears that this was popular in the past, but gaskets have gotten much better and tolerances have gotten much tighter, so if you are doing this and think you need a sealant applied to one side or the other (or both), double-check with the manufacturer of the gasket. It didn't look like the old gasket had any goop on it (not that you could really 100% tell), and the head and block surfaces were totally clean when separated, so I concluded that no sealant was used before. So, I also did not use any.


I did, however, need to clean the block mating surface. Using a razor-blade, I scraped off every tiny anomaly. I verified that the surface was clean by running a new blade around the surface, holding it completely perpendicular to the surface. Any resistance or even a change in the sound lead me to something that needed to be removed. This takes a long time. You can't rush quality, and spending the time here will pay dividends later.

Head Meet Engine
With a heavy head in hand, I thought about how to get the head to line up without damaging the head gasket. I tried putting a chop-stick into one of the head bolt holes in the block, but that only sort of helped. It did help me get my bearings, but, ultimately, I pulled the stick out before the real attempt. I found that random hoses or other things kept getting in the way, so I had to remove a couple of things I had not removed for the head removal in order for me to get the head to set. Most memorable was the removal of another coolant hose and pipe that wrapped around the rear side from the driver side to the front passenger side.

With the impediments removed, I could ease the head onto the gasket/block. I set it down gently and wiggled it lightly to make sure it set squarely on the 2 pegs on the exhaust side. I checked 2 of the head bolt holes with a chopstick, and confirmed that the stick set all the way through to the block. Knowing the head was aligned, I set new bolts through the head-holes. I finger-tightened them, then torqued to 30 pounds, 40 pounds and finally 44 pounds. With my breaker bar, I did 2 1/4-turns (spec says 1/2 turn past 44 pounds). I put a little dribble of oil on the cam lobes and a little dribble into each spark plug hole. I replaced the vale cover gasket and installed the valve cover.

This post got super long, and the content creation has taken most of the last few months. Why? Well, remember how I'm slow? Yeah, someone says it will take 2-1/2 hours. For me, we double their number and increase it a measure, so 2-1/2 hours means 5 days for me. That's 5 full days, like 40 hours. If you spread that out across weekend days, it takes a long time. I know that sounds like hyperbole, but have you read this blog? Anyway, I'll post back on this soon.

Thanks, as always, for following along-

TDI install retrospective: Physical Mounting

Continuing the process of back-documenting what I did to put a TDI engine into an old (1972) air-cooled VW bus. Today I cover physically attaching the engine to the bus. This may sound simple, and if you only saw the end result, you might still think so. That's the beauty of what Hal built. Elegant in it's simplicity.

Transaxle Mounts

early mount attempts

The stock bus has three mount locations for the engine/transaxle combo: the front of the transaxle, the bell-housing in the middle and the "mustache bar" near the rear. I decided to reuse the transaxle mounts without modification. The transaxle wasn't really getting modified, so it should just fit. And it did. There were some who questioned the vibration through the mounts at the bell-housing, though. I believe there have been experiments by others where the bell-housing mount was eliminated, but I have also heard that this put stress on the output shaft and other components because all of the weight was being handled on the ends. Looking at the front-most transaxle mount, I'm not sure it really should hold up more weight than it was originally designed to hold. Your bus; your choice. Now, years later, I am fairly convinced that the bell-housing mount holds most of the weight of both the transaxle and the engine.

Motor Mount Attempts

bar to rear frame, 1st model

I looked at and tried a few different things to mount the engine. First, we tried building mounts that would attach to the sides of the engine to threaded bolt holes that were unused in the stock vehicle install. These bolt holes were on the block for industrial applications of the engine so, they should have been stout enough to do the deed. The idea was to reuse the original bolt holes for the mustache bar. The holes in the block and the mustache bar didn't line up very well and since the plan didn't include anything to absorb vibration, the bus would have shook like a paint mixer. I tried to accommodate this issue with some vulcanized rubber mounts, but it looked complicated and not terribly strong. So, I threw the whole thing out, and got to talking with Hal about it.

box being developed

Hal knows bus. He recognized that the rear bumper brackets are held to the bus with thick carriage bolts to the rear frame. The same rear frame that the mustache bar attached to, just more rearward. We figured that if we leveraged the mustache bar concept with a thick, straight bar between the bumper brackets, we could re-use the stock TDI passenger-side motor mount. To the right, here, you can see an earlier version. Hal and I agreed that something less complex would work.

Motor Mounts
Seeking something simple, we started with a piece of 1" angle-iron, onto which we marked and drilled holes to align with the bumper bracket holes. The bracket was then bolted to the rear frame so a straight bar could be measured, cut and welded on. Bar done.

Then, Hal welded together a simple box with a central support that would join the mount to the straight bar with bolts. The whole thing bolts together: the bar to the frame (re-using the bumper bracket bolts), the bar to the box (with a basic bolt-nut pair) and the box to the mount (again with a basic bolt-nut pair).

I have removed and re-installed the engine both with and without the transmission numerous times. This part of the system is a total non-factor in the process. I have also removed the rear bumper a few times independently of the motor mount, and while it is slightly more difficult than before, it is not really impacted: I simply support the bar with a floor jack, remove the 4 bolts (2 per side), remove the bumper, put the bolts back in and lower away the jack. Super simple. Of the things that were done for me for this build, this is the single greatest contribution. I said it then, but haven't said it enough since: THANK YOU, Hal. You really crushed it.

Measurements of the straight / cross bar:
length of main span, from bracket to bracket: 42"
angle-iron "bracket" depth (front-to-rear): 6"

Measurements of the box:
across the top: 9"
height: 7.75"
space between vertical bars: 3"

Prior related posts:
Preparation
Fuel System

Next related posts:
Vacuum System
Air, Intercooler and Exhaust
Primary Electrical
Cooling
Secondary Electrical
ECU, dashpod and Sensors

Sadist Engineering

I'm taking a quick break from the TDI install retrospectives for a rant, and an update on the A4 efforts. We'll be back to our regular posts next week.

With modern cars, the effort to wedge more features and power into a steadily shrinking engine compartment makes maintenance increasingly difficult. Back when we had the Saturn (Dude), we learned that in order to change the light bulb in the speedometer, the dashboard needed to be removed. What? Yep, that's right; we didn't bother. There are plenty of examples of this, but today's post comes from the efforts to get to the head gasket on a 1997 Audi A4 1.8 Turbo.

1.8 Turbo Head Job
Once I ruled out all other possible causes for the coolant losses in T's Audi A4 1.8 Turbo, I started trying to find someone else to do the work. After 4 months of attempts, and numerous rejections, I finally got a quote from someone: $1250US. Since we paid less than that for the car, I couldn't do that, but we are unable to use or sell the car in it's current shape either. So... into the darkness we go. We start with the usual internet searches for steps and find very lengthy descriptions. Unfortunately, many how-to's neglect to mention key points in a timely manner (I'm sure I'm guilty of this too). For example, the best step-by-step I found did not mention that you needed to disconnect the fuel pump relay and run the engine to depressurize the fuel rail. So, once you got through 15 or so steps and you were told to dismantle the fuel rail... you couldn't. There's 100psi of fuel in there, and no way to safely depressurize it now because the timing belt is no longer on the engine, and the front end has been completely removed.. Sweet. So, rather than follow bad directions, or release fuel all over the engine bay and shop space, we get to go our own way where the fuel rail is not disturbed.

This is when we discover just how sadistic the German auto engineers really are... or were in 1997.

1.8 Turbo Intake
The intake manifold is held on with 10 fasteners: 2 10mm nuts, 2 10mm bolts and 6 5mm Allen bolts. The 10mm nuts and bolts are on the ends and the Allens are between the intake runners. Of course, they are also mostly obstructed by the fuel injectors, so, again, if you want to remove the intake, it is best to dismantle the fuel rail. Grr.. With a wobble-head attachment on my 3/8" ratchet, I was able to crack all of the bolts except one. That one bolt eventually broke free after throwing various bits at it. I will need to replace at least that one bolt for re-install.

1.8 Turbo Exhaust
With the timing belt off, the evil coolant flange removed and the intake separated from the head (with shop towels covering up all openings), we shift to the exhaust side. The exhaust manifold is held on with 2 13mm nuts, 2 12mm nuts (yes, that's right. we've got both kinds of nuts) and, on this car, 3 6mm Allen bolts and 3 13mm bolts. I say that because someone swapped out some of the 6mm Allen bolts and replaced them with hex bolts. Now, the car is usually equipped with a heat shield which is held on with Allen and hex bolts. This car did not have the heat shield but all of the bolt holes were occupied with hex bolts which were in place nice and super-tight. Ugh. Love PO's. Regardless, the nuts are on studs at either end of the head, making them hard to get to, driving you to using old-skool spanners especially near the firewall. Once removed, I shifted focus onto the Allen bolts, and things got interesting. The exhaust manifold is very short, wrapping into a turbo that sits almost directly in front of the lower mounting bolts. So, from above, you need to thread a long Allen key between everything and torque the bolts free. Penetrating oil and long-stemmed Allen sockets recommended. Even then, you can't reasonably get to all of the Allen bolts under the exhaust manifold.

The best solution is to remove the turbo from the exhaust manifold and pull the head with the manifold still attached. Then, you can get after those Allen bolts. The turbo is held to the manifold with 3 17mm bolts. These bolts point straight up and have a slight recess around them in the manifold. Water likes to build there when the car is parked outdoors, so what do you get when you leave water standing on iron? That's right. Rust. Enter the penetrating oil again while you think about how you're gonna get a socket onto those bolts. Realizing you can't, again, you go old-skool with a spanner. I was very grateful for my penetrating oil comparison (See Hammered Rims Part 1) when I discovered the Kroil. After a few days of spray-and-wait, those bolts came off with very little extra coaxing.

Head Out
With the intake off, all the other little bits remove, the outlet flange off and now the turbo released from the exhaust, there was nothing more in the way. Except the valve cover and the 10 mystery head bolts hidden within. The valve cover is held on with a bunch of 10mm nuts and bolts. Once removed, the cover should just lift off, but nah. I applied some convincing with a rubber mallet and then with a pry bar on each corner. With the pry bar, I made sure to apply very light pressure so I didn't crack anything. With the valve cover off, we could see what king of fastener held on the head. After the swap-about on the exhaust, I couldn't just assume based on make-model-year. Turned out the bolts were the 10mm RIBE-style, so they could have been original. Regardless, these are typically sold as part of head gasket kits. With a bit in hand, I made to remove them and this was my first indication that the head had been removed at some point. I put my torque wrench on, and they required very little effort to loosen. By "very little effort" I mean I put my torque wrench on there and gave it a good bump with my hip and they broke free. I should have needed a breaker bar.

With the bolts all loose, I was able to wiggle the head. So, I put fingers into the coolant outlet, and my other hand around the belt tensioner and pulled straight up. It came free, and dropped the head gasket out as the head lifted out. What I saw was very interesting. The coolant passages rising up from the engine block met the gasket, but the hole in the gasket wasn't the same size as the hole in the block. The gasket has a little pin hole and the block was a decent-sized triangle. While I thought this was the application of the wrong gasket, according to Fel-Pro, it isn't. This is by design to correctly meter the coolant through the engine.

I will be taking the head to a shop to get pressure tested, decked and otherwise refreshed while I clean up the block face. The coolant passages look like some build up occurred behind the gasket, so I'll clean that up with a pick and a shopvac (so no bits fall deeper into the engine). I'll return to this effort blog-post-wise once I have the head back and the install is happening.

Thanks for following along. More next time-