Project Potential... variable vane turbo on 1.6l

[fspGTD]

My dad's '84 Rabbit (1.6l NA Diesel, 2-door, atlas grey) ran low on coolant late last week and seems to now have developed a leaking headgasket.  He said there was excess white smoke out the exhaust, and then he found a low coolant level - it like sounds like the coolant was leaking into the combustion chamber to me.  The plan is that I'm going to fix it for him, and do other work that it needs that makes sense at the same time (like motor mounts and timing belt), and maybe a turbo upgrade.  Since it sounds like he caught and shut down the motor before the overheat got really serious, I'm optimistic there are good prospects that we can get by with a headgasket replacement and not a head rebuild.   I think there was a slow coolant leak out of a cracked expansion chamber which was caused by overpressure from a faulty coolant pressure relief valve.  I'd plan on checking the head with a straightedge to verify no permanent warpage after I get the head off.  This motor is a good rebuild still with relatively low miles for what these are capable of; it uses one quart of oil per 6 tanks I think dad said.  Not really good, but not bad either.  Always easy to cold start so seems to have good compression.  I will be putting the head back on there probably with some raceware head studs I have been saving for it.

We have been toying with the idea of upgrading this motor to turbo for quite a while now, and this time when the head comes off, it might be the right time to go the whole nine yards and do the turbo upgrade while it's down.  Since it doesn't have piston squirters, my plan would be to run the fueling conservatively, and/or use a big free flowing exhaust.  We have already prepared the car in this department by installing a 2.25" techtonics exhaust which is on there, which is what I'd run with the turbo.

The question on my mind at the moment is turbo selection.  This car is dad's daily driver, and I'd love to set him up with a sweet performing ride, and maybe experiment a little bit in the process if it is a calculate risk.  We've got a complete 1.6lTD turbo external bolt-on setup collected here (with the big KKK turbo), which I know would work and is generally a "bolt on" option.  But, we also have collected a used 40k mile old"ALH" TDI variable vane turbo that I know to be in good condition.  I am considering the ALH variable vane turbo vs the factory 1.6lTD turbo.  With the ALH turbo, torque would be monstrous, especially if it were coupled with a long-runner gasser intake.  With either turbo, the downpipe will need some work if we want it to be free-flowing, the stock 1.6lTD downpipe is not an option.  So it might not be a lot of extra work to adapt the different turbine outlet position of the ALH.  The other factor is that the air intake system with the ALH variable vane turbo would need to be changed (and custom).  But to install the 1.6lTD airbox would require welding the airbox moutning bracket, which isn't so easy for me to do anyway, so again in the department of the airbox and air intake, it might not be incrementally that much more work to adapt the ALH turbo.  In either case the fuel filter will need to be relocated (I think) and I haven't decided if factory TD position will be used welding on the bracket to the firewall, or to adapt to the in-fender fuel filter location that I have been very happy with on my GTD Autocrosser.  Regarding ALH turbo boost control, I worked out a few weeks ago on a napkin I put somewhere a scheme for controlling the boost of the "ALH" wastegate involving vacuum, manifold pressure, and some variable bleeder valves.  It is completely untested, but it might just work!  I need to go and find it.  I am thinking I would want to run unintercooled to keep the project simple and not dragging on too long, at least initially.

Just throwing out ideas at this point for your guys consideration / comments...  whatever the plans are, the project will be starting very soon, in order to try and minimize turn-around time and get dad's daily driver back on the road with minimum of delay.  I might go to the shop to rip the head off of it tonight or tomorrow...

[jtanguay]

I'd go with the TDI turbo.  Instant boost is awesome

[lord_verminaard]

Sweet.  Scan that diagram of the VNT controller and post it!  I'd love to see what you came up with.  It seems the TDI VNT is a tad small for the TDI when more performance is desired out of it.  Running stock-ish boost levels (for the TDI) on a 1.6, which I think is somewhere around 14-16 psi, would probably be a pretty sweet ride, and the VNT's size should keep it fed with plenty of air at the top end too.  That much low-end boost should make it fly off the line as well.

I'm still toying with what I actually want to do with the Rocco, (although I am still just getting it back together in gasser form!) as 1.6T's are REALLY hard to find anymore, a 1.9IDI would be sweet but even more rare, a TDI would be nice, but I want a little more rpm potential than I think a TDI could provide.  Pretty much all options are out of my budget now, but the prices of salvage TDI's are coming down, and finding a mechanical pump to run it shouldnt be too hard either.  

Depending on how this experiment comes together, it might help my decision some.  Who knows, maybe the VNT turbo will really make these 1.6's into hot-rods.  

Good luck, we will be waiting.   :twisted:

Brendan
84 Scirocco 8v
00 Camaro L36 M49

[QuickTD]

In the "normal" configuration the VNT vanes are pulled in to the "closed" or max boost position by vacuum, the modulating solenoid then bleeds in atmospheric pressure as commanded by the ECU to reduce the boost. It would stand to reason that you could just spring the VNT actuator fully closed and then use boost pressure at the vacuum fitting to open the vanes as the pressure rises. Springs are cheap, simple and reliable like an anvil.

  I don't think it would be too difficult to fab up some kind of bracket to use a spring on the actuator, then hit the local industrial supply for the proper spring. You can get a rough idea of what you need by calculating the area of the diaphram, from there you can establish the force required to hold the vanes closed up to your desired boost level. It would take some fiddling but I'm sure it could be made to work. I use a "helper spring" of sorts on my (K03 external actuator) wastegate to increase the boost pressure without resorting to bleeds and other messy plumbing, works quite well.

[DVST8R]

sounds great, I have been toying with the idea of going to a VNT 17 or 20 instead of twins, but I think the bling factor and the cheapness factor will keep me to twins.

However back on topic QuickTD mentioned "In the "normal" configuration the VNT vanes are pulled in to the "closed" or max boost position by vacuum, the modulating solenoid then bleeds in atmospheric pressure as commanded by the ECU to reduce the boost. It would stand to reason that you could just spring the VNT actuator fully closed and then use boost pressure at the vacuum fitting to open the vanes as the pressure rises. Springs are cheap, simple and reliable like an anvil."

As I reacall it seems that there is another smaller port on the back side of the actuator that you can just run boost into to have the vanes move, however you "MAY" need to be able to contorol how much boost, and how fast it comes on, as it may only take a few pounds to move the vanes there full cycle or it may be somewhat linear, things that need to be worked through, but should be simple enough to solve. My vote is definalty VNT.

If there is anything that you need fabbed for the project that wouldn't be really costly to ship, or wouldn't need trial and error fitment let me know, I will do my best to help you out. (I have CNC plasma, and MIG setup for mild steel, our TIG is on the road for the next bit, so no aluminum or stainless welding right now, but I can still cut it.)

[fspGTD]

In the "normal" configuration the VNT vanes are pulled in to the "closed" or max boost position by vacuum, the modulating solenoid then bleeds in atmospheric pressure as commanded by the ECU to reduce the boost. It would stand to reason that you could just spring the VNT actuator fully closed and then use boost pressure at the vacuum fitting to open the vanes as the pressure rises. Springs are cheap, simple and reliable like an anvil.

  I don't think it would be too difficult to fab up some kind of bracket to use a spring on the actuator, then hit the local industrial supply for the proper spring. You can get a rough idea of what you need by calculating the area of the diaphram, from there you can establish the force required to hold the vanes closed up to your desired boost level. It would take some fiddling but I'm sure it could be made to work. I use a "helper spring" of sorts on my (K03 external actuator) wastegate to increase the boost pressure without resorting to bleeds and other messy plumbing, works quite well.

Hmm!  I am liking that idea!  I am going to take a closer look at the diaphgram mechanism.  Maybe I can open it up and change the spring from one side to the other of the diaphgram.  And I agree in theory at least, that it would work!

I found the sketch of my other idea.  It connects vacuum from the vacuum pump to the vane control diaphgram through a restrictor... let's call that restrictor "R1".  It also has a "T" connection at the vane control diaphgram so it is also connected through another restrictor ("R2") to manifold pressure.

My conjecture is that by varying R1 and R2, boost control could be tuned without any modification being required to the diaphgram housing or vane control system.

I worked out a few scenarios... here is how the numbers would work out, as well as some things I determined...

(Note: resistance of zero means there is no resitrction to airflow.  A greater resistance means a smaller orifice that restricts airflow more.  Negative pressure numbers mean pressure is vacuum, zero pressure means atmospheric pressure, and positive pressure means that pressure is above atmospheric.  Assuming vacuum pump pulls -14psi.)

Things we know about the vane control mechanism:
* When vane diaphgram pressure is low enough (far enough negative) the vanes will more into a closed position, causing turbo to speed up and boost pressure to increase.

* When vane diaphgram pressure is high enough (which is a negative number but may be close to 0) the vanes will be in an open position which will cause turbo to slow down and boost pressure to decrease.

* When R1 = R2: If boost = 0, vane diaphgram pressure = -7 psi. (possibly completely or partially closed?)  If boost = 15 psi, vane diaphgram pressure = 0 psi (completely open.)

* To increase boost pressure, increase R2 and decrease R1.
  Example: When R2 = 2 x R1: If boost = 0, vane diaphgram = -10 psi.  If boost = 15psi, vane diaphgram = -5 psi.

* More restriction (R1 + R2) makes less flow (load) through the vacuum pump, but slower reacting vanes.

How to tune it depends how quick you want it to react (I would want as quick reacting without fluttering while not causing too much burden on vacuum pump), how much max boost pressure is desired, and how much vacuum is necessary to get the vanes to open and close.  But if 2 variable bleeder values were used, it would be fiddled with pretty easily it seems.  I can say that the vane mechanism is completely open at atmospheric pressure, and seems to need some amount of vacuum before it starts moving towards closed (I just don't know how much.)  I also don't know how much vacuum is needed to completely close the vanes.

The OEM TDI N75 electronic solenoid valve is similar to the concept proposed above in that R1 is fixed and it connects vane control diaphgram to vacuum pump, but is different in that R2 is variable and connects vane control diaphgram to constant atmospheric pressure source (pressure = 0).

[fspGTD]

As I reacall it seems that there is another smaller port on the back side of the actuator that you can just run boost into to have the vanes move

I thought of that possibility also, but wasn't able to make it work.  It would require the linkage coming out of the turbo side of the VNT actuator diaphgram "can", which moves in and out of the can, would need to be sealed airtight to the opening of the "can".  I couldn't think of a way that the setup could be easily modified at least that would achieve this.

If there is anything that you need fabbed for the project that wouldn't be really costly to ship, or wouldn't need trial and error fitment let me know, I will do my best to help you out. (I have CNC plasma, and MIG setup for mild steel, our TIG is on the road for the next bit, so no aluminum or stainless welding right now, but I can still cut it.)

Nice... thanks for the help, that is some sweet equipment you've got access to.  I will keep it in mind!  

[DVST8R]

Good to know about the linkage ect... hopefully either QuickTD's idea with the spring or your restrictors, or a combination will work! I'm now having dreams of a twin setup with a VNT on the small side and a ball bearing garrett or a pdr holset on top. :twisted: Monster bottom end torque massive topend flow.

[fspGTD]

I got the VNT diaphgram can off and took some measurements and some pics:





It takes 3 pounds force before it begins to move.  It reaches the end of it's travel when 40 pounds of force are applied, and the length of its travel is 5/8".  The inside diameter of the can is about 2 7/16", which gives the diaphgram an area of approx 4.66 square inches.

That means it would take -.6 psi (vacuum) to make it start to move and -8.6 psi (vacuum) to make it reach the end of its travel.  I am able to make it reach the end of its travel by applying suction with my mouth by the way.

Note that the actual travel of the VNT mechanism is probably somewhat less than the full travel of the diaphgram can.


-----------------
re: QuickTD's idea of opening the can and changing the spring

Unfortunately, the can is crimped together, and I can't figure out how to undo the crimp without damaging the can.  It looks like it is swaged together and is not designed to be disassembled.  You can see in the pics the band around it which is where the crimp is.  Does anyone know how to disassemble it so that it can be reassembled afterwards?  If I have confidence I can disassemble it without destroying it, I would be willing to try it.  But right now I can't figure a way it can be done.

[jackbombay]

Does anyone know how to disassemble it so that it can be reassembled afterwards?

  Maybe dremmel around the crimp with a thin cutting wheel a little a t a time so it does not get to hot and melt stuff inside.

   then reassemble it with 2 large home made washers with the diameter of the hole in the "washers" the same as the OD of the whole actuator, 3 or 4 nuts and bolts around those two two large "washers" (one on each side of the drimp) and mabye a bit of sillycone before reassembly to ensure a good seal.

[fspGTD]

Does anyone know how much vacuum our vacuum pump pulls?

[fspGTD]

That idea to clamp two washers together around the outside of the can might just work!  But I wonder where I would get washers with inside diameters that large?  The inside diameter would need to be about 2 9/16".

I think I might re-install the VNT diaphragm can onto the turbocharger, and measure the amount of travel it requires to the "full closed" position so I can calculate the require vacuum to completely close the vanes (or if I can find a vacuum gauge, I would hook it up and directly measure the vacuum required.)

I whipped up an excel spreadsheet I am using to calculate the wastegate can pressure from various manifold pressures and various R2 to R1 relationships.  I am hopeful there is a setting that is workable... the ideal goal would to have the VNT vanes be in a full closed position at 0 psi manifold pressure, but full open at not too high (IE: not much over 10psi) positive manifold pressure.  This could be a viable scheme if the vacuum pump pulls strongly enough, and/or if the vacuum required in the VNT diaphgram can to move the vanes completely closed is weak enogh.

[jackbombay]

Here is a drawing of what I was trying to describe earlier,

 

[fspGTD]

Great pic!  

[jackbombay]

That idea to clamp two washers together around the outside of the can might just work!  But I wonder where I would get washers with inside diameters that large?  The inside diameter would need to be about 2 9/16".

  It would be a chore, but with a cutoff wheel in a dremmel then a sanding drum in a drill press you could make them, or dremmel then a carbide bit in  adie grinder maybe?

    :idea:  Bi-metal Hole Saw would make some fairly quick and easy as well, but will probably be 15-20 bucks for the hole saw and hole saw mandrel.