Project Potential... variable vane turbo on 1.6l - page 15 - IDI Engine - VWDiesel.net The IDI, TDI, and mTDI source.

• #210 by Jetman on 20 Aug, 2005 12:06
• My new engine project
  
  click
  
  click 2
  
  click 3
  
  click 4
  
   Welcome
• #211 by fatmobile on 20 Aug, 2005 21:10
• I saw some 3 piece thrust bearings. 1st click third picture up from the bottom.
   I've heard of them tearing up cranks. Getting off the VNT topic but had to mention it. Ooops, it's the 1 piece thrust bearing that messes things up.
   Great pics. Nice car for your new engine.
• #212 by Maarten on 21 Aug, 2005 05:33
• Phat ride.. but you need to clean the enginebay to match the outside a bit
• #213 by vwmike on 02 Oct, 2005 13:01
• The best method of control I've been able to come up with so far is to run an electronic boost controller like a Profec A as it integrates "fuzzy logic" and ought to be able to anticipate boost spikes and such. Instead of hooking it to manifold pressure, I would hook it to the vacuum pump.
• #214 by fspGTD on 03 Oct, 2005 01:31
• In defense of my VNT controller: it really did do everything I was hoping it might have: it gives excellent performance - both in terms of broad powerband and excellent throttle response (I believe superior than stock TDI, which lags for a split second after throttle application while the vanes close), and it gives good fuel economy (> 50mpg was measured on several consecutive tanks.)  I am not convinced that a few pounds excess boost at lower throttle ranges (AKA "hyperactive" boost) is a bad thing.  What happened after driving it more, was that you just get used to it boosting nearly all the time, and that it doesn't mean anything bad is going to happen.  An unexpected side benefit of the extra boost with the efficient intercooler, was that it really helped the engine stay at very stable operating temps (in fact, the temp gauge was reading 2 "clicks" cooler than when it was naturally aspirated.)
  
  One thing I also like about the mechanical VNT controller - is that like the engine it's attached to, it doesn't have any complicated electronics, wires, or sensors.  There is something to be said about a self-contained package without even any lines from the vacuum pump even.
  
  I agree it is important not to blow a boost tube with it though!  That is perhaps its achilles heel, and it prompted a few precautions.  For example, I proactively beaded every boost tube connection.  I also tuned the fuel injection LDA to severely reduce fueling (more than a standard turbo-diesel) if the boost pressure drops.  So the fueling is as boost pressure sensitive as I could make it.
  
  I think every VNT-15 turbo installation has one or more potential failure modes though.  The stock TDI setup can have the vanes stick closed and have the boost spike to very high levels.  Since my design has a wastegate (not to mention the pressure relief valve, both of which were not present on the original design) I'm not too concerned about the vanes getting stuch closed from carbon build-up... or even the VNT controller catastrophically failing somehow! :lol:
  
  Re: thoughts of your proposed setup:
  * VNT in a vanagon 1.6lTD sounds like a great idea.  The low boost will really help get the vanagon moving!
  * If you want to run a full dose of fuel right up to the smoke level at your target boost pressure of 15 psi, a wastegate will be required.
  * I can't think of any elegant way to implement a mechanical (IE: non-computerized) controller with all the inputs you describe.
  * If you try something out, let us know how it works.  Good luck!
• #215 by fspGTD on 03 Oct, 2005 15:09
• I sincerely wish you luck in coming up with an effective, safe, alternate design, and in executing it.  This will however be no easy task.
  
  But if you are not able to make the design foolproof, I'd advise you to not loose sleep over it as long as you are dealing with at least a half-intelligent driver.  Why not worry too much?  These VNT turbos are a lot tougher than you'd probably think from reading about the failures on TDIclub.
  
  For our purposes, they only need to be tough enough to last long enough for the driver to notice the funny noises, sudden lack of power, "0 psi" boost gauge reading, and pull his foot off the accelerator and limp over to the side of the road or parking lot to check things out... as happened once when I blew a boost tube early on in testing.
  
  The exact VNT-15 turbo used in this Rabbit 1.6D was previously misdiagnosed as a bad turbo when it was installed on a TDI.  The boost controls were installed incorrectly by a VW tech, with full vacuum signal straight from the vacuum pump to vane control diaphgram, meaning the ECU boost controls were totally bypassed and the vanes were put in constant full closed position during operation.  It was run like that for at least 2 test drives that I did myself, each of which eventually ended up with the engine going into a "limp home" mode only after several minutes of driving (the ECU was not quick to put the engine into this mode.)  Each full throttle burst resulted in pulsation sounds coming through the intake, which I believe was compressor surge.  Yet though all the extreme RPMs the turbo must have been turning to do that through those wide open bursts without a wastegate even to open and divert exhaust gasses around the turbine, the rotating parts held together and didn't instanteneously burst.  In fact, they seem to be working as good as new now, happily boosting the 1.6 IDI motor.  What I take away from this: is that I believe what kills these turbos is when they are run at too high RPMs for too long of a long period of time, long enough for the metal to failure from cyclic fatigure (aluminum in particular has a particularly low resistance to this kind of cyclic fatigue abuse.)  I think the wheel burst failures we read on tdiclub about must have been the result of the turbo overspinning for a quite extended period of time.
• #216 by jtanguay on 11 Oct, 2005 20:04
• well time to bust out the rpm sensor for the turbo!  just add another tach to your dash  :lol: problem solved!
• #217 by xud9te on 10 Mar, 2006 05:51
• Hi,
  
  Have found this thread through the TDI board, very interesting reading!!
  
  I am also carrying out a VNT conversion at the moment and have gone through surprisingly similar processes and problems!
  
  Libbypapa re. your poste above:
  
  "The control system I imagine for the vanes would be as follows.
  I would first have the vanes be normally open (spring pressure).
  I would have a throttle arm controlled spring to exert pressure to close the vanes. I would run another cable off the trottle arm, to a spring at the vane control. Whenever throttle was increased it would increase pressure to close the vanes.
  I would add a control that would be rpm dependant. With increase in rpm, there would be an increase in pressure to open the vanes. I am toying with the idea of using a vacuum diaphragm from a cruise control unit. I imagine that engine vacuum could be used to control the diaphragm. Hopefully, a happy relationship to the three controls could be found. I am wondering if vacuum from the vacuum pump is linear with rpm.
  I imagine that I would still want to use a wastegate system from the EGR port similar to the one you used in order to control boost spikes.
  I imagine that the control system outlined would have the effect of keeping the vanes open when not demanding more boost (in order to lower EGT's, and increase engine efficiency), eliminate the hyper boost response at cruising, and be far more likely to fail to a safe(er) mode (vanes open). "
  
  There is no boost dependancy here so the system has no feedback for such things as air temp, altitude or any other factor that a pressure referenced actuator accounts for.  Also the 3 springs would have to be so accurately balanced to get the correct boosting characteristics that it would be almost impossible!  Remember you would have to include the forces exerted on the turbo vanes by the flowing air in your calculations, which measuring would be no mean feat in itself!!  You are talking carburettor style accuracy!
  
  The rpm dependant method may yeild VERY strange boost characteristics, as although airflow into the engine may be linear with respect to rpm, air flow accross the compressor is not proportional to these for the boost curve.  An engine load referenced spring may work better.
  
  How can a cruise control work? There is no engine vacuum on a diesel?
  
  Also, the problem of the turbo overspeeding if a boost hose blows off is alot less catastrophic than if the vanes stick closed when the boost hoses are still on!  Zero p2 pressure means that the compressor wheel has effectively no forces accross it and as such is in no danger.  The problem is that the shaft speed may become high enough to cook the bearings or wobble onto the housings at either end.  This will NOT happen after 5 secs (at most) before you notice you have popped a hose!  In fact last time a hose popped at full boost myself and passenger actually noticed quite quickly, it was apparrent by the huge !!POP!! (and passenger cursing) that nearly made me swerve into oncoming traffic!
  
  FSPgtd:
  
  Have to say that your setup is looking pretty sweet, and such nice fabrication on the exhaust!  
  
  I had designed a very similar setup to yourself, although the only extra trick i am using is a ballvalve connected to the throttle arm that allows vaccum from the servo to pass to the -port on a 2 port actuator when the throttle is closed, meaning that when the throttle is closed beyond a certain point, the VNT vanes will be opened, allowing less backpressure at idle.  The ball valve has to be completely closed at cruise speeds however as this loses too much driveability.
  
  It also means that when you lift off suddenly or brake the vanes open up.  This gives a slight lag, but nothing like a normal turbo, and avoids using the compressor past its surge point.  Have you compared your boost at certain rpm’s to the flow map for the GT15?  It may be producing boost at the wrong point (but probably not by the sounds of how well it is running!!).  There is a map here: http://not2fast.wryday.com/turbo/maps (very good site) or try http://www.turbobygarrett.com/turbobygarrett/products/turbochargers.html  if you have the part number (vnt has a slightly different comp spec to wastegate GT15).
  
  There are so many adjustment points now though that testing is becoming a heavy task!
  
  There is:
  
  1) Vnt start point adustment (on threaded rod)
  2) Vnt finishing point (small allen stop on turbine body) (gt20)
  3) Bleed valve to boost + side of actuator to determine vane movement (boost) curve
  4) Ball valve active point
  5) Thinking of installing another bleed in the vacuum line to see if can use more proportionally!
  
  
  A better solution to the economy/overboost situation may be to use a tps module (or similar potentiometer) on the throttle arm that will vary voltage output for different throttle openings.  This can then be used as an input to a small circuit to control an amal solenoid bleed valve (through PWM) connected to the vacuum side of the 2 port actuator.  Maybe a very simple digital boost controller uses this same circuit (in reverse)?
  
  Anyway, very interesting reading guys, will keep following this one!!
  
  Cheers
  Greg.
  
  Citroen ZXVNTD (GT20)
• #218 by fspGTD on 10 Mar, 2006 08:44
• Greg - interesting idea about using a ball pneumatic valve to add some throttle control to the VNT system.  Do you have any pictures of your ball valve installation and/or throttle linkage?
  
  In this thread we discussed compressor maps (including the VNT-15) and how well they match to IDI engines:
  http://www.vwdiesel.net/phpBB/viewtopic.php?t=1576
  
  Driving the VNT Rabbit, which is actually tuned to deliver less boost pressure than it delivers in the stock TDI application I believe, I have not heard any sounds of compressor surge.  I also know the sound of a surging VNT turbo because I've heard it when driving an A4 Golf TDI before.  This particular car had it's VNT controller incorrectly hooked up to full vacuum position (this was the local VW dealership's fault), and the sound it made I would describe as an odd, shuttering "sh sh sh sh sh...." whenever the throttle was quickly lifted after brisk acceleration.  It didn't hurt the turbo though.  I think they are a lot more durable than most give them credit for.
  
  I am also kind of in denial about whether there really is any problem with leaving the vanes closed.  If the added load were significant, I'd think the engine RPMs would drop when the vanes were flipped from open to closed at idle, and I don't remember the RPMs changing very much when I tried that.  The extra airflow even at low/partial loads might actually help aid in combustion efficiency.  Also, any energy recovered from the pre-turbine backpressure in the form of boost pressure would be returned to the crankshaft on the intake stroke.  The only efficiency "loss" would be that energy lost to the turbocharger's aerodynamic/mechanical inefficiency.  The VNT turbo is pretty efficient, I think.  If the VNT 1.6lIDI Rabbit were not seeing as impressive fuel economy figures as it has shown (which are IMO, above and beyond what a "common" Rabbit 1.6lTD could do) I may be inclined to make a change, but it has just been excellent.
  
  There is a small potential problem that the added boost from the closed VNT vanes could be hard on the headgasket sealing properly when the engine is cold.  But I hope that a multi layer steel headgasket (in the plans next for this car) will address this reliability concern.
• #219 by Benjamin on 23 Apr, 2006 11:53
• i got a gt2052v wich came from a vag 2.5tdi v6.
  i readed the whole topic,
  but i am asking myself now,
  can you use a VNT turbo on a TD, if you use a wastegate and control the vanes by pressure? will it work good, an most off all, is the job worth it?
  
  tnx, greetz, Benjamin
• #220 by foxracer1 on 20 Feb, 2007 11:41
• So Hows the long term on the VNT rabbit? Any updates?
• #221 by RabbitJockey on 20 Feb, 2007 17:00
• rode in a vnt caddy today, but he just has vacuum applied from the brake booster, so it's always shut.  unfortunately i couldn't talk him into disabling the bov he's scured a little lol.  btu man that thing builds boost like you wouldn't believe. he doesn't have any fueling mods and it spools faster than anything i've ever ridden in, it is almost instant lol.  but then at 10 psi the bov wigns open and all you hear is woooooooooosh, if he stays on it he can bring the boost and hold 10psi with the bov open haha.
• #222 by SMOKEYDUB on 21 Feb, 2007 04:38
• thats kind of a bad setup. I think he is really overspeeding that turbo if hes just opening a dumpvalve and still hammers into it.... :?  Cool post though.
• #223 by jackbombay on 21 Feb, 2007 07:24
• Quote from: "libbybapa"

  Quote from: "SMOKEYDUB"

  thats kind of a bad setup. I think he is really overspeeding that turbo if hes just opening a dumpvalve and still hammers into it.... :?  Cool post though.

  
  
  Yup, I agree.  I'd be far more concerned about grenading the turbo just keeping the vanes closed and dumping through the BOV.  Cripes, can you say pumping losses and overspeed turbo...   :roll: The boost wouldn't kill the motor, its the shrapnel in the intake that might be a problem.  Hopefully he has an intercooler that will catch the big pieces.
  
  Andrew

  
  
     I must be missing something, but, IMO, if he uses an external wastegate to control boost there would be no overspeed issue as once max boost is reached exhaust is let out of the exhaust mani so it doesn't go through the turbine wheel.
  
    I plan on doing a similar system when I VNT my friends Caddy, vanes set partially open with a controller like "Named tintin" made and an external wastegate to set max boost.
• #224 by RabbitJockey on 21 Feb, 2007 16:30
• no they mean the bov on the intake.  and yeah thats what he's doing,  i'm trying to point him into the right direction of getting an external waste gate, since the only real way to control boost.  the other problem like you all said is that it has vacuum all the time.  but i think he's going to fix that