K26 efficiency

JShadzi at aol.com JShadzi at aol.com
Sat Oct 27 23:54:14 EDT 2001


Richard, I have talked about this a little in the past, to see one of my 
posts check out:

http://www.80tq.com/thoughts/turbo_upgrade.html

IME, the K26 is a reletively inefficient turbo, quickly tapering out in 
performance past its limits.  As a general rule of thumb, don't run a K26 
past 15psi, anything higher creatins very high intake air temps, and 
aggrivating detonation and producing less overall power then at lower boost 
pressures- especially if you're running the MC intercooler.

The K26 is a small/meduim sized turbo.  The guys out there running 20+ psi 
out of their K26's would probably be much quicker if they turned the boost 
down about 5psi.  My experience with 80tq was that the car was faster at 12 
psi than at 18psi - mostly a function of air temps, which would rise at a 
much higher rate at the higher boost levels.  

Pressure does not make power, the flow of pressure does.  Pressure by itself 
is not a good indicator of performance.  Just because the boost gauge is 
saying 22psi doesn't mean the motor is making (stock NA power level) times 
1.49.  Flow and temp are big players into this, and much of it is at the 
heart of the turbo's efficiency.

Considering the K26 was designed and built in the late 70's/ealy 80's, it's 
safe to say that technology has come a ways since then.  Granted, the Garrets 
were too, but thier impeller technology has continued to progress, I am not 
aware of any impeller/turbine options for a KKK turbo...

In a message dated 10/27/2001 7:23:46 PM Pacific Daylight Time, 
billzcat1 at hotmail.com writes:

<< Just a quick question on the K26 turbos.
 
 How much boost can you efficiently run in a 5000/200 turbo?  ...The K26 
 turbo cars.  At some point turbos just heat air and don't really boost any 
 more?
 
 Any comments?  I know there are a few of you out there (JShadzi) that run 
 (ran) K26s at high boost, and I'm sure he's just one shining example of all 
 you who like boost ;)
 
 Thanks,
 
 Richard
  >>



More information about the quattro mailing list