RE: K24 and boost

[QSHIPQ@aol.com]

A couple of thoughts following this thread.  The k24 is meant to develop more 
torque, audis effort to make a I5 feel like a v8.  Quite successful IMO.  For 
track use the k24 tends to heat soak on extended runs (just a straightaway 
will see a massive drop in boost, I recorded a 7pis drop on the front 
straight at RA in an S car - stage II).  Exhaust energy is what drives the 
turbine, so larger displacement motors use larger turbos.  The k24 is smaller 
and lighter, so it spools up quicker than the k26 (and it's maximum output is 
at 2000rpm vs ~4500 for the audi k26).

WG springs on 10vt.  M. Peterson hasn't presented this concept quite right.  
The WG frequency valve either does nothing, or adds boost pressure above 
spring pressure to the top of the WG.  You can test this by putting a LED on 
the WG plug.  In fact, when experimenting with springs, the best spring load 
for a given setup, will show the LED just lighting up (usually in the 
midrange - highest torque).  A stiffer WG spring will indeed give more torque 
down low, for more on this, see Maximum Boost by Corky Bell, the wastegate 
cracking section.  Basically, the rule of thumb is you want a WG spring to be 
about .5 of the maximum boost pressure of the system.  I tend to go a little 
higher than that with the k26, since it can ramp up quicker without surging.  
The k24 OTOH, can't take a faster ramp, lot's of folks find the undesireable 
side of the k24 when they try to do this (also one of my main objections with 
the addition of the aftermarket boost controllers).

WG springs on 20vt - Be careful, the k24 won't take a stiffer WG spring well. 
 The difference between the 20vt and the 10vt is that the WG regulation is on 
the feed side of the WG.  A percentage of actual boost pressure is sent to 
the bottom of the WG, the spring itself is really weak, like .3 bar or so.  
The problem with this setup, is that if the WGFV fails (and the modded cars 
valves do regularly), the baseling spring pressure is the default.  Also cars 
at altitude with the k24 need to be careful, cuz audi reduces the boost 
pressure at 3200 ft to the k24, a stiffer spring won't really allow that 
safety feature.  

As you can see from the above, computer programming is the key to making the 
turbo car right.  The WG springs, tho somewhat effective, are rather crude.  
To really get a given setup to work right, some tweeking of the boost profile 
in the chip/s is the way to fly, with a small increase of the WG preload 
(under the plug in the WG cap) during winter months.  IMO, the MAC-11> 
computers are better boost controllers than any aftermarket hardware.  Fine 
tuning it may take some time, but more parameters are monitored, and more 
maps are included, than the aftermarkets.

Single pass IC - do it.  You don't really gain all that much in terms of 
efficiency, but the flow really helps the low end on a 5kt.  Single pass IC's 
can be made with a stock IC with new end caps (the core of the single pass 91 
200 and the older 5kt are the same).  To use the 91 200 setup, you need to 
get the crossmember from the 91 (or do some cut and welding), and it's not 
cheap.  B4B, the IC is the *best* upgrade next to your computer mod.  In 
fact, you will find fuel to be a problem with high boost pressures.  I wrench 
on the exCarl Jerritts "beast" (5ktq with massive IC), and what that car does 
with 15psi is incredible.  

The k24 is a very driveable turbo, and a great upgrade for the older WX 
motored Urq's as well (Bob Dupree installed one in his 84, and it's very well 
matched at 13psi).  For track use, the stock or a modified k26 and/or hybrid 
is the way to go.  IME, the trick to real power is cold air, period.  Boost 
pressure is secondary.  A whole bunch of power out of a 10vt I5 can be made 
under 18psi.  

Before you swap turbos, consider what the car will be used for.  And, if it's 
mostly street, you may want to consider addressing the IC instead of the 
turbo.

My .02

Scott Justusson
'87 5ktqwRS2 -10vt
'84 RS2URQ - 20vt
'87 4Runner turbo - RESQ