Porsche 911UK Forum

Welcome to the @Porsche911UK website. Register a free account today to become a member! Sign up is quick and easy, then you can view, participate in topics and posts across the site that covers all things Porsche.

Already registered and looking to recovery your account, select 'login in' and then the 'forget your password' option.

Cat Bypass

PPBB, excellent mate that was what I was looking for, with the cats removed does it increase sound levels by much. 20bhp is some gain, if anywhere near that it has to be worthwhile.

So basically the cars management can cope with the fact there's no cats, might be a stupid question but with no lambda sensors how can it alter fuelling ?
 
Drew1209 said:
PPBB, excellent mate that was what I was looking for, with the cats removed does it increase sound levels by much. 20bhp is some gain, if anywhere near that it has to be worthwhile.

So basically the cars management can cope with the fact there's no cats, might be a stupid question but with no lambda sensors how can it alter fuelling ?

Cats of the late 90's weren't anywhere near as good as modern stuff in terms of flow so whilst on a modern car you won't generally see much of a gain on something like a 3.4 996 the cats are fairly restrictive! There are quite a few dyno plots kicking about showing the effects of a decat. They do get a fair bit louder without them in though so it's worth considering if you have neighbours that aren't car fans. I have a single silencer per bank and that's it so mine is fairly loud, although from inside you can barely hear it it's over the noise limit for the circuit where I work :D

I can't speak for the later 3.4 and 3.6 cars with post cat lambdas as I don't have any experience of them, aside from to say you will get a CEL as that is the purpose of the post cat lambdas - however the early DBW throttle cars really do not care at all. The pre cat lambdas are purely used as fuel trim so without them it just runs from the MAF. The ECU uses a mathematical model of the engine to determine the correct amount of fuel to inject based on feedback from the MAF i.e it knows with good accuracy exactly how many grams of air are flowing into the engine at any moment in time so it knows exactly the volume of fuel to inject to achieve the air/fuel ratio the model tells it to target.

For example lets say we have 100gs of Air per second entering the engine and for the given torque request (a function of pedal position) and engine speed the ECUs model tells it that it needs to achieve lambda 1 or 14.7:1 AFR (14.7 parts of air to 1 part of fuel). Some quick math in the box then tells it that for the next injection event it needs to inject 6.8(and a bit) grams of fuel, it knows the specific gravity of pump fuel is about 0.76 (or it was in 1999) so it knows that it needs to inject 5.168ml of fuel. As the ECU also knows the injector size and the operating pressure in the fuel rail it can then determine exactly how long to open the injector for to provide this amount of fuel on the next injection event.

The lambda sensors are just used as a trim to this calculated value and then only in engine operating regimes where Lambda 1 is targeted (to keep the cats happy and operating), once you get past 4500RPM and 60% or so torque request then the ECU will operate purely from it's model based on a measure of inducted airmass.

It's because of this system you can bolt a load of mods onto an engine 300CC bigger than that which the ECU was originally calibrated for and find it still achieves 0.91L at WOT and peak power RPM exactly like it did on the standard engine because the MAF is still able to accurately measure the increased air flow and the value is still within the model the ECU uses to calculate fueling.

What's really clever is that if the MAF fails the ECU then reverts to a speed/density operating regime where it uses throttle position as it's primary load sensor and operates from another model that ties throttle position and RPM to the expected amount of air mass, it's still pretty accurate on fueling even at that! Unless you've done the above load of mods and a larger engine thing.... if you have the model is quite a way off and it's properly lean under load :lol:
 
To expand on this slightly here is some data from my car taken during testing today:



The top green line is GPS speed, The second blue line is Lambda on the left bank, the third purple line is MAF (this is an arbitrary reading unfortunately rather than an SI unit), the fourth dark blue line is engine RPM and the lower red line is TPS.

From the graph you can clearly see the relationship between throttle position, MAF reading and lambda. With the throttle wide open and the engine RPM around 5000RPM it is running purely from the model in the ECU to determine how much fuel to inject and we can see it's pretty good at it by how stable the Lambda is at 0.9 +/- 0.1. Note how the MAF reading reflects the shape of the throttle position trace, as throttle position directly relates to the inducted airmass and therefore the MAF sensor output.

Also quite interesting to see it's done just over 100mph to 125mph in top gear with a slow throttle roll on in a little over 5 seconds and I came off the throttle about 3/4 of a second before peak speed was reached! :eek:
 
I have a single 100 cell cat in the 850R- which you can see through, so it's not what I'd call hugely restrictive. Of course there will be a defreeenof restriction but I can pass an MOT anytime, anywhere. What's the performance that I am leaving on the table, and does the convenience out weigh it?
 
This is turning into a very interesting thread :thumb: thanks for the input gents. :thumb: :thumb:
 
Of interest to me too, any more info on this one anyone may have, has I'm embarking on a Kline exhaust, it came with 200 cell and bypass pipes but do not know which to fit?
My two major concerns being noise and MOT passing 😁
All this to install after the Dinslaken jaunt :thumb:
 
My limited experience of 200 cell cats is they will pass an mot if hot enough and no running faults .. 100 cell and no .. not the ones i have seen which to be fair isnt many !
 
diverzeusy said:
Of interest to me too, any more info on this one anyone may have, has I'm embarking on a Kline exhaust, it came with 200 cell and bypass pipes but do not know which to fit?
My two major concerns being noise and MOT passing 😁
All this to install after the Dinslaken jaunt :thumb:

What model car?
 
My car a 99 C4 3.4, one of the early egas cars.

As commented on above, I'd also like to thank everyone for there input especially PPBB very informative.
 
996ttalot said:
diverzeusy said:
Of interest to me too, any more info on this one anyone may have, has I'm embarking on a Kline exhaust, it came with 200 cell and bypass pipes but do not know which to fit?
My two major concerns being noise and MOT passing 😁
All this to install after the Dinslaken jaunt :thumb:

What model car?
Hi My car is 996 turbo 2000
 
I thought I might chime in to give a little more info for early 3.4 NA owners. As has already been mentioned, there is no such thing as a requirement for backpressure. If you look at motorsports where there is no noise restriction, such as drag racing, frequently open stubs are used because these offer the least resistance (the least backpressure if you will). The thing with exhausts on four stroke engines is that they need to be designed to give the greatest exhaust velocity. The faster you can get the exhaust gasses out of the engine, the more power you will make.

In order to achieve this an exhaust needs to be designed with a diameter that is sympathetic to the volume of gas produced. If the bore is too big then you will actually slow down the gasses in much the same way that a river slows down when it widens. Obviously, at different revs then engine produces different volumes of gas, so the ideal exhaust would vary in diameter in line with revs. Until this fantasy pipe is invented we need to settle on a compromise diameter of pipe.

If you are going catless, you will see bigger gains than going with 200 cell cats all across the rev range. This is not just because the cat matrix isn't in the way, but also because the bulb in the exhaust that holds the cat takes energy (velocity) out of the gas by slowing it down where it widens at the start of the cat and then another restriction is faced at the other side of the cat where it tapers back to the normal exhaust diameter.

To get maximum power out of a 996 NA exhaust, the mufflers would also be changed to something that used a straight through design, such as the After Hours one:
https://www.6speedonline.com/forums...ted/232933-996-after-hours-sport-exhaust.html

The only problem with this is the very tight radius of the bend between the mufflers. Although the design is much less restrictive than the normal 996 muffler, more power could be released by increasing that radius.

I'm currently toying with getting something made in the after hours style, but replacing the first muffler with a 200 cell cat, and running the manifold direct to the cat & muffler on it's own side. If I can get the radius of the bends in the exhaust large enough (basically 6" or greater) then I could lose a collossal amount of weight from the rear of the car, and gain somewhere in the region of 15-20 hp.

As regards lamda sensors, I strongly advise you run them. Without the fine tuning of the mixture that they provide you may have trouble with rough running. I'm speaking from recent personal experience with my early 3.4 car here. Even with a decat pipe, I would still keep the lamda sensors. With one of my lamda sensors out of action recently, the bank it was on failed the emissions test for the MOT, and the car only passed because the tester then tested the bank with the working lamda sensor.
 
Also, if you decide to install decat pipes and intend to put in the cats for the MOT, make sure you use all stainless nuts & bolts. This will preserve your sanity when you come to swap them around.
 
The After Hours mufflers are much louder than the stock ones, so that's something to bear in mind. Also, I reckon that you wouldn't see much change from £1200 if you bought & imported them, whereas I think you could probably get a pair made for about £500 at a custom exhaust fabricators in the UK.
 

New Threads

Forum statistics

Threads
124,353
Messages
1,439,439
Members
48,708
Latest member
JLav211
Back
Top