EA827 16v exhaust manifold tuning for cast, vs, LHD 4-2-1 vs MK3 Kit car

They say knowledge is power. BHP type power?
So how do you acquire that power?

Well back in the day when I had open access to the Theo Thompson-Quartey's Garage Streamline dyno in London, I did many educational tests to gauge the effect certain components had on engine performance. I shared them with you the readers here.



Using my then buddy's car as a reference, we experimented many different things on the ABF engine, before I created my own test rigs in the form of JORAN/JENVEE.

This car, later shown here in the ITB thread.



ABF 16v exhaust tuning.



From long before I started to contribute to this forum in 2006, many of you owners of 16v engines who wanted to go faster, fitted a typical generic 4-1 "four branch".
There were many well published Club GTI dyno tests from the good ol days, some with near 200 bhp that was guaranteed to have one of those typical 4-1 parts fitted.
This often became a very copied path, even if the enthusiast had a standard tuned engine.
Those systems are still made today, and with folks simply copy fitting without testing. These systems found their way on ABF converted cars, resulting in a bit of fouling in the limited space between the bulkhead and rear crossmember. This lead to the production of the Golf MK2 ABF system to suit the ABF's height increase in a MK2 chassis.
Most folks never tested the benefits in detail, but simply believed the price they paid for the system must have some benefit to performance.

Having studied competition components Volkswagon Motorsport chose for GrpA and kit car engines ( see below), I was not convinced that a generic part 4-1 made by the aftermarket was directionally suitable for regular street engines with just stock engine tunes and hardware.
Why I say that? Because a GprA ABF motor was capable of ~ 230ps@ 7000 with usinmg a combination of production parts and some competition parts, such as a merged down-pipe and the more extreme kit car engine more than 270 bhp with a well matched tuned exhaust system.
After reading and the spec sheet over and over, I always wished I had the opportunity to test all different extremes of exhaust components back to back.
G3 GrpA combination of EU production 4-2-1 header and competition exhaust components



G3 Kit car competition exhaust components



Back in the day of tuning megasquirts, I had the opportunity that could not be refused when a Mr Hugo Pauwels (HPR), lent me a Golf 3 kit car exhaust 4-2-1 manifold and said I could test it back to back on the vehicle I was developing a calibration for.



Obviously the kit car part was optimised with long primary tubes, short secondaries and a tapered collector to tune the scavenging effect at up to near 9000 rpm those engines maxed out at. So what delta at WOT would be seen on an otherwise std engine? Read on...


That car was BenS' MK2 Golf that I encouraged him to fit an ABF engine to, having had a 1.6 8v engine. In general the engine used was a stock ABF motor with stock 8v exhaust and OEM cast manifold and an early iteration of a Megasquirt calibration to match.
Torque was significantly higher than the 140 lbft I would usually tested on a factory Digfiant 3.x engine at the Garage Streamline dyno.
However at the time we were still attempting to understand why the plateau of expected torque between 4000-5500 was not there and slowly dropping off post 4800 rpm. Torque in the region of 4000-5500rpm or so is something we had come to expect from an engine with a maximum ECU tune and as a result ~ 170 bhp@ 6000-6200rpm and 160+ lbft of torque.
As we developed the calibration, which was pretty stable, we elected to fit a Golf 3 Kit car exhaust to observe the effect.
As this car had constant progression, by the time the KC manifold was fitted, a Jetex 2.5 inch system was also fitted to replace the 8v cold end. On the OEM cast exhaust manifold with the Jetex, the torque profile had the same shape, ramping to 4500rpm and slow decaying at 5500 rpm.
With the calibration adjusted to take into account the scavenging effect of the new exhaust we observed what is seen in the dyno plot below.



Bearing in mind there was more to come from that engine on existing hardware later on, ultimately, the kit car exhaust seemed to have a top end benefit that would have contributed significantly with an engine with greater CR, open induction and matching camshafts. There an overall increase in torque, however on this application it would result in a small effect for the driver sub 4500 rpm. Throttle response to load changes was the same around city driving.

I should note, there was quite a bit of jiggery pokery involved as the engine location in the engine bay had to be altered to allow fitment of this specialised part. This size of the 4 primary tubes left little room between the steering rack and firewall.









Due to significant fouling encountered on the Kit car manifold in this MK2 car, I elected to fit an understated OEM part to this vehicle that also has VW 16v Motorsport applications i.e the 220bhp GrpA Mk3.



The so called LHD ABF tubular manifold.
The car on test benefitted from this fitment as did several other friends in our circles at the time and ultimately is fitted to JENVEE, where I develop close to 100bhp/litre from my ABF motor.

Here it is fitted to the test car with a modified 8v downpipe.





OEM fitment means not messing with bulkheads, engine mounts or fouling and you ave a system more suited to production OEM engine hardware. Due to the lower flange location of the ABF EU manifold, you need to modify your existing down pipe or use the OEM part it would come with. This difference in flange height is shown in the following picture.



The car in the pictures was fitted with a 02A/J type gearbox. However if you are running a 020 gearbox some robust linkages from Hotgolf would be mandatory.


jo jo's picture from FB

The final plot on the EU 4-2-1 manifold looked this way after the rest of the exhaust was optimised.



You can see from this plot that torque is at the quantity I would usually expect to test on this rolls along with a typical plateau. This engine always had a slight drop in torque nearing 5000 rpm. However a marked improvement since the OE manifold was fitted.

Now to walk the result to capture if there was any improvement between the kit car and EU 4-2-1 exhaust manifolds on the Jetex system, we used G meter acceleration pulls to note the difference on an engine that was now better optimised. As the dyno was not available at the time the correlation was required.
Because we did not have the opportunity to revert to the OE UK ABF cast manifold we chose to capture the data this way.

G trace compare:



NB
Black trace is with the EU 4-2-1 manifold fitted.
Red HPRs 4-2-1 kit car exhaust.
Noise does affect the smoothness of the data compared to the dyno plot. Use the data as an indication of drive feel differences that would be difficult to subjective quantify by the seat of the pants.

Unlike mechanical mods done back in the 90s and early 00s, access to the ECU meant engine control calibration was optimised and monitored to match each stepped changes.



Observations.
Mid range torque at WOT a tad bit more torquey with the EU 4-2-1 (not that you would pick up as a driver input)
Top end torque and power enhanced with the kit car exhaust.

Conclusion.

While the changes seen on the engine at WOT in this post were small on this stock engine, that was mapped for maximum response, the use of a generic off the shelf 4-1 manifold on an ABF fitted with an OEM mapped digifant 3.X ECU or on a Kjet, would possibly not yield the same result. In fact I would expect a similar result with OEM exhaust parts.
Increasing the bore on the exhaust to ~ 2.25 to 2.5 inches in bore does help overall engine output at higher revs, However to maximize engine character, you need to calibrate the ECU.
Directionally in this exercise, there was a benefit from fitting the EU 'LHD' 4-2-1 that can be used as a future proof part in the event the engine was to be significantly modified.

Further developments on Toyotec's JENVEE



Since this exercise was carried out many years ago, my own car JENVEE was acquired, allowing further ABF developments to continue, with components tested to limits, without the risk of damaging someone else's pride and joy. In addition many miles have been accumulated on various components I share information about.
JENVEE had similar characteristics as Ben S' Golf MK2 ABF, achieving 170 bhp @ 6000 rpm and 160 lbft or so from 3800 rpm to 5500 rpm. Naturally it's engine is controlled by Megasquirt.
All of what I knew on ABF engines at the time, features in JENVEE.
Some time later, a further development on exhaust tuning was carried out on JENVEE, in line with the GrpA system pictured above plus a few words confided to me from a German tuner on dimensions.
Style Dynamics in London made up a prototype downpipe to test.
The 6 bolt flange was laser cut and removed the OE restriction found in the OE downpipe, merging much further back at an undisclosed distance as per my specific instructions.



Due to life commitments, not all of the developments on JENVEE were often always reported on this forum platform and it was impossible to test back to back to the old 170 bhp days, as the spec of the engine changed significantly with headwork such as:
CTM Performance cylinder head with HPR/Toyotec port dimensions.
HPR/Toyotec Cat Cams.
Everything else was left standard, including the intake manifold.
The net some of these compoents, with respective tuning along with the factory tubular exhaust and a Style Dynamics/Toyotecwerke downpipe was tested on the dyno, albeit not at Garage Streamline.
The results of several years of ABF tuning on a stock 10:1 bottom end was 193 bhp @ 6300 rpm and 168 lbft from 5400 rpm to 5900 rpm.



Hope this post has been useful to some of you before you spend that hard earned on parts that could turn out to be useless if the foundation is not put into place.

Many thanks to Ben S at the time for the use of his vehicle and to Hugo, who trusted me with his MK3 kit car 16v exhaust manifold and some of his expertise.

All powered by enthusiasm and the quest to have first hand knowledge.

Thanks for reading.

Toyotec.

 

Thanks interesting reading

 

No problem Carey.

I am happy to encourage constructive discussion on a topic like this so feel free to ask any questions on the test work was was carried out some years ago.

 

Do you have a plot with all 3 manifolds overlaid ?

I`m curious why there is less torque on the final plot compared to the previous 2 manifolds.

Thanks for sharing.

 

The last dyno plot with the 4-2-1 EU manifold shows a trace with 174bhp and 160lbft spread across an rpm range of 4200-5100rpm. That is what I would expect from a stock ABF engine with this dress kit.

I am not sure how that is less than the other tests, which show the car in another state.

 

as always; superb skills and and dedication.

 

Good read, thanks ,

I got one of those manifolds from a french ABF I bought ,

Just got to find where I’ve put it .

On 16v’s , rather than go the 4 branch all the time, I used 8v down pipe, delete the silencers and then just use the back box , and it also save me lots of money too, £20 for the pipe and £20 for the 2 x 90* bends

 

I have made some edits to the first post to bring up to date.

I hope this thread is useful to making decisions around stock, 4-2-1 or 4-1 header systems.

Ed.

 

Thanks for updating this Ed.

Your original ABF conversion thread and this one have been really useful reading as I source parts for my MK2 upgrade.

 

Hi Ed, thanks for writing this post very informative and nice to keep the forum going rather than on the more disposable Facebook page...

For my understanding, are you suggesting that the benefit seen from the tuning length on a 4-1 is too high up (or is it in fact out of) the Rev range for a typical street engine redlining at say 7300rpm?
Whereas the 4-2-1 tunes for a broader mid range and as you show this breadth can be extended with longer secondaries?

Quick summary
4500 rpm peak with cast OEM setup
4000-5400rpm with LHD tubular and stock, short secondary
4000-5500rpm (with improvement in torque over LHD tubular after 5500rpm) on the GrpA with long secondaries

Would you be able to suggest tuning peak of one of the common 4-1 manifolds and pressuably this would be a lot more ‘peaky’ Than the 4-2-1s as mentioned?

 

The generic 4-1 crap is not what I would use on the street on the basis of the testing done with legit race parts.
You are likely to loose more low down torque in the areas you drive, so less response.
On an ABF engine it is pointless as the 4-1 stuff is not really tuned to suit stock camshaft profiles and the intake manifold plenum lengths.
The 4-2-1 will increase the torque between 3000 and 5000 rpm and maybe more if you shift around your plenum lengths or you run different cams and cam timing.
The stock one will still work, with an ECU tune as I used to do and the car will accelerate faster. However as you push on, you find out it is tuned to optimize to the performance envelop most production cars are likely to be operating in such as sub 6000 rpm.

 

What stock downpipe did you connect to the 'LHD tubular' manifold? and do you know what length this was?
When you have since made your own custom pipe what length did you use and was this iteratively sized, dynoed and measured or were you aiming to match the Kit car system?
Thanks

 

Originally my car was on the original matching EU MK3 16v downpipe. I bought both bits from germany. Ben was on a modified 8v GTI downpipe.

For the custom downpipe, I consulted several european sources to create a system similar to a GprA vehicle, not kit car. Lengths were set to match my custom cams and any future proofing.
I was not able to have the car re dyno'd as yet.

 

I have mentioned in the main post

.

This picture illustrates, how much lower the manifold 6 bolt flange is between the LHD EU tubular and the conventional cast manifold found in UK 16v vehicles.



For 020 equipped MK1/2 vehicles, the long linkage with the plastic ball socket, is in close proximity to the LHD manifold joining flange, and can be at risk of melting.

Gas N Gears in Wales, makes a stainless steel linkage kit (see MK2 kit below), created for durability, which for this specific component replacement, also resolves any issues with linkage failure.



I ran one some of the first prototype kits. They functioned brilliantly after a couple 10k miles with the heat of the LHD 4-2-1 system having no effect on function. Fitment and removal was pretty straight forward, once the retaining clip was removed.

 

I am glad to hear.
While I was not the first to carry out an ABF conversion, this is one of the more popular conversion threads on the site for sure.
Explaining the conversion in some detail for various paths to enthusiasts all over.
Precisely my intent for creating it.

 

This is a typical generic 4-1 that folks have been shoehorning into their MK2 16v cars for quite a few decades.


I was unable to test this manifold back to back with the real MK3 kit car job on loan from HPR as pictured below.



My data suggests the above assisted cylinder filling of the engine at higher engine speeds, but as the engine was not really spec'd to fully utilise the tuned characteristics of that system, it was not a well matched component.
I cannot see how the generic shiney part would be any better than the kit car part, with all the fouling that tends to occur between the bulkhead and steering rack. You might as well leave it stock to reduce the hassle.

Your peaks refer to torque on a stock engine yes?

 

I have also added my dyno plots in the main post, so demonstrate this OE tubular manifold supporting my 193 bhp ABF engine.

 

Your peaks refer to torque on a stock engine yes?[/QUOTE]

Yes I thought you had stated throughout your testing these were done with stock cammed cars?
For your custom 2-1 that you made, what ID pipe are you using please? I don’t have a stock 2-1 to refer to

 

The thread has been merged with another thread I wrote.
Certainly the experimentation and development was done on Ben S' stock ABF with a Megasquirt ECU.
As I proved all I needed to know from that exercise, my car @ 193 ps has been added to the post, demonstrating the EU exhaust manifold working as a component to support performance beyond stock hardware.
The exhaust manifold is also likely to stay as I look for ways to improve the performance of JENVEE's engine, beyond the dyno plot I shared.

It all must be in context.

Dyno time between these tests were uncontrolled and often at great logistical pressure, based on where the test car was located and how busy we were outside of working time.
At the time when the KC manifold was tested on the dyno, the reference was the cast manifold. The calibration was still maturing.
The delta captured revealed, the KC manifold and tuning resulted in a benefit to torque everywhere, with a clear increase beyond 4500 rpm.

Knowing the KC manifold had to be returned, to allow continuity to the experiment, I captured it's performance on an accelerometer local to the owner, after some further calibration optimisations. We could later return to the same place with other hardware.
The KC car manifold was removed with all the changes to make it fit to the car and the LHD manifold I owned at the time, along and Ben's original 8v downpipe was fitted, possibly as a permanent feature moving forward.
It was impossible to test on the dyno at that time, so we used the accelerometer at the same place to do the back to back with the KC manifold previously. There were little calibration changes.
The data revealed kit car part had a bit less acceleration though to 5000 rpm, but was superior beyond that. The reality is you would be hard pressed to tell the difference. So we did that to remove any placebo effect.

We later tested the EU manifold. The plot for that shows on the dyno, 160 lbft @ 4200 to 5200 rpm plateau and 174 bhp @ 6000 to 6600 rpm. Other engines might differ by an the odd lbft of torque here or there, and the range of the plateau.

So walking back that dyno plot back to the acceleration test result, we can draw conclusions on the KC part, suggesting it may have achieved torque increasing to say 158 lbft @ 5900, with that torque slowly dropping off to result in a 180 ps at 6600 rpm.
I can express this on a graph, for later.

Yes, we can also superimpose the longer secondary tubes into this for the LHD exhaust manifold, but remember on a stock engine the effects will be small, but certainly head into the direction you suggested, regarding a bit more top end maximum torque.

I ran 2 in diameter pipes as an experiment. However, it was what was recommended and a start point, based on a few experts I spoke to, based on my engine spec and what my plans were.
The engine changed too much and had little dyno development between, to reference a stock LHD 2-1 or 8V GTI 2-1 downpipe.
I sort of knew what I was doing so just pushed through without reporting on here.
For my engine, There is still work to be done there to alter away from the dimensions initially used, which might involve a change in hardware.
It is the nature with prototype parts when you know what the end game should be like.

 

As someone that is rediscovering these older motors, I am grateful that info like this is available. The work is systematic, persistent, and generously offered. Thank you for taking the time to not only do this, but to share what you have learned. I'm in the US, and obviously, we didn't get the ABF. Building a clone now and drinking up what this forum has learned. It's destined for my rally car, a mk2 golf, that is now running an ABA.