[xj-s] Increased performance for AJ16 Engines - DIY modification

XJ-S
1

I am a regular contributor to the X300 forum, but have not
previously posted on the XJS forum.

I have designed a revised crank sensor bracket which advances the
ignition timing of the AJ16 engine by 5 degrees at all operating
conditions. I have fitted this bracket to both my X300 XJR6s and a
number of other X300 owners have them fitted to their cars too. All
have remarked how noticeable the improvement in engine performance
has been. I conducted some back-to-back tests measuring in gear
accleration times, both before and after fitting the bracket. This
has confirmed the performance benefit in terms of objective
meaurements.

This modification is equally suitable for any XJS fitted with the
AJ16 engine. If you would like any more information about my
modification, then please contact me by email directly.–
Andy Stodart - 1995 & 1996 man XJRs and ex-Jaguar engineer
Northampton, United Kingdom, United Kingdom
–Posted using Jag-lovers JagFORUM [forums.jag-lovers.org]–

// please trim quoted text to context only

2

XJRengineer wrote:

I have designed a revised crank sensor bracket which advances the
ignition timing of the AJ16 engine by 5 degrees at all operating
conditions.

Do you need to start using higher octane fuel?

– Kirbert

// please trim quoted text to context only

3

In reply to a message from Kirbert sent Mon 31 Aug 2009:

The AJ16 has knock sensors Kirby, so ‘need’ isn’t quite the right
word.

With Andy’s mod you can exploit the higher octane better, but if
you don’t run top grade fuel the car won’t come to any harm as the
sensors will prevent trouble by backing off the timing util the
knock ceases. It just won’t go as well and I think it’s one of the
reasons the AJ16 cars respond so well in terms of performance and
economy to Shell Optimax and similar > 99 RON fuels.

Pete–
The original message included these comments:

Do you need to start using higher octane fuel?
– Kirbert

–
68 E-type OTS, 96 X305 XJ12
Cambridge, United Kingdom
–Posted using Jag-lovers JagFORUM [forums.jag-lovers.org]–
–Support Jag-lovers - Donate at http://www.jag-lovers.org/donate04.php –

// please trim quoted text to context only

4

In reply to a message from PeterCrespin sent Mon 31 Aug 2009:

Peter is right. You don’t need to run the engine on anything higher
than 91pump octane (95 research octane) with my bracket fitted, but
the performance gains will be greater if you do. However, there
isn’t any point running a std engine on anything higher than 91
pump octane because the std EMS can’t advance the ignition timing
to take advanntage of the higher octane fuel. That’s where my
bracket comes in.

I thought Shell Optimax was 98RON rahter than 99RON. There are some
99RON fuels available in the UK including Tesco’s ‘‘own’’ super-
unleaded, achieved through the use of low levels of ethanol. That’s
super-unleaded that I use.–
Andy Stodart - 1995 & 1996 man XJRs and ex-Jaguar engineer
Northampton, United Kingdom, United Kingdom
–Posted using Jag-lovers JagFORUM [forums.jag-lovers.org]–
–Support Jag-lovers - Donate at http://www.jag-lovers.org/donate04.php –

// please trim quoted text to context only

5

In reply to a message from XJRengineer sent Mon 31 Aug 2009:

Why not just rework the firmware for the ignition map in the ECU??
Would it not be cheaper than spoofing it with a physical shift of
the crank position sensor??–
lockheed
–Posted using Jag-lovers JagFORUM [forums.jag-lovers.org]–
–Support Jag-lovers - Donate at http://www.jag-lovers.org/donate04.php –

// please trim quoted text to context only

6

PeterCrespin wrote:

The AJ16 has knock sensors Kirby, so ‘need’ isn’t quite the right
word.

Exxxxxcellent!

I have always presumed that, when an engine is fitted with knock
sensors, the baseline timing is advanced to the point of optimum
performance as though you’ll be running 200 octane fuel. Hence, the
better the fuel you pour in the tank, the better the performance and
fuel economy. Why would Jaguar have done such a lousy job that the
owner has to advance the timing 5 degrees to get better performance?
And why doesn’t he advance it 10 degrees?

I have been told that my 2002.5 Mazda Protegé5 has a knock sensor,
although I’ve never located it. I can tell you that you can fill it
up with premium and it doesn’t perform one iota better than on
regular, and the fuel economy doesn’t change. So it gets regular.

– Kirbert

// please trim quoted text to context only

7

I installed one of Andy’s brackets on my XJR. Works just as advertised !

No affiliation, of course

Cheers
Doug Dwyer
Longview Washington USA
1995 XJR

// please trim quoted text to context only

8

In reply to a message from Doug Dwyer sent Mon 31 Aug 2009:

Lockheed,
Changing the ignition advance in software is clearly the most
elegant solution. This is precisely what I used to do when I worked
for Jaguar. However, I don’t have the necessary software skill to
reprogramme the ECU without access to the original development aid
software that I used, when I worked at Jaguar. Whereas, I can
design and make a bracket that adds 5 degrees.

Kirbet,
I don’t think my colleague did a ‘‘lousy’’ job calibrating the
ignition timing. He was under instructions to calibrate it in such
a way as not to exercise the knock sensing system. At that time
knock sensing was a relatively new technology and so the decision
was taken only to use it to protect the engine in the instance that
it was run on poor quality fuel. My bracket exercises the knock
sensing system more actively to ensure each cylinder runs closer to
the optimum ignition timing.

During development I ran engines with the nominal ignition timing
set 10 degrees beyond borderline detonation and let the knock
control system find the optimum level for each cylinder.–
Andy Stodart - 1995 & 1996 man XJRs and ex-Jaguar engineer
Northampton, United Kingdom, United Kingdom
–Posted using Jag-lovers JagFORUM [forums.jag-lovers.org]–
–Support Jag-lovers - Donate at http://www.jag-lovers.org/donate04.php –

// please trim quoted text to context only

9

In reply to a message from XJRengineer sent Tue 1 Sep 2009:

I can’t speak to the AJ16, but this is a common modification
on the 87-93 Mustang to improve performance. The factory
timing is pretty conservative and on those cars you can
increase initial timing from 10 degrees to about 14 degrees
and gain several hp by resetting the distributor so that the
sensor is in a different position relative to the crank.
You need premium fuel to take full advantage and the knock
sensor will pull out timing if it detects pinging. But on a
Mustang they say its worth 10-12 hp.

However, be aware that if the Jag ECU operates like the
Ford, the knock sensor has some ‘‘hysteresis’’ in the
programming, so if it detects a knock, it will take out more
timing than necessary to stop the pinging, and you actually
lose horsepower with regular fuel.–
MustangSix
Orlando, FL, United States
–Posted using Jag-lovers JagFORUM [forums.jag-lovers.org]–
–Support Jag-lovers - Donate at http://www.jag-lovers.org/donate04.php –

// please trim quoted text to context only

10

In reply to a message from XJRengineer sent Tue 1 Sep 2009:

‘‘At that time
knock sensing was a relatively new technology and so the decision
was taken only to use it to protect the engine in the instance that
it was run on poor quality fuel.’’

Andy,

If the technology wasn’t that great then, and there was fear for
the health of the engine, what components have changed in the
system since the date these cars were manufactured, that it is now
possible to have geater confidence in the system to now protect the
engine against damage? Or is it just a tonque in cheek approach,
without adequate laboratory data, hoping the system might give
adequate protection because of some unknown factor? A field trial??–
lockheed
–Posted using Jag-lovers JagFORUM [forums.jag-lovers.org]–
–Support Jag-lovers - Donate at http://www.jag-lovers.org/donate04.php –

// please trim quoted text to context only

11

XJRengineer wrote:

During development I ran engines with the nominal ignition timing set
10 degrees beyond borderline detonation and let the knock control
system find the optimum level for each cylinder.

How’d it work? Were there any downsides at all to this scheme?

– Kirbert

// please trim quoted text to context only

12

In reply to a message from Kirbert sent Tue 1 Sep 2009:

Lockheed,
I didn’t say the knock sensing system didn’t work when it was
introduced on X300 for 95MY, as a new technology. In fact it worked
very well, as all the testing I did at the time demonstrated.
However the company took a very cautious view and decided not to
use it actively to control the ignition advance. Sometimes people
just aren’t confident with new technology, and prefer to keep doing
things the way they have previously done them. What I have done is
not a tongue in check approach. I was there. I did the testing. I
know it works. Ask anyone if they would advance the ignition timing
of an engine 10degrees beyond borderline detonation without knock
control.

In Europe premium engines are now routinely mapped on 98RON fuel
and the knock sensing system is relied upon to protect the engine
byreducing the ignition advance when operating on 95RON or 91RON
fuel.

Kibert,
There aren’t any downsides. If fact the fuel economy is also
improved as well as the performance by 1-2mpg–
Andy Stodart - 1995 & 1996 man XJRs and ex-Jaguar engineer
Northampton, United Kingdom, United Kingdom
–Posted using Jag-lovers JagFORUM [forums.jag-lovers.org]–
–Support Jag-lovers - Donate at http://www.jag-lovers.org/donate04.php –

// please trim quoted text to context only

13

In reply to a message from XJRengineer sent Tue 1 Sep 2009:

Jaguar really were conservative! My 1983 Renault Turbo had a knock
sensor, 12 years before the AJ16, and I was never aware of pre-
ignition. Hardly new.

Andrew–
AJME, 90 XJ40, 94 XJS 4.0, 89 Benz 190E, 75 Moto Guzzi
–Posted using Jag-lovers JagFORUM [forums.jag-lovers.org]–
–Support Jag-lovers - Donate at http://www.jag-lovers.org/donate04.php –

// please trim quoted text to context only

14

In reply to a message from XJRengineer sent Tue 1 Sep 2009:

Andy,

Didn’t mean to be so abrupt, nor did I mean to suggest that the
knock system didn’t work at all, but there are some serious
questions here involving potential engine damage. Even though you
had confidence in the system while you were there, there were
obviously those within the Jaguar organization who didn’t have full
confidence in it, and that’s why the AJ16 spark map was apparently
biased such that it wouldn’t normally use the knock detection
system as a buffer unless there were some abnormally bad fuel
present. So, unless some of the components in the knock detection
system have changed from the original production spec. sheet, those
factors which affected the judgement of those who had doubts, must
still be present. I don’t know what those factors may have been,
but you might.

Although a knock detection system may have been a new
implementation to Jaguar at that time, domestic USA and a number of
other foreign manufacturers have been using the technology since
the early to mid '80’s.

I’m a bit familiar with GM’s implementation of spark knock systems
and their Power Control Modules (PCM) which have a learning
capability over time. As I understand it, the PCM ignition map
starts with a baseline spark advance map and then slowly ramps up
the timing until a knock is sensed, it then backs off a number of
degrees before again beginning a time based spark advance till it
again hears a knock. It begins building a data base of the knock
occurance points as it repeats this process X number of times.
Over a period of time with the history in the data base, it
essentially builds somewhat of a ‘‘custom’’ ignition timing map for
the particular individual’s driving habits. The knock sensor
system is always present and functional, and if a more aggessive
driver jumps in the car, or a lower octane gas is used, or some
other factor starts causing a knock, it will still protect the
engine and begin remapping. This ‘‘custom’’ map remains in PCM
memory as long as it is not unpowered (battery power) for a maximum
time, for example 5 minutes. If it is unpowered for longer than
that, that ‘‘custom’’ map, or any others for other functions, are
lost and the whole process starts over again when it is repowered.

I don’t know how sophisticated the AJ16 ECU is with regards to
knock sensing, or just what it does when it senses a knock other
than it retards the timing. Perhaps you could enlighten us as to
just how it functions with regard to how much retard, for how long,
or is it a momentary retard with no time delay function. Does it
then slowly advance the spark timing again? Keep a history for
remapping, etc.? Inquiring minds would like to know.

Thanks!–
lockheed
–Posted using Jag-lovers JagFORUM [forums.jag-lovers.org]–
–Support Jag-lovers - Donate at http://www.jag-lovers.org/donate04.php –

// please trim quoted text to context only

15

In reply to a message from lockheed sent Tue 1 Sep 2009:

The knock sensing system in the Lucas/sagem GEMS EMS can only
retard the ignition from the mapped value, not advance it. This
retard is done on an individual cylinder basis. Ignition timing is
retarded by 3degees at a time, up to a maximum of 10degrees. I
think the re-advance rate was about about 0.5degrees/sec, but I
can’t remember for sure. The software was able to learn the back
ground noise level in the absence of knock. If the noise level
exceed that level by a cylinder specific, speed and load mappable,
amount then knock was recognised.

I had complete confidence in the knock system to protect the
engine. Other people weren’t prepared to rely on it to protect
90,000 AJ16 engines in the field. There wasn’t a test we could
think of to PROVE it would work on all 90,000 engines, so knock
control wasn’t used agressively. These days agressive use of knock
control is much more widespread. Whilst other cars may have had
knock control in 1995, I can’t comment if they were used any more
aggressively that was ultimately the case in the Jaguar
applications.–
Andy Stodart - 1995 & 1996 man XJRs and ex-Jaguar engineer
Northampton, United Kingdom, United Kingdom
–Posted using Jag-lovers JagFORUM [forums.jag-lovers.org]–
–Support Jag-lovers - Donate at http://www.jag-lovers.org/donate04.php –

// please trim quoted text to context only

16

In reply to a message from XJRengineer sent Wed 2 Sep 2009:

‘‘This retard is done on an individual cylinder basis.’’

Andy,

Sounds like a great concept to tailor the knock retard timing to an
individual cylinder. Were there multiple knock sensors - one for
each cylinder - to identify which cylinder had the knock? Or, was
it done with a single knock sensor and by reference to crank
position and firing order?–
lockheed
–Posted using Jag-lovers JagFORUM [forums.jag-lovers.org]–
–Support Jag-lovers - Donate at http://www.jag-lovers.org/donate04.php –

// please trim quoted text to context only

17

XJRengineer wrote:

The knock sensing system in the Lucas/sagem GEMS EMS can only
retard the ignition from the mapped value, not advance it. This
retard is done on an individual cylinder basis.

That’s actually pretty impressive. If it finds itself having to
retard one particular cylinder a lot more than the others, does it
flag a fault code?

Ignition timing is
retarded by 3degees at a time, up to a maximum of 10degrees.

Ah. Hence the bracket that advances the timing by 5 degrees.
Definitely don’t wanna go beyond the system’s capabilities to
correct.

The software was able to learn the back ground
noise level in the absence of knock. If the noise level exceed that
level by a cylinder specific, speed and load mappable, amount then
knock was recognised.

Most impressive. So, if you had a loose manifold bolt or something
that was causing a rattle, this thing could figure that out and
ignore it? I guess by retarding the timing and noticing that the
sound didn’t change?

I had complete confidence in the knock system to protect the
engine. Other people weren’t prepared to rely on it to protect
90,000 AJ16 engines in the field. There wasn’t a test we could
think of to PROVE it would work on all 90,000 engines, so knock
control wasn’t used agressively.

What a bunch of dipsticks. I’ll bet the guys who initially promoted
putting the knock sensor in were spitting mad. What’s the point of
having a knock sensor and then not using it?

– Kirbert

// please trim quoted text to context only

18

In reply to a message from Kirbert sent Wed 2 Sep 2009:

Locheed,
One knock sensor is in line with cylinder 2 This one detects knock
on the front 3 cylinders. The 2nd knock sensor is in line with
cylinder 5 and detects knock on the rear 3 cylinders. The signals
coming from each knock sensor is only analysed between 12 and 60
degrees after TDC firing on each of the 3 closest cylinders. This
is known as the knock window. The timing of the window is
calculated from the crank sensor and half engine speed sensor
signals.

Kirbert,
The strategy does not compare the levels of retard between
cylinders.

I set the bracket at 5 degrees to leave 5 degrees of authority for
the knock sensing to cope with any other causes of knock, though if
the engine is running on 91pump octane fuel it shouldn’t really
even need this much range.

The knock sensing system only analyses the signal between 12-18kHz
(top end of the audible range), so will not respond to low
frequency rattles. Vibrations generated by anything not mounted
directly to the block will be very attenuated by the time they
reach the sensor. Out of interest, if you mounted the sensors on
the head instead of on the block and looked at the signals from the
them plotted against crank angle, it is possible to make out the
vibrations caused by the valves closing on the seat inserts in the
cylinder head.

I think you now appreciate why it was so frustrating for those of
us who were closely involved in the development of the knock
sensing system, not to have it used more actively. This is why I
developed my bracket. I’ve been running one on my manual XJR6 for
about 3 years.

Jaguar weren’t the only OEM to fit knock sensors that weren’t
really used to run the ignition close to knock on each individual
cylinder. These days with so much pressure on all OEMS to maximise
performance for any given engine capacity, these systems are used
more aggressively.–
Andy Stodart - 1995 & 1996 man XJRs and ex-Jaguar engineer
Northampton, United Kingdom, United Kingdom
–Posted using Jag-lovers JagFORUM [forums.jag-lovers.org]–
–Support Jag-lovers - Donate at http://www.jag-lovers.org/donate04.php –

// please trim quoted text to context only

19

In reply to a message from XJRengineer sent Wed 2 Sep 2009:

Slick!!!

Carl–
Carl Hutchins 1983 Jaguar XJ6 with LT1 and 1994 Jeep Grand
Walnut Creek, California, United States
–Posted using Jag-lovers JagFORUM [forums.jag-lovers.org]–
–Support Jag-lovers - Donate at http://www.jag-lovers.org/donate04.php –

// please trim quoted text to context only

20

In reply to a message from XJRengineer sent Mon 31 Aug 2009:

please let me know how to order the bracket for my 96’ XJS, Lord
knows they need every extra bit of power they can get.
Thanks
Mike Haberthur–
The original message included these comments:

This modification is equally suitable for any XJS fitted with the

–
Bullhab
–Posted using Jag-lovers JagFORUM [forums.jag-lovers.org]–
–Support Jag-lovers - Donate at http://www.jag-lovers.org/donate04.php –

// please trim quoted text to context only