In reply to a message from lockheed sent Mon 16 Sep 2013:
I hear you, but, like Kirbert, I am still wary that the
input is the issue. Gene has a continuous spark at the A
side. We currently know we have enough input to drive the
A spark…so the B output is still very suspect.
Now that Gene has the scope, it only takes 5 minutes to
verify that the B output of the ECU is indeed good prior
to the harness. 5 minutes with the cool scope could save
hours of splicing unknown coax into the myriad of other
unknowns.–
John. '95 XJS 6.0L convertible. Southlake, TX
–Posted using Jag-lovers JagFORUM [forums.jag-lovers.org]–
–Support Jag-lovers - Donate at http://www.jag-lovers.org/donate04.php –
In reply to a message from CJ95 sent Mon 16 Sep 2013:
''I hear you, but, like Kirbert, I am still wary that the
input is the issue.’’
That’s what I said: ‘‘the speed signal output from the flywheel
sensor arriving at pins 3 and 16 of the PCMI has been degraded by
improper wiring in the harness. Likewise, we can suspect that the
crank sensor signal to the PCMI (pins 1 and 2) may be degraded for
the same reason,’’
GIGO - Garbage in, Garbage out.
‘‘Gene has a continuous spark at the A side.’’
During cranking yes, but as of yet, we know nothing about idle or
above when the crank signal rises to a higher frequency and goes
thru the same sort of suspect wiring as the flywheel speed signal.
Anyhow, it will be interesting to find out how Gene is doing with
his ‘‘O’’ scope.–
lockheed 92 XJS Cpe/97 LT1 Miami FL/ 96 XJS Cv 4.0 Austin TX
Austin, TX, United States
–Posted using Jag-lovers JagFORUM [forums.jag-lovers.org]–
–Support Jag-lovers - Donate at http://www.jag-lovers.org/donate04.php –
In reply to a message from CJ95 sent Mon 16 Sep 2013:
To be honest with you guys, I feel like I have the deer in
the headlights look. I am getting some great info, but I
lack the confidence to KNOW that replacing the shielded wire
with coax will actually fix the problem. I suspect it
strongly, but I agree with the idea of having A side to
continue firing means that the shielded wire must be doing
at least something right.
I have followed advice given here, and must be doing
something really wrong. This oscope is neat, but I still
don’t know what I am looking at. When I change the settings
to the point that I can ‘‘see’’ a signal on the rear sensor,
the same setting won’t allow me to ‘‘see’’ anything really on
the front sensor. I know this has to do with time. The front
sensor is giving 3 signals per rev. If the engine is turning
300rpm with the starter, then that is 900 signals per
min(?). Likewise, the rear sensor should be sending 160
signals per rev, and 48000 signals per min?
Obviously, if I don’t change the scale on my meter, then I
won’t see anything that even remotely looks like what I see
from sensor to sensor. And so far, I have YET to see
anything that looks exactly like what I see in the manual.
And if I don’t verify that I am getting a proper signal TO
the wiring, and then AT the ECU, then how in the world is it
going to help to change my wiring? I may do all of that, to
still have my problem. And then, even though I know I am
getting a signal TO the ECU, what kind of signal should I be
seeing OUT of the ECU, on either side? (meaning what sort of
signal should be going to the amp(s). So far, all I really
KNOW, is that the analog volt meter shows that a signal is
going into the ECU from each sensor, but B side output drops
out. But even then, I am reading from one pin to the other,
not pin to ground, as has been suggested–
89 XJS convertible Marelli ignition
–Posted using Jag-lovers JagFORUM [forums.jag-lovers.org]–
–Support Jag-lovers - Donate at http://www.jag-lovers.org/donate04.php –
In reply to a message from Gene Holtzclaw sent Mon 16 Sep 2013:
‘‘but I agree with the idea of having A side to continue firing
means that the shielded wire must be doing at least something
right.’’
Gene, The A side spark timing is directly from the front crank
during initial starting, and then shifts to the timing map at
around 350 rpm according to documentation. The flywheel speed
input to the PCMI is apparently not needed initially to generate
the A spark. But, it may be needed for the timing map lookup when
the engine goes from the crank to ‘‘run mode’’. That’s something we
don’t know yet because your engine has not started yet, although
you indicate it did fire some, but never accelerated to idle speed.
The documentation also tells us that the B timing is always, repeat
always, computed from the A timing by the PCMI software which needs
the flywheel speed input to do that. The frequency of the crank
pulses are much lower than that of the flywheel - 3 vs 160 pulses
per revolution. So it would seem reasonable to guess that the
lower frequency crank pulses are degraded less than the higher
frequency flywheel pulses by the lack of proper coax wiring. And
therefore, you still have A sparks while cranking but the B sparks
go away as the speed signal becomes unusable when the engine begins
to turn faster, and the PCMI software cannot compute the B spark
without a useable flywheel speed signal.–
lockheed 92 XJS Cpe/97 LT1 Miami FL/ 96 XJS Cv 4.0 Austin TX
Austin, TX, United States
–Posted using Jag-lovers JagFORUM [forums.jag-lovers.org]–
–Support Jag-lovers - Donate at http://www.jag-lovers.org/donate04.php –
In reply to a message from lockheed sent Mon 16 Sep 2013:
Gene can you entertain us all here for a
minute and first charge your battery
before going any further. Then swap the
front to back sensors, left to right amps
and front to back coils. Also make sure
you dig out your ignition ecu and it’s to
hand. Then with a fully charged battery
and leads going to the top pf the
distributor removed and positioned 1cm
away from say air rails have someone
cranking the engine while you try and
wiggle first the rear sensor wire which
can be reached just behind the
‘‘turntable’’ and then the front sensor
wire which is just left and forward of
the crankcase ventilation pipework (that
mushroom at the front of B bank, you got
your banks right, right? Sitting in the
car right is a left is b, right) and see
if that changes anything in sparks coming
from those two main leads. Unplug and
plug the sensors while cranking see what
happens.
You are way off in suspecting the type of
wiring used here as this is a really
primitive computer that only cares if it
sees 12v going on and off on a particular
input (or wire) coming from your sensor.
Even a coat hanger could provide it with
a good ‘‘signal’’. Maybe if nothing but the
connector on the ecu that’s dirty or not
seated properly? The long one with 20-
something pins. I think you need to have
a break and think about the problem a
little. Go back to basics as they say.
Also don’t forget to try the 3rd spare
sensor you have. In both locations. But
if you already tried swapping sensors
then that can be eliminated as your
problem. Maybe the little vacuum tube is
disconnected from the ecu? Now that your
ecu is out and at hand you can verify
that by the person who was cranking the
engine sucks or blows on it on one side
and you trying to feel the air flow on
the other by holing the end of the tube
to your cheek or lips. And then well go
from there.–
'91 xjs v12 5.3 early facelift in BRG
–Posted using Jag-lovers JagFORUM [forums.jag-lovers.org]–
–Support Jag-lovers - Donate at http://www.jag-lovers.org/donate04.php –
In reply to a message from Pele sent Mon 16 Sep 2013:
Do Gene a favor and go back and read the history - you will see
that he has already done all of what you ask. He has spent enough
time repeating answers he has already given.
''You are way off in suspecting the type of
wiring used here as this is a really
primitive computer that only cares if it
sees 12v going on and off on a particular
input (or wire) coming from your sensor’’
Obviously, Jaguar and Marelli (with their electronic engineers) do
not agree with you. For one thing, the signals from the sensors
are not 12V going on and off, but much smaller, probably less than
1 volt with waveforms - can you tell us the voltage exactly? And,
knowing how cheap Jaguar was with the XJS and how near to going
broke they were, what makes you think they would wastefully spend
extra money to put in coax that wasn’t needed? Furthermore, they
added additional coax lines, beginning with the facelift in '92, to
the signal lines between the PCMI and the Power Modules. Have you
written to tell them how you know more than they do about this
subject, and how foolish they were?–
lockheed 92 XJS Cpe/97 LT1 Miami FL/ 96 XJS Cv 4.0 Austin TX
Austin, TX, United States
–Posted using Jag-lovers JagFORUM [forums.jag-lovers.org]–
–Support Jag-lovers - Donate at http://www.jag-lovers.org/donate04.php –
In reply to a message from lockheed sent Tue 17 Sep 2013:
Alan, I follow your logic, except I have never seen a starter turn
an engine above 350 RPM. 150 to 200 rpm would be about
right…unless maybe when all the plugs are out you may push 300 or
so.
Gene, your assessment of the difference between the scope readings
on the front vs rear sensors is right on. To make them look the
same on the scope, you will have to reduce the ‘‘ms’’ value of the
scope by a factor of about 53 times (160 teeth divided by 3 teeth)
when you change from reading the rear sensor to the front sensor.
I would think the auto function should get the scales close.
Once you take a snapshot of the pattern, you can adjust the scale
values on the snapshot to get a better picture of the pulses.
Larger ‘‘ms’’ numbers will compress the pulses into spikes.
Small ‘‘ms’’ values will stretch them into longer patterns. The
voltage (vertical) scale can be adjusted on the snapshot to bring
the height of the pulses into a convenient size on the screen.
Just to play with the scope and get familiar with it, you can read
the wall outlet voltage. It will give you a nice sine shaped wave
you can use to play with the different settings on the scope. Once
you feel comfortable taking holds from the steady wall voltage, you
will be confident on the car.–
John. '95 XJS 6.0L convertible. Southlake, TX
–Posted using Jag-lovers JagFORUM [forums.jag-lovers.org]–
–Support Jag-lovers - Donate at http://www.jag-lovers.org/donate04.php –
In reply to a message from CJ95 sent Tue 17 Sep 2013:
‘‘except I have never seen a starter turn an engine above 350 RPM.’’
John, you are probably right about that - I’ve never watched it
that close. But even then, 200 rpm of the flywheel would be giving
(200*160) 32,000 cycles per minute or (32000/60) or 533.3 Hz. It is
still conceivable that the signal could still be screwed up even at
some rpm below 350 such that the PCMI is getting a speed signal it
cannot use to compute B spark just above the initial cranking speed.
‘‘you can read the wall outlet voltage.’’ Don’t know that you want to
tell him that, unless you know the input limits of his scope.
There is info. on it at: www.velleman.eu Look for the link to
HPS140i. Maybe suggest checking the output ripple of his cell
phone charger, or something of lower safer voltage.–
lockheed 92 XJS Cpe/97 LT1 Miami FL/ 96 XJS Cv 4.0 Austin TX
Austin, TX, United States
–Posted using Jag-lovers JagFORUM [forums.jag-lovers.org]–
–Support Jag-lovers - Donate at http://www.jag-lovers.org/donate04.php –
In reply to a message from lockheed sent Tue 17 Sep 2013:
As fate would have it, I have received a bad oscilloscope. I
talked with Velleman today, and they want me to send it
back, as they have not ever seen one of their oscopes do
what mine is doing. According to the directions, their
website videos, as well as what I could see on youtube,
there is something wrong. The tech guy at Velleman asked me
to send it right back. Just when I thought I was gonna
figure this out…–
89 XJS convertible Marelli ignition
–Posted using Jag-lovers JagFORUM [forums.jag-lovers.org]–
–Support Jag-lovers - Donate at http://www.jag-lovers.org/donate04.php –
In reply to a message from lockheed sent Tue 17 Sep 2013:
I can tell you the voltage exactly - it
is DC as specified per the manual. DC on
this car is 12V and that’s what the
sensor will give off but since they say
DC that means anything will work. Servos
are like that, they just see and count if
something triggers their input or not and
the spec on ye olde ‘‘computers’’ would be
something like 3-12V DC - just like your
bog standard LED. You will of course need
a resistor in series with that to get any
meaningful current out of it but that’s a
different story altogether. And I don’t
need to suspect anything as he has the
exact same wiring on his car as I do on
mine with twin conductors and a shield -
that is the original marelli wiring on
the sen8 s sensor. Arguing whether
marelli installed the correct wiring on
their own product is what you lot have
been arguing on here since forever. And I
'm saying that there is no need to. And a
coat hanger would demonstrably carry a Dc
voltage to where it needed to get to.
If you have an intermittent problem in
the wiring that means wiggling the wires
in question could restore function. He
doesn’t have an intermittent problem as
his coils fire for a few seconds and then
one goes berzerk. That could mean the ecu
gets reset. Maybe power or ground for the
ecu are not ok? All I’m saying is that he
needs to stop and give it a rest as
debating about cables has stopped being
productive ages ago.
And just because I didn’t quote my degree
and professional membership info doesn’t
mean I don’t have the stamina or
intellectual and technical prowess to
analyse the issue currently at hand. I
thought we were supposed to keep this a
hobby and fun.–
'91 xjs v12 5.3 early facelift in BRG
–Posted using Jag-lovers JagFORUM [forums.jag-lovers.org]–
–Support Jag-lovers - Donate at http://www.jag-lovers.org/donate04.php –
In reply to a message from Pele sent Tue 17 Sep 2013:
I sincerely appreciate any and all
input here. I also wish my problem
were as simple as a loose wire, or
even a broken wire. I feel that I
have eliminated that possibility,
due to the fact that I have ohmed
out every single wire, every
ground, voltage tested every 12V
lead, switched components, had my
ECU checked by a fellow lister, and
now purchased an oscilloscope. The
only thing I know for sure at this
point, having my wiring harness
laid wide open, is that my problem
is still here.The ECU drops out the
signal going to the B side amp, for
whatever reason. It is down to
either a bad signal getting to the
ECU, or improper wiring, or too
much resistance causing the ECU to
stop sending a signal.I do know for
a fact, that the wire going to the
Lucas ECU is coax. The workshop
manual shows this wire being a
shielded wire, but its not. Whether
all of the other wires are shielded
vs coax, I don’t know, as this car
does not have the original harness,
nor do I know where to see a
original harness. Either way,
something is wrong, and I can’t
figure it out, yet. Thank you.–
89 XJS convertible Marelli ignition
–Posted using Jag-lovers JagFORUM [forums.jag-lovers.org]–
–Support Jag-lovers - Donate at http://www.jag-lovers.org/donate04.php –
I was reading about Zytec systems and thay switch from a start mode to a run mode. The static timing is close to TDC and the pick up has 6 rotors. The system switches to run and goes to the next trigger point putting the signal well before the fireing point so there is enough time to calculate the needed advance of up to 35 degrees.
In reply to a message from Geoffrey Green sent Wed 18 Sep 2013:
Geoff,
Interesting - you are saying the crank timing disk has 6 fingers
instead of the 3 that the Marelli has?
Caught the point about the Zytec system advancing to the next
finger in the rotation when it goes to the run mode. There is
speculation about the Marelli doing that or not. A letter has been
sent to AJ6 Engineering (since Marelli has not responded to any of
us) asking clarification on that point, and some other items such
as timing maps and suitable (coax or otherwise) wiring for the
crank and flywheel sensors. Roger Bywater said he will respond
when he gets back from a few days off.–
lockheed 92 XJS Cpe/97 LT1 Miami FL/ 96 XJS Cv 4.0 Austin TX
Austin, TX, 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
Kirbert
(Author of the Book, former owner of an '83 XJ-S H.E.)
215
I was reading about Zytec systems and thay switch from a start mode to
a run mode. The static timing is close to TDC and the pick up has 6
rotors.
Is this the Zytec system used with the TWR cars? We’re talking the
Lucas distributor, not the Marelli, so it had to develop 6 sequential
sparks per revolution rather than two sets of 3. And I thought the
pickup was inside the distributor, which would have called for 12
fingers on the rotor, not 6. It would have been 6 if it was on the
crank.
And I’m pretty sure the Zytec had the centrifugal advance operating.
I believe the vacuum advance was locked down, but you pretty much must maintain the centrifugal advance operation to keep the rotor
sufficiently aligned with the spark.
The system switches to run and goes to the next trigger point
putting the signal well before the fireing point so there is enough
time to calculate the needed advance of up to 35 degrees.
IOW, it operates using the mechanical advance to start, but once
running it switches to an electronic map – and has to switch to one
tooth earlier because it cannot predict the future, it works by
applying a calculated delay to the previous pickup signal. This
makes sense when using a Lucas distributor and generating sequential
sparks, but it doesn’t make any sense on a Marelli because you’d have
no way to generate a starting spark for the B bank.
– Kirbert
// please trim quoted text to context onlyOn 18 Sep 2013 at 10:35, Geoffrey Green wrote:
In reply to a message from Kirbert sent Wed 18 Sep 2013:
Guys, I haven’t sent the oscope back yet, and started the
series of tests again, writing down what I saw with the
oscope on automatic. Mind you, I didn’t pick up a good
signal on my front sensor, and changed it out with the one
Mike sent that he said was good. Yet today, I couldn’t get
it to give me a decent signal even with the oscope on
automatic. Now I’m beginning to wonder if I have a bad front
sensor, or maybe something wrong with the timing fingers, if
having bad fingers is possible.
Anyway, for those who know what I might be looking at, with
the oscope on automatic, the rear sensor was tested first.
Unplugged. If I put the oscope sensor to one output, and
clipped the ground to the engine, the scope read:
1ms/div 10mv/div .03V
I then switched the leads, and got:
1ms/div 50mv/div 3V
I then checked the sensor across the two prongs, and got:
1ms/div .5mv/div 5V
I couldn’t get much of anything from the front sensor, as
well as the battery of the car started getting pretty weak,
so I quit.–
89 XJS convertible Marelli ignition
–Posted using Jag-lovers JagFORUM [forums.jag-lovers.org]–
–Support Jag-lovers - Donate at http://www.jag-lovers.org/donate04.php –
In reply to a message from Gene Holtzclaw sent Wed 18 Sep 2013:
Gene, I think I would have started with the front since it has a
slower signal (3 vs 130) and would have been easier to resolve and
sort out - but, with scope in auto it probably should not make any
difference. The waveforms given by JDHT stipulate that the
connector be hooked up to the sensor - at least when you are taking
the readings at the relevant PCMI pins - duh - I guess it would be
kinda hard to get the signal at the PCMI unless the sensor were
plugged in to the harness. Despite that, you should still be able
to see the pulses at the sensor as the teeth or fingers pass the
sensors. The reason you see different values when you swap the
leads might be because there is a diode in the scope. Are you in
DC or AC mode? Check it in both to see what you get. I don’t know
if it is possible to have bad timing fingers at the front or not.
If they were bad, why would there be any A spark at all? Since I
have never had the front crank assembly apart on a Marelli system,
as some listers have, I don’t know if it is possible for the timing
disk/fingers to slip out of proper sync or not. Someone that has
taken this thing apart can provide that information.–
lockheed 92 XJS Cpe/97 LT1 Miami FL/ 96 XJS Cv 4.0 Austin TX
Austin, TX, United States
–Posted using Jag-lovers JagFORUM [forums.jag-lovers.org]–
–Support Jag-lovers - Donate at http://www.jag-lovers.org/donate04.php –
In reply to a message from lockheed sent Wed 18 Sep 2013:
I left the charger on, and will try the front sensor
tomorrow. If I still don’t get anything, I’m thinking I will
swap the sensors altogether, and see what occurs.
As far as I know, when the oscope is in auto, it determines
voltage automatically. I’m thinking that’s what the last
number is, in DC. Whats weird, is that both sensors read the
same on the DVM as well as the analog volt meter. Now, with
this oscilloscope, I am getting this problem with the front
sensor. Of course, I now know that you simply can’t read the
sensors with anything but an oscope.
I really need to send this thing back, because I know it has
a problem. It is supposed to record your reading with the
hold button. On mine, when I push the hold button, the
screen goes blank. It also won’t get what is called an
extended menu. Other than that, I think it is reading
correct. If it means anything, the pattern shown on the
screen during cranking on the rear sensor looks exactly like
its following the teeth of the flywheel. A consistent smooth
motion.–
89 XJS convertible Marelli ignition
–Posted using Jag-lovers JagFORUM [forums.jag-lovers.org]–
–Support Jag-lovers - Donate at http://www.jag-lovers.org/donate04.php –
In reply to a message from Gene Holtzclaw sent Wed 18 Sep 2013:
Gene:
I am just now catching up on this thread.
The sensor I sent to you is what I believe to be a sound sensor,
but I never verified it on a running configuration. It was
something I changed on my engine when I was working through the
igntion troubles I had at the beginning of getting this car. But,
the resistance measured about right, and, I found the problem to be
elsewhere in the harness on my car, so, the conclusion is that I
had changed the sensor out in vain. In any case, I suspect this
sensor is good.
On the scope signal display, and its correspondence to (or lack
thereof) the OEM manual figures: The figures in the shop manual
are idealized waveforms. Real world waveforms will have many
hitches or other artifacts on them. The things you want to look
for are overall amplitude of the signal, and its period or period
between pulses. The Lucas Fuel ECU signal output from the Marelli
should give something closer to the idealized pulse figure in the
manual, but even that one will probably have some artifacts
(overshoot on the leading edge, possible ringing during the hold
part of the pulse, etc). Don’t let these kinds of things throw
you. I say this because when I put a scope on my front and rear
sensor signals, I was shocked at their signature, but after a few
minutes of reflection and deliberation, I came to understand what
was on the display.
Sorry you had trouble with the scope, and hope to read about your
progress when the replacement comes!
Kirbert
(Author of the Book, former owner of an '83 XJ-S H.E.)
220
Dumb question: Are those sensors at the crank and flywheel the type
that GENERATE a voltage with a passing metal tooth, or do they
require a voltage input to operate on? I’ve been presuming they
generate their own voltage, but I don’t really know.