[rrudytoo]

Last September my 2003 999 began idling just below 1,000 rpm from its usual 1,200 - 1,300 rpm and would simply quit running while either idling or clutching off so I added some fuel system cleaner thinking the injectors may be getting gummed up. After a couple of weeks and about 100 miles it was getting a bit worse so I thought I would give it an "Italian tune-up". I rarely operate my bike above 6,000 rpm and thought wringing it out might restore things a bit. What happened was when I reached around 8,500 rpm the engine would experience a pronounced miss and a loud, single backfire out the exhaust. I repeated this three times with the same thing happening each time and this was accompanied by more frequent stalling and sluggish off-idle throttle response.

When I got home I checked for fault codes and the one I had said the speed sensor was faulty. That sounded odd but I ordered then replaced the speed sensor but the problem was still there. This time the fault codes showed not only a faulty speed sensor but a faulty crankshaft position sensor, too, so, like before, I ordered and installed a new CPS but, again, the problem was still there only now the fault code showed a faulty barometric pressure sensor. Ugh! I ordered one and soldered it onto the printed circuit board inside the instrument display and, again no improvement but now the fault codes are spilling out like water from a faucet and the bike is running worse. In talking with Ducati-knowledgeable friends we figured the ECU was at fault so I acquired another one, installed it and, guess what, no improvement but now the instrument display no longer shows the coolant temperature. Where the temperature value should be it was only showing horizontal lines that, like with actual values, flashed until the engine reached operating temperature and then remained on and steady. As the engine has always started right up and idled nicely until it reached proper operating temperature before it began acting up I thought that the coolant temperature sender may be at fault so I bought a new one, put it in and no joy.

The overview is simply this: I've given up reading fault codes as I don't believe they are generating actual faults but, rather, glitches in the system. The poor idle and low speed throttle response, stalling and higher rpm back firing is worsening. The headlight no longer works nor does the side stand switch (it assumes the side stand is always up) and if the OBD reader is connected the engine coolant temperature IS displayed but reverts back to bars instead of numbers if the reader is disconnected (faulty ground somewhere?). Even if there was a short or a break in the electrical harness somewhere I doubt that things would get progressively worse and I don't have any blown fuses. I did look the primary harness over closely, cleaned up any and all oil from the wires and connectors and rewrapped much of it but I saw nothing out of the ordinary.

I am at a loss as to what could be the cause of all of this. Is it possible that the printed circuit board inside the instrument display has gone bad? Should I install a different primary wiring harness? Obviously, I am overlooking something but what? The engine only seems to run poorly when the proper operating temperature is reached which has to be a huge clue. I have been unwilling to rev the engine while it is below proper temperature but it would be interesting to see if it backfired at the upper revs.

Any and all opinions, suggestions, directions or resources would be very much appreciated. This is a great bike in both looks and performance that has garnered a good many compliments while I've owned it and I would hate to see it parted out over something that must be obvious to someone other than me.

Thank you all in advance.

Rrudytoo

 

[BOTT]

Anyway to clear the coded and see who pops up first?
Fuel filter was last changed when?
Starts right up cold? Stalls when warm, correct?
Compression is where?
Valve adjust is up to snuff?
Vapor canister clogged some?
Intake manifolds supple and not cracked around the boot rubber?
Spark plugs were changed every 7,500mi?

 

[TOP-DUC]

You mentioned it yourself. Remove and clean the #1 ground on the bike. I’m not familiar with the 999, but there should be only one ground point and all grounds in the various components are run back the the one ground into a multi wire crimp. You can use the wiring diagram and check every circuit with a multimeter for voltage and resistance.

 

[rrudytoo]

Hey BOTT,

Thank you for your reply. Below are the answers to your questions.

Anyway to clear the coded and see who pops up first?
I use a BAFX OBDII Interface in conjunction with an Mx5 App. I am currently unable to get them to "talk" with the bike. When it was reading about a month ago it spat out several codes, all of which I have seen before but as for which was the first one listed, I could not say. If I get it to link up again I'll pay attention and post it.

Fuel filter was last changed when?
June of last year with only 73 miles put on it since then.

Starts right up cold? Stalls when warm, correct?
Yes, that is correct.

Compression is where?
I have not checked the compression.

Valve adjust is up to snuff?
The next Desmo service would be performed in another 4,520 miles. Prior to the electrical hiccups the engine ran as strong as ever.

Vapor canister clogged some?
The vapor cannister was removed by the previous owner.

Intake manifolds supple and not cracked around the boot rubber?
Try as I may, I cannot get to the screw in the area of the steering head to remove the air box lid so I cannot accurately answer your question. I can say that the other OEM rubber components on the bike are fine, though.

Spark plugs were changed every 7,500mi?
I have only owned the bike for one year and have put just 1,978 miles on it. The spark plugs were changed when the fuel filter, among a host of other things, were changed, June of last year, 73 miles ago.
I would like to add that the spark plugs read perfectly, light brown with zero build up or spotting on the center porcelain.


As a reminder, the issues with the bike are electrical as noted in my post concerning erroneous coolant temperature displays on the instrument panel, the side stand switch no longer working, the headlight not functioning and the OBD reader no longer connecting. I believe that the engine's issues are faulty signals coming from the ECU but as the replacement ECU is doing exactly what the original ECU did it must be a case of "garbage in" rather than "garbage out". In short, the ECU is getting bad information somewhere......but where?

Thank you for your input, BOTT.

Rrudytoo

 

[rrudytoo]

You mentioned it yourself. Remove and clean the #1 ground on the bike. I’m not familiar with the 999, but there should be only one ground point and all grounds in the various components are run back the the one ground into a multi wire crimp. You can use the wiring diagram and check every circuit with a multimeter for voltage and resistance.

I will check into this TOP-DUC. This sounds feasible.

Thank you,

Rrudytoo

 

[edgar999s]

Yup, electrical. Pull and strip the entire loom, top to bottom. Lord knows I chased electrical gremlins for 5 months and had to address the loom twice. Check every last wire, pin, junction, whatever it takes. No other way really.

 

[Shrek]

I don’t think you can use obd to read . Most common is guzzidiag or something

 

[rrudytoo]

Yup, electrical. Pull and strip the entire loom, top to bottom. Lord knows I chased electrical gremlins for 5 months and had to address the loom twice. Check every last wire, pin, junction, whatever it takes. No other way really.

edgar999s,

Sounds like I have what you had. Where exactly did you find your problem? Was it a broken wire, a bad connection, a short? Was this in the main harness?

I will look over what I have and if I don't see anything I may order a used harness off of eBay. What would you recommend?

Thank You!!!

Rrudytoo

 

[BOTT]

I'll try to explain a code occurring out of the blue. A self diagnosing computer bike will trigger a code via 3 ways: Signal out of range/short/or open.

Signal out of range is the component failure. Input are many different signals once say the throttle position sensor moves up and down in that range of open and closing. Call this analog input. ECU is sort of in AC mode [a ping back an forth] even though it's DC driven.

Short to ground is another variable. This may be short internally in the sensor, or a wire path from sensor to ECU. A short is now sending in a digital signal, or the same signal over and over. The ECU recognizing this single junk in, sets a code, then sends out a backup signal from the ECU. Sometimes this cuts about 2 HP off the engine, or limps it to save the engine components from damage.

Open is something like a 'work hardened' wire that is vibrating and has a wire loop out of the other laid down wires. This breaks internally and sets the ECU to code its backup number to calc off of.

For any of 3 variables to fail in the 'signal out of range' is watching the change of good known signals like swapping the ECU and it reads the same, does not clear, nor did the swapping of the other [good known] sensors change or clear the code. WATT does that tell us?

The can't be's are as follows:
Can't be a sensor that is out of range, short, or open inside the sensor. It does not clear changing out the component. It says it's not short, not open, but the signal is faulty in some sort of grounding or open path between the harness connection somewhere? It shows it's not the new good known sensor, it's not the good known ECU swap, it has to be a wire connection.

I'm not sure about the duc, but there are two [ground] running wire harness paths. One is the 12v grounds for lights, winkers, etc. Call this a junction point of all the tied together 12v wires. The other is a 5v ground with all the grounds tied together with one eye-loop to ground. Codes are the 5v ground to look for.

If you remember the first code recognized, you'd find the 5v eyelet with all the grounds wired together, (-) probe to the ground harness eyelet; (+) posi probe to the sensor wire's harness side to see a complete loop reading at the ohmmeter. No read, then that wire ground is 'open' somewhere?

Did I describe a set code from a phantom code? Phantom code is the comes and goes code. Or here today, gone tomorrow. It shows the sensor is not out of range, shows the right input, not junk. Meaning a single-constant-same-digit over and over is not happening. Is not short or open if it comes back online and shuts the light off fandango.

I go by the definition of E to think things out. "Magnetism: You cannot separate heat from the chemical reaction."
I see that white powder on the battery posts, it says E is in it's chemical reaction-PUSHING volts {volts means push} and that means heating up molecules.

That also means a condensation or corrosion, or call it electrolysis, which produces a degradation of having a good ground due to condensation between ground cable and frame. This rust sets in is one variable. This layer blockage electrically between two touch points. Bike long in the tooth? Connector pins look all green looking?

No harness needed. It's usually the one wire. Say Red-Green-Black are the sensor outputs. 3 wires, 6 contact pins of male/female. Red is hot to the sensor, green for ground, black back to the ECU. Drop any wire, it spits a code. This would read something like: Connector not connected/wire out of connector/pin to pin in the connector.

Make sense?

 

[rrudytoo]

I'll try to explain a code occurring out of the blue. A self diagnosing computer bike will trigger a code via 3 ways: Signal out of range/short/or open.

Signal out of range is the component failure. Input are many different signals once say the throttle position sensor moves up and down in that range of open and closing. Call this analog input. ECU is sort of in AC mode [a ping back an forth] even though it's DC driven.

Short to ground is another variable. This may be short internally in the sensor, or a wire path from sensor to ECU. A short is now sending in a digital signal, or the same signal over and over. The ECU recognizing this single junk in, sets a code, then sends out a backup signal from the ECU. Sometimes this cuts about 2 HP off the engine, or limps it to save the engine components from damage.

Open is something like a 'work hardened' wire that is vibrating and has a wire loop out of the other laid down wires. This breaks internally and sets the ECU to code its backup number to calc off of.

For any of 3 variables to fail in the 'signal out of range' is watching the change of good known signals like swapping the ECU and it reads the same, does not clear, nor did the swapping of the other [good known] sensors change or clear the code. WATT does that tell us?

The can't be's are as follows:
Can't be a sensor that is out of range, short, or open inside the sensor. It does not clear changing out the component. It says it's not short, not open, but the signal is faulty in some sort of grounding or open path between the harness connection somewhere? It shows it's not the new good known sensor, it's not the good known ECU swap, it has to be a wire connection.

I'm not sure about the duc, but there are two [ground] running wire harness paths. One is the 12v grounds for lights, winkers, etc. Call this a junction point of all the tied together 12v wires. The other is a 5v ground with all the grounds tied together with one eye-loop to ground. Codes are the 5v ground to look for.

If you remember the first code recognized, you'd find the 5v eyelet with all the grounds wired together, (-) probe to the ground harness eyelet; (+) posi probe to the sensor wire's harness side to see a complete loop reading at the ohmmeter. No read, then that wire ground is 'open' somewhere?

Did I describe a set code from a phantom code? Phantom code is the comes and goes code. Or here today, gone tomorrow. It shows the sensor is not out of range, shows the right input, not junk. Meaning a single-constant-same-digit over and over is not happening. Is not short or open if it comes back online and shuts the light off fandango.

I go by the definition of E to think things out. "Magnetism: You cannot separate heat from the chemical reaction."
I see that white powder on the battery posts, it says E is in it's chemical reaction-PUSHING volts {volts means push} and that means heating up molecules.

That also means a condensation or corrosion, or call it electrolysis, which produces a degradation of having a good ground due to condensation between ground cable and frame. This rust sets in is one variable. This layer blockage electrically between two touch points. Bike long in the tooth? Connector pins look all green looking?

No harness needed. It's usually the one wire. Say Red-Green-Black are the sensor outputs. 3 wires, 6 contact pins of male/female. Red is hot to the sensor, green for ground, black back to the ECU. Drop any wire, it spits a code. This would read something like: Connector not connected/wire out of connector/pin to pin in the connector.

Make sense?

Bott,

Thank you for the most informative explanation.

In going over the wiring harness today I did find some exposed bare wire at the end of a lead that links the ECU to the Crank Position Sensor lead. The bare section was exposed due to the rubber boot where it connects to the CPS lead being torn. This was the Black wire. I cut the protective insulation that encases the three wires back about 1 1/2" and was surprised to see that the Black wire was not insulated but had its strands spread out around the other two wires similar to the way a coax cable is designed. About 1" from the opposite end where the lead enters the ECU connector the protective insulation stops and the end of the Black wire is crimped onto a conventional wire as it enters the ECU. Awkward explanation but maybe the pictures will help.

I doubt if that this exposed wire had anything to do with the glitches but it still needed attending to.

Rrudytoo


This is the torn boot that shows the bare Black wire. This is the end that connects to the CPS lead connector.

I removed a short length of the protective insulation and saw the Black wire's strands wrapped around the other two wires. (??)

This is the opposite end where the wires enter the ECU connector. Note: the Black wire's strands are rejoined and crimped onto a conventional insulated wire as it goes into the ECU connector.

 

[TOP-DUC]

The black wire and outer mesh appears to be the shield around the two signal wires. If that shield circuit is not solid and connected to ground at both ends, it can allow outside signals into the ECU and create havoc.
It looks cobbled together, so someone has likely had it apart before. You may need to replace the entire cable and sensor yo fix it. That area could be the source of your problems.

 

[rrudytoo]

Thank you TOP-DUC.

 

[rrudytoo]

All,

Thank you so much for your input on this most perplexing problem. After four months, several hundred dollars, many hours of my time and a whole lot of frustration and still not finding anything I have elected to throw in the towel. I sold the bike to an independent Ducati shop where they will either repair it or part it out. It is a shame as this bike had personality, good looks and was a genuine performer and I will miss it but it is time to move on.

I wish you all many, many miles of smiles!

Rrudytoo

 

[edgar999s]

Ah bummer, sorry to hear that Rrudytoo 😞

Many agree that Ducati electrics of the 90s and 2000s were of questionable integrity... And it just takes one wire failure to mess the whole bike up. On these models its usually the wires feeding into the fuel pump, the fuel pump relay, the stator oil wick issue, under-specced starter wires, fried rec/regs or bad ECUs.

I know exactly what you mean with the internal wiring of the CPS - I dissected mine too at one point. Was quite perplexing.

I had a multitude of issues that caused the bike to not start. Things like this came up (front coil connector from loom side) -

Corroded internals of the fuel pump relay connector -

This is pretty much what it takes -

I replaced a multitude of old wires, re-soldered weak junctions, optimised lengths and routing, overhauled the internals of all switchgear, and re-wrapped everything.. Twice..
The main loom was the culprit. Secondary loom only needed a re-wrap.
It got so bad at one point I opted to swap my ECU like you, and a forum member had posted theirs over to me. Before it arrived, I had solved my issues.

My electrical troubleshoot also included -

Fuses check
New rec-reg
New CPS and gapped correctly
Fuel pump overhaul
Starter overhaul
New stator
New battery
Upgraded starter wires and ground
New coils
New plugs
New starter solenoid
Even went as far as ultrasonic clean of main ECU connectors

The final fix came during the second loom overhaul where I had found that the new pins on the inside of the fuel pump relay connector weren't engaging fully with the relay itself.

I realise it's a lot of hassle and not everyone has time for all this.

Anyway, may you find a good bike that treats you well!

 

[Stumpy]

I had a similar problem with a 2003 GSxR always thinking electronics but found out that water was laying in the bottom of the tank, and produced a backfire and reduced power but only at high revs. For how long it was sitting there I have no idea but long enough to end up having to rust treat and plastic line the tank. Lesson to me was do not always presume electronics.