907ie misfire - maybe a cure???

discussions specific to the 907IE
Post Reply
Laddie
Posts: 56
Joined: Fri Sep 21, 2007 12:00 am
model: 907 I.E.
year: 1991
Location: Australia

907ie misfire - maybe a cure???

Post by Laddie »

Hi Guys,
I think many of us have experienced the 3,500 - 4,000 RPM misfire / stutter at cruise...the one that nobody can pin down...???
I've owned a 907ie for around 2 years now.
The bike has been cursed with the dreaded "misfire" - it played up on the day I bought it, even though the previous owner swore blind it ran like a charm...yeah, right.

Typical symptoms - runs fine at idle or cold, warms up and plays up.

so I called the bike shop that did the last service for the previous owner - they were a general shop...their response was that the 907 was too complicated and they did not have the correct equipment to tune it.
good start, huh?

...so I take it to the local Duke dealer...(bad idea).
they do an injector clean and adjust the steering head bearings (not asked for) and $800 later it's just the same. I take it back; they fart around; I take it back; they get pissed off...usual stuff.

so I try a Duke race mechanic...
the fuel tank is full of rust; the filter is fully clogged - solid.
the filter is replaced - but that's not the problem.
The tank is pitted in places and will leak soon if not replaced - so I order a new tank from the USA.
there is a tiny leak in the outlet from the fuel pump - the pump is repaired - but that's not the problem...
the bike is fully tuned; the cams and belts were way out of tolerance - but that's not the problem
we try replacing the P7 and the chip
I even had the P7 tested by FIM and it was fine...
the water temp sensor dropped out at around 90 degrees; this was replaced - but that's not the problem...
the bike would not hold steady throttle at 3,000 - 4,000 RPM unloaded - even careful throttle control the tach would dive 1,000 - 1,200 RPM on the way down or jump from 3,000 to 4,500 on the way up - very touchy....

I change the rear sprocket to lower the ratio.
The idea was to cruise at 100km/hr at around 4,300 instead of 4,000 RPM. This made a difference - but didn't fix the misfire...and lead to another problem in traffic...

I try another local Duke service agent.
we look at the coils, the ignitors and leads...
we swap the two crank pick ups with operational ones...no change
the misfire is finally observed on a computor
the timing (spark & fuel) are all over the place...what could cause this???

In hot weather the bike develops a horrendous flat spot at 1,500 RPM. The lower ratio sprocket means that if I'm idling in traffic (yuk) that the bike is chugging at idle or zing - trying to crack a wheel-stand....what a pain in the arse....

the new tank is fitted - no change...not that I expected it...

a new fuel pump is fitted...and tested...3 bar; spot on.

so I take the bike to an EFI technician...
he goes over the thing front to back...
new rectifier / regulator....new MAP sensor...re-crimped wiring...all bits n' pieces re-checked - same problem

many hours pondering and running out of options I try something...

I stuff a cigarette butt in the tube between the MAP sensor and the air box. The idea is to damp the air pulses to the MAP sensor. The ciggie butt will gradually allow the MAP sensor to adjust to the ambient pressure and throttle vacuum - it just takes a little longer.
The bike now has a controllable throttle. I can creep all around 3,000 - 4,000 RPM just like a Jap bike. The short test rides so far have not changed the performance under acceleration - it still takes off just fine. There is no flat spot.

I'll be pissed off if this "free" fix does solve the problem.
I'd hate to add up the time, money and strtess this bike has had spent on it.
User avatar
Finnpaso
paso grand pooh-bah
Posts: 3090
Joined: Mon Jul 26, 2004 12:00 am
year: 0
Location: Finland

Post by Finnpaso »

When my 907 old engine valve seats went to bad condition, it function well cold, but after warmed up, it started to make such, that very difficult to start and didnt function, like should....

Have You checked valve clearances and compression pressure ??????? Pressure test gives more info, are valve seats worn out. That was problem in those early 90s Ducatis, like 907, bad quality from factory...
Antti http://www.youtube.com/watch?v=HeKOh3XoXPg&NR=1
KTM 990 Adventure -08 metal dark grey
Paso 750 -89 red/metal grey
907IE -91 red/metal grey
907IE -91 red
2xST4S -02 red/metal grey
ST2 -01 red/metal grey
Volvo V70 Bi-Fuel Classic/titanium
Laddie
Posts: 56
Joined: Fri Sep 21, 2007 12:00 am
model: 907 I.E.
year: 1991
Location: Australia

Post by Laddie »

Finnpaso wrote:When my 907 old engine valve seats went to bad condition, it function well cold, but after warmed up, it started to make such, that very difficult to start and didnt function, like should....

Have You checked valve clearances and compression pressure ??????? Pressure test gives more info, are valve seats worn out. That was problem in those early 90s Ducatis, like 907, bad quality from factory...
when the race mechanic did the tune we tested compression and it was around "standard" both cylinders. Valve clearances were re-set, too.
I would expect if it was a valve problem that the bike would back-fire.

to explain a bit more, the "miss" is intermittent...commonly the bike needs to run (say) 15 minutes before it appears...this is why it's hard to get a mechanic to experience it - they might need to ride for a while until it appears. the road has to be flat and somewhere that a steady throttle can be held for a while. Even very very slow acceleration will cause the misfire to disappear. My guess is the TPS takes over in the MAP/ECU??? I dunno.

Temperature does come into the equation somewhere - the bike is OK below 80 degrees. A few times the mechanics have ridden the bike with the fairing off and it doesn't get hot enough to cause the misfire.

if I blip the throttle the miss disappears for maybe 5 - 15 seconds - then returns at cruise - go outside the 3,500 - 4,200 RPM range and the miss is not present. A guy from Weber -Marelli suggested to check for manifold air leaks (in case this was leaning out the air/fuel mix) - typically blipping the throttle will make more fuel available to the cylinder for a short time...

I checked the balance between the manifolds with ye olde analogue vacuum gauges. Typically, the gauges "pulse" on the intake stroke. to remove the pulse I similarly stuffed ciggie butts into the tubes - very high tech. The balance at idle was fine but there was a difference at 4,000 RPM of about 10%.
I adjusted the butterflys and then re-set at idle.
This seemed to settle the misfire regularity - but it was still present - just less often.

The air box does have a pulse - not severe as each manifold but still present. The MAP sensor adjusts for fluctuations in manifold and absolute pressure...this is pretty fine adjustment in absolute - so I estimated that the output signal to the ECU would possibly be fluctuating. I know we had observed this fluctuation with an oscilloscope - at the time we were having regulator problems and we put some of the signal variation down to power supply.
I just didn't really know how steady the signal would need to be.
I was not convinced that the MAP signal was steady so I thought I'd try to damp the pulse from the air box.. It could only delay MAP pressure response times.

since I stuffed the ciggie butt in the MAP tube I have done about 1 hour of riding. The bike is more tractable and much smoother at steady throttle. There is a slight change to the exhaust note on deceleration. The running temp has dropped about 5 degrees. The bike looks better, too (lol).

I have a sneaking suspicion that the P7 ECU does not cope well with receiving fluctuating data from the sensors. I'm not sure with the P7 but I do know that some ECUs "learn" and that un-plugging them can re-set the drive responses. Maybe too much "clutter" into the ECU results in confused outputs. It would be handy to know acceptable tolerances for the various signals...

My thoughts are that the 907 with EFI is designed to be a higher performance bike (in response to the poor response / performance of the 906) and that the designers have engineered it to be ridden harder - which is probably why the throttle control in more like a racing bike than a tourer...I expect they concentrated on the horsepower, response and acceleration rather than the cruise aspects????

I'll stick with the trial for a while but at this stage all indications are positive.
Tamburinifan
paso grand pooh-bah
Posts: 1527
Joined: Thu Jun 21, 2007 12:00 am
model: 907 I.E.
year: 1991
Location: Gothenburg, SWEDEN

Post by Tamburinifan »

Wow, that`s a lot of work/testing! Thank you for sharing!

Have sort of same problems but still haven`t checked all basics yet, dunno if injectors are clean f e.


>The idea is to damp the air pulses to the MAP sensor<
You mean from the air sensor in the airbox to the P7 ECU?


One thing I`ve read comes in mind: The placing of air sensor at rear of airbix is maybe not a good idea? It`s located behind fairing on later models.
Here`s a little about that:
After a little more investigation, however, I came to realise something even more influential - the fact the air temp sensor on the 851 (and 888) is inside the airbox. Even on a cold autumn night, the airbox can get pretty hot when sitting at traffic lights (40 to 50 degrees), with all the heat coming off the engine. Given the large variation in trim for air temperature in the 851 map, in this case, leaning off quite a lot for high air temps, this was provoking everything from backfiring on overrun to a huge stumble on takeoff. This changing trim due to what the air temperature sensor was reading at a given time is where my ‘sometimes rich/sometimes lean’ problems were coming from, but autumn nights in particular, for some reason, really showed them up.

In contrast, the air temp sensor on a 996 or 900SS is under the headlight/dashboard area, and much less effected by engine heat. Which is where the air temp sensor on my 851 has now been relocated. Simply for the sake of consistency.
Whole story: http://www.moto-one.com.au/performance/ ... npro4.html

Maybe try to relocate air sensor? It`s a longshot, I know, but who knows!
Gert

907 I.E. -91
M900 -97
MTS 1100s -07
User avatar
Finnpaso
paso grand pooh-bah
Posts: 3090
Joined: Mon Jul 26, 2004 12:00 am
year: 0
Location: Finland

Post by Finnpaso »

Laddie, hows Your TPS adjustment and does it function well in hole its area?

Do U have any friends, who have P7 ECU to change and test, if Your ECU make some "mistakes". My original ST4S ECU failed and it didint want to function well after engine warmed. It dropped bach sylinder down. New FIM ECU function now very well... :thumbup:
Antti http://www.youtube.com/watch?v=HeKOh3XoXPg&NR=1
KTM 990 Adventure -08 metal dark grey
Paso 750 -89 red/metal grey
907IE -91 red/metal grey
907IE -91 red
2xST4S -02 red/metal grey
ST2 -01 red/metal grey
Volvo V70 Bi-Fuel Classic/titanium
Laddie
Posts: 56
Joined: Fri Sep 21, 2007 12:00 am
model: 907 I.E.
year: 1991
Location: Australia

no expense spared...

Post by Laddie »

I have spent a #$%#ing fortune on this bike...there were times I was going to get rid of it...I could hardly walk past it without spitting...

the original TPS would not zero - that's a sign they are failing.
it would stop at about 150mV - (which is actually "zero" on shut throttles).
The TPS was replaced and set as per FIM instructions to match the ultimap chip. The TPS has been zero-ed a few times to make sure it's still OK

The P7 ECU was bench tested OK by Duane at Ultimap.
The bike has an Ultimap chip to start with when I got it.
we even tried another P7 ECU
the race mech borrowed a Duke OEM 907 chip to try - no difference.
then Duane made me a few "special" chips to try.

I can tell you I've tried and tested EVERY bloody component in the EFI circuit...and found some that were faulty...replaced them and no difference.
We even tried replacing the WTS and ATS with resistors to "lie" to the ECU.

Honestly - so far after spending more money than the bike's worth on pieces I can tell you the BEST result I've had has been the ciggie butt in the map sensor air line. Crude but effective. I rode the bike today and it's a different beast all together...tractable, smooth, predictable...and still accelerates very well.
I used to get cramps in my arms on twisty roads from holding the throttle so tight ready to shut it down on the flat spot response.

today I really enjoyed the ride. I'm still buzzing because it behaved so well.

it took 10 minutes to fit the filter into the hose - you only need to take off the side cover and force someone with small hands to do the job...lol

all I can say to those with jerky response or the dreaded misfire - try the filter - it's FREE. (And that's a lot cheaper than ANY Ducati part.)
Laddie
Posts: 56
Joined: Fri Sep 21, 2007 12:00 am
model: 907 I.E.
year: 1991
Location: Australia

the are two sensors on the air box...

Post by Laddie »

the ATS is the air TEMPERATURE sensor - and may be affected by heat because it's just near the vertical cylinder...

Like I said elsewhere - I replaced the ATS with a resistor which would have told the bike it was at between 30 - 40 degrees air temp.

the sensor on the outside of the air box is the MAP [PRESSURE]sensor.
it has a rubber hose connecting the air box to the inlet to the sensor. The MAP sensor tells the ECU what the ambient air pressure is (like if you are at altitude) and it reads the airbox internal pressure for engine vaccuum

and yes, I have tried a spare MAP sensor to make sure it works...MAP sensors are hard to test properly because you need controlled conditions and vaccuum.
I was lucky and picked one up cheap - they're about $500 from Ducati. It is the same unit from a Peugeot.
Tamburinifan
paso grand pooh-bah
Posts: 1527
Joined: Thu Jun 21, 2007 12:00 am
model: 907 I.E.
year: 1991
Location: Gothenburg, SWEDEN

Post by Tamburinifan »

OK, got it. Thx again f sharing1
Will keep the cig butt trick in mind... :thumbup:
Gert

907 I.E. -91
M900 -97
MTS 1100s -07
User avatar
Finnpaso
paso grand pooh-bah
Posts: 3090
Joined: Mon Jul 26, 2004 12:00 am
year: 0
Location: Finland

Post by Finnpaso »

Sometimes these injection faults are very difficult to fix... :evil:
Antti http://www.youtube.com/watch?v=HeKOh3XoXPg&NR=1
KTM 990 Adventure -08 metal dark grey
Paso 750 -89 red/metal grey
907IE -91 red/metal grey
907IE -91 red
2xST4S -02 red/metal grey
ST2 -01 red/metal grey
Volvo V70 Bi-Fuel Classic/titanium
Laddie
Posts: 56
Joined: Fri Sep 21, 2007 12:00 am
model: 907 I.E.
year: 1991
Location: Australia

hasn't missed a beat...

Post by Laddie »

another hundred++ kms and still great...
I'm actually smiling while I ride it!!! I'm letting it sleep in my room now...my 900SS is jealous.

I'd be interested to test the MAP output to the ECU on an oscilloscope to see what the chart looks like before and after...
I expect the line would be a lot flatter - without the pulses.
I don't know enough about the ECU to understand how it processes the MAP signal but my guess is that a wider tolerance makes things more difficult for it to cope with...

I thought of doing this six months ago - I don't know why I didn't try it then...

It would be even better to put the thing on a dyno to determine if the power output is affected. There is a dyno near me but they keep trying to sell me an aftermarket box to put on the injectors...
I don't expect any power increase (how could it?) but the dyno may show up response lag times that you don't feel thru your ass!!!

I'm sure if I said to you that I had used some fandangled expensive electronic device to "condition" the MAP signal that there would probably be more acceptance of the method.

Maybe I can sell the idea to Ducati...
Tamburinifan
paso grand pooh-bah
Posts: 1527
Joined: Thu Jun 21, 2007 12:00 am
model: 907 I.E.
year: 1991
Location: Gothenburg, SWEDEN

Post by Tamburinifan »

I replaced the ATS with a resistor which would have told the bike it was at between 30 - 40 degrees air temp.
What value on that resistor? What type of resistor?
Do you know to what percentage of fuel this modification alters?
Gert

907 I.E. -91
M900 -97
MTS 1100s -07
Laddie
Posts: 56
Joined: Fri Sep 21, 2007 12:00 am
model: 907 I.E.
year: 1991
Location: Australia

Post by Laddie »

All the temperature sensors are just NTC resistors that change resistance at temperature...
the "fixed" resistor only replicates what the ATS or WTS would output as a resistance at a particular temperature.

you just use a "standard" resistor - the kind you find in any electronics shop. these only cost a dollar or less each - so you can buy a few different ones and try any temperature [resistance] you like...I would try to chose (say) 80 deg for the WTS and 30 - 40 or lower for the ATS. If you ride in a hot place or ride in slow traffic where the bike gets hot maybe try WTS 100 / ATS 50-60 too...

be careful not to damage the plug when you push them in - preferably use similar size spade connections.

NOTE: DO NOT LEAVE THESE IN FOR AN EXTENDED TIME. This is just to diagnose a potential fault and eliminate a variable. If you leave them in the bike will think it's at the selected operating temp even if it's stone cold or boiling hot - you might feed it the wrong fuel mix - probably won't kill it unless you overheat...TAKE CARE WITH THIS. Keep an eye on your engine temp.

here is a link to the ATS and WTS and graphs for temperatures
http://www.magnetimarelli.com/racing/Sensors_2005.pdf
http://www.cosworth.hu/content/hasznos/ ... -05-07.pdf

suggest for the ATS use a 2k ohm +/- 500 ohm

you can test the ATS sensor with a multimeter and a hair dryer on heat.
you can even put it in the fridge if you live somewhere cold and need to test below zero...
you can test the WTS with a multimeter and a pot of water on the stove - when it boils you should have 100 degrees
watch the multimeter doesn't drop out during the heating - this might indicate a dead spot in the sensor.

the ECU reads these signals [resistance] and adjusts the fuel amount according to the map in the chip.

it is not uncommon for auto manufacturers to fit "piggy-back" resistors in parallel or series to EFI components to adjust the mix. This also goes for the sensors in the exhaust.
User avatar
Finnpaso
paso grand pooh-bah
Posts: 3090
Joined: Mon Jul 26, 2004 12:00 am
year: 0
Location: Finland

Post by Finnpaso »

Good job, Laddie! :thumbup: :thumbup: :thumbup:
Antti http://www.youtube.com/watch?v=HeKOh3XoXPg&NR=1
KTM 990 Adventure -08 metal dark grey
Paso 750 -89 red/metal grey
907IE -91 red/metal grey
907IE -91 red
2xST4S -02 red/metal grey
ST2 -01 red/metal grey
Volvo V70 Bi-Fuel Classic/titanium
Tamburinifan
paso grand pooh-bah
Posts: 1527
Joined: Thu Jun 21, 2007 12:00 am
model: 907 I.E.
year: 1991
Location: Gothenburg, SWEDEN

Post by Tamburinifan »

Thx f the info & links, Laddie! :thumbup:
Gert

907 I.E. -91
M900 -97
MTS 1100s -07
blakduc1

Post by blakduc1 »

Laddie,

I am so glad to see your posts about this stuble problem!!
I have two 907ie 1992 models, both charcoal color. I have experienced this same dreaded hiccup/ stumble poor running condition. Ducati mechanics do not even want to try to fix it. None seem to know what it is exactly. Nothing seems to cure it. Although at times it has been pretty good and fairly smooth. What mine usually do when in good tune, is run like banshees when running them hard or at a track day. But try to ease around town and you will be jerking, blipping and cursing. And as it gets hotter, it gets worse. They like cold dense air evidently, or either the computer can not deal with the temp changes.

I am adjusting valves on one now, that has been down for some time, and I will look forward to trying this cig butt method. Any way you could take a pic of exactly where it goes and post it or send it to me??? That would be really helpful.

There has to be a more scientific fix. But hey, if this works in the meantime until we figure the reason it works and how to make a better remedy I will gladly do it.

Thanks for your input.
Post Reply