31 replies to this topic

[Hector]

Last week I took my 2009 TDM for a tour of Spain,where it performed brilliantly. Except for a mystery cold start problem one morning. We'd stayed at a rural hotel in the mountains and it went down to 2 degrees overnight. In the morning the bike wouldn't start. The battery was strong but the engine wouldn't fire up. I tried several bump starts but nothing. After numerous attempts to start it up the battery was starting to weaken so I phoned my breakdown company who took me to a Yamaha dealer in Santander. They took the bike off the tow truck and it fired up FIRST time!! Their mechanic checked it over on a diagnostic machine (and found no faults), charged the battery for 3 hours, charged me 25 euros and off we went. After that it started up every morning, but the night temperatures were higher at about 8 degrees. Any ideas what this mystery random fault could be? My thoughts are the cold mountain air created condensation which affected something, which then evaporated while it was on the tow truck to Santander. Any ideas or suggestions greatly appreciated. Thanks for reading this long question.

[dapleb]

Cold Start Mechanism

Most likely needs a fettle.

[Studley Ramrod]

Condensation on the HT leads/plug caps perhaps ?

[TKH]

I'm with Studders, Condensation on ht leads (due to cold mornings) which then go to earth. Self amalgamating tape works wonders if that's the problem  - see other posts.

[dapleb]

If it's leads then replace leads then wrap if you still feel the need.

[fixitsan]

There is a possibility that it is the battery. This is a graph showing expected battery voltages when testing it at half rated cold cranking amps, CCA http://www.carquestp...batteryTemp.gif

 

The voltage of the battery drops with temperature. Additionally the oil is thicker than when running and seals and bushes are probably tighter with a cold engine.

 

It is possible that the battery is close to failure. If you were riding at very steady rates for a while before turning the engine off then when the bike was parked the battery might not have been at full capacity.

 

The effect of temperature on capacity is also known - the capacity of a battery at 0 degrees Celsius is about 80% of the capacity of the same battery at 25 degrees Celsius  https://www.e-educat.../less3_fig7.jpg

 

A known phenomenon is that due to lots of engine cranking, the battery is put under load, which causes some internal heating and also movement of electrolyte around the plates, enough for the voltage to rise a little after a few minutes rest following a long cranking period.

 

 

The ECU in the TDM has voltage regulators which will not operate if the voltage drops below a certain level. It won't just work badly, it will simply not work at all. The shake around on the back of the van combined with a likely increase of temperature during that time might have  been enough to make the difference.

 

The only other thing I can think of which hasn't been previously mentioned would be a poor connection in one of the many plugs under the tank, or to the relays

Edited by fixitsan, 09 April 2018 - 08:03 am.

[Bjørge]

The ECU in the TDM has voltage regulators which will not operate if the voltage drops below a certain level. It won't just work badly, it will simply not work at all. The shake around on the back of the van combined with a likely increase of temperature during that time might have  been enough to make the difference.

 

Wow ! That would be a disastrous feature - the ECU shutting down around freezing point......

[fixitsan]

 

Wow ! That would be a disastrous feature - the ECU shutting down around freezing point......

 

Yah, or it would be a safety feature to prevent riding on ice

Edited by fixitsan, 09 April 2018 - 09:57 am.

[dandywarhol]

If it's leads then replace leads then wrap if you still feel the need.

 

Go to an old school car accessory shop and buy a metre (old school might be in feet  ) of non resistor, copper core HT lead. The old leads unscrew from the coil and from the plug cap and just screw on the new ones. 10 minute job, unless I'm talking pish and the newer 900s have sealed coil and lead - if so then disregard everything I'm just said.

[dmmsta]

 

Go to an old school car accessory shop and buy a metre (old school might be in feet  ) of non resistor, copper core HT lead. The old leads unscrew from the coil and from the plug cap and just screw on the new ones. 10 minute job, unless I'm talking pish and the newer 900s have sealed coil and lead - if so then disregard everything I'm just said.

Accurate for the 9er!

[JBX]

The ECU core & the sensors work at 5V.

Two reasons for that : 

- all µC work at either 5V or 3.3V, other ic work generally at 5V.

- 5V will allow a very weak battery to still get the ECU/sensors to work with a large safety margin.

 

All actuators are activated by the ECU with ground switching so they can work at any voltage level, usually 12V.

Most relays (mainly relays involved in the engine management) are not 12V relays but 9V. this to allow them to work with a weak battery during engine start.

 

Oddly enough, the Vss (ground) for the sensors is different than the battery/chassis ground !

[fixitsan]

The ECU core & the sensors work at 5V.
Two reasons for that : 
- all µC work at either 5V or 3.3V, other ic work generally at 5V.
- 5V will allow a very weak battery to still get the ECU/sensors to work with a large safety margin.
 
All actuators are activated by the ECU with ground switching so they can work at any voltage level, usually 12V.
Most relays (mainly relays involved in the engine management) are not 12V relays but 9V. this to allow them to work with a weak battery during engine start.
 
Oddly enough, the Vss (ground) for the sensors is different than the battery/chassis ground !

What is used for the critical injector switching .... relays there make no sense so are they Mosfets ? a low Vgs would drastically reduce peak current through the injectors. (a charge pump could be used but it too will have a strict minimum Vin ) A regular BJT instead would suffer similar problems with it's Vbe , and heat disipation

What is the Vmin of the injectors ? I can't imagine them working as well at 8V as they do at 14V. Just wondering.

I'm not ruling out the HV lead idea but I haven't come across a situation where 2 leads break down together completely at precisely the same climate conditions and then both start working together again simultaneously.

Im sure theres a logical reason

Edited by fixitsan, 10 April 2018 - 06:19 am.

[itchyfeet2]

I recently had quite a few cold start / damp start issues with my bike, (4tx) After going around in circles for a while the fault was traced to a poor connection on the low tension side of the HT coils. At first glance and feel both the small wires that connect to the coils seemed perfectly tight and secure. However although the small rubber boots of the electrical connections were tight, underneath the actual metal part of just one of the connectors on just one of the coils was hardly making contact. It was probabaly stretched when perviously removed. i missed it for ages as the small rubber boots had shrunk with heat? or just hardened and attatched themselves to the coil tabs which masked the fact that the metal underneath was allowing condensation to prevent starting on cold mornings.

 

Hope your prob is as simple to fix.

[thelodger]

I don't think this is the best explanation for your problem but I put it forward as an alternative idea for what went wrong . It is possible that you filled up with petrol which had a lower octane rating and/or higher ethanol content than you are used to ( or it could have been bad petrol ) Once you had tried to start the bike a couple of times it probably fouled the plugs so wouldn't start . The journey on the pickup warmed up the petrol to increase its combustibility and dried out the motor which is why it started later on. Food for thought perhaps !

[dandywarhol]

I don't think this is the best explanation for your problem but I put it forward as an alternative idea for what went wrong . It is possible that you filled up with petrol which had a lower octane rating and/or higher ethanol content than you are used to ( or it could have been bad petrol ) Once you had tried to start the bike a couple of times it probably fouled the plugs so wouldn't start . The journey on the pickup warmed up the petrol to increase its combustibility and dried out the motor which is why it started later on. Food for thought perhaps !

 

It's a thought but having run a 900 on Bulgarian wolf piss fuel, I don't know if that is the problem here.

 

I don't have a wiring diagram to hand but the low tension affecting both is a possibility as previously said - I think the coils low tension share a common earth.

Edited by dandywarhol, 10 April 2018 - 09:03 am.

[fixitsan]

 

It's a thought but having run a 900 on Bulgarian wolf piss fuel, I don't know if that is the problem here.

 

I don't have a wiring diagram to hand but the low tension affecting both is a possibility as previously said - I think the coils low tension share a common earth.

 

 

That's right, my money is on low voltage, just out of spec. My 900 didn't want to start a couple of times in winter, despite cranking strongly enough and not slowing down readily. But then I just connected another battery with jump leads, didn't move the bike at all, and it cranked a little bit faster and fired up after a couple of seconds.  The same battery is on it now and it works fine in the warmer (above freezing) temps

[JBX]

The injectors work well at 9V, the main problem is that they work slower than usual. To prevent this, the ECU monitors the battery voltage and uses the value to compensate the injection timing when required.

 

This may be a problem at engine start-up : if the battery voltage drops too much, the ECU cannot compensate and does not open the injectors at all, even if the starting motor turns.

 

There may be also a problem with the fuel pressure which may not be high enough as the fuel pump cannot works with the right voltage.

To prevent this the ECU during the start-up procedure first launch the fuel pump alone to get enough fuel pressure inside the common rail.

[fixitsan]

Voltage dependent injection timing ? Are you sure ? I know a lot of people test injectors with 9V but I note that most of them don't do so with normal fuel pressures. I've seen people claim a test at 5V is reliable......and when they put pressure on the pintle it suddenly doesn't work !

[JBX]

Yes I'm sure, this is what happens during start-up on the injection system.

Remember the Diag Mode #9 is not there just to give a nice display !

During normal working condition the charging system / battery delivers the right voltage so the injectors do not need timing compensation.

 

Btw 5V is obviously part of the "internet claims". 

[fixitsan]

Yes I'm sure, this is what happens during start-up on the injection system.
Remember the Diag Mode #9 is not there just to give a nice display !
During normal working condition the charging system / battery delivers the right voltage so the injectors do not need timing compensation.
 
Btw 5V is obviously part of the "internet claims".

ok. I assumed an old Denso ecu from 2002 predated low voltage pwm compensation but live and learn.

check it for yourself, some folk playing with injectors at 5V .... a misunderstanding of what Vdd means !

Edited by fixitsan, 10 April 2018 - 06:12 pm.