Starting issues

[suzyj]

Hi guys,

My 695 has intermittent starting problems.  Approximately 30% of the time (generally when the bike is cold, but not necessarily), the solenoid doesn't pull in.

The battery is good, and when it won't engage the solenoid I can start it very easily by bridging the solenoid contacts.

So I looked at the other side of the solenoid.  When I energise the solenoid coil, it pulls in fine.  At this stage I was thinking something must be up with the ECU, so I cut the line to the ECU that drives the solenoid and wired it up directly to the starter switch as a temporary measure, while I save pennies for a replacement ECU.  Unfortunately the ECU really dislikes not having a solenoid coil.  The bike starts perfectly, but the check engine light stays on, and even weirder, when I stop the bike the ECU periodically pulses the fuel pump until I pull the fuse.

So I removed the wire direct to the starter button, and I'm back to bridging out the starter contacts when it refuses to start.

I know the battery and starter are fine, and I've ruled them out.  It starts reliably when the solenoid is bridged.

The starter solenoid pulls in reliably when I energise it's coil, so it's fine (I think - it's possible there are smarts in the ECU that limit coil current or something).

The ECU is clearly seeing a connection to the starter solenoid, as it gets very upset when it doesn't have one, so that rules out the wiring from the solenoid back to the ECU.

So that leaves me with???  I'm at wits end.  Has anyone else come across anything similar?

[Howie]

Could the ECU be smart enough to see an open circuit but not smart enough to see high resistance?  I don't have a wiring diagram in front of me, but I think I recall one wire from the solenoid going directly to the ECU and the other wire branching off to the kill switch, ECU and ummm dunno.  Anyway, since the bike fires up bypassing the solenoid I think we can assume the kill switch is good.  I am also going to guess the solenoid coil is energized when the key is turned on and when you hit the starter button the computer grounds the solenoid through a transistor, allowing the starter motor to operate. 

Assuming my guess is correct, using your volt meter you should have battery voltage at one terminal at the solenoid winding with the key on.  When you hit the starter button you should get battery voltage at both.  One will probably read less.  Back probe from the terminal that reads less and you should find your problem. 

[suzyj]

Further diagnosis makes me suspect the solenoid.  When I disconnect the low current side of the solenoid and substitute a resistor load, the ECU powers the load appropriately.  I suspect the solenoid coil is drawing too much current and the ECU is clever enough to sense this - certainly the load drivers for the indicators etc have current sense.

So the next logical step is to replace the solenoid.

Oh, for reference, according to my trusty Fluke clampmeter, the starter motor stall current is 302 A with a freshly charged 2P4S stack of A-grade A123 LiFePO4 cells.  That's a pretty hefty current.  With dodgy ebay cells, I was only getting 175 A.

[Ddan]

I really wish I understood WTF you just said   

[Howie]

Further diagnosis makes me suspect the solenoid.  When I disconnect the low current side of the solenoid and substitute a resistor load, the ECU powers the load appropriately.  I suspect the solenoid coil is drawing too much current and the ECU is clever enough to sense this - certainly the load drivers for the indicators etc have current sense.

So the next logical step is to replace the solenoid.

Oh, for reference, according to my trusty Fluke clampmeter, the starter motor stall current is 302 A with a freshly charged 2P4S stack of A-grade A123 LiFePO4 cells.  That's a pretty hefty current.  With dodgy ebay cells, I was only getting 175 A.

Interesting.   Your diagnostics make a lot of sense, but I wonder if you are giving the computer too much credit for smarts.    Cheaper solenoid, same as OEM except for the connector.  Shows out of stock so you should call.  Or you could pick one up at a local Yamaha dealer and buy your own connectors.
http://www.ca-cycleworks.com/ca-cycleworks-yamaha-starter-relay-battery-and-charging

Gotta ask, how did you stall the starter motor?  In gear with the brake on?

[suzyj]

Interesting.   Your diagnostics make a lot of sense, but I wonder if you are giving the computer too much credit for smarts.    Cheaper solenoid, same as OEM except for the connector.  Shows out of stock so you should call.  Or you could pick one up at a local Yamaha dealer and buy your own connectors.
http://www.ca-cycleworks.com/ca-cycleworks-yamaha-starter-relay-battery-and-charging

Thanks for that.  I'm tossing up between buying a used one from ebay, buying a new replacement, and engineering one.  Hence measuring the current - it gives me an idea of what sort of current an engineered one would have to put up with.

FWIW, it looks like a couple of these: in parallel would manage the current.

Gotta ask, how did you stall the starter motor?  In gear with the brake on?

I used the max hold function on the clampmeter to capture the initial start current (with the bike in neutral), so it's not strictly the same as a stall current, but close enough and a lot less exciting.

[DarkStaR]

...
Unfortunately the ECU really dislikes not having a solenoid coil.  The bike starts perfectly, but the check engine light stays on, and even weirder, when I stop the bike the ECU periodically pulses the fuel pump until I pull the fuse.
...

So that leaves me with???  I'm at wits end.  Has anyone else come across anything similar?

Weird.  Don't think the ECU can detect whether or not the solenoid is plugged in...

I recently resurrected a 620 with a partially fried ECU (that I helped fry), by bypassing some/most of the start circuit from the ECU.

http://www.ducatimonsterforum.org/index.php?topic=49635.msg907221#msg907221

[Speeddog]

Is there any condensation or water in your gauges?

Inspect the sidestand switch wiring all the way from the switch to the wiring loom.
I've had a couple customer bikes get water in the gauges, and for some reason throw enough current to the sidestand switch to barbecue the wiring to the switch.
One of those bikes ('04 M1k) would refuse to energize the starter solenoid, had to do the 'direct-wire' trick as you did.
Didn't notice a fuel pump cycling phenomena.

<alternate theory>
Your ECU is shot.
Please create a functional Monster ECU using a MicroSquirt module.
That way I can borrow your work and replace the FIM U59 I've got.


Seriously, I'm close to pulling the trigger on a MicroSquirt ECU.
If you want to do one, I'm interested in having someone that understands the specs to talk to.
My U59 works fine,but I can't tune anything besides the trim, and if it burps I'm on foot.
</alternate theory>

[J5]

Further diagnosis makes me suspect the solenoid.  When I disconnect the low current side of the solenoid and substitute a resistor load, the ECU powers the load appropriately.  I suspect the solenoid coil is drawing too much current and the ECU is clever enough to sense this - certainly the load drivers for the indicators etc have current sense.

so this test was when it refused to start ?

[suzyj]

Thanks for the heads up - I'll go checking other parts of the loom - I had noticed some condensation in my guages a month or so but pretty-much ignored it.

As for the microsquirt - I'm drawing ideas at the moment.  Removing the stock battery gives me piles of room where the old battery went to do all sorts of things.  I figure the largest Hammond instrument case (see http://www.hammondmfg.com/1455V2.htm would fit nicely there.  Inside there's room to squeeze an A123 battery, switching rectifier/regulator, MOSFET based solenoid, and replacements for the other relays.

Making the case just a tad longer gives space to mount a couple of 35 pin ampseal connectors (these are what's used in the microsquirt, and also on the monster gauges).  There's gobs of space inside to put a microsquirt with interface electronics to all the Ducati sensors and actuators.  Of course you'd have to cut the existing connectors off the loom and replace them with ampseal ones, or build a completely new loom.

Here are my doodles thus far:



This is wild speculation though.

[Langanobob]

<Snip>

Seriously, I'm close to pulling the trigger on a MicroSquirt ECU.
If you want to do one, I'm interested in having someone that understands the specs to talk to.
My U59 works fine,but I can't tune anything besides the trim, and if it burps I'm on foot.
</alternate theory>

I've seen the subject of a MicroSquirt Ducati system come up before but so far I haven't seen anyone actually do it.  If you and SuzyJ end up working together on this it would be great to have a dedicated blog type of thread  (maybe in the Accessories and Mod's section?) so that we can follow along with our own projects.  I've been close to getting a MicroSquirt myself, but for a relatively mundane project on my '87 Toyota pickup.  I've talked to the MicroSquirt developers and they seem to be very cooperative and helpful.

Anyway, hope you can go ahead with your Duc MicroSquirt project.

[Speeddog]

I googled "microsquirt ducati", and it lead right back here:
http://ducatimonsterforum.org/index.php?topic=25962.0
Funny. 

I'm still plowing through the Microsquirt manual and info on the web.

The vocabulary is very EE, and I'm an ME, so it's slow going.

[Langanobob]

I googled "microsquirt ducati", and it lead right back here:
http://ducatimonsterforum.org/index.php?topic=25962.0
Funny. 

I'm still plowing through the Microsquirt manual and info on the web.

The vocabulary is very EE, and I'm an ME, so it's slow going.

Speeddog, SuzyJ,

Lots of ECU problem threads lately made me think of the Great Ducati Microsquirt Project.  Just wondering if you've come to any decision?   

[Speeddog]

Speeddog, SuzyJ,

Lots of ECU problem threads lately made me think of the Great Ducati Microsquirt Project.  Just wondering if you've come to any decision?   

I've done some poking around on the Microsquirt forum, and there's at least one person that's got a Microsquirt on a Duc.

So there's some groundwork done, and a bit of the learning curve may be trimmed.

[ozs4]

I suggest that you don't rule out a wiring fault.  I experienced a very similar issue.  I ended up being a partially severed wire from the starter switch in the loom where it passes the headstock.  It was just making a contact enough to pull in the solenoid intermittently.