Hi,

I'm having cold starting trouble with temperature in the low 40sF or around 5C or so. My normal method of using the "choke" half or full open no longer works. The only way I can start it now is with no choke and holding the starter button in for 5 or 6 seconds or more. I usually just bump it once at the beginning to get fuel pressure up. I arrived at this with multiple other variations. Clutch lever pulled in to reduce drag. It will then start at low RPM, maybe 5 or 600 or so. I then can add the choke a little while holding the start button in to enrich, tweak the throttle slightly to raise the RPM and it will idle faster. Any attempt at this point to go full choke kills it.

A year ago I was unable to start my bike and went through the whole trouble shooting guide of Frog's and eventually changed the fuel pump. Problem was solved then and it ran and started well. I have put very few miles on the bike since then. This was my thread on that past problem, https://www.k100-forum.com/t13459-another-hard-starting-problem

I just completed switching out my K100 2-valve TBs to K1100. Once I get the bike started now, has noticeable more and smooth power.

Status now:

- K1100 from low mileage bike. All cleaned out with new needle valve O-rings. All blue paint was intact and looking through closed butterflys showed all closed with throttle off. Smooth operation of TBs through the stroke
- Fuel and air filters all good
- new fuel pump
- fresh rebuilt fuel injectors
- All new rubber boots and rear vent hose with the TB change. I did have a bad crack in one boot on cylinder #4 that I could never balance well
- vacuum hose to K100 fuel pressure regulator good
- I have manifolded all the vacuum lines together with two taps, one between 2 and 3 for my cruise control and then on the end at #1 for the K100 fuel pressure regulator
- synchronized the TBs three times, balancing was easy but did it to tweak the idle RPM up
- idle RPM once warmed up is about 900
- TPS (K100) is set so just clicks with throttle just starting to open. disconnecting it made no difference in starting.
- fresh 89 octane ethanol free gasoline
- battery new
- Once warmed up it starts immediately with a touch on the start button

I also checked the AFM setting this way and got the "ideal" situation per below:

Air-flow meter vane check.

To set it correctly - engine idling, press the starter button.
That causes a 10% enrichment of the mixture. Listen to
the RPM:
- if RPM increase and stay increased - the mixture is lean.
Adjust screw in.
- if RPM decrease and stay decreased - the mixture is rich.
Adjust screw out.
- Ideally - the RPM should increase about 100RPM, then
drop down to the same speed or lower as it started at


Looking for suggestions please.

Thanks,
Glenn

Sounds like you might have a problem with reading the coolant temperature.  I would check the connection to the coolant sensor behind the radiator, if that is good, then see what the resistance is.  Should be about 2500 ohms when the engine is cold. 

If that sensor is bad the mixture will be lean on start up.

Did you disturb the TPS?

I find that doing one throttle blip ahead of time helps when starting.

Despite the factory spec of 950 +/- 50, I set the idle on my Ks at about 1,100 when warm.  Not sure how that impacts starting but the bike gets off the line quicker that way.

Point-Seven-five wrote:Sounds like you might have a problem with reading the coolant temperature.  I would check the connection to the coolant sensor behind the radiator, if that is good, then see what the resistance is.  Should be about 2500 ohms when the engine is cold. 

If that sensor is bad the mixture will be lean on start up.

Well this may be it. i just measured the resistance and it is about 75k, way too much. I'm going to pull the lower right fairing now and see if it connected first.

What I'm not understanding though is if it is a very high resistance, the mix should be richened but I have to hold the starter button in to make it rich that way?


Thanks!
Glenn

If the coolant temp sensor is disconnected it will flood and piss gas all over the ground.  At least that's what my K100 did when the connector fell off. (I think a prior owner lost the clip because it was missing when I went in to investigate and found that the connector had fallen off.)

Well it is connected so it must be bad. Fan comes on when hot so never suspected this. Too much trouble now to replace it ....

Thanks guys

Per Frog's guide:

"The water temperature sensor contains 2 resistors: 1 feeding the EFI and one feeding the temperature switching unit. One resistor may be defective when the other one may be OK. If only the one feeding the EFI is bad, you won’t see the temperature light glowing. If the water temperature light stays on, and the engine doesn’t start, this is a good indication of a bad temperature sensor."

My light doesn't stay on so I must have a defective resistor to the EFI I guess.

glennpm wrote:Smooth operation of TBs through the stroke

Does through the stroke mean twist of the throttle or sweep of the choke lever or something else entirely?

Actually both. I checked this to make sure that there was no sticking with for example, a sticky throttle plate. The unit was advertised and looked like a low mileage one. I completely cleaned everything. Action installed with the throttle and the choke are smooth.

Have you checked the temperature sensor resistance through the pins on the EFI connector?  That would check the cable as well.

Yes, resistance between pins #13 (ground) and #10. That is how I measured the very high resistance of 75,000 ohms. I was thinking maybe the plug wasn't attached but it is and wire casing looks fine.

Have you checked this dimension?

Hi Laitch,

You can't measure this gap on the K1100 conversion. I have the choke cable set by RPM which I did when thoroughly warm.

just a note
something I found was the thread to the temperature sender sensor was badly corroded  where it bolts into the water stub

I removed and cleaned the threads on both parts

after doing this I got the most economical fuel mileage that the bike ever had , believe it or not

I also installed a copper washer (tinned) as the sealing washer between the sensor and the water stub instead of the corroded aluminium washer , which allowed me to take an earth wire directly to the under tank earth point

hope this helps

Thanks charlie,

That makes a lot of sense and could be a factor in my high resistance reading. Right now I'm not up for pulling it. Too much trouble to drain, pull air box out, etc and once I get it going, it is fine. However, I'm not good on leaving things alone!! I'll probably buy a sensor to have on hand and go with the Cu washer sometime.

glennpm wrote:You can't measure this gap on the K1100 conversion. I have the choke cable set by RPM which I did when thoroughly warm.

This instruction from the BMW K1100 service manual seems to imply that the setting is established by position of the cable and adjustment components. RPM isn't mention. I wonder if the position of Screw (2) was changed by the previous owner. Certainly, another K1100 owner will chime in here again.

Hi Laitch,

Any position of the choke other than off, and I can't get it started. The choke cable is set so that off has no tension on the cable. The choke is just ineffective until I start it as mentioned above, no choke, no throttle and enrichment via holding the starter button in. I set via RPM that I know is in the range for the k100 in middle and full on position. Its not a precise thing, just an aid to increase starting speed, aka RPM.


Thanks,
Glenn

glennpm wrote:Any position of the choke other than off, and I can't get it started. The choke cable is set so that off has no tension on the cable. . . . Its not a precise thing, just an aid to increase starting speed, aka RPM.

I've read your history of this moto's quirks from February of last year to the present. The only items not under investigation in the interim seem to have been the coolant temperature sensor and the increased starting speed assembly's linkage.

The increased starting speed function might not be precise, but on your moto is isn't operating at all and to me that bears investigation. If the sensor itself is faulty, maybe after it is replaced the increased starting speed assembly will work correctly as you have presumed to adjust it. Please let us know.

Hi Laitch,

Re your reply, "The increased starting speed function might not be precise, but on your moto is isn't operating at all and to me that bears investigation"


My choke lever/cable does work, however any throttle opening during a cold start with a throttle twist or use of the choke lever, prevents starting.

glennpm wrote:My choke lever/cable does work, however any throttle opening during a cold start with a throttle twist or use of the choke lever, prevents starting.

What I intended to convey was that your moto's cold start, throttle increase assembly doesn't work as intended—not at all—and that bears investigation.

You're fortunate that the moto apparently does run well once started. All it will take is the inclination on your part to continue the investigation now that causes have been reduced to such a small number.

I was getting some large fluctuations in one of my cheap VOM meters yesterday so measured the coolant sensor's resistance again this morning with a better meter. My resistance is 3.78k ohms at 9 °C so not that far off the curve. Curve at this temperature is 4.3k.

Looks like it may not be this sensor after all. I have a replacement on order anyway.

Glenn