i'm trying to help a friend with a non starting 1990 K100.
the bike has been converted to a "Café racer" and some changes to electrics were made (not by me), however it did work for a little while, but always had difficulties starting, some time ago it's wouldn't start anymore.
i'm trying to work with the troubleshooting guide but have some basic questions first:
1. There are places in the guide that have Volts~ symbol - does this mean i need to setup the multimeter to AC volts? - i think its mostly on the ignition module and jetronic module plugs.
2. from what i could see so far it seems the injectors are not "Pulsing", i hear them "click" once when pushing the start button and the squirt a single dose of fuel (fuel pump can be heard working so i believe the relay and fuel pump are ok - pump is new) 
3. there's a spark on all 4 plugs - does this indicate the hall sensor is ok?
4. looks like the injectors are getting a constant +12V on startup - again i understand the pulsing is on the negative side.

any additional advise will be greatly appreciated

Eran

You're right, earth to the injectors should be pulsing, not constant. Only one click sounds like you have constant earth which would cause flooding. If you still get a click with the fuel injection control unit unplugged, that might point to a short with the earth wires to the injectors. If not, the fuel injection control unit might be the cause of the fault.

If a guide only says Volts, you can assume it's DC Volts.

Thanks, 
what i meant is that on this table for instance (in the "type" column there are both "Volts" and "Volts ~")
So does the "Volts~" refers to AC or it just means approximate value?

That would be approximate Volts DC.

The earth to coil tests should be 9, 12 and 14. Twelve only applies to the k75 which has three coils, so you only have to test 9 and 14.

It's not Volts~ really, but the meter will go nuts if you try on DC.
The injectors get a straight 12 Volt DC, from there the wire goes to the computer switching the power on and off very fast.
Everyone has a digital meter, cheap as hell mostly.
Better would be an analog meter, or osciloscope.

My suggestion is to follow the article with a digital meter

Arlina wrote:It's not Volts~ really, but the meter will go nuts if you try on DC.
The injectors get a straight 12 Volt DC, from there the wire goes to the computer switching the power on and off very fast.
Everyone has a digital meter, cheap as hell mostly.
Better would be an analog meter, or osciloscope.

My suggestion is to follow the article with a digital meter

I Have a digital multimetr, so your saying to set it for AC voltage measuring for those places in the guide?

Yes, wherever it reads "Volts~", use the AC volts setting on your multimeter.  The meter will measure and display  the RMS value of the time-varying signal at those test points, thus the chart has a possible range published instead of a single value.

ok, so i made sure all contacts are cleaned with contact cleaner (ignition unit plug, EFI Plug, 4 wire connector under tank) and started to take some measurements
Some numbers seems to match but there are few exceptions as you'll notice.
1. may not be a thing - the idle switch only shows voltage both on pin 2 & 3 only when the starter button is pressed (the table doesn't say the starter need to be running for this test)
2. pin 4 gives 10.80V and not 12
3. the injectors gives different resistance values then listed
4. Resistance between pin 13 and ground is higher than 0, it gave me varying values between 3.2-1.6 ohms
i tested one injector with a led, it flashes once or twice when i press the starter and once again when i let go of the button but not constantly while the starter is working
any ides where to go next?

Sounds like your injectors may be gummed up.  The flash is probably back EMF that gets generated when the FI relay applies power to the injector coil, and then releases it when you stop pressing the start button.  Are you sure of the injector conditions?  Can you test them on the bench (or bike) off the harness?

Pull the harness on one injector; 
temporarily hook up some alligator jumper leads to the injector; 
test the coil resistance. it should be 16 ohms or close to that.
apply 12V from the battery directly to the leads and listen for clicking in the injector using a stethoscope (or screwdriver resting against the injector to the ear)
no click with the battery means injector is stuck

Repeat for all injectors

robmack wrote:Sounds like your injectors may be gummed up.  The flash is probably back EMF that gets generated when the FI relay applies power to the injector coil, and then releases it when you stop pressing the start button.  Are you sure of the injector conditions?  Can you test them on the bench (or bike) off the harness?

Pull the harness on one injector; 
temporarily hook up some alligator jumper leads to the injector; 
test the coil resistance. it should be 16 ohms or close to that.
apply 12V from the battery directly to the leads and listen for clicking in the injector using a stethoscope (or screwdriver resting against the injector to the ear)
no click with the battery means injector is stuck

Repeat for all injectors

Thanks robmack, will check this.
as for all the other issues i mentioned, you don't think they can point to a problem?

Update:
so today i continued to work on the bike.
tested continuity between each injector connector and pin 12 of the efi plug - all good
so i decided to check the pulse once again with the 12V Led, and lo and behold...i got pulses on on all connectors when pressing the starter button - a good sign.
i also connected the injectors and could feel they were "ticking" when pressing the starter button.
at this point i decided to put everything back together, so i re-installed the injectors, the fuel rail and the fuel tank...when i tried to start the bike it just cranked but didn't start
i disconnected the connector from one injector again it was not pulsing the led, just a single blink when pressing the starter button, very frustrating!
I guess there's a bad contact somewhere in the wiring under the tank (i did disconnect the fuel pump connector to see if that's the problem but it had no affect)
hope it wont be to difficult to track this bad contact, this sort of gremlin can be very sneaky to track down.
is there a way to have the tank connected but not positioned on the bike so i can try to debug and see if it start while being able to check the spaghettis bundle under the tank?

Check the 4-position plug that joins the injector harness to the main harness.  It's situated on the left side of the bike near the top of the air box.  That might have come loose.

4 pin plug seems ok, cleaned it again anyway just to make sure.
i turned my attention to the hall sensors as everything else seems more or less ok.
could it be possible for Both of them to fail? doing the test according to the troubleshooting guide the led does not turn on on either of them?

This is how i tested the sensors, the sensor plate was hanging on its wire, outside the metal "CUP" i suppose like this both sensors should have the led "ON"
of course i tried both pins with led negative wire

I'm thinking the hall sensors are good cause you have spark at the spark plugs. 

Your test light will remain off with the hall sensor test if the sensors are blocked. If the crankshaft turns until the opening is in line with the sensor, then the test light will glow.

If you read robmacks last post again, I don't think he was referring to the four pin tank connector plug.

i took off the sensor plate when testing, it's still connected to the wires but none of the sensors is blocked.
the bike owner said it has spark, i haven't verified this myself.

btw, how hard is it to replace the sensors? i understand the sensors themselves are not expensive

I'm a tight ass, so I don't replace bits until I know they're a problem. I'd check the plugs for sparks. The hall sensors are easy easy to replace though.

eran_k wrote:This is how i tested the sensors, the sensor plate was hanging on its wire, outside the metal "CUP". . .
i took off the sensor plate when testing . . .

The "CUP" is called the rotor. Explain exactly how you "tested" the sensors after disassembling them. Your method doesn't seem to conform to the one in the troubleshooting guide.

i took of the cover screws, then took off the two small Allen screws that hold the sensor plate in place, so i was able to pull the plate away from the rotor while still connected to the wires. then i tested on the connector as the picture i posted shows.

later on i also installed it back and made sure the opening in the rotor is located IN front of a sensor but still none of the contacts lit the led.
i then went and tested for spark, no. 1 and 4 - no spark, no 2 and 3 had spark but it look week (blue and small)

eran_k wrote:i took of the cover screws, then took off the two small Allen screws that hold the sensor plate in place, so i was able to pull the plate away from the rotor while still connected to the wires. then i tested on the connector as the picture i posted shows.

The picture shows the plug test sockets but only the text explains how to test it. It doesn't need disassembly to be tested. Where did you put the feeler gauge when the parts were dangling?

i didn't get to the feeler gauge test as touching either of the contacts (2 & 5) with the (-) didn't light up the led.
to my understanding as far as nothing obstructs the air gap of the sensor the led should have been working. since the led didn't light up i assumed the sensors are defective

eran_k wrote:. . . since the led didn't light up i assumed the sensors are defective . . .

How have you determined that the plug isn't defective?

Swap the position of the LED bulb terminals, then test again. The attached diagram is from the BMW testing manual. You'll note the differences in plug polarity between the diagram in the manual and the one in the troubleshooting guide. Swap the positive and negative leads at the plug and do it again.

I will repeat, are you suggesting there's an error in the guide?
can i test he the way i did?  (plate off the rotor)
the original problem (no negative pulse to injectors) plus the no spark on two of 4 cylinders do sound like classic symptoms of bad hall sensor aren't they?

Something doesn't look right, looking at the connector connected (see image)
i seems the red (+) and black (-) do go to where i first tried - and according to the troubleshooting guide.
so what am i missing here?

I don't think you are missing anything. According to your posted photos and the troubleshooting guide instructions, + and - are the power pins which supply power to the Hall Sensors, and the negative lead of your test light goes to the Hall Sensor output. If the Hall Sensor is unobstructed (as it would have been if you removed the plate and let it dangle, then the LED should have been lit. If you introduce a metal feeler gauge in the slot of the hall sensor, the led would extinguish. Since you never got to the stage of trying a feeler gauge, you can't say anything definitive about the state of the hall sensor but it's not looking too good. Your diagram shows testing one hall sensor; did you test the other? You said 1&4 has no spark and 2&3 have weak spark. Maybe the one you tested was for 1&4.

Yes i tested both pins and the led didn't light on both

If you're getting spark to two plugs, and fuel, but no light with the hall sensor tests, I think your test is faulty. 

Previously I did a hall sensor like you have, but with a computer safe test light, and the light came on. Did your wires make solid contact, were they insulated against each other, and how many Amps is your led test light?

After the tests and a sticky beak, I've noted on my wiring diagram that the orange wire (that you show in your test photo) is between the top sensor and coil 1 (for cylinder 1 and 4) the brown wire is from the bottom sensor to coil 2 and to pin 8 of the ignition control unit which goes to the fuel injection control unit.

Well, i think i found the problem, and as always it was a simple thing.
i added a ground wire directly from the battery to the frame ground under the tank and everything seems to work now, although i still think it will be a good thing to rpelace the hall sensors as the tachometer is a bit erratic.

eran_k wrote: i still think it will be a good thing to rpelace the hall sensors as the tachometer is a bit erratic.

I don't understand your reasoning. You already have proven the system has faulty ground connections. Perhaps checking and cleaning the tach's connection at the instrument plug, the coil and the ignition control module for oxidation and fit would be effective.

Adding a dedicated ground wire makes me wonder if you have you cleaned the ground connections on the frame tube and the side of the transmission.  Clean grounds are critical to making the electrical system operate reliably.  

Wouldn't hurt to pull the starter motor to clean the commutator and check the brushes.  That is another important part of the goofy electrical system on these bikes.

Laitch wrote:

eran_k wrote: i still think it will be a good thing to rpelace the hall sensors as the tachometer is a bit erratic.

I don't understand your reasoning. You already have proven the system has faulty ground connections. Perhaps checking and cleaning the tach's connection at the instrument plug, the coil and the ignition control module for oxidation and fit would be effective.

Well, the new hall sensors were already ordered by my friend+ the gasket which was missing so i might as well install them anyway

If you don't install them, you'll never need them. 

I would as so check the resistance of the water temp sensor. When I rewired mine to motogadget it too would not start turned out one side was toast. Replaced it and all was good