Hi everyone!
   This is my first post other than my intro, so forgive me (and correct me) if I mess up any etiquette.

   My wife gave me a non-running 1985 K100rs for my birthday a couple weeks ago. I've been going down the list fixing the countless things wrong with it. I'm pretty good on mechanical stuff, but the electrics have me stumped.

   My current big problem is the fuel pump. As I got the bike, the fuel pump was doing nothing. I measured 12 volts at the electrical connections directly at the pump, but it didn't spin at all. I figured that pointed definitively to a dead pump, so I ordered a new one. That arrived today and I installed it. The first time I tried starting the bike, the pump spun intermittently a few times. As it did the engine started to catch, which I take as a good sign. Then the pump stopped altogether. No amount of starter cranking would get the pump to move again at all.

   I've been using this guide (http://hrsbstaff.ednet.ns.ca/bvogel/K100/download/bike-wont-start1.htm) to try to figure out what's wrong. Here's what I've found:
   The fuse is fine.
   There is power from the ignition to pin 86 of the FI relay.
   The relay itself seems to be dead. I tried swapping it with the headlight relay, which made the headlight stop working.
   There is no connection (infinite resistance) between pin 7 of the ECU connection and pin 85 of the FI relay. Actually, pin 7 at the ECU side (testing on the wire, not the ECU pin itself) seems to connect to ground.

    This makes me suspect a break in the wire that's shorting to ground. Does that seem reasonable? Could that be what killed the relay? I'd try rewiring that connection (ECU 7 to relay 85) but I'm terrified of frying something expensive. Also, I noticed that what should be a single wire looks (maybe) different at each end. At the ECU end, it looks like it's yellow with a brown stripe. At the relay end, it looks like it's brown with a yellow stripe. I might have that reversed, and I also might just be mistaken. Brown/yellow looks a lot like yellow/brown. Most of the wire, of course, is inside the harness. But as far as I can see the harness seems intact. I don't see anywhere where the wire could be broken.

    Any thoughts? Am I on the right track? Is there something else I should check to figure out a diagnosis? Should I just try connecting ECU 7 to relay 85 directly? Will that hurt anything to try?

   Oh! It also might be relevant to mention that I just replaced the hall sensor. I think I did so correctly, but I certainly could have messed it up. If I did mess that up, could it create the symptoms I've described?

    Thanks so much for your help!
    -Jon

Bikesmith wrote:Then the pump stopped altogether. No amount of starter cranking would get the pump to move again at all.

Do you still get power(12v) at the pump when cranking the engine?
Before looking at a very difficult problem, check the power at the pump and let us know what you find.
Your new pump may be defective and have died after a short time.
If you remove the pump from the tank and apply (for a very short time) 12V, does it turn?
You cannot run these pumps dry as they will overheat but a one second run without any fluid won't arm them.

I'll check tomorrow to confirm the pump is functional (Bike is at work, I'm at home).

 But that wouldn't explain the relay issues, would it? I'm pretty sure the relay is dead. If it were good it would work in the headlight relay socket, right? Plus there's the lack of connection to the ECU. Am I wrong in thinking there should be connectivity through the harness between ECU pin 7 and Relay pin 85?

If you isolate the Yellow/brown wire at both ends and replace it with a temp connection you should not harm anything and try it with a good known relay.
That y/b wire puts an earth to the Fuel injection relay for 1.5 seconds when the starter button is pressed and continues to have the earth when the engine is turning and the ignition control unit is recieving pulses from the hall sensors.  If the relay was faulty or the ICU was not extending that earth then the PO may have deliberately put a permanent earth on the y/b wire to make things work which would tie in with what was said about the hall sensors.

Jon, you mention that the wire from the ECU(actually it is the Ignition control unit ICU) pin 7 reads to ground.  Did you measure this with the relay removed.  If the relay was in when you took the measurement you may be seeing something on the other side of the relay coil.

I would suggest that you check continuity of the wire from pin 7 of the ICU connector to the socket of the fuel pump relay.  Then I would check to see that there is no continuity to ground at each end of the wire.  All these tests should be done with the connector and the relay unplugged.  This will confirm if the wire is good or bad.

Okay, here's the update:

Crazy Frog, I checked the pump. I connected 12V directly to it inside the tank so it wouldn't have to run dry. It ran fine and pumped out gas into my official BMW Gas Receptacle Bucket. So I've also learned that the filter is passing gas and that I hate working on things inside a gas tank.

RicK G and Point-seven-five, I'm checking the brown/yellow wire at each end, from the ICU connector end (disconnected from ICU, of course) to the relay socket end (with relay removed). There's no continuity from one end of the wire to the other. There is continuity from the ICU end to ground, but no continuity from the relay end to ground.

Sounds like that wire is fubared, right? Should I snip it just outside of the ICU connector and relay socket and splice in a new section of wire?

Also, just as a quick idiot check, can you guys confirm that this really is pin 7?

Yes that is pin 7 just check if it has a y/b wire attached.
There is an earth point on the left side of the upper frame tube not very far back from where you are working, have a look to see if there is a y/b wire on that. They should all be brown maybe some black for after market items.
I would be disconnecting the y/b wire at both ends and making a temp replacement and try to start the bike.
Many people are very quick to blame the ICU or Hall sensors and maybe they put the earth on that wire and on the wrong part and when the bike still didnt start just pushed it in a corner. I have seen it many times.

Bikesmith wrote:.......
RicK G and Point-seven-five, I'm checking the brown/yellow wire at each end, from the ICU connector end (disconnected from ICU, of course) to the relay socket end (with relay removed). There's no continuity from one end of the wire to the other. There is continuity from the ICU end to ground, but no continuity from the relay end to ground.

Sounds like that wire is fubared, right? Should I snip it just outside of the ICU connector and relay socket and splice in a new section of wire?

JUST HOLD UP THERE FOR A TICK. Before you go snipping any wires!!!

It looks to me as tho you may have incorrectly identified the FI relay. This is from your initial post #1 above...

Bikesmith wrote:......
The relay itself seems to be dead. I tried swapping it with the headlight relay, which made the headlight stop working.
......

First of all, there is no headlight relay as standard on a K100. That's gotta be the Load Shedding relay you've swapped.
Second, the FI relay has a different connection to the other 2 relays at the front of the electrical box so I doubt thats the FI relay you are swapping out. I'd say you've been probing the horn relay connections as that's the only other relay in there.
Someone correct me if I'm wrong on this please.
Refer to the diagram below and make sure you have the correct relay and run your tests again. The FI relay is in the front/left of the electrical box.

I would expect all 3 relay bases to have provision for the 87a connection but the load shed and horn relays dont have an 87a terminal. If the horn or load shed relay is inserted in the FI base then the fuel pump wont run but the EFI will be functional but for fuel pressure.
It may be that the incorrect relay has been used but that would not stop the relay operating when the earth is extended from the ICU.
The reason the Load shed relay could be mistaken for a light relay is that the headlight is always on as it is a US model without the right hand switch and the headlight would be what is noticed as not working when the relay is removed.
That is the way I read it

Just trying to prevent possible/needless damage to the wiring by "snipping" that wire.
Rick, correct. I can see what you mean about the relay bases. But something does not add up in Bikesmith's description of the wiring and results he's getting and I think it's worth investigating further. For example, from post #1....

"Also, I noticed that what should be a single wire looks (maybe) different at each end. At the ECU end, it looks like it's yellow with a brown stripe. At the relay end, it looks like it's brown with a yellow stripe."

The yellow/brown is the return wire from pin85 on the FI relay to pin7 at the ICU where it is also yellow/brown. Bikesmith says the wire at the relay looks like brown/yellow. This happens to be the colour of the wire at pin85 on the horn relay in the front/right of the box.

He says he refered to the info in the troubleshooting page. If you check the other smaller diagram of the electrical box layout there you might see there is a discrepency in the description imo.
This diagram puts the FI relay in the front/right of the box. Just my interptretation. keep in mind that in this diagram you are viewing the box obliquely from the front.

Update:
   Did I mention that I'm an idiot when it comes to electronics?
   Kaptain Holister, you were exactly correct. You spotted all the clues that I didn't. That diagram of the electrical box that you mentioned shows the fuel pump relay as being all the way to the right. I didn't even question it. You mentioned the yellow/brown versus brown/yellow discrepancy. Totally right. (Also, that diagram has the middle relay labeled as headlight. That's what threw me there.)
   That also explains why I was getting continuity from ICU pin 7 to ground. I thought I had removed the fuel pump relay, but I hadn't. So I was reading ground through the relay.
   You're all wondering if I caught the mistake before snipping a wire. Nope. I rewired ICU 7 to horn pin 85. After doing so, I was getting 1.5 seconds of ground on pin 85 of the relay socket. Just the wrong socket. It was when I noticed I didn't have power to 86 or 30 that I looked closer. I saw pins 86 and 30 wired together with green/brown wires that then went into the temperature switch. I checked the wiring diagram and thought, "Why, that looks just like how the horn relay is supposed to be wired." Then it clicked.
    So I put the wiring back how it had been (with two more butt splice connectors than there used to be). Then started testing the actual fuel pump relay. It works fine. Though I did notice the power output seems more like 10-11 volts than a full 12. Could be my $5 voltmeter.
  
    So I have a relay that's putting out power, and I have a fuel pump that works when I put power straight to it. Narrows the problem down nicely, doesn't it? So now I think the problem is this:

It looks like the previous owner got creative. The connector does connect to the bottom of the tank, but it's not correct, is it? Is it a re-purposed anti-theft connector?
   Regardless, I suspect if I redo all those soldered and taped connections with splice connectors and heat shrink tubing, then I should be good to go. But I don't want to assume that the PO mapped those wires correctly (especially since the low-fuel light hasn't been working either).
   So I was wondering if one of you fine folk could tell me which of the original harness wires (brown, brown/black, brown/white, green/white) was supposed to go to which pin on the bottom of the tank?

   Thank you so much for all your help and support in this. I'm so excited to ride this bike for the first time. It feels like I might be close.

Don't sell yourself short. Sounds like you are doing a good job.

Just make sure you have the FI relay in the FI relay socket. It's the odd one out of those 3 at the front (see Blakey's photo above) and has an additional connection.

Cheers

I took the photo away because the relays were identified incorrectly.

I didn't want to add to the confusion.

I thought it was accurate blaKey but may have been mistaken. Cheers

The bike is running!

   I finally found the real problem. After rewiring the connector at the tank, I found that I had power at the pump, but only about 8 volts. Actually it was varying voltage. Usually 10ish volts, but sometimes as low as 8 and sometimes a full 12. I expected it to be a huge deal to track down the voltage drop, but it was actually really easy. I found that the load line into the relay socket had a strong 12 volts, but that the output was variable. I asked my dad (an electrician) if that made sense. I always thought relays worked or they didn't. He told me that was completely typical of an old relay. The points in them corrode and create a voltage drop. Also, he said that since the points contact in a slightly different way each time, that explains the variability.
   So I ordered a new relay. Then, because any time I have a dead part I am compelled to disassemble and study it, I started to cut open the old relay. I noticed that it can be opened non-destructively, so I did, saw the clearly corroded points, cleaned them, reinstalled the relay. Perfection! The bike started instantly. Sounds good, too.
   It's funny that all that work I did was for nothing. If I'd known what to do I could have had the bike running five minutes after I got it (maybe an hour). It's not wasted time, though. I learned a ton about the electrical system of the bike, which is what's unfamiliar to me. Thank you all for your help with this issue. On to the next one!

Great work Jon!

Agreed, it's definitely not wasted time when you learn so much about these bikes from working on them!

Keep the questions coming.

great result

cheers

I love a story with a happy ending!  

That's great news, Jon!  It's never a bad thing to be mucking about in the wiring of your brick as long as you do it gently with a little love and leave a trail of clean connections.  The knowledge and confidence you got from this experience will stand you in good stead in the future.