12v+ on pin 86 doesn't mean much if you don't have a connection to ground on pin 85.  Pin 85 should be receiving a ground connection from the ECU.  Both have to be present at the same time for the starter relay to work.

Erratic starter COULD mean brushes or worn starter. This can be a completely random fault depending on where the starter comes to rest after it has been used.

But, erratic could be caused by low battery, either bad battery or underlying non charging issue.

Kioolt, I've got continuity from pin 86 to the gearbox.  If that's not the right test, please let me know what I need to do to test instead.

Olaf, I have a new battery, about 3 weeks old, and it shows 12.8v.  I'm confident that I'm charging.  Can the starter brushes be seen/evaluated without disassembly?  My bike does have 87k miles on it.  I don't know if the brushes have ever been changed.  If not, should I assume its time?

I've never read the troubleshooting guide that you referred to but I would just put my meter on pin 85 and pin 86, push the start button and see what voltage I have at that point.  Also,  an easy way to check the starter is to put the meter on the starter lead at the starter, press the start button and see what voltage you have.  If you've got 12 volts there and all of your main grounds are ok then your starter is the problem.

I've got 12v across pins 85 and 86 when I hit the starter button.  I also have 12v at the starter lead (red wire to starter) both with the ignition in the the off position, in the on position and when I hit the starter button.  I also put my old relay back in, just to check if my new relay is the problem.  My old relay behaved exactly like my new one.  So, all of that would point to a starter problem.

However, what I don't understand is why my starter turns when I bridge pins 30 and 87 at the starter relay with a wire?  Could this mean that I've got a bad ground somewhere?  Or some other wiring problem elsewhere?

Assuming that the starter is the problem, what's the best fix?  A whole new starter?  New brushes only?  Or do I have to take it out first and see how it looks inside to make a call?

What needs doing to the starter very much relies on what you find when it's dismantled. The most common thing is brushes but these things are getting old and time is the big factor now and not miles covered. Expect almost anything.

OK, this is tremendously embarrassing, but I just realized that I wasn't taking my reading at the starter.  I was taking it at the alternator. I can't believe that I was doing such an idiotic thing!

I just tried testing at the black wire leading to the starter.  That is, the wire running from pin 87 to the starter.  I got no power at the starter when hitting the starter button.  However, as mentioned above, I can engage the starter if I bridge pin 30 (feeding power into the relay) across to pin 87.  So, this must mean that, for some reason, I am not getting power through the starter relay to engage the starter.  Is that right?  Assuming so, where do I look for the solution?

Probably also should mention that I have another issue that might be related. I'm running Robmac's neutral enable start board, v.2.  I've been having some issues with it and he has a new board on the way to me....should arrive this coming week.  First the neutral light stopped lighting up, though I could still start with the clutch open.  Then I could only start with the clutch in.  And now no start either way.

Sure does look like your starter relay is the problem to me.  You said you had 12volt across 85 and 86 which should be the two small wires on the starter relay. If you've got the correct voltage across the starter relay coil and it doesn't close the contacts then it is bad.  Do you hear it click when you press the starter button?

KJustin wrote:OK, this is tremendously embarrassing, but I just realized that I wasn't taking my reading at the starter.  I was taking it at the alternator. I can't believe that I was doing such an idiotic thing!

Ha Ha we have all done those things, at least you admit to it.

Yes, I have 12v across 85 and 86.  And the new relay does click when I hit the starter button.  I also tried my old relay (which worked before, but had gotten stuck in the on position twice before, which is why I replaced it) and I had the same result...clicking but no engagement of the starter motor.  FWIW, I opened up the old relay and cleaned the contact with 400 grit sandpaper before trying it this most recent time.  

Guess its possible I have two bad starter relays.  Any other theories before I spring for another relay?

I would say it's probably a good idea to check out the starter because it may be the cause of the failing relays.
If the bearings are worn the armature can contact the magnets and this will cause a much bigger current drain than it should. If the armature is shorting to earth on one or more of it's segments it will have the same effect.

Rick, what you say makes some sense to me.  If I understand it correctly (and, as always, feel free to correct me), this might explain why my (old) starter relay stuck in the on position even when I had a brand new, fully charged, battery.  That is, I was having excessive current draw due to a defective starter, which caused a dip in voltage supplied and then caused the relay to stick in the closed position.  Perhaps this also explains why I can make it work by force feeding the starter by bridging pins 87 and 30 manually.

Anyway, I'll pull the starter and see how it looks.  Looking a the bike, it does appear to be old (i.e., original to he bike).  I suppose after 30 years it wouldn't be surprising to find that it's worn out.  

I'm short of time today and have a busy week coming up, so it might be next weekend until I can report back.  I'm not sure exactly what I'm looking for once I have it out, so I'm sure there will be more questions.

Thanks to both of you (Rick and kioolt) for your help with this.

I think Rick is spot on the money with that. I had a similar problem back in January. Went thru a couple of starter relays before I deduced some time after that it was probably my starter motor drawing too much current. I ordered a new Starter Repair Kit  but haven't had a chance to install that yet. I did pull my starter apart back then and gave it a clean out. Brushes were in spec. I put it back together. Not had any problems so far but the new kit will go in shortly.

But as for your relays... sounds like they're kaput. Try pulling the new set apart and if the contacts haven't been blasted away you might be able to clean them up to get you going or use as a spare.
Cheers

Thanks guys. I actually did pull the starter tonight and take it apart after all. It was a bit dirty inside, but didn't seem terrible. Cleaned it all with deoxit. Used a toothbrush and gently a scotchbrite pad to get the carbon off the copper armiture (I think that's the right term). Carbon was pretty caked on there. The brushes measured 0.35 and 0.43 inches. My clymer manual says new is 0.47, so they were in pretty good shape. But I will get replacement brushes installed soon, per Kaptain H's suggestion. I'm out of time for real today and I haven't had a chance to put the starter back in.


The question I have is whether there's any harm in trying to start with my existing relay or if I should get another before trying?

Also, is there any way to check how much current my starter is pulling? And, the related question of how much it should be pulling?

KJustin wrote: Any other theories before I spring for another relay?

have you confirmed that he earth wire to the gearbox is good .....a slight high resistance on this wire when the starter is trying to operate can cause a huge voltage drop to the starter whilst attempting to rotate

typically the start of turning requires huge amps whilst building to speed

also the ground connection voltage drop is hard to diagnose

many of us have replaced the widdle flimsy cable from the negative of the battery for something almost twice the size

just a suggestion

Justin, you will probably need more than brushes. Rick mentioned bearing replacement. There's a link in my post above for the Starter Repair Kit which contans brushes, bearing and o-rings. Its not all that expensive.

Charlie, I think a small car starter will draw about 100amps and around 3 or 4 times that while spinning up. Our Ks a little less maybe.

Another area I like to clean to make sure the starter is properly earthed is the bed where it's bolted down to the GB.

Good point.  I'll install the whole kit.  After cleaning things up and putting the starter back together, it appears to spin freely by hand (though I could hear and feel the brushes contacting the armiture) and I couldn't feel any play in the bearing.  I didn't try to clean the bearing...thought leaving the existing grease in there was probably best since I wasn't sure if I could effectively repack it.  Both before and after disassembly and clean-up, I noticed some slight fore and aft play/movement (but not side-to-side) on the gear sticking out of the front end of the starter.  Is this normal?

Justin, I'll second charlie99's notion. Ensure you've solid, immutable grounding(s). Starter relays go nutz when they don't have ALL the juice that they require. And shouldn't your initial voltage read higher than 12v? Mine reads closer to 14.

Will do another round of grounding clean-up, just to be sure.  Pretty sure that I've cleaned all grounds in the past few months or so.  But never hurts to check/do it again.

12v was just shorthand.  My battery usually reads 12.8, minimum, sometimes a little more.  And most places I test are close to the reading at the battery...though there is often a few tenths of voltage drop.  I assume that this is due to the headlight being on or the like.  I assume that's normal, but if not, let me know if I have some other issue to chase down.

Had a few minutes to work on this problem tonight.  Reinstalled my cleaned-up starter motor.  I also had my old (OEM) relay open, out of the protective box.  Since it was open, I put that in to see if it was actually making contact when the starter button was depressed.  Turns out that it is not.  The clicking I'm hearing when I hit the starter button is something else (l'm guessing the load shed relay, is that right?).  As a test, I pushed down the plunger and when it made contact, my starter fired right up.

I'm thinking that this means one of two things.  First, that the relay is somehow fried.  However, It doesn't show any signs of discoloration, arcing, etc.  I don't see any broken wires or anything that looks disconnected.  Can it be bad without looking bad?

Second, maybe I'm not feeding pins 85 and 86 with 12v (to ground).  But I've checked that and when I hit the starter button I get 12v there.  

Is there something else I'm missing?  All ideas for next steps welcome.

OK, performed one more test.  I manually grounded one of the start enable pins (can't recall if it was 85 or 86) and hit the starter button.  And, the starter fired!  So, I've diagnosed the problem as a bad ground.  

That means that I have to track down where the ground is failing.  That's going to have to wait for another day, but at least I know what the problem is.  I'm going to give the wiring diagram a hard look and trace the ground wire through the loom.  But if anyone has an idea where to look/how to test, I'd appreciate hearing it.

The starter relay gets it's earth from the ICU via the brown/red wire so the problem could be with the ICU or the earth to it. Check the plug is in firmly there has been a few lately that we have seen where the plug is not connecting properly.

Yeah, I'm one of those guys with a wonky ICU plug/connection. Caused all sorts of strange problems and massive frustration, detailed in a few recent threads. I cleaned it wit deoxit recently and that cured my other issues. Then this happened. I'll give it a more careful cleanup and check the ground from the ICU to the frame also.

Yeah so many with many problems. The external ground connection only grounds the metal case to block interference to the hall sensor signals. It is important to have it grounded properly but the grounds you need to check are the brown wire on pin #1 of the ICU which should as I read the circuit diagram go to the ground point under the tank. There is one other brown wire on pin #13 which is the ground going to the hall sensors.

For what its worth I have now figured once a year go take a sunny day and a six pack of beer, park the bike on the patio, take off the gas tank and some fairing panels and do all the electrical connections  including the ICU and the FICU under the seat. Particular attention to the ground/earth points, connections under gas tank, even take out bulbs and clean the bulb holders.

Unless you live in a very dry climate.

92KK 84WW Olaf wrote:For what its worth I have now figured once a year go take a sunny day and a six pack of beer, 

Don't you only get sunny days when the Aussies visit?  
You need the bike running then. 

Bill

Bill

That's why we like Aussies to visit! So we can take them up the Sally Gap in the sun.

In summer we get very long days of daylight......4am to 11pm so time to do both.

Finally got back to this after a busy week at work.  From my starter relay to pin 11 at the ICU I have continuity.  I actually attached a jumper at pin 11 on the plug and was able to get it to ground, got the starter to fire.  That much is good.

As I read the wiring diagram, pin 4 is the ground on the ICU.  I assumed that this went from the ICU to the main under tank ground point to the frame.  However, the wiring diagram on the portal page just shows it hanging out, not attached to anything.  I checked continuity to each wire that comes to the under tank grounding point and found no continuity.  So, either I have a break in the ground from ICU to grounding point, or that's the not where the ICU grounds.  I can't trace the wire from pin 4 without doing serious cutting up of my wiring harness.  I'm not willing to do that unless I am absolutely certain that this is the problem.  

Can anyone tell me where ICU pin 4 grounds for certain?  Also, what is the advisability of grounding my starter relay to the frame (if I can't figure out this ICU pin 4 issue) and not running it through the ICU?  Bad idea or no harm, no foul.  I don't like the idea of an ungrounded ICU....

Pin #4 is the shield wire for the hall sensor cable to prevent electrical interference. It doesnt actually ground.
Pin #1 (brown) is the ground wire which goes to the ground point under the tank.
There is one other brown wire which is on pin #13 which goes to the hall sensors and is the negative wire of the feed to the hall sensors.
The ICU alloy case is usually grounded but that wire is not the electrical ground for the internals that is on Pin #1

KJustin wrote:Can anyone tell me where ICU pin 4 grounds for certain?

As Rick pointed out, ICU Pin 4 is the electrical ground for the HES cable shield (to prevent RFI).  The ICU safety ground is  Pin 1 which connects to the ground point on the frame under the tank.

ICU Pin 11 is the starter relay "ground".  I'm using quotes because this signal acts in a peculiar way, which will lead to an answer your second question.  This Brown/Red wire provides a path to electrical ground through circuitry in the ICU.  If the start button is pressed, the ICU senses this action and also reads the engine RPM at the same time.  If the RPM is less than 700, the ICU connects Pin 11 to Pin 1 (and the start relay engages).  When the engine fires up, the RPMs increase above 700 and the ICU internally disconnects Pin 11 from Pin 1.  The starter relay drops out when this happens and the starter stops spinning.  

This is a safety feature.  If the operator presses the start button while the engine is running, the starter will not engage because the ICU has broken the ground continuity internally.  Once the RPMs drop below 700, only then will the ICU reconnect Pin 11 to Pin 1.

KJustin wrote: Also, what is the advisability of grounding my starter relay to the frame (if I can't figure out this ICU pin 4 issue) and not running it through the ICU?  Bad idea or no harm, no foul.  I don't like the idea of an ungrounded ICU....

If you connect the starter relay directly to frame ground, you will defeat the safety feature. The starter will begin to work.  However, remember  if you press the start button with the engine running, the starter motor will kick in.  Who knows what damage may or may not result.

Thanks guys. Immensely helpful. I tested and have continuity from pin 1 to the frame ground. Also with the ICU plugged in I have continuity from the brown/red wire at the relay to the frame ground. But the reading on my multimeter is .13, whereas on all other readings on all other wires/test points, it comes up as 0.0. Not sure what to make of this. Bike starts when grounded directly to the frame (defeating the safety feature of the start button) but not through the ICU. Any chance this means a bad ICU? I've cleaned all grounds and plugs. Any other ideas before I just wire it to run without the starter disable safety?

Also, I'm thinking again about the starter motor...given the above and that the brushes had lots of material and I already did a good clean in there, is there any chance that replacing brushes, bearing, etc., would solve my problem?

hmm .13 ohms could be the silicon junction  ( read switching device ) to ground ...depending on which way round you had the leads of the multi meter

the black lead on a digital is the usually the more + lead

have you tried reversing the leads and measuring the exact same points again ?

I will be going through all of these suggestions and steps when I can. Suspect elecctrical but will follow Burt's flow chart.

Have been stumped by non start and failing to restart.

If the starter does not keep spinning when pressed what does this tell me? Voltage drop I guess and not enough to turn the starter motor. I have been through the basics and have cleaned the starter motor.

Q: what grease on the gearbox ground lead? ... not dielectric I assume because it's non conductive.

I have an intermittent (but new ... 7 years ago ... very little use) fuel pump. I have pulled the relay but apart from replacing it, is there an test I can do?

good luck with your bike and keep posting.

Guy

Charlie, it's .13 no matter which way I attach the leads. I assume the silicon junction/reed switching device is the device inside the ICU that controls the break between pins 1 and 11 as described by to mac, above? If that's my problem I think it's time for a new ICU. Cracking that open to fix its internals is above my skill set.

Guy, Berts flowchart is great. Definitely follow that first. Sounds like you are already doing this, but taking apart and cleaning all plugs and grounds is the essential first step. Even if they look ok, do it anyway. I highly recommend Deoxit. Not cheap but very effective.

I've not heard of using any grease on the gearbox ground. Just fine sandpaper and deoxit. Others can correct me if I'm wrong or there's a better approach.

You might want to start a new thread on your issue. I think it's likely to get more looks and input that way.

if you could borrow an icu unit ...to test the theorys  of the failed start function

might be proof of the failure - repair

yeah .13 both ways doesn't sound encouraging....perhaps it has received a short circuit on this pin ...perhaps from shorting the terminals to the 12 positive  during start button testing ...where as it would normally be current limited by the relay coil  to just a small amount ?

I would have thought that it would be a pretty common switching transistor ...mounted on board the icu circuitboard.... but as you say  out of skills set for many

Doesn't really seem like it could be anything else at this point.  I was just able to buy a used ICU for $25 on ebay...don't have a spare or a friend with a K.

This problem first arose after many successful starts and after I drove to the gas station and filled the tank.  Then it wouldn't start after that.  In other words, I don't think that it occurred during my troubleshooting process.  Assuming that's so, I wonder if there's any other cause?  Is there anything else I should look at before installing the new ICU to prevent the same occurrence?  Would sure hate to have some other issue and fry another one.

The Suzuki starter relay could be the reason your ICU is bad now.  If I recall correctly the OEM relay has a CEMF diode installed internally. The Suzuki one may not have the same and the counter electro motive force created by the starter relay when turned off could have damaged your ICU.

That could be a very real possibility but before jumping to that conclusion....

Possibly the temp sensor may have gone 'open circuit' which will create an over temp warning, preventing the engine from re-starting. Try testing the sensor impedance at the FICU plug pin #10 and #13(earth) as well as at the temp controller socket across pin #E and pin #31(earth) for good measure. If this is the case then it would read less than 150ohms but anything other than 2.5 kilohm when cold would indicate the sensor is kaput.
Failing that.... I'd have to agree that it's looking like the ICU.

I think that my temp sensor is OK.  I believe that it measured 2.65 kilohms at pins 10 and 13.  Forgive my ignorance, but I want to be sure did the measurement correctly.  My multimeter said 2.65 and the scale was on 20k. That means 2.65 kilohms, right?  I'm guessing that the difference between 2.65 and 2.50 referenced by KH is due to it being slightly chilly in my garage.
 
Assuming that the temp sensor is OK, then it looks like my ICU is the problem.  If the Suzuki relay caused this problem for lack of a diode, it would be a simple matter to put one in between the relay and the ICU.  I already have some from another project.  If I read the specs right, I would use a 1N4001 or maybe a 1N4007?  Kioolt, can you confirm that this would work (or, if not, what would)?   Relatedly, if there were already a diode in the Suzuki relay would there be any harm in adding another?

hmm now the issue arises

an in4007 might be the best choice  ...the voltage realised on a 12 volt system of a relay collapsing is somewhere near 10 - 12 times the applied voltage
so that would make it a tad over 100 volts ...so a 1000 volt diode (in4007) will fall within its safety margin quite comfortably where as the in4001  would be exceeded

the diode is wired across the relay coil ...not in series to it
you might measure the Suzuki relay or open it to see which way the diode is polarised ...so that that you could confirm the correct terminations of positive and negative  but you must have it the right way around ...else it should blow a fuse on activation to earth


hope that helps ..


but doesn't explain the failure really,  I think that there would already be enough piv protection inside the icu already ...but weird things do happen

again just make sure that you have significant earth return connections in place ....

cheers

The diode should be applied across the coil terminals, in a reverse bias configuration (anode to ground terminal, cathode to positive terminal) in series with a 12V zener diode.  This is called a "snubber circuit".  

Source: https://www.element14.com/community/servlet/JiveServlet/download/11344-5946/13c3264.pdf

A 1N4007 is more than adequate to handle the transient voltage; just make sure the forward current rating of both diodes can handle the current of the coil.

All very interesting.  Robmac, to be sure I understand, the snubber circuit would simply bridge between the brown/red (earth) wire and the black/yellow (12v+) right before it enters the relay?  Is that right?  Sorry if this is obvious from what you already said.  I'm quite weak on electronics and want to avoid any missteps (though my knowledge has improved with tutoring from the folks on this forum....really appreciate the education).  

I also appreciate the need to get the diodes the correct direction.  That will be easy as they are clearly marked. 

I need to look at it more carefully, but I'm not sure there's any easy way to open the Suzuki relay and check what's in there.  

One final question, what is the likelihood that missing/incorrect diodes in the Suzuki relay are the cause of this (assuming that the ICU is the problem, and keeping in mind that I had quite a few successful starts with the Suzuki relay before this happened)?  Reviewing the posts, it seems like reasonable minds may differ on this point....I'm not trying to start an argument, just trying to increase my knowledge and understanding.

KJustin wrote:All very interesting.  Robmac, to be sure I understand, the snubber circuit would simply bridge between the brown/red (earth) wire and the black/yellow (12v+) right before it enters the relay?  Is that right?  Sorry if this is obvious from what you already said.  I'm quite weak on electronics and want to avoid any missteps (though my knowledge has improved with tutoring from the folks on this forum....really appreciate the education).  

Yes, that is correct.  Between the Brown/Red and Black/Yellow wires.

KJustin wrote:I also appreciate the need to get the diodes the correct direction.  That will be easy as they are clearly marked. I need to look at it more carefully, but I'm not sure there's any easy way to open the Suzuki relay and check what's in there.  

I doubt that there is room inside the relay for you to put the circuit.  It's a tight squeeze in there.  

Here are some good things to remember about diodes.  The cathode on a diode is marked on the body of the diode with a band.  This band corresponds with the bar in the symbol for a diode.  That's how you remember which is which.  

Also, the current flows in the direction pointed to by the triangle on a diode symbol.  That's another good thing to remember about diodes.  When someone says to "reverse bias a diode", what they mean is hook up the circuit so that the current attempts to flow against the direction of the arrow.

To use this new found knowledge, in the circuit I presented earlier, the 1N4007 is reverse biased and the 12V Zener diode is forward biased. In the circuit above, the band on the 1N4007 will hook up to the Black/Yellow, the band on the 12V Zener diode will hook up to the Brown/Red wire, and finally the remaining unconnected wires of both diodes will  hookup each other.

KJustin wrote:One final question, what is the likelihood that missing/incorrect diodes in the Suzuki relay are the cause of this (assuming that the ICU is the problem, and keeping in mind that I had quite a few successful starts with the Suzuki relay before this happened)?  Reviewing the posts, it seems like reasonable minds may differ on this point....I'm not trying to start an argument, just trying to increase my knowledge and understanding.

It's likely.  there's a lot of stored energy in the back EMF that gets generated in the windings of the relay coil when the ICU cuts ground.  All this energy dissipates across the transistor in the ICU, causing it to eventually fail.

That makes it abundantly clear.  Thanks so much!

As far as I can make out the earlier starter relays including those for the K series OEM #61311459008 didn't have the diode snubber circuit but only a resistor. The later Tyco (Bosch) relays introduced the diodes.
There's more info and a circuit diagram at 
http://www.euromotoelectrics.com/BOSCH-Starter-Relay-BMW-K-61-31-1-459-008-p/rel-k008.htm
If the current relay is from a non-ecu engine there may be no built-in protection from back EMF.
Before you go swapping out the ICU make sure you have a properly functioning oem starter relay installed and all your connections check out OK re voltages and continuity etc

One more thing: which zener diode do I need? I'm guessing it's a 1N4742, but I'm not sure if I'm reading the specs correctly. TIA.

a resistor is not totally a bad idea rob ...it gives a significant path ...and load for the back emf of an open circuit type switching condition ....which creates the higher back emf to start with.

even though we are doing things differently these days...lets call it catch up engineering . not that long ago these things were more common practice  rather than exceptional practice.  

I'm reminded that a capacitor and resistor combination is also a very good snubber circuit ....if designed right ....even a fairly significant variation seems to work very well

but I digress  ...yes a 13 volt zenner in4743  would work well, making the clamp voltage about 13.6  round abouts

remember we are talking about startup and the battery voltage wont exceed 13 or so under start conditions  the reverse biased diode in the circuit stops conduction until back emf is to be passed
.

Either suggestion - 1N4742 or 1N4743 - will work. Maximum surge current is about 370 mA, nominal current is about 75 mA. The 1N4007 has a forward current rating of 1A. All of this current is passed in the few milliseconds after the ground is switched off anyway.