• Welcome, Guest! We hope you enjoy the excellent technical knowledge, event information and discussions that the BMW MOA forum provides. Some forum content will be hidden from you if you remain logged out. If you want to view all content, please click the 'Log in' button above and enter your BMW MOA username and password.

    If you are not an MOA member, why not take the time to join the club, so you can enjoy posting on the forum, the BMW Owners News magazine, and all of the discounts and benefits the BMW MOA offers?

roger 04 rt

New member
Back when I had my red ‘04RT (miss that bike), I spent a lot of time working out what it takes to get a dual-spark Oilhead to start in less than a second when cold: https://forums.bmwmoa.org/showthread.php?70495-Starter-or-Battery-2004-R1150RT . Bedsides the usual suspects (fuel, compression and spark), the ‘04 bikes need a clean electrical system, a fully charged non-sulfated battery and a good starter that doesn’t draw too much current.

Fast forward to the past week: my 2001 R1150GS is turning over a bit slower than usual and takes a full second to start first time in the morning, and the starter sounds a little rough. That’s not bad but it was faster a couple years ago. On top of that I’ve had three occurrences where I get ABS errors that don’t clear after riding off (blinking in tandem sometimes or both flashing alternately other times). The ABS errors can be cleared by restarting the bike.

Yesterday I hooked up my GS-911, selected all realtime values, put the 911 in turbo mode for fast data collection, and collected a realtime log. The log confirmed it was starting okay but the voltage readings went 13V to 12.6V at key on (that’s fine) and then to about 8V for the first couple samples when I hit the starter. The 8V dip explains the slightly slower start and the ABS errors (btw, an 8V dip on a dual spark would really mess up the start).

The question is, do I have a bad battery or starter or something else?

The bike has 42,000 miles on it, the starter is original and has never been serviced, and the bell housing area shows a lot of clutch dust. The battery is a PC680, several years old, but well maintained and still holding 100% SOC even after the surface change is bled off by turning on the headlight for 30 seconds. The cables to the battery, ground, alternator and starter are tight and clean as a whistle.

I hate throwing parts at a problem so I’m going to begin by servicing the starter, we’ll see ...
 
This pair of Blue-Point gauges are may favorite for checking starter draw and alternator output.

s-l225.jpg


I don't know where the spec is for the starter draw when it is brand new but they due draw more as they age. I will buy a DC amprobe at some point (Fluke= $350.00) but until then, the BP set gets me by.
You may be able to check the difference with a battery charger (capable of a jump start) hooked up to the bikes battery and watch for a drop in voltage. IMO, 8-Volts is too low. Ideally, a test with a fresh battery as sometimes the older batteries will cause a false read.
OM
 
Thanks OM, Here's a better view of what's going on. You can see the dip to 8V and it may be even lower since the GS-911 only samples every 0.4 seconds, even in Turbo mode.

Screen Shot 2019-08-04 at 10.18.44 AM.png
 
Another example of the “starter drain theory” is with my Yamaha TW200. Battery is still reasonably good. It’s not good enough for when a long period of cranking sucks down the voltage low enough so the ignition sequence doesn’t have enough voltage to fire.
If the bike doesn’t start in 10 seconds or so, 1 kick of the kick starter will bring it to life as the starter isn’t using the bulk of available battery.
Hope that helps.
Gary
 
Load test battery-pass/fail. Marginal, charge battery, retest, still marginal=fail

Current draw starter, pass/fail. If pass: voltage drop test. If it exceeds 2.0V, grounds,cables, relays to be checked.

Test starter/battery/alternator by voltage and ignore current, you have about 50% of the picture.

Testing does not require scan tools, spreadsheets or complicated and irrelevant assumptions.

Common sense is required...........................................
 
Load test battery-pass/fail. Marginal, charge battery, retest, still marginal=fail

Current draw starter, pass/fail. If pass: voltage drop test. If it exceeds 2.0V, grounds,cables, relays to be checked.

Test starter/battery/alternator by voltage and ignore current, you have about 50% of the picture.

Testing does not require scan tools, spreadsheets or complicated and irrelevant assumptions.

Common sense is required...........................................
I sorta agree.....but- most go with what they know.....or how they have done it in the past. It's also important to be familiar with the test equipment and that takes usage on a regular basis. I have some basic test equipment and some high end test equipment so sometimes I use one to check the results on the other to confirm proper operation- on my end.
OM
 
Load test battery-pass/fail. Marginal, charge battery, retest, still marginal=fail

Current draw starter, pass/fail. If pass: voltage drop test. If it exceeds 2.0V, grounds,cables, relays to be checked.

Test starter/battery/alternator by voltage and ignore current, you have about 50% of the picture.

Good advice, thanks.

Testing does not require scan tools, spreadsheets or complicated and irrelevant assumptions.

Common sense is required...........................................

You seem to have a spreadsheet phobia and at the same time manage to sound condescending and arrogant. Your post would have been better without it.
 
Last edited:
You seem to have a spreadsheet phobia and at the same time manage to sound condescending and arrogant. Your post would have been better without it.

Routinely, I get called worse and I have never apologized for it. That make me bad man being Canadian and all.

Spreadsheets are just a way to collect and organize data for dissemination. I hate them. I call them the "devil's method of keeping me from doing my job." I have a lot of spreadsheets.

Last time I checked, a scan tool and a spreadsheet don't have jumper cables to get you home.

Demonstrating how smart you think you are, generally, po's those that know you aren't.

If you are allowed to come to Canada and your car won't start at -40C, you gonna create a spreadsheet and hook up a scan tool or you gonna go KISS?
 
I get it now, thanks. Your approach to problem solving is to be a rude jerk if you don’t agree with someone’s way of doing things. And then when you realize you’ve been an idiot, why apologize? I get that too. I’m a sh1tkicker from way back myself.

When you feel diminished by the idea someone might be smarter than you, I agree that you’d be better off not writing those long posts trying to “Demonstrating how smart you think you are,” because it “generally, po's those that know you aren't.”

Next time you start to get all twisted up in your knickers by how I explain my problems or answer a question from another rider, go get a beer and then scream out the window if it’s not frozen shut. If that doesn’t work, just block my posts, then you won’t get threatened and feel yourself start to shrivel. Or just don’t bother reading them.

The funny thing in this case is that without a gs-911 I’d have never known that the voltage dipped to 8V for half a second. I’d have just had the flashing ABS and the sense that my starter was rough. When you upgrade from the 1100 series to a bmw that produces data, get yourself a GS-911, and when you want help, PM me like so many others do.
 
Last edited:
Back to work ...

I had a chance to make several measurements this morning.


1) I pulled the fuel pump relay rather than F5 so I didn’t lose my adaptive values but could crank without the bike starting. The cranked and measured the battery voltage. Although there is a half-second drop to 8V, it will then crank for 20 seconds (as long as I cranked) at 10.6V.


2) While cranking, the voltage drop from battery (-) to engine case is 100 mV


3)The voltage drop during cranking measured from battery (+) to the stud on the starter was 125 mV.


The cable drops mean their resistance is a few milliohms which seems fine.
 
The starter’s out, the drive end was disassembled, cleaned, lubed (not the solenoid) and reassembled. It was pretty dirty and the planetary gears and nose bushing were dry but there’s no smoking gun and the gear grease cover was tight and intact.

Tomorrow I’ll get at the back end
 

That was lot of work when two NSFW words would have said the same.

Not everyone has access to a scan tool or can read a spreadsheet but the problem remains very simple and simple tools will get you to the root of the problem.

I have tools that would blow your mind and I can re-program any ECM/controller on any machine. I have a $4000.00 battery/alternator/starter tester that can tell me the sp/gr of any cell in a battery, pull in/hold in current, draw of any starter and it can even tell me that I might have a leaking diode in my rectifier.

Stand in line and take a number if you want to take a piece out of me. #30 is available, I'll bump my Mom to #31.

I'm on this forum to provide information in such a manner that everyone/anyone can use. I do it day in and day out with my staff.

I knew your problem from the description in the first post. Good luck finding it.

As to riding a slightly older antique than yours, I can troubleshoot my bike with $2.00 test light on the side of the road.

I can't say the same for my service truck.
 
The starter was cleaned and lubed. It looked dirty but there wasn’t any smoking gun. I reinstalled it and it doesn’t seem like the slightly slow start has improved.

I moved on to the battery and as directed by odyssey I am running 3 amp discharge cycles to 10V and then recharge. This will often recover a battery that hadn’t been properly charged over part of its life. It worked well on my ‘04RT.
 
Last edited:
My first discharge cycle yielded about 13 amp-hours out of 16 amp-hours for a new battery. That is about 80% of new capacity. Odyssey considers this a failed/worn-out battery. I’m recharging it now. It will be interesting to see if the next discharge cycle is improved. If it is only sulfated it may improve. Odyssey says to run up to 5 discharge cycles.
 
For anyone interested, here's my discharge network of three 6 ohm resistors and an sad pigtail, which creates a 4 ohm network, which yields a 5C (3 amp) discharge current. It gets hot!

IMG_4865 2.jpg
 
Last edited:
For anyone interested, here's my discharge network of three 6 ohm resistors and an sad pigtail, which creates a 4 ohm network, which yields a 5C (3 amp) discharge current. It gets hot!

Photo not uploading yet

Had similar issues yesterday while answering a post with a lot of questions. Uploading a photo was taking a very long time and it finally just hung. I lost a half hour of post and photos and just gave up. It was a fairly large photo so maybe the system won't allow that but the exit was not graceful and nothing in the cache to return to was disappointing.

I've never done the mod on my oilhead alternator to properly charge the Odyssey battery but whenever I return home from a ride I always attach my Odyssey specific charger and top it off. It seems to work OK but I should run a capacity test as you are doing and see where it's at. For my airheads I install a digital regulator that has a trimpot allowing me to set the charge voltage. I run them around 14.5 but I believe 14.7 is optimal to keep them fully charged and prevent sulfation. I worry a bit about running those old alternators with the fragile rotors too hard.
 
It was a fairly large photo so maybe the system won't allow that

Dollars to donuts that's the issue. Always good to resize pictures down to the 1-2Mb range before trying to upload. More bytes in the picture doesn't make it any better to see on the computer monitor. After a certain size, it's all a waste of time...as you found out! :wave
 
Dollars to donuts that's the issue. Always good to resize pictures down to the 1-2Mb range before trying to upload. More bytes in the picture doesn't make it any better to see on the computer monitor. After a certain size, it's all a waste of time...as you found out! :wave

Thanks. I resized and went back and added the picture.
 
I just finished the second discharge. It took 5 hours and 20 minutes and yielded 15.3 Ah, which is around 95% capacity.

There was another change too after the first discharge cycle, when I attached my programmable power supply (set at 14.8V) the current started at 4.5 amps, and climbed to 5.5 amps in the first hour. The battery then charged overnight. After the second discharge cycle, when the power supply was attached, the charging current was about 9.5 amps right away and stayed there for a couple minutes and is now slowing coming down. It seems a good sign based on Odysseys data which says to charge it correctly you need to charge with a minimum 6 amp supply.

The battery will now recharge for the rest of the day.
 
Interesting side development ...

I've been unable to get the GS-911 to record values fast enough since I first discharged the battery. Hexcode processed a trouble ticket for me but said they couldn't duplicate it. Further study of the data showed that every time I pressed the start button, there was a 3 second delay from the Motronic so I couldn't see the transient voltage dip during the first second.

This morning I realized why that might be happening. I have been discharging the battery while installed in the bike. The voltage has dipped to 10V twice, and the first time it reached 9.54V for several minutes before I got back to check. That was enough to reset the Motronic apparently, and I never reregistered the throttle (key on, rotate the throttle twice, key off). Apparently, without the key on/rotate/key off after disconnecting the battery, the Motronic does some housekeeping every time the starter button is pressed and it keeps doing that until the throttle is reregistered.

After resetting the Motronic and reregistering the throttle, the GS-911 now logs at full speed, even during the startup transient.

On the third start of the day (while I figured out the GS-911 problem) the 300 mS transient voltage was 9.54 volts. Much better than the 8.05V recorded at the start of this project. However, since it was the third start of the day, I don't think we can rely on the 9.54V data point.

The battery is now discharging for the third time, and I will reregister the throttle before my next test!
 
Back
Top