• 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?

Starter or Battery 2004 R1150RT

Well done Roger!
Thanks for writing this up.

I think we should all get together and sort out what we'd like you to work on next. :D
 
Thanks Mike, not quite done yet. Next project is to enjoy riding ...

Here's a quick update. I discharged my PC680 again last night this time it took an additional hour which means there was 13 aH of output, up from 11 hours on the first discharge. It's headed in the right direction. Bike continues to start well.


1st time: 6 hours to discharge (11 aH), 16 hours to recharge
2nd time: 7 hours to discharge (13 aH), 22 hours to recharge

Goal is 16 amp-hours, cycle can be repeated up to 5 times. Each cycle removes sulfate from undercharging.
 
Roger, what level of recovery did you reach with your battery? Also, how did you load the battery during the discharge? What do you use to determine when you have discharged to goal?

For the PC680 for the first two runs i used a 6 ohm 50 watt resistor. I monitored the battery voltage and stopped discharging at 10 V per Odyssey's instructions.

For the third run today I used a 4 ohm resistor (50 W) which is for he 5 hr. discharge rate. Estimated charge for the third discharge was 87%.

So the summary is 68%, 80%, 87%. This means the battery has improved.

I called odyssey today about the slow charging rate. After some discussion we agreed the charger is defective and they're sending another. Great customer service.

The charger's problem was it was tripping to a low current initial charge. In its place today I used my car idling for an hour and a half which charged the battery at an 8 amp rate, finishing with the odyssey unit.

I'm hopeful that when I get a working charger later in the week I can add some more capacity, maybe getting to 100%.
 
While waiting for the charger to arrive, I ran a test yesterday. After pulling the Motronic and Fuel Pump Relays, and the Motronic fuse, I powered all those circuits with jumpers from a second battery.

The idea of this test was to see how well it started if the voltage supplying those circuits was not glitched by the starter load. The home alarm 7 aH battery I used resulted in a steady 11.4 volts with the 5-6 amp fuel pump load.

It turned over once or twice and started; the sensors were better behaved than I have ever seen. It ran very smoothly immediately.

1. The RPM jumped from 0 to 1350 RPM in 1.1 seconds.
2. The spark advance kicked in immediately and stayed at about 25 degrees. (This often moves from zero to 14 back to 7, all over the place during a cold start.)
3. The dwell moved smartly to 1.6 mS and stayed there reliably. (This often moves from 0.6 to 0.3 to 1.6, etc.)

RB

Chart of Start Sequence (RPM on Right Axis)
secondbatterystart1.jpg
 
Last edited:
I went ahead an am trying the 6amp version

I ordered, and the Brown Truck dropped off the 6am Odyssey Ultimizer. I'll try it out, and get back to you. I hope this is a good one and fails not.

Dadrider.
 
I ordered, and the Brown Truck dropped off the 6am Odyssey Ultimizer. I'll try it out, and get back to you. I hope this is a good one and fails not.

Dadrider.

Good, I'll be interested to hear how it goes. I had a Battery Tender harness so made a 4 ohm, 50 watt resistor to plug in to discharge and test the battery's capacity. They get searingly hot.
 
To try and further see what parts of the power system were being affected by starting transients I ran three more tests:

1. Connect only the Motronic to the second battery. Leave pump/injectors on the main battery.

Results: Not a bad start but longer and not as clean as the test in the prior post, when both the Motronic and Pump/Injectors were on a second battery.

2. Power everything as usual but use a second battery for starting.

Results: Okay, but kicking in the starter solenoid took the voltage on the main battery to 10.5 volts, and then the alternator came on line bringing the voltage to 13.6 volts. That was enough voltage changing to affect the fueling and maybe the Motronic. As a result it took longer to get a stable idle.

3. Run the test in the prior post--Motronic and pump/injector on a second battery.

Results: same good results as in the prior post.

Still waiting to see how things will be when my battery is recovered. However there is no doubt that a second small battery for the injectors and Motronic would produce a great starting R1150. A bit theoretical since I doubt I'll add a second.battery.
RB
 
Roger I think you've solved the mystery behind a very old saying that a new battery breathes life into an old bike and makes it just like new again. :)

I think you're right though. It's not really practical to add a second power source. Might make a great project for RTP owners though. They could re-purpose that second battery in the glove box to a whole new civilian life!
 
Roger I think you've solved the mystery behind a very old saying that a new battery breathes life into an old bike and makes it just like new again. :)

I think you're right though. It's not really practical to add a second power source. Might make a great project for RTP owners though. They could re-purpose that second battery in the glove box to a whole new civilian life!

If I just did the injectors and Motronic, I'd need about a 1 aH battery, maybe less. It would require a disconnect relay that left it off-line for maybe the first minute of operation. We'll, after I've worked out the battery issue.
 
If I just did the injectors and Motronic, I'd need about a 1 aH battery, maybe less. It would require a disconnect relay that left it off-line for maybe the first minute of operation. We'll, after I've worked out the battery issue.

While experimenting with keeping the LC-1 alive during startup so I could log every second I ran into a snag due to the heater element in the O2 sensor which was pulling around 3 to 4 amps during warmup. My small homemade power supply normally used for breadboard projects was not up to the task. A 1aH battery might not be enough? Power for the O2 sensor heater is coming from the same source and fuel pump power on my setup though. Yours may be different.
 
While experimenting with keeping the LC-1 alive during startup so I could log every second I ran into a snag due to the heater element in the O2 sensor which was pulling around 3 to 4 amps during warmup. My small homemade power supply normally used for breadboard projects was not up to the task. A 1aH battery might not be enough? Power for the O2 sensor heater is coming from the same source and fuel pump power on my setup though. Yours may be different.

You would have to break into the wiring harness and power the injectors alone.

I just did some measurements and each injector draws .760A @ 12.8v so the load is 1.52A.
I then tried to keep a pair up on a 35,000uF capacitor (5"h x 3" dia) - It would only last for less second.
A battery seems to be only solution with an anti backfeed diode (preferably a shottkey type with low forward v drop)

Realistically, a properly charged and maintained main battery is the simplest solution to the undervoltage problem during startup.
 
While experimenting with keeping the LC-1 alive during startup so I could log every second I ran into a snag due to the heater element in the O2 sensor which was pulling around 3 to 4 amps during warmup. My small homemade power supply normally used for breadboard projects was not up to the task. A 1aH battery might not be enough? Power for the O2 sensor heater is coming from the same source and fuel pump power on my setup though. Yours may be different.

Good point, I currently have the LC-1 running on the fuel pump relay along with the injectors.

If I did this, I'd add separate wires to power the injectors.

You would have to break into the wiring harness and power the injectors alone.

I just did some measurements and each injector draws .760A @ 12.8v so the load is 1.52A.
I then tried to keep a pair up on a 35,000uF capacitor (5"h x 3" dia) - It would only last for less second.
A battery seems to be only solution with an anti backfeed diode (preferably a shottkey type with low forward v drop)

Realistically, a properly charged and maintained main battery is the simplest solution to the undervoltage problem during startup.

Although the back diode would be my favorite approach because it would be automatic, I've taken it off the drawing board for now. It seems like the up transient when the alternator comes on line also jumbles the fueling.

I'm going to take more data because when I powered everything but the starter on the main battery. The voltage went 12 at key on 11 when the starter was pressed (solenoid kicking in) and then 13.6 when the alternator came on line. Those three voltage steps occurred over a couple seconds. The bike started and ran but the Motronic responded with funny timing.

Although the injectors are 3/4 amp each, the duty cycle during starting is only about 10-20%. (at wot it gets to about 80% or more.) So the Motronic runs a 1/4 amp plus another 1/4 for the injectors means 1/2 amp total during starting. I would relay connect the two batteries after 60 seconds which would give the Motronic enough time to get settled.

I won't do anything though till I have a battery that meets spec.
 
Realistically, a properly charged and maintained main battery is the simplest solution to the undervoltage problem during startup.
I learned this a long time ago on bikes with a kick start. A bike that had an iffy battery, when the electric starter was pushed, frequently would not start. An attempt with the kick starter would bring it right to life as the "draw" of the starter had been taken out of the equation and all the voltage went to the spark circuit.
The explanation of the interaction of all the sensors in regards to voltage draw is a good reference for those with computer controlled motorcycles. The importance of matching a charger to the specific battery in the vehicle is sometimes more than it seems. :clap
OM
 
Today I ran a cold start test using a 13.7 volt source to power the Motronic and fuel pump/injectors. I found a very interesting result regarding dwell time. Keep in mind that I have new stick coils installed.

Before getting to today's result, yesterday when I cold-started with a stand-alone battery whose voltage was 11.7 volts powering fuel pump/injectors and Motronic, the after-start stick coil dwell was about 1.7 mS.

Today (same config) using a 13.7 volt source--my car's battery with the alternator running--the dwell time was 0.5 mS, that's 1/3 of the dwell compared to using the 11.7 volt source. And the start was not as robust as when I used the 11.7 volt source.

The difference is that the coils had 13.7 volts on them today (note to self: check voltage on the other side of the key which powers the sticks) and the Motronic read my second battery also 13.7. Yesterday, the Motronic read the second battery at 11.7 but the coils had actually 13.7V. So the combination of a high voltage at the sticks but the Motronic thinking it was a low voltage leading to longer dwell produced a spark that better combused the mixture.

I'm floored! Who would have guessed that the Motronic sets dwell partly based on battery voltage? And that a stronger spark would lead to a better start!!

RB
 
That is indeed interesting Roger. Kind of makes me wish for a good old set of points and a feeler gauge. Well, almost.

Jim
 
Hi Jim. Kinda feeling that way too.

Below is a chart of the starting parameters for 11.7 vs 13.6 (Motronic and injectors). One of the reasons that 11.7 works so we'll is that the sticks have 9-10 volts during cranking then 13.6 volts once running but the Motronic only sees 11.7 so calculates a long dwell.

Notice in the chart how much faster the rpm builds in the 11.7 volt case compared to the 13.6. That seems to be due to inadequate dwell at the higher voltage case.

Also note that the Motronic has no way to adaptively learn low battery voltage corrections for coils and injects during cold start.

secondbatterycompare.jpg
 
Voltage corrective tables, Cranking Dwell settings/ fuel enrichments/ priming pulse widths/ etc. are all there if you want them (well, with MS that is).

Here is a link to Phil Tobins "Tuner Studio" tuning software and other stuff. All free on the lite edition, the advance product (and subsequent new releases) for a one time very low $ number. This is just a newsletter (for a new release of MS 2.5 on Tuner Studio) but you can see some of the tools that are available (pretty well anything you want to do, log, or change on bikes operation. The new "Horse Power and Torque" gauges are beckoning me, I will have to add a GPS.

http://www.efianalytics.com/email/2012-10-10-NewsLetter.html

Just another way to skin the cat as they say.
 
It would be great to be able to adjust these older motors. The data I've taken strongly suggest that the injection and dwell voltage correction tables are inadequate. Get them right and my motor starts from dead cold in about a second. Get them wrong due to low battery (or my manipulation) and the engine labors for 10-15 seconds.
 
Back
Top