How much current...

[guitardad]

... should the starter draw on my '94 R1100RS?

OK, some background. I've been getting intermittent ABS faults - the alternate flashing lights that means low voltage at startup. Basically, I have to make sure EVERYTHING is turned off when I start the bike - driving lights, heated grips, even my XM Radio! Battery is only 6 months old. Now, as the temps drop into the 50's, the fault is starting to come back. So I'm starting to suspect the starter is the real issue - it's drawing too much current. I borrowed a clamp-on current meter from work, so I can measure what it takes to spin the starter up. But what's a good number, and what's too much?

 

[manicmechanic]

Not intending to sound trite, but you really don't have a problem, unless your battery is getting old. As the temperature drops, several things occur - your oil gets thicker (including the oil in your transmission), your battery cranking ability goes down, and therefore the current required by the starter to rotate the engine increases. When this happens your battery's terminal voltage goes down, in this case to the point that the ABS faults on low voltage. My K-bike did it, my old 1100GS did it, and my 1150GS does it. I expect it. What to do about it? Well, keep the battery charged. Perhaps use a lighter-weight oil in the engine. Pull the clutch in when starting to minimize the amount of drag on the starter. Keep the bike in a heated enclosure. Move someplace warmer. Install an Interstate Megatron battery. All of these should work, but which are practical, for you? I think someplace on the IBMWR site is a thread on how to do some rewiring of the ABS power circuit so that it comes on after the engine is running. Check it out.

 

[nrpetersen]

Sounds like a weak battery to me also, but my guess on the starter current is maybe 200 Amps. Strictly a guess. What do you measure?

And - as the voltage in your battery goes down, the starter current won't go down at all. If anything it will go up.

 

[globalrider]

... should the starter draw on my '94 R1100RS?

Depends how much load you put on it. And that changes with the viscosity of the oil (OAT).

Notice the current reading isn't a steady value when you crank.

But I remember reading 100 amps in one of my BMW service manuals.

On a Dash 8, it would be 1200 amps to get the props to start to spin initially, tappering off to 200 amps as you light up the engines.

 

[jingdog]

Hmmm everbody has a story. But nobody really KNOWS!

 

[ANDYVH]

One thing that is known, is that the ABSII (or ABS2) system in GuitarDad's and my 94 R1100RS was known to be very sensitive to low voltage battery conditions. The low-voltage threshold value in the ABS modulator computer was in effect, too high. Or said this way, if after a second or three of cranking, the battery voltage dropped to 11 volts, the ABS would report a "soft fault" in the system.

My bike did exactly the same thing. The bike would always start (I always ran Mobil I synthetic). But the ABS would soft-fail as indicated by the alternating flashing lights. I'd stop about a mile up the road, shut off and restart the bike, and everything was fine, no ABS fault. It did this already back in 96, but I still used that original Mareg battery all the way up to 2002 with no other problems (uh, other than a fried ABS computer). My dealer in 98 said I had done too many "low voltage starts" and that it had permanently fried the soft-fail into the computer. Bye Bye ABS.

Back in 96 to 2002, I had a 1/2 hour minimum highway ride no matter where I went because I lived out in the country, and I rode nearly every day. So I never used a charger overnight cause the ride home kept the battery up, or so I thought. These days I no longer have the 1/2 daily highway ride, so I do use a Battery Tender. Same bike for 13 years, with the same starter in it, 125,000 miles. No ABS start faults anymore.

I'd get a Battery Tender first, plug it in each night, and then monitor the start up conditions. I'm also inclined to modify my bike's wiring so the ABS powers up after the engine has started. The newer ABSIII has a much lower "low voltage" threshold to avoid this exact frequent soft fail condition.

 

[nrpetersen]

My dealer in 98 said I had done too many "low voltage starts" and that it had permanently fried the soft-fail into the computer. Bye Bye ABS.

Yah sure...........! Low voltages don't fry solid state electronics. Over voltages, reversed polarity, and heat do.

 

[guitardad]

Not quite done, but here are the measurements so far:

On my wife's '04 R1150R80:
____Current drawn by the starter: 250-260 A;
____Voltage at the starter: ~10.5V
On my '94 R1100RS:
____Current drawn: 240-255A
____Voltage at the starter: ~8.5-9V
____Voltage at the battery (when starting): ~9V
____Voltage drop across the positive cable: 350 mV

Currents were measured using a Fluke 337 clamp-on current meter, set to its "Inrush" function. Voltages are my little handheld DVM - no min or max function, so I'm reading it by eye. Still need to measure the voltage drop from negative terminal to the engine case near the starter, since that's the current return path from the starter. But I don't think there's anything wrong with the starter, and the drop across the positive cable seems perfectly reasonable to me. Given what I measured at the battery, I don't think the negative side will be a problem either. So I'm concluding this Toyo SLA battery has a higher internal resistance than a more typical motorcycle battery, and so it has a higher internal voltage drop under the large load of starting. Which would mean I can either:
a. Just live with it - if I restart after the bike is warmed up, it starts easier and the fault goes away.
b. Look up the ABS power mod on the IBMWR site.
c. Buy a "real" motorcycle battery.

For now, I'm leaning toward "a." I'll keep ya'll posted.

 

[jingdog]

I agree about the "fried in" problem to the ABS computer! Sounds like hooey to me!

 

[ANDYVH]

Ya know, I still have my doubts about the "fried" computer also, but I had no means to prove it otherwise, other than installing the old modulator to see if it worked with my new battery.

Oh, I'm also using the Toyo SLA battery, and have had no problems.

But this winter I plan to rewire my abs so it only fires up after the bike has started. Then there are no low-voltage problems to be concerned about.

Hey Chaz, 9volt at the battery? Was the bike sitting for a while or was the battery not at full charge when you started the test? By reading 9-volts I'm not at all surprised you are getting ABS faults. At 9-volt that may be close to the threshold value for the engine computer. How about trying after a good charge on the battery. Also, I have read that just because you can't see corrosion on the battery terminals doesn't mean there isn't any (doesn't have to be green).

An old trick I heard about for battery terminal corrosion: find an old dime with high silver content, and expoxy it between the terminals on top of the battery. Its supposed to act like a anode.

 

[guitardad]

OK, last measurement was 250 mV from engine case (near the starter) back to the negative battery terminal while starting. Which leads me to conclude there's nothing wrong with Clifford.

To be clear, If I ride the bike every day, and remember to shut off all the farkles before starting it, I get no ABS faults. But when I measured 9V at the battery, it was after 3 or 4 sessions of cranking the bike (to start it) without riding a while to charge the battery back up. What all this tells me is, on my bike at least, the Toyo SLA battery is barely adequate for my use.

I think the near-term solution is at hand. I started the winter tear-down of my airhead today. Since the PO had installed an Odyssey PC680 battery recently, I may just have to transplant that battery into Clifford. And if the Toyo doesn't survive into the spring, when Hans (the R65) is ready to roll again, at least I haven't lost an expensive battery, right?

 

[guitardad]

Ya know, I still have my doubts about the "fried" computer also, but I had no means to prove it otherwise, other than installing the old modulator to see if it worked with my new battery.

This is a situation where I'd like to understand better what the ABS and/or Motronic software is doing. Flash memory of early-90's vintage did have a limited number of cycles before it failed. And the typical failure mode was that new data could no longer be written over old - in essence the fault would be burned in permanently. If the software didn't read the flash - to see if the fault was already burned in - then I could see the flash memory failing. I know how I'd write the software, but I'm not quite so sure about a German Engineer!

 

[ANDYVH]

"Und, vhut are you zaying about zah German engineer,....hmmmm?"
"Perhaps you haff zah impresshion vee only zee things one vay?"

What you are saying does have some meaning. But to clarify my ABS issue further, the ABS on my bike always reset itself if I stopped about 1-mile from the first start of the day and restarted the bike.

But back in May of 98 I had an accident on the bike, where the bike fell on my and broke my leg, after I rode the bike into a ditch, and it bounced off a rock our culvert. The only damage to the bike was the sidestand safety switch was busted. I took care (got rid of the switch) by just connecting the two wires at the switch. Ever since that accident the ABS would never reset itself.

So I always thought the accident somehow caused the ABS modulator to permanently malfunction. The bike landed on me, snapped my lower left leg, with the bike laying on me on its left side. After that crash the ABS never worked again. And I checked all the obvious, battery, wheelspeed sensor, connections.

 

[globalrider]

9 volt at the battery? Was the bike sitting for a while or was the battery not at full charge when you started the test?

I believe that was "under load" as I read it.

Healthy batteries have internal resistance of about 10 to 15 milliohms. Anytime you draw a current, their voltage will drop from the open circuit voltage of approximately 12.65 to 12.85 volts (sometimes a bit more).

Now you can see why it is imperative to have near zero resistance electrical connections anytime you are drawing any current, especially when we are talking high currents. Any resistance along the way within any connection creates a voltage drop (and heat) that isn't helping the starting process.

Every have a slow turning starter in the car? You then pop the hood and give the starter cable a good yank and it resolves the issue quite often...a corroded high resistance connection is what you have.

 

[guitardad]

OK, the final measurement was 250 mV from engine case, right where the starter is bolted, back to the battery negative terminal while cranking. Using the 250 A current I measured, that makes the negative cable resistance 1 mOhm - which seems perfectly reasonable. The drop on the positive cable - 350 mV - means the resistance there is about 1.4 mOhm. Again, seems quite reasonable. I didn't try to measure the resistance directly, because my little handheld DVM wouldn't measure a small resistance like that with any accuracy. So the voltage measurement under load is better.

But I've got another experiment going now. I started tearing down my project bike last night - my '79 R65. And what did I find inside? An Odyssey PC680! So if one bike is in pieces and one is running, which one gets the Odyssey? Simple choice, really.