Ducati.ms - The Ultimate Ducati Forum banner

1 - 20 of 20 Posts

·
Registered
Joined
·
90 Posts
Discussion Starter #1
I have spent a lot of time reading through the hallowed archives here at Ducati.MS. I've noticed that starting/charging issues are very common. There is no reason for this.

Step 1.


That's around 3lbs 15oz


No superbike should be hauling around 4.5kg(10lbs) of lead. It just seems silly to spend any money on titanium until you buy a bit of the lightest metal, Lithium.
In this case- Lithium Iron Phosphate (LiFePO4), these cells are 3.3v 10ah, $17/each. I will start with 4 wired in series to make a 13.2v 10ah battery.
I have just received 16 of these cells with a battery management system(BMS) for powering an electric bicycle. I will be ordering some additional cells to use for the Ducati.



These cells are the same type of chemistry used in Speedcell's A123 cells, however the A123 cells have a very high discharge rate over 50C, this means that their 4.6 ah battery can put out 240 amps max pulse. Those are very good batteries and I'm sure that anyone who buys them will be happy for a long time. The cells that I am using are rated for 15C pulse, which is 150amps. For $70 I'm willing to experiment. Besides the friendly price, these batteries have screw terminals, this means no spot welding the interconnects.




Step 2.


Replace the 1.7kg(3.75lbs) starter motor with a 220gram(7oz) brushless motor controller and use the 3-phase alternator as a 3-phase brushless motor.




I'm not sure how many amps it will take to spin a 1 liter motor with a high compression ratio, but I will find out. This will determine the final battery sizing. I may end up running a higher voltage for the starting system and use a buck/boost circuit to provide the computer and the rest of the bikes circuits with the voltage they need. It is possible to program a soft start function into the motor control, this will ease the burden on the battery, although other means such as ultra-capacitors may be needed. I will also need to contemplate effects on the flywheel attachment from starting torque being transmitted through it.

There is potential to save ~4.5kg(10lbs), plus some gear train weight, and have a starter/charging system that just does it's job without problems.




to be continued...
 

·
unM0derator
Joined
·
2,853 Posts
The reason that starting and charging issues are common has more to do with poorly built / situated regulators and undersized wiring than it does with the starter motor or the battery.

While I'm all for the DIY approach, the risks of damaging the bike by using the alternator as a starter motor are not appealing. I'm not even sure you could put enough current though the alternator without frying it, to start the engine.

There is quite a bit of gear reduction going on from the starter motor (30:1). I'm thinking that you will simply not have enough torque to do it reliably without burning out the armature the way you are planning. The wiring in the stock alternator is only rated for something like 40A, and that's across the three wires. Look at the gauge on the starter lead, and compare it to what comes out of the alternator, and remember it takes almost 60A *with* reduction to crank the engine.

Mathematically, if the starter consumes 700W, that's roughly 1 HP. 1 HP ~ 2 ft lbs. / second of torque (at a one foot radius), at a guess of 2500 rpm (what speed the starter turns). With a 30:1 mechanical advantage, it now takes about 60 ft lbs. / second at 83 rpm and a one foot radius to overcome static compression. At 10cm the flywheel would need about 6 times that amount (5cm radius), or 360 lbs. / second exerted on it to have the same effect. That would be close to 6 HP, or around 4400W. If you can get 360A through the alternator, you are my hero.

But, good luck. I'm waiting to see the results, and I hope I'm wrong.
 

·
Registered
Joined
·
477 Posts
I had used the speedcell batteries in my 749; this battery was able to start my bike twice and that was the end of that. I had to have friends at the track bump start me.

The reliability of a strong battery outweighs the weight savings for me :). However, this was a very interesting Sunday morning read :).
 

·
Registered
Joined
·
1,031 Posts
Excellent idea. I think the data on the alternator limitations is right on. Not to mention the mechanical coupling limitation that may exist. Lithium is the way to go and there is some real break thought just around the corner. The average DIY will probably burn down is bike at a minimum, ever see a lithium ion battery explode from an over amp situation. Not pretty. If you are hard core racing weight reduction is obviously important. A good AGM/RGM of adequate size is a much more practical and safe approach. This being my second Ducati and maybe 15th bike that acid bomb on the left side is one of the first things that have will go. As I look through countless sites and offerings I have to ask myself, if I install the ti axel and nut who will I be able to beat now, oh thats right no one. If its just a bling or bragging right thing well some bits are just cool. I love the concept and the project for what it is, an improvement over factory technology. Will be following this one for sure.
 

·
Registered
Joined
·
300 Posts
I had used the speedcell batteries in my 749; this battery was able to start my bike twice and that was the end of that. I had to have friends at the track bump start me.

The reliability of a strong battery outweighs the weight savings for me :). However, this was a very interesting Sunday morning read :).
I build and sell A123 battery packs for my mates in the swedish ducati club and all have had great results, even the ones with hard starting racing engines that didtn want to start with stock battery. I ran my bike 36 runs in a dyno with my packs and no problem so it seems strange that speedcell dies on you like that.
But the diffrence is that i build mine with balance tabs so you can balance the batteries and that gives you 1000 cyckles aprox that s full discharge and full charge not like us just scraping on the top a little.
And i allso divide my packs in to groups of max 4 batteries so if you race and one or two group dies on you the pack still keeps the injection etc going so you dont have to break the race as you would if the pack was one big unit.

I love theese packs and will keep using them.
 

·
Registered
Joined
·
90 Posts
Discussion Starter #9
Yes I have too much time, and thanks for the support :)



The reason that starting and charging issues are common has more to do with poorly built / situated regulators and undersized wiring than it does with the starter motor or the battery.
I agree, it seems that a new approach to the reg/rec is part of this. That's what I was getting at with the buck/boost circuit. Something that can regulate the charge and make DC at various voltages, and not be mounted next to an exhaust header!

While I'm all for the DIY approach, the risks of damaging the bike by using the alternator as a starter motor are not appealing. I'm not even sure you could put enough current though the alternator without frying it, to start the engine.
The only damage that I would worry about is spinning the flywheel off the crank or something like that, but the start torque is being transmitted to the flywheel hub currently, so this may not be an issue, I'll just go with a double nut system and thread locker. I'm not worried about burning the coils, they can be rewound (maybe with some improvements, more turns or a different pattern). I will most likely buy a used one to experiment on.

There is quite a bit of gear reduction going on from the starter motor (30:1). I'm thinking that you will simply not have enough torque to do it reliably without burning out the armature the way you are planning. The wiring in the stock alternator is only rated for something like 40A, and that's across the three wires. Look at the gauge on the starter lead, and compare it to what comes out of the alternator, and remember it takes almost 60A *with* reduction to crank the engine.

Mathematically, if the starter consumes 700W, that's roughly 1 HP. 1 HP ~ 2 ft lbs. / second of torque (at a one foot radius), at a guess of 2500 rpm (what speed the starter turns). With a 30:1 mechanical advantage, it now takes about 60 ft lbs. / second at 83 rpm and a one foot radius to overcome static compression. At 10cm the flywheel would need about 6 times that amount (5cm radius), or 360 lbs. / second exerted on it to have the same effect. That would be close to 6 HP, or around 4400W. If you can get 360A through the alternator, you are my hero.

But, good luck. I'm waiting to see the results, and I hope I'm wrong.
I definitely have considered the gear reduction that the OEM starter uses, and wonder if I will be able to make enough torque to spin a large displacement high compression motor on a freezing cold morning :) If it can't do that, then it isn't a starting system.

For the weight savings alone it is worth messing with. The cost per gram will destroy almost any other mod. The batteries are a very safe chemistry, there is no worry about violent thermal events due to overcharge/discharge or physical damage. I would rather have these cells near me in a crash than lead acid or gasoline, for that matter. I'm going to start with a $70 investment that will do a much better job than the lead acid of the same price, and do it for 2.75kg(6lbs) less. Maybe I'll end up adding a BMS to lengthen the life of the cells, this will add a little more money(~$50) and very little weight(<100g).

The second part of this project is obviously more complicated, I am going to try and make it work and will have fun in the process. Maybe ditching the starter won't be worth it for commuter/tourer oriented bikes, but loosing that extra weight, including weight in the power train (engine sensors and accounting for removed components with spacers will need attention) would be worth it to many. The cost for a brushless motor control should be under $200, just to put a rough figure on it. So let's round it up to $500 for a total system, battery, motor control, charge/reg circuits that saves big weight and introduces reliability. To me that is worth dicking around in the shop, maybe melting some wires here and there.

Thanks again for all the input folks, I'm ordering parts now, so when I have new info I will share it.
:abduct:
 

·
Registered
Joined
·
90 Posts
Discussion Starter #10
+1 lechef.

rapter, it sounds like something is wrong with your pack or charging system, did you contact Speedcell? Those packs should work for years, even the 4.6ah should give you many starts before running down. Convert the capacity to amp minutes (4.6ah*60=276am), you don't want to completely drain any battery, so just figure half this capacity to be safe, that means (276am/2=138am), if your starter uses 60amps then you could power it for over two minutes(138am/60a=2.3minutes) not continuously of course, you would over heat the motor, the pack could handle it though.
 

·
Registered
Joined
·
71 Posts
Photon

I look forward to your results and think its a neat idea. I dont know if it will work or not but if we avoided trying something new just because someone said "it'l never work", then I guess we would still be walking everywhere and hunting with stones. Whether it works or not it should be a fun project.
 

·
Registered
Joined
·
97 Posts
Great stuff Photon - I killed about 3 hours last night trying to figure out how to build myself a LifePo4 pack on the cheap - Looks like you're doing it. Thanks for explaining the 'C' rating - That had me puzzled.

So if I only wanted to replace my lead-acid battery and leave the alternator intact, would I need a PDB to manage charge/discharge?

If you can say, where did you buy your threaded-end cells?

thanks
 

·
unM0derator
Joined
·
2,853 Posts
The only damage that I would worry about is spinning the flywheel off the crank or something like that, but the start torque is being transmitted to the flywheel hub currently, so this may not be an issue, I'll just go with a double nut system and thread locker. I'm not worried about burning the coils, they can be rewound (maybe with some improvements, more turns or a different pattern). I will most likely buy a used one to experiment on...I definitely have considered the gear reduction that the OEM starter uses, and wonder if I will be able to make enough torque to spin a large displacement high compression motor on a freezing cold morning :) If it can't do that, then it isn't a starting system.
I think you'll be OK with the stresses that are placed on the flywheel and crank end. Remember that the flywheel's main purpose is to store kinetic energy to overcome compression and keep the engine running under no load. In this respect the flywheel has to absorb and re-transmit as much torque as it takes to keep the engine running. Also, when you accelerate and engine brake, the stresses go many times higher than an idle condition. The amount of torque necessary to start the motor should not pose a problem here. The maximum output of the 996 motor is nearly 90KW after all.

But, what I was getting at with damage was more to the vehicle's charging system. You would 100% have to replace the armature of the stock alternator with one that would allow you to run the high amperage current necessary to have the torque to start the motor. I think, however, that this will be space prohibitive, and without some other gear reduction you simply can't cram that much winding into the space you have. The dimensions of the electric motor you are trying to create will most likely also add a few pounds of metal back into the system with increasing the girth and carrying capacity of the coil.

Take, for example, a state of the art pancake motor that will run a small motorcycle: something like the Manta PMG226. It has a rating of 10 HP at 48V, and a maximum inrush of 480A for 45 seconds. This is probably a bit bigger than necessary for this application, but it's in the ballpark. This motor will produce 0.1 ft lbs. per Amp. So to get the required 60 ft lbs to start the motor is 600A. Obviously, with this setup, gear reduction is required, since the nominal current exceeds the maximum inrush. The motor also runs at 72 rpm per volt, so with a 2:1 reduction, you could get by with 2V, 300A (the motor is designed to run in these ranges) which will be closer to 700W, for the same 1 HP required. Again, this motor is close to the ballpark of what we are talking about.

Now, what are the dimensions? Weight 22 lbs. - 8"X 9"X 9"

A flat motor needs significant radius to generate torque, whereas a starter, or cylindrical motor, uses windings along the length of the motor or gangs of windings to generate the same torque in a smaller radius. The flywheel, having a small radius and surface area makes the input requirements grow geometrically. As seen previously, a 10cm diameter flat motor needs 6 times as much energy to generate the same torque as the 1HP starter and 30:1 reduction.

Put simply, there are engineering reasons why the starter motor is the preferred solution to the problem. Trust me that it has been refined over 100 years, being about as light as possible for the requirements.

Of course, with new materials and higher-density batteries, a solution like you are presenting might be possible. However, I don't think we're there yet, and the cost will be pretty high (think room temp. superconductors to get the right torque / sq in).

Another interesting approach would be to use something similar to an older-type spring-style starter. You could capture the kinetic energy on shutdown as store it as potential in a coil spring. When starting the motor, you simply unload the spring. The downside is you'd only get one shot at starting the motor. A light DC motor with gear reduction could be employed to recharge the spring, at the cost of the time it takes to input that much energy (maybe around 30 seconds). It would certainly save weight, but most people might consider it a nuisance to wait for the coil to recharge if the engine didn't fire the first time.

Ideally, I think what you could do for the parameters of this situation, is to use a gear driven flywheel, that can use a mechanical clutch to disengage from the main drive and turn a reduction cascade when driven directly. This would entail coming up with a new geartrain to fit to the crank end, and connect to the original starter gear. Made from high-strength and lightweight materials, you might eek out a net savings in weight. But, now you'd have to accommodate lubrication into the system.

Keep moving forward, though.
 

·
Registered
Joined
·
90 Posts
Discussion Starter #14
So if I only wanted to replace my lead-acid battery and leave the alternator intact, would I need a PDB to manage charge/discharge?

If you can say, where did you buy your threaded-end cells?

thanks
Ideally a battery management system would be good for any battery, the LiFePO4 chemistry seems to do fine with low level lead acid chargers, that is why I am willing to risk some cells by trying the simplest approach first. The A123 cells seem to tolerate the Ducati charge system. On the discharge side these cells should not have a problem powering an OEM starter, just put a big fuse and use high quality heavy gauge wire and connectors.

With any series string of cells it is best to keep the cell voltage equal among them, that is part of the job of a BMS. With only four cells it would not be that hard to check and balance them every now and then, or they might last for years without doing a thing.
I would weigh the expense, as long as the parts aren't costing a fortune, then I don't mind testing to failure, but if another $50 spent on a BMS will gain me more years of use, out of one $70 pack, then that's easy math. If you want to be safe wait until l have some results using the cells with no BMS.

evcomponents.com is who I got these cells from, they are made by Headway(Chinese)#38120S 10AH, but check around there may be other good choices out there.
 

·
Registered
Joined
·
90 Posts
Discussion Starter #15
SE84,
I do appreciate what you are saying, and I assure that I do not imagine this will be a simple goal, but the idea of removing an entire mechanical system from a machine and getting one component to do two jobs is engineering gold.

Did you calculate the 60ft/lbs to initiate rotation figure from the 700w load the OEM starter uses? Because only some portion of that gets to the crankshaft. I have spun my motor by grabbing the flywheel and turning it with my hands, it is not easy, but I don't think it requires 6hp. We need to compress 500cc of air 12 to 1, pump 500cc in or out through the valves, overcome the inertia and friction of the engine parts(very light valve springs, thanks desmo), spin the water and oil pumps(at slow speed). All for the grand total of one or two seconds, and that is why this has a chance of working, ultra low duty cycle. Here are some pictures of a bicycle hub motor

This is the heavy duty stator that you are talking about, I have put 3kw through this motor. The big difference is the duty cycle, a motor needs to be very robust to power 300lbs up hill for long durations. A starter motor may need to put out a lot of power, but it is only for a very short time.

As for the diameter of the coil wires, look at the wire on the hub motor stator compared to the Ducati, those wires have had 3000watts run through them-

Even though the conductors are small, they are very long. The heat is being shared by a lot of wire. Maybe I'll add a little bit of aluminum to sink some heat from the stator, it could also function as a skid plate for the alternator cover. All those little refinements will come, the motor timing can be tuned for the job. The job of a starter motor is short and defined, it does not have to cover a wide rpm or load range. This makes fine tuning and optimizing much easier.

Speaking of timing, notice the three Hall effect sensors that are epoxied around the rim of the hub motor stator-

Each of the sensors has three small wires, these are run to the brushless motor control, which uses this info to time the three phases, these could also provide an rpm signal to the engine management computer, as the stock rpm sensor reads the starter gear teeth, and hopefully that gear can be removed.

Long story short, this project is about starting a superbike not a BMW GS that tours the arctic. I like the idea of reducing down a machine to a simple pure state, and most importantly not being stuck doing things the same way they were done yesterday. If the starter motor is removed space will be freed up to move other components around. If I build a new reg/rec then it will be possible to try things like running a higher voltage to the coils, etc...
 

·
Registered
Joined
·
511 Posts
You guys are pretty,........electrical. I have a hard time with the light switches on the wall, and electrical plugs.


Good luck.
 

·
unM0derator
Joined
·
2,853 Posts
Did you calculate the 60ft/lbs to initiate rotation figure from the 700w load the OEM starter uses? Because only some portion of that gets to the crankshaft. I have spun my motor by grabbing the flywheel and turning it with my hands, it is not easy, but I don't think it requires 6hp. We need to compress 500cc of air 12 to 1, pump 500cc in or out through the valves, overcome the inertia and friction of the engine parts(very light valve springs, thanks desmo), spin the water and oil pumps(at slow speed). All for the grand total of one or two seconds, and that is why this has a chance of working, ultra low duty cycle. Here are some pictures of a bicycle hub motor
I did calculate it from the starter motor draw. Your statement that all the power doesn't get to the crank is slightly misconstrued. Over 98% of the power transmitted is available at the crank, but the necessary torque to start the motor might be lower than the 60 ft lbs. But, since I have encountered hard starting on my 2002 ST4s, I think at times you will need every last bit to get the motor to crank. Remains to be seen what the actual value may be.

This is the heavy duty stator that you are talking about, I have put 3kw through this motor. The big difference is the duty cycle, a motor needs to be very robust to power 300lbs up hill for long durations. A starter motor may need to put out a lot of power, but it is only for a very short time.
OK, from your first picture of the 16 cells for the bike, I would assume that this is the entire kit. If you made 4 packs @ 13.2V in parallel, you would have 40Ah to work with. If the bike is meant to be ridden around for 1 hour, that means that the avg through the armature is only 528W, or 0.7 HP. This would follow why the wiring is similar in gauge to the alternator leads. What is the max current that this motor is rated for? Just because you can put 3KW through the motor, doesn't mean it will tolerate it forever. Heat-cycling the wires will oxidize them and increase resistance, which is a positive feedback loop, until you turn them into just heating elements, and something catches fire. Duty cycle is still considered to be under the max inrush, and when you exceed that, you have no meaningful metric.

As for the diameter of the coil wires, look at the wire on the hub motor stator compared to the Ducati, those wires have had 3000watts run through them-

Even though the conductors are small, they are very long. The heat is being shared by a lot of wire. Maybe I'll add a little bit of aluminum to sink some heat from the stator, it could also function as a skid plate for the alternator cover. All those little refinements will come, the motor timing can be tuned for the job. The job of a starter motor is short and defined, it does not have to cover a wide rpm or load range. This makes fine tuning and optimizing much easier.
Where you should be more concerned is the diameter of the wires, not as much the length. Thinner wires have more resistance, and will heat-cycle more readily.

Long story short, this project is about starting a superbike not a BMW GS that tours the arctic. I like the idea of reducing down a machine to a simple pure state, and most importantly not being stuck doing things the same way they were done yesterday. If the starter motor is removed space will be freed up to move other components around. If I build a new reg/rec then it will be possible to try things like running a higher voltage to the coils, etc...
Whether you are starting a 996, a big block Chevy, or a weed-whacker, the laws of physics don't change. I still stand by my earlier assessment that if you try to put that much current through the stock alternator, that it will eventually short, or catch fire. At a minimum, you are adding stress to the armature wiring, that also serves as the coil that charges your bike, possibly compromising charging ability. At the worst, I hope you will have a fire extinguisher handy in all cases. There was just a thread around a couple days ago about a guy with a SS who almost caught fire because the R&R shorted.
 

·
Registered
Joined
·
90 Posts
Discussion Starter #18
OK, from your first picture of the 16 cells for the bike, I would assume that this is the entire kit. If you made 4 packs @ 13.2V in parallel,

16s 53v 10AH is how that pack will be configured, that motor won't be run at 3kw, my point is that it can take that much power for a lot longer than two seconds.




Whether you are starting a 996, a big block Chevy, or a weed-whacker, the laws of physics don't change. I still stand by my earlier assessment that if you try to put that much current through the stock alternator, that it will eventually short, or catch fire. At a minimum, you are adding stress to the armature wiring, that also serves as the coil that charges your bike, possibly compromising charging ability. At the worst, I hope you will have a fire extinguisher handy in all cases. There was just a thread around a couple days ago about a guy with a SS who almost caught fire because the R&R shorted.
I bring up the BMW GS not because I think the laws of physics have changed, but to illustrate the different design parameters between a starter system for a big bike that goes into the middle of nowhere and a superbike . In order to reduce weight and simplify some compromises in function are often made on race machines. However, be assured, no one is going to be running any current through your stator, but mine just might get melted and then my whole bike might also burn up, and then you can rest easy knowing that you were right all along. But I will just build a new bike even better than before. So what is the problem?
 

·
unM0derator
Joined
·
2,853 Posts
I bring up the BMW GS not because I think the laws of physics have changed, but to illustrate the different design parameters between a starter system for a big bike that goes into the middle of nowhere and a superbike . In order to reduce weight and simplify some compromises in function are often made on race machines. However, be assured, no one is going to be running any current through your stator, but mine just might get melted and then my whole bike might also burn up, and then you can rest easy knowing that you were right all along. But I will just build a new bike even better than before. So what is the problem?
No problem. I'm just as excited to see you fry your bike as you are. ;)

In all seriousness, I think what you're doing is a good learning exercise, and there might be some good surprises to come from it.

After all, it wouldn't be science if you don't test the theory.
 
1 - 20 of 20 Posts
Top