Is the rectifier/regulator really as crude as it looks?

[Some Dude]

Any progress on this? 

[jerryz]

the ducati regulator is also used by many Hondas  and has issues with them too just look at VFR  forums.

[suzyj]

No, none, save a few simulations.  I'm sorta tied up playing with suspension (sbk forks) at the moment - I'll probably return to things electrical in a few months.

[Duck-Stew]

Don't know how the hell I missed this the first go-around, but heck yeah....I'm interested! 

[mitt]

Definitely a cool project.

mitt

[suzyj]

Okay, this is the sort of thing I'm playing with.



The issue with A123 cells (or any LiFe cells) is voltage control and charge current control.  According to the datasheet, to get good cycle life, one doesn't want to go above 3.45V float (13.8V for four cells), and shouldn't charge at more than 3A (6A for a 4s2p pack).

The only way I can think of to do that and maintain high current capacity for the rest of the bikes electronics is to have two separate DC-DC converters, one for the battery, and one for the loads.  I'm liking the Linear Tech LTC3703 controller, as it's able to drive big MOSFETs and will run to 100V input, plus incorporates current limiting, which is important for setting charge current.  The circuit I'm running sims on is the same as the application note circuit, but I'm using Infineon IPD122N10 MOSFETS and Vishay IHLP-6767GZ 10uH inductors, to significantly boost maximum current.  I'm using the same circuit (except for the current set resistor) for both battery charge and load supply.

So the principle is as follows.  The six diodes on the left rectify the AC from the alternator.  The zener/MOSFET across the HV supply is just there to protect the following circuitry from overvoltage - it'll shunt when the HV voltage gets to 90V.  Then both DC-DC converters are in parallel.  The one charging the battery goes across the battery, with balancing circuits across each cell.  The one for the loads goes straight out.

A simple diode from the battery + to the loads allows the loads to function before the bike is started.  Also, if the loads draw more than 12A, the charging converter may also source current to the loads via the diode, up to a max of 18A or so.

What are people's thoughts?  The only part that'll need heatsinking is the protection circuit, as in the event of overvoltage it may have to dissipate significant power.  Otherwise everything is tiny.  The inductors are 17mm square, the MOSFETs are TO252 (small), etc.  The whole shebang will fit on a circuit board around the same size as the existing rectifier/regulator, but only ~12mm thick.

[mitt]

So, does the ducati alternator not work like a modern car where the regulator adjusts the rotor field current to produce the only power needed? 

mitt

[suzyj]

So, does the ducati alternator not work like a modern car where the regulator adjusts the rotor field current to produce the only power needed? 

mitt

No, the alternator has permanent magnets in the rotor, and three-phase windings in the stator.  There's no way to adjust field current - the only thing limiting output power is saturation in the stator.

[mitt]

No, the alternator has permanent magnets in the rotor, and three-phase windings in the stator.  There's no way to adjust field current - the only thing limiting output power is saturation in the stator.

That is crude, but maybe they saved a few pounds by using magnets instead of a rotor winding and claws?


mitt

[suzyj]

That is crude, but maybe they saved a few pounds by using magnets instead of a rotor winding and claws?

Yes, it is.  In their defence though, the application isn't like that in a car, where there are lots of things turning on and off and drawing current (stereo, fans, headlights, etc.)  The load for a motorbike is relatively constant, so you can get away with a much simpler setup.

[Howie]

The rotor and stator also are an integral part of the engine, saving a lot of space and eliminating the need for a drive system.  All you really (IMO) loose the low speed efficiency since you cannot control field or run the alternator at higher than engine speed, which would not be a good idea on a sport bike engine anyway.   

[sbrguy]

now you want to use this with a speedcell type battery, how will this setup you are proposing work for a reguarar sealed stock battery that came with the bike? and what if anything can you do to get the packaging even smaller than the stock one?

[Howley]

And will it work for non three phase systems?

[suzyj]

It'll work okay with the stock lead acid battery, but it's overkill for that application.  You don't really gain anything over the standard circuit (well maybe you'll get a cooler bottom), for quite a bit of effort.  The PCB dimensions ended up 102mm long (4") by 65mm wide (2.5").  I designed the board to fit neatly alongside an A123 pack, with the battery connections in convenient places to connect to the cells.

I can't see why it wouldn't work on single phase alternators - I went a little overboard on the rectifier diodes, and used MUR1620s.  These come in a TO220 package and are able to pass 16A each, so it will be okay for single phase as well as three.

In any case, after a couple of late nights, here's an image of the PCB (a crappy screen grab from my CAD tools):



The rectifiers are down the left hand side - next along is the balancing circuitry, then the two DC-DC regulators.  The largest parts on the board are the inductors for the converters.  Finally on the right are the pass diode to allow the battery to power the load with the bike not running, and the overvoltage protection FET.  The loads connect to the right hand side.  The solenoid/starter would bypass this board and connect directly to the battery.

I haven't added up the cost for all the parts yet, but I'm thinking something like $200 per unit in modest quantities.

If anyone wants the PCB and schematic files (I use Protel), shoot me a PM and I'll email them to you.

[erkishhorde]

To answer the first question, yes, the stock regulator sucks. That's why CA-cycleworks doesn't sell it and sells stuff by these guys instead. Ah the memories. Nothing like having your bike catch on fire when the stock voltage regulator fails while riding through rush hour traffic. It's so nice when traffic parts for you even if it is because they're afraid you're bike is going to blow up and you're enveloped in smoke.

http://www.electrosport.com/street/street-models-ducati.php#Anchor-DUCATI-47857