Microcontroller based voltage regulator

[christensen]

Hi,
I'm planing to build a new voltage regulator for my bike based on microcontroller.
What i'm expecting from it:
- cut the rotor power if the engine rev is below 600rpm.
- make it more reliable, faster, low lost and precise.
- easy diagnostic/communicate with dashboard instrument panel thru SPI(Serial Peripheral Interface).

The classic regulator is a linear, i don't want a linear because of the power lost on the power transistor for that i was thinking to use a puls PWM(Pulse-width modulation)for rotor coil power.
The frequency will be the same(at this moment i don't know .. maybe 200hz .) but i vary the duty cycle and i control the current. on the power transistor will be a small amount of power lost which will be caused by the internal resistance of the transistor.

The question: how will affect the pulse control of rotor the bike voltage stability after rectifier bridge?i should use an RC snubber for the pulse which work at the same frequency with the coil pulse? what should be the optimal frequency for powering the rotor coil?

thanks

[kutt]

for 6 or 12V?

[christensen]

for 12V but it doesn't matter in the software is needed to make some modification and add a LDO(Low-dropout regulator) for the microcontroller. the normal 5v stabilizer(7805) need at least 2V higher input voltage than output. or can be changed the microcontroller with a low power one(3V).
Atmega 8 (4.5...5.5V).
Atmega 8L(2.7..5.5).
With lower voltage than 5V could be problems with the REV pick-up senzor(hall sensor which usual don't work under this value). the point pick-up isn't good for REV count because the generated noise.

[net-harry]

Hi Christensen,

In my opinion I see no need to use a microcontroller for the simple task of regulating the voltage in a motorbike...

But, a few weeks ago, I have a discussion via PN with a member (xs-wolle) of this great forum about a switching regulator used in a XS (Yamaha ?) bike. In the XS-forum the simple circuit I append below was posted. The green cable is connected to the excitation winding, 12V is connected to the accumulator and, of course, the output of the stator winding(s) is/are rectified and connected to the accu, too)

The only (I guess, because I dont test the circuit) problem with this circuit will be that the "engine-start-without-batterie-function" used in the MZ is no longer available.
In this case the input voltage of the regulator is to low and regulation won't start.

Greets
Harald

[lothar]

I'm planing to build a new voltage regulator for my bike based on microcontroller.

Nice idea.

What i'm expecting from it:
- cut the rotor power if the engine rev is below 600rpm.
- make it more reliable, faster, low lost and precise.
- easy diagnostic/communicate with dashboard instrument panel thru SPI(Serial Peripheral Interface).

I think you should better cut the rotor power if generated voltage by maximum rotor current goes under board
voltage minimum limit. It is not only dependiung from the minimum rev limit. Important in my opinion is
to pay attention to the temperature influenced optimum charge voltage of lead batteries.

The classic regulator is a linear, i don't want a linear because of the power lost on the power transistor for that i was thinking to use a puls PWM(Pulse-width modulation)for rotor coil power.
The frequency will be the same(at this moment i don't know .. maybe 200hz .) but i vary the duty cycle and i control the current. on the power transistor will be a small amount of power lost which will be caused by the internal resistance of the transistor.

About 1990 MZ already used this method by applying the IC L9480. This IC is out of production since years. Unfortunately
a similar simply appliable follower does not exist. This is the chance of your project ...
I have got one IC and tried to investigated its behaviour within an ETZ150. See here:
http://pic.mz-forum.com/lothar/ELEKTRIK ... -L9480.htm

The question: how will affect the pulse control of rotor the bike voltage stability after rectifier bridge?i should use an RC snubber for the pulse which work at the same frequency with the coil pulse? what should be the optimal frequency for powering the rotor coil?

Applying this IC they recommended a capacitor of 0,1 ... 1 uF at 51 although the circuit has internal filters.

In this case the input voltage of the regulator is to low and regulation won't start.

Harald, we talked about a solution while having a cup of coffee - you remember ... ?

Regards
Lothar

[christensen]

Thanks to both!

Harald, i need microcontroller because it is flexible, can communicate with my other project (digital dashboard), can be found everywhere and is cheap. if for some reason this model of microcontroller is not longer build by Atmel i can choose another and rewrite the software.
The "engine-start-without-batterie-function" i can solve simple using a switch or a jumper and bypass with a relay(NC), when the uC boot than release the relay contacts and the driver transistor take the relay control (but this is not my concern, i always have a battery and if not than i can use a laptop battery to start).

Lothar,
I've read your description about L9480 and thanks a lot, almost all was clarified, the frequency will be around 400hz and for the filtering i will use a capacitor 0.1 ... 1uF for noise filtering.

i forgot to mention the whole list of feature:
- cut the rotor power if the engine rev is below 600rpm.(I use a 4Ah battery for space reason, see the picture and i need to make economy with the power especial when 2 hungry consumer are powered: Ignition coil and rotor coil, so i prefer to keep as low is possible the power consumption, because this i've change the bulbs to LED).

moto.jpg

- make it more reliable, faster, low lost and precise.
- easy diagnostic/communicate with dashboard instrument panel thru SPI(Serial Peripheral Interface) (send an error message to the dashboard instrument panel if: voltage is to low, to high, no power on the regulator, communication problem on SPI between 2 microcontrollers, rotor coil drag out too much current ).
1. in case of voltage to high the microcontroller will try to decrease the duty cycle till reach the charging voltage if this isn't possible cut the rotor power.(in the dashboard instrument panel will be another voltmeter which will warn me if the voltage is still to high and maybe the coil driver transistor is feeding the coil constantly).
2.in case of rotor coil drag out too much current, than the uC will cut the power from it and warn me.