Propellor clock question

[jerryd]

Electro Tech forum,

I'm starting to work on a propellor clock, an old
project I know but new to me.

I'm using mosfets to drive 5mm 3000mcd red leds with
an interrupt driven 16F88. The motor runs at 1725 rpm
so doing a little trig it looks like I'll have to have
them on for about 100-150us.

Experimenting on a breadboard turning them on works
great but it seems to take them a long time to turn off.
I'm not using a current limiting resistor since the on time
is so short. Maybe my approach is all wrong.

There are many web sites on this subject but none get
into this kind of detail.

Anyone have any ideas on this?

Thanks,
jerryd

 

[crutschow]

You still need a current limiting resistor.

How are you driving the MOSFETs? Their gates can have a lot of capacitance and you need a driver that can rapidly discharge that capacitance to achieve a fast turn-off.

 

[jerryd]

Crutschow,

I'll use a 100 ohm current limiting resistors for the LEDs.

I'm driving the MOSFETs directly off one of the pic ports.
No series resistor to the gate. Vdd is 5V.

Coss is 100pf, Ciss is 150pf.

jerryd

 

[crutschow]

So how slow is the turnoff?

 

[OlPhart]

Are you going to put "stuff" on a blade at 1725 RPM? The angular momentum will make life Very Interesting, let alone the balancing tolerances.. My message unit is added to a ceiling fan ~180 RPM on high. This also makes the LED switching times a Lot more manageable. Good Hunting, the software is "interesting"... <<<)))

 

[KeepItSimpleStupid]

You always have to limit the peak current to the LEDs.

 

[jerryd]

Electro Tech,

OlPhart:
I can see where a slower motor speed would solve
lot of my problems.

Your message fan proves that and looks great.

I couldn't find anything around my house or on
the internet that would work.

Any suggestions?

jerrd

 

[carbonzit]

This guy made a propeller clock from an old VCR/floppy drive motor. Not sure how fast it spins, but I'll bet faster than 180 RPM.

 

[Diver300]

Experimenting on a breadboard turning them on works
great but it seems to take them a long time to turn off.

You should short out the LEDs to make them turn off quickly.

If you are driving them from a logic output that drives high and low, they will be shorted out. If they are wired to turn on when the logic output is high, then when the output drives low they will be shorted out and they will turn off quickly.

If you only turn the current to them on and off, any capacitance on the LED lines will cause the light to fade rather than stop.

 

[crutschow]

If you only turn the current to them on and off, any capacitance on the LED lines will cause the light to fade rather than stop.

There is not enough capacitance on the LED lines (for any reasonable line length or type) to add a noticeable delay to the LED turnoff. If there is a delay it is from the driver electronics.