/elec/propeller-clock : contents of notes at revision 91

: (revision 91)

To get this branch, use:

bzr branch
http://bzr.ed.am/elec/propeller-clock

DISPLAY

The display is split up in to seconds, each with 5 subdivisions
(segments).  That's a total of 300 segments per revolution.

If the propeller spins at 2000RPM, that's 33.3 revolutions per second,
or 30ms (30,000μs) per revolution.  That means we'll be drawing 10,000
segments per second, which is 100μs per segment.  With a clock speed
of 16MHz, this is 1600 cycles per segment, which is plenty.


SCEMATIC NOTES

The diode (D14) across the fan's power connections is there because if
the power across the fan breaks (due to the unreliable nature of the
brushes), the motor in the fan has coils, which act like an inductor
and will produce a back EMF (a huge negative voltage across the power
connections) as the magnetic field collapses.  This won't be good for
the Arduino and could cause sparks on the brushes.  The diode simply
shorts the negative voltage.

The capacitor (C1) and resistor (R14) are there to smooth the power
supply from the unreliable brushes.  The capacitor would discharge
fairly slowly (due to the resistance of the circuit), but will charge
very quickly.  Potentially, it will charge so quickly that it'll pull
too much current from the power supply (i.e., short the power supply
and trip it).  So the resistor limits this.  Unfortunately, the
resistor will also have a potentiometer effect (with the resistance of
the main circuit).  10Ω was chosen as a value due to these rough
workings: Lets say the Arduino circuit takes 500mA.  If we aim to lose
1V across the resistor, that's 1V / 0.5A = 2Ω (from V=IR).  The 100μF
was a guess (from Dad), but "PCB" Mat suggested something larger, like
2200μF.  So we went with 1000μF, which appears to power the board
after power-off for a couple of revolutions.  The factors here are
that a capacitor that is only able to hold a small charge won't be
able to maintain a current for a reasonable amount of time when the
power breaks.  If it's too large, it will take ages to charge and
effectively short the power (save for the resistor) while it does.

39 by edam updated notes and scematic (adding capacitor+resistor and diode)	1	DISPLAY
	2
	3	The display is split up in to seconds, each with 5 subdivisions
	4	(segments). That's a total of 300 segments per revolution.
	5
	6	If the propeller spins at 2000RPM, that's 33.3 revolutions per second,
	7	or 30ms (30,000μs) per revolution. That means we'll be drawing 10,000
	8	segments per second, which is 100μs per segment. With a clock speed
	9	of 16MHz, this is 1600 cycles per segment, which is plenty.
	10
	11
	12	SCEMATIC NOTES
	13
	14	The diode (D14) across the fan's power connections is there because if
69 by Tim Marston cleaned-up notes	15	the power across the fan breaks (due to the unreliable nature of the
39 by edam updated notes and scematic (adding capacitor+resistor and diode)	16	brushes), the motor in the fan has coils, which act like an inductor
	17	and will produce a back EMF (a huge negative voltage across the power
	18	connections) as the magnetic field collapses. This won't be good for
69 by Tim Marston cleaned-up notes	19	the Arduino and could cause sparks on the brushes. The diode simply
39 by edam updated notes and scematic (adding capacitor+resistor and diode)	20	shorts the negative voltage.
	21
	22	The capacitor (C1) and resistor (R14) are there to smooth the power
	23	supply from the unreliable brushes. The capacitor would discharge
	24	fairly slowly (due to the resistance of the circuit), but will charge
	25	very quickly. Potentially, it will charge so quickly that it'll pull
41 by edam modified schematic and updated notes	26	too much current from the power supply (i.e., short the power supply
41 by edam modified schematic and updated notes	27	and trip it). So the resistor limits this. Unfortunately, the
69 by Tim Marston cleaned-up notes	28	resistor will also have a potentiometer effect (with the resistance of
41 by edam modified schematic and updated notes	29	the main circuit). 10Ω was chosen as a value due to these rough
69 by Tim Marston cleaned-up notes	30	workings: Lets say the Arduino circuit takes 500mA. If we aim to lose
41 by edam modified schematic and updated notes	31	1V across the resistor, that's 1V / 0.5A = 2Ω (from V=IR). The 100μF
	32	was a guess (from Dad), but "PCB" Mat suggested something larger, like
	33	2200μF. So we went with 1000μF, which appears to power the board
	34	after power-off for a couple of revolutions. The factors here are
	35	that a capacitor that is only able to hold a small charge won't be
	36	able to maintain a current for a reasonable amount of time when the
	37	power breaks. If it's too large, it will take ages to charge and
	38	effectively short the power (save for the resistor) while it does.