Subversion Repositories svnkaklik

EQ6 Mount controller analysis (v1 - 10 Jan 2004)

Darren Hutchinson
dbh@gbdt.com.au

$Id: eq6-mnt.txt,v 1.3 2004/03/07 07:44:59 dbh Exp $

Overview:

CPU:            AT87F51
Clock speed:    3.52 MHz (measured at pin 18 of CPU)
Firmware:       MotorDrive DW3505010

RA connector
1       RA coil A       Driven RA=2,8,16x
2       RA coil C       Driven RA=2,8,16x
3       RA coil A       Driven RA=1x, 22R to 1, 178R to 4
4       RA coli B       Driven RA=Any, 157R to 1, 178R to 3
5       RA coil C       Driven RA=1x 173R to 6, 155R to 2
6       RA coil D       Driven RA=Any 173R to 5, 22R to 2

DEC connector
1       DEC coil A      A/B are coil with 16R3 resistance
2       DEC coil B
3       DEC coil C      C/D are coil with 16R3 resistance
4       DEC coil D

Drive waveforms:

RA 0X:  Coils not driven
RA 1X:  Driven +12V for 114 ms, 0V for 114 ms, not driven for 38ms b/w
        driven states, overall time is 302ms [OSC only accurate to 2ms]

        Each coil driven at 90deg to other coil.

        Timing correct for half step mode

RA 2X:  Driven +12V for 75ms, 0V for 76ms, overall time is 151 ms, each coil
        at 90deg to other coil
RA 8X:  Overall period is 38ms, no undriven time, 
        coils at 90deg
RA 16X: Overall period is 12.8 ms, no undriven time, coils at 90deg

- Timing indicates full step mode
- No evidence of velocity ramp in x2, x8, x16 mode
- RA 1X timinig is correct assuming 180 tooth worm and 1:132 transfer gear
  & gearbox reduction ratio

DEC 0X: Outputs not driven
DEC 2X: Overall period 151 ms, no undriven time
DEC 8X: Overall period 37.8ms, no undriven time, coils at 90 deg
DEC 16X: Overall period 12.6ms, no undriven time, coils at 90 deg

- Timing indicates full step mode
- No evidence of velocity ramp in x16, x8, or x2 mode

Drive circuit:

The drive circuit is a fairly plain "H" driver with one quirk that may, or
may not, be intentional design from Synta.

The half-"H" is built from a pair on PNP/NPN transistors operating in
common-emitter mode. This is somewhat unusual as circuits intended to
drive high currents typically operate in common collector mode.

Anyway, the PNP transistor that pulls the output high has a base current of
about 11 mA (depending on the supply voltage). With a transistor hfe of
40 this should give an output current of at least 500 mA.

The other part of the bridge is another issue. The identical circuit is used to
drive the base of this NPN transistor, but in this case it will only give
the transistor 1.2 mA, leading to a lower current. 

The rated hfe of the transistor only guarantees a current of about 50 mA, but
the hfe is typically much higher. In the unit measured the coil current seemed
to be limited to about 250 mA.

This may be an attempt to limit the current in fault conditions where
both the NPN and PNP transistors are active, or it may just be bad circuit
design.

In any case there are two practical effects:

First the transistor may not be saturated, so it will run hot.
Second the coil current is limited, giving less torque.

Both of these effect mean that the existing electronics would be
a poor choice as the basis of a GOTO design (but I guess you knew
that ....)

MCU Pin Label   Use
1       P1.0    1 = RA at 1x, 0 = RA != 1x
2       P1.1    Same as P1.0
3       P1.2    Same as P1.0
4       P1.3    Same as P1.0
5       P1.4    Same as P1.0 but connected to relay (1 = relay off, 0 = on)
6       P1.5    Same as P1.0
7       P1.6    Same as P1.0
8       P1.7    Same as P1.0
9       RST     R/C reset circuit, C = 22uF, R = 5K1, active high
10      RXD     Data from controller via 1K resistor
11      TXD     No data
12      INT0    0V
13      INT1    +5V
14      T0      +5V
15      T1      +5V
16      *WR     +5V
17      *RD     +5V
18      XTAL    xtal out (freq meas point)
19      XTAL    Xtal in
20      GND     0V
21      P2.0    1 = RA coil A to +12v, 0 = No effect
22      P2.1    1 = RA coil B to +12v , 0 = No effect
23      P2.2    1 = RA coil C to +12v , 0 = No effect
24      P2.3    1 = RA coil D to +12v, 0 = No effect
25      P2.4    1 = DEC coil A to +12v, 0 = No effect
26      P2.5    1 = DEC coil B to +12v, 0 = No effect
27      P2.6    1 = DEC coil C to +12v, 0 = No effect
28      P2.7    1 = DEC coil D to +12v, 0 = No effect
29      *PSEN   +5V
30      ALE     Toggle at 602 KHz (high 540 ns, low 1.1 us) [~6 clk / ALE ]
31      *EA     +5V
32      P0.7    1 = No effect, 0 = RA coil A to 0V
33      P0.6    1 = No effect, 0 = RA coil B to 0V
34      P0.5    1 = No effect, 0 = RA coil C to 0V
35      P0.4    1 = No effect, 0 = RA coil D to 0V
36      P0.3    1 = No effect, 0 = DEC coil A to 0V
37      P0.2    1 = No effect, 0 = DEC coil B to 0V
38      P0.1    1 = No effect, 0 = DEC coil C to 0V
39      P0.0    1 = No effect, 0 = DEC coil D to 0V
40      Vcc     +5V

Controller connector port:

Pin numbers from RIGHT looking into socket with tab up (as numbered on
controller board)

1:      Switched +12V from hand controller
2:      Ground
3:      Data from hand controller via 1K resistor
4:      Short to pin 1  
5:      +12V from source via diode
6:      Short to pin 5

Notes:

- Northern hemisphere selected during measurements
- Relay is 5V, but appears to be driven via a resistor from the 12V supply
- Finder LED has buffering (Q2)
- Controller is series diode protected from reverse power connection

Mechanical details:

14mm from PCB to inside of cover plate
5mm from PCB to top of MCU socket
9mm from PCB to top of MCU