Znalazłem chyba błąd. Sterowniki źle działają. Są to sterowniki Mardus-Kreuts.
.
Problem w tym, że nie wraca do swojej pozycji, tylko zjeżdża niżej, a potem podnosi się do góry normalnie. Czyli np ma zjechać 6mm, a zjeżdża 8 i podnosi 6.
1. Przy ustawieniach programu na maksymalną prędkość i 1/16 kroków to za każdym razem zjeżdża nawet o 1cm. Przy wolnej prędkości tego nie widać.
2. zamiana polaryzacji sygnału DIR w programie sprawia, że podjeżdża za każdym razem. Zamiana cewek miejscami po zmianie sprawia, że znów zaczyna zjeżdżać.
Może coś nie tak w wsadem do Attiny? Albo źle zaprogramowałem fusebity? Programowałem ISP programmerem Adama Dybkowskiego i eXtreme Burner, mogłem coś pomylić....
Wysyłam kod Bascoma i schemat. Schemat ma niepotrzebnie dorobione wejście Enable, ale nóżkę 12 układu 74HC14 wyjąłem, żeby nie pszeszkadzała.
Kod: Zaznacz cały
'+++++++++++++++++++++\\\\Rev01P+////+++++++++++++++++++++++++++++++++++++++++
'++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
'Mardus-Kreutz project for Unipolar Microstepper driver
'11/2006
'REMEMBER TO SET THE FUSE FOR EXTERNAL XTAL. DO NOT CHANGE THE WATCHDOG-ON FUSE
'CODE WRITEN INITIALLY IN BASIC (BASCOM AVR) BECAUSE IT IS INTENDED TO BE A LEARNING TOOL.
'Disable the Div/8 fuse'
'Enable the Brown-out detection (4.5volt) fuse when all tests are done....!!!
'
'Stepper motor phase A PWM reference signal output pin PORTB.3
'Stepper Motor phase B PWM reference signal output pin PORTB.4
'Phase A PORTB.0
'Phase A/ PORTB.5
'Phase B PORTB.6
'Phase B/ PORTB.7
'Portb.1 output is used for error LED, and "Output Disable" on the Power board (active Low)
'PortD.5 used for hardware standby current reduction.
'Rev 00 START 11/24/2006 10:28 p.m. FINISHED 11/25/2006 9:35 p.m.
'Measured Interrupt service time worst case: 14 microSeconds.
'tested at 28KHZ step frequency
'------------------------------------------------------------------------------
'To Do:
'-Standby current reduction after 3 seconds
'-Use of Maximum torque table
'-Step Reference to full-Step after certain speed.
'------------------------------------------------------------------------------
'Rev00a
'Corrected:
'There was a measured Big step around Step 10 on the 16th Step table,
'Changed value 221 (wrong) to 212 on the table.
'------------------------------------------------------------------------------
'Rev01a 11/27/2006 7:54 p.m.
'More Corrections in this revision:
'Prior revision occupied 94% of Flash, not including the missing options.
'Some code optimization, saved 4% of the space. 12/23/2006 Compiling
'under "optimize code" saved another 3%
'------------------------------------------------------------------------------
'Rev01b 12/25/2006
'added Maximum torque tables and associated logic,further optimization,
'flash used 93%
'added speed sense, flash used 104% , Disabled home position selection
'in order to make space for code on the flash memory. 12/29/06
'
'To do: Add Watchdog timer as interrupt every one second
'------------------------------------------------------------------------------
'Rev01c 12/29/2006 Flash used 93%
'added Watchdog interrupt for speed sensing and standby time counter
'To do:
'add hardware power down (TO SCHEMATICS,Translator Rev 2.1) {added on 12/30/2006}
'------------------------------------------------------------------------------
'Rev01d 12/30/2006 14:11 hrs Flash used 87%
'added Chopper PWM oscillator frequency 39.2 Khz, time high = 1uS (Blanking time)
'Modified Schematics (to Schematics , Translator Rev 2.1a)
'Hardware power_down was added during this modification.
'------------------------------------------------------------------------------
'Rev01e 12/31/2006 Flash used 87%
'CHANGED speed sensing algorithm for a more reliable watchdog interrupt every 16 ms
'the step routine keeps track of how many FULL-step pulses occur within 10 watchdog
'interrupts (160 ms) and changes tables to Full Step at about 250 Full-steps /sec
'Step interrupt routine takes 24 uSec to execute, limiting theoretical step
'rate to about 40 KHz.
'To Do: More optimization is required.
'------------------------------------------------------------------------------
'Rev01f 1/02/2007 Flash used 83% Optimized Speed sensing and reduced Step
'interrupt routine time to 19.7 uSec.
'To Do: More optimization
'------------------------------------------------------------------------------
'Rev01g 1 / 02 / 2007 Flash Used 83% interrupt routine time is now 16.1 uSec
'after replacing the tables by two arrays resident in RAM.
'To Do: Optimize interrupt request code in assembly.
'-----------------------------------------------------------------------------
'Rev01h 1/5/2007 Flash used 90%, interrupt routine worst case time is now 15.1 uS
'Converted the variable S into byte (from Integer) and changed the comparison to 128
'in order to find out if it is a negative number (over 128). F was converted into a
'Byte type and assigned the value 64 when Forward and 192 when Reverse. This save us two
'comparisons and some execution time too.
'Added the three sequences Microstep - HighTorque - Full-Step dependent on speed.
'To do: Testing on the board driving a motor.
'Adjust Blanking time depending on board.
'Do the Interrupt routine code in assembly using AVR registes R1, R2, R3, R8, R9, R11, R12, R13 AND R14
'to hold variables like S, T, Q, Spd, Speed, modeshift, 2ud_module, ud_module and 2modeshift.
'From now on the source will not be published since it will lose readability as a learning tool.
'we will post the firmware in Hex to be programmed using any programmer of your choice.
'----------------------------------------------------------------------------
'Rev01i 1/6/2007 Flash used: 86% interrupt time 13.8 uSec added S and T variables into
'the registers r8 and r9, (iRam),
'Changed compare1a and compare1b to load the value shown on the table
'Changed the tables to avoid substracting 255 - tablevalue when loading the compare registers.
'Reduced speed sensing timebase to 80 mS to improve behavior at higher acceleration values.
'----------------------------------------------------------------------------
'Rev01j 1/9/2007 Mardus Found A Bug On The Firmware , The Bug Is A Bascom Compiler Problem
'Related To Timer0 Pwm And Portd.6 Not Going Down To Zero Or Up To + 5 Volt When Used As Input
'So I Decided To Write The Timer0 Configuration For Pwm Directly To The Registers.
'----------------------------------------------------------------------------
'Rev01K Last revision level in Basic. Flash used 87%. From now on only parameters changes for tune-Up.
'are going on revision01xx. Assembly versions will be called Rev02.
'----------------------------------------------------------------------------
'Rev01L 1/15/2007. Flash at 87%. Part of the code on the interrupt Int0 service routine has been optimized.
'Added dead time (300nSec) on the switching sequence and corrected a bug pointed out by Wojtek
'on the reading of the jumpers before comparing for 1/5 or 1/10 step mode.
'----------------------------------------------------------------------------
'Rev01M 1/19/2007. Dropped the high-torque jumper option, reason: not smooth micro-stepping at low rpms.
'Will keep the same waveform as an intermediate morphing stage between micro-step and Full step at high speeds.
'S1 7-8 Jumper is not being read in this version. Changed Full-step behavior to have a DC reference all the time.
'Introduced automatic torque compensation during the whole speed range in order to have the maximum available
'torque present all the time. USER ADJUST TRIMMERS TO REFERENCE = 1.41 * Rs * FULL-STEP MAXIMUM CURRENT (R.M.S. MAXIMUM)
'reference table is changed to Full-torque reference.
'1/27/2007 New version 2.1F SMD of the PCB includes the translator stage. After Rev01M current reduction
'becomes mandatory, no jumper selection is available anymore. User select the % reduction by changing R39
'and R40 values R39=R40=1Kohm => 50% reduction. Current reduction is done after 1 second.
'----------------------------------------------------------------------------
'02/08/2007 REV01M++
'added zero reference output current inhibit,in order to avoid the minimum current issue due to the forced conduction
'during the blanking time, when output current is supposed to be zero..
'Corrected the speed references for waveform switching. Maximum Step frequency tested is 75.8 Khz
'---------------------------------------------------------------------------
'04/25/2007 REV01M+++
'minor correction of a constant used for speed changing when decelerating.
'---------------------------------------------------------------------------
' 06/11/2007 Rev01N. worst Case Interrupt routine time = 10.2 uSec. Maximum tested Step rate = 90 Khz
' @ 1 usec pulses, 110 Khz with 2 usec step pulses.
' Flash used 86%. Optimized Speed dependent Table selection.
' Added "Home" Signal output at Pin 4 of Port D.
'---------------------------------------------------------------------------
' 06/13/2007 Rev01N++ tested to 100Khz 1 uSec Step pulse width. Worst Case Interrupt routine time = 9.7 uSec
' changed the Speed variable into a byte, limited the value of the SPD variable (word) to 254
' before transfering its Low byte value into Speed
'+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
' Rev01O replaced variable name: Time by Time1, due to conflicts with BASCOM reserved variable name in version 1.11.9.0
'+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
' 10/14/2008 Rev01P Reduced Speed calculation timebase to 48 mSec in order to improve acceleration behavior.
' Modified speed reference values to keep the same transition rpms between microstep and high power reference waveforms.
' code change has been tested with bipolar power boards and was found to be correct.
'--------------------------------------------------------------------------------
' 12/28/2008 Rev01P+ flash used 87%. Added a reset to spd variable on the speed calculation routine in order to avoid the speed related morphing to hang
' with the wrong table after a crash ( motor stall ) or e-Stop event at high speed. Thank You, Tivoidehuong!
' 12/31/2008 Added Table reset to microstep mode ( to the idle current handling code ) for the same purpose.
'
'$sim
$regfile = "attiny2313.dat"
$crystal = 20000000
$framesize = 2
$swstack = 2
$hwstack = 32
Config Portb = &B11111111 'CONFIGURED FOR OUTPUT.PORTB.2 (AS CHOPPER PWM OSCILLATOR )
Config Portd = &B00110000 'ALL INPUTS EXCEPT PORTD.5 (output used for power reduction)
'and PortD.4 (output used for Home signal S = 0) (added in Rev01N)
Portd = &B11001111 'Configure Pull ups
' Config Timer0 = Pwm , Compare A Pwm = Clear Up , Prescale = 1 'for 39.2 Khz @ 20Mhx Xtal
'Configured registers directly because of the above statement's bug in BASCOM
Tccr0a = &HC1
Tccr0b = &H01
'//////////////////////////////////////////////////////////////////////////////
'BLANKING TIME ADJUSTMENT BY CHANGING THIS REGISTER'S VALUE
Ocr0a = 12 'LOAD OUTPUT COMPARE REGISTER FOR TIMER0 IN ORDER TO GET 1.2 uS Low Sync pulse
' (that is:10 = 1 uS Blanking time)
'//////////////////////////////////////////////////////////////////////////////
'\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\
Config Int0 = Rising 'ACTIVE DURING Low TO High TRANSITION to be used when using optocoupler
'Config Int0 = Falling 'ACTIVE DURING HIGH TO LOW TRANSITION for STK500 testing
'\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\
Config Timer1 = Pwm , Pwm = 8 , Compare A Pwm = Clear Up , Compare B Pwm = Clear Up , Prescale = 1
'Configure Watchdog timer
Wdtcr = &B01000000 'interrupt enabled, 16 mS timeout
On Wdt Watchdoggy Nosave 'on watchdog overflow execute interrupt routine: Watchdoggy
Dim Step_mod As Bit 'Step_mod = 1 => full to 1/16 mode, Step_mod = 0 => 1/5 or 1/10 mode
Dim Jumper As Byte 'Variable holds jumper settings
Dim F As Byte 'Direction, Forward = 1, Reverse = 0
Dim Q As Byte 'Quadrant, values 1,2,3 or 4
Dim Ref1 As Byte 'PWM value for sine output
Dim Ref2 As Byte 'PWM value for Cosine output
Dim S As Iram Byte 'Sine channel Index HAS TO BE AN INTEGER BECAUSE IT CAN HAVE NEGATIVE VALUES
Dim T As Iram Byte 'Cosine channel Index
Dim I As Bit 'Up/down counter state =1 => up, 0=> count down, when in Forward motion (inverted in Reverse)
Dim Mode_shift As Byte ' ( 1/16 =1, 1/8 = 2, 1/4 = 4, half = 8, Full = 16)
Dim Home_pos As Byte 'home microstep number, typically 0 or 45 degrees (Mstep= 0 or 8)
Dim Speed As Byte ' pulse per unit of time (uses WATCHDOG TIMER AS TIMEBASE )
'Used for microstep to Fullstep at mid rpms, limited to 8 bits.
Dim Power_down As Byte 'inactivity power down counter
Dim Full_step As Bit 'Use Full_Step table instead of normal microstep table, when =1 => Use fullstep table
Dim U_d_module As Byte 'Up /down counter module
Dim 2u_d As Byte 'Twice the Up-down Module value
Dim Spd As Word 'temporary variable for speed calculations
Dim High_torque As Bit 'when =1 use max-torque table instead of microstep or full-step table
Dim 2mode_shift As Byte 'twice the mode_shift value for UP-Down counter code speedup
Dim Time1 As Byte 'variable to measure timebase for speed measurement
Dim A(17) As Byte 'arrays will hold the proper Sine, High-Torque and Full Step table values in order to speed up lookup.
Dim B(17) As Byte
Dim C(17) As Byte
'================================================================================
Portb.1 = 0 'TURN ON THE ERROR LED DURING INITIALIZATION, WILL BE TURNED BACK OFF AT THE END OF THE INIT CYCLE
'Disable all outputs
Portb.7 = 0 'B\
Portb.5 = 0 'A\
Portb.6 = 0 'B
Portb.0 = 0 'A
Portd.5 = 0 'no current reduction
'===============================================================================
Compare1a = 255 'SET COMPARE REGISTES FOR 0 OUTPUT WHILE IN INITIALIZATION
Compare1b = 255
'Initial conditions
'------------------------------------------------------------------------------------------------
Q = 0 'the Home position is always on the first quadrant
'------------------------------------------------------------------------------------------------
'Here read jumper settings and define initial conditions
Jumper = Pind And &B00001011 ' PortD reading masked to find out step mode values
'==============================================================================
'CHECK FOR SERVICE MODE
'SERVICE MODE IS USED TO ADJUST THE CURRENT LIMITS, ENTER SERVICE MODE, IT WILL PLACE MAXIMUM VOLTAGE
'AT THE REF1 AND REF2, ADJUST POTENTIOMETERS TO PROPER VOLTAGE AS PER REQUIRED CURRENT.
'THE ONLY WAY TO LEAVE SERVICE MODE IS BY POWERING DOWN AND DISCONNECTING ANY OF THE 3 JUMPERS.
'While in SERVICE mode all outputs are disabled, No power is delivered to the motor coils.
'==============================================================================
If Jumper = &B00001011 Then 'SERVICE MODE
Compare1a = 0 'SET COMPARE REGISTES FOR OUTPUT Maximum PWM Voltage
Compare1b = 0
Do 'ENDLESS LOOP
Waitms 500
Toggle Portb.1 'ERROR LED BLINKING ONCE PER SECOND
Loop
End If
'==============================================================================
'END OF SERVICE MODE
'==============================================================================
'Finding out which options were selected by the User (jumper programming).
Select Case Jumper
Case Is = &B00000000 : Mode_shift = 16 'Full
Case Is = &B00000001 : Mode_shift = 8 'Half
Case Is = &B00000010 : Mode_shift = 4 '1/4
Case Is = &B00000011 : Mode_shift = 2 '1/8
Case Is = &B00001000 : Mode_shift = 1 '1/16 mode
Case Is = &B00001001 : Mode_shift = 2 '1/5
Case Is = &B00001010 : Mode_shift = 1 '1/10 mode
End Select
'Assign variables to selected options
'==============================================================================
'===============================================================================
'Re-read Jumpers (thank you Wojtek!!)
Jumper = Pind And &B00001011 ' PortD reading masked to find out step mode values
'-------------------------------------------------------------------------------
If Jumper = &B00001010 Or Jumper = &B00001001 Then
Step_mod = 0 ' 1/10 or 1/5 modes
U_d_module = 10
'transfer table content into arrays
For S = 0 To 10
A(s + 1) = Lookup(s , S10) 'use Sinus table
B(s + 1) = Lookup(s , F16) 'use FullStep table
C(s + 1) = Lookup(s , M10) 'max torque used as transition
Next
Else
Step_mod = 1 ' 1/16 to Full Mode
U_d_module = 16
'This part of the code transfers table content into arrays in RAM
If Mode_shift = 16 Then 'It is in Full step mode, use Fullstep all the way
For S = 0 To 16
A(s + 1) = Lookup(s , F16) 'use FullStep table
B(s + 1) = Lookup(s , F16) 'use FullStep table
C(s + 1) = Lookup(s , F16) 'use FullStep table
Next
Else
For S = 0 To 16
A(s + 1) = Lookup(s , S16)
B(s + 1) = Lookup(s , F16)
C(s + 1) = Lookup(s , M16)
Next
End If
End If
2u_d = U_d_module * 2
2mode_shift = Mode_shift * 2
'Write The Startup Values Directly To The Compare Registers.
Home_pos = 0
S = Home_pos
Compare1a = 255
Compare1b = 0
'==================================================================================================
'Energize motor IN HOME POSITION
Portb.7 = 0 'B\
Portb.5 = 0 'A\
Portb.6 = 1 'B
Portb.0 = 1 'A
I = 1 ' Q=1, sinus counter state to UP in Forward Movement (or Down in Reverse)
'-------------------------------------------------------------------------
'Enable the STEP interrupt
On Int0 1step Nosave 'jump to STEP on INT0, Do not save registers on Stack.
Waitms 500 'wait 500 milli-seconds
Power_down = 0 'no standby power down on startup
Spd = 0 'zero speed
Full_step = 0 'use normal microstep table at startup
Portb.1 = 1 'initialization finished, turn off the ERROR LED
Time1 = 0 'initialize 3 second power_down timer
Enable Interrupts
Enable Int0
'+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
'+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
' END OF INITIALIZATION
'BEGIN MAIN PROGRAM ENDLESS LOOP WAITING FOR INTERRUPTS
'+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
'+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Do 'endless loop
nop
Loop
End
'+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
'+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
'INT0 INTERRUPT SERVICE ROUTINE
'+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
'+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
'Worst case duration of this routine was 11.40 uSec. Now it is 9.70 uSec (Rev01N+).
1step:
Disable Interrupts
'Portb.1 = 0 'Used to measure delay with scope
Portd.5 = 0 'disable current reduction
Power_down = 0 'reset Power_down counter
'-------------------------------------------------------------------------------
'MODIFIED IN REV01L
If Pind.6 = 0 Then 'Read the Direction Pin
F = 192 'Two complement of 64 (8 bit) REVERSE
'CALCULATE INDEX VALUES TO THE STEP TABLE
If I = 1 Then 'F= -64 AND I =1
S = S - Mode_shift
Else 'F= -64 AND I=0
S = S + Mode_shift
End If
Else
F = 64 'FORWARD
'CALCULATE INDEX VALUES TO THE STEP TABLE
If I = 1 Then 'FORWARD And I = 1
S = S + Mode_shift
Else 'FORWARD AND I=0
S = S - Mode_shift
End If
End If
'F = 192 'FOR TESTING PURPOSES ONLY (USING SIMULATOR)
'Portb.1 = 1 'FOR TIMING MEASUREMENT
'Now from the beginning of the interrupt to here it takes 1.151 uSec (WORST CASE) (Rev01N+)
'-------------------------------------------------------------------------------
'SELECT MICROSTEPPING TABLE DEPENDING ON SPEED
'Do not move this part of the code from here!!
If S = 0 Then
Portd.4 = 1 ' activate the S=0 marker, used for Full Step period calculation.
If Speed > 19 And Speed =< 27 Then 'OVER 4 RPS (=>200 Full Steps/sec) =240 rpm
'but less than 288 Full Steps/sec = 345 rpm
' use high torque table, Modified in Rev01N
High_torque = 1
Full_step = 0
Else
If Speed >= 29 Then 'over 6 rps (300 FullSteps/Sec)= 360 rpm
' use Full Step Table Modified in Rev01N
Full_step = 1
High_torque = 0
Else
If Speed =< 18 Then 'ABOUT 3.8 RPS (about 190 FullSteps /Sec) go back to microstepping
'use microstep table Modified in Rev01N
High_torque = 0
End If
End If
End If
Else
Portd.4 = 0 'Reset S=0 marker
End If
' Portb.1 = 0 'for testing timing
'===============================================================================
'LOOKUP ON THE TABLES
'-------------------------------------------------------------------------------
' CHECK THE UP-DOWN COUNTER FOR MINIMUM AND MAXIMUM COUNTS
'Portb.1 = 1
'from the beginning of the interrupt to here it takes 1.651 uSec (worst case 4.502 uSec) (Rev01N+)
'-------------------------------------------------------------------------------
' ZERO CROSSING IS COMMON FOR THE TWO STEP MODES
If S > 128 Then 'We are Decreasing "S" going "negative"
' since "S" is a byte type variable, will not take negative values but will get up to S=240 when Mode_shift=16
'hence the 128 to decide if it is a negative value or not
S = S + 2mode_shift 'add two times the Mode_Shift value
'-------------------------------------------------------------
'Increment Spd twice PER FULL STEP for speed measuring purpose on the zero crossings
Spd = Spd + 1
'--------------------------------------------------------------------
'Now Change The Quadrant
Q = Q + F
End If
If S > U_d_module Then 'REACHED MAXIMUM COUNT, GOING DOWN NOW
S = S - 2mode_shift 'Subtract two times the Mode_Shift value
'Now Change The Quadrant
Q = Q + F
End If
'----------------------
T = S + U_d_module 'U_d_module = # of steps shift to find cosine value on sine table (90 degrees)
If T > U_d_module Then
T = 2u_d - T 'If it goes past the UP-Down counter's module, then rollback to the beginning of the table
End If
'-------------------------------------------------------------------------------
'-------------------------------------------------------------------------------
'Portb.1 = 1
'from the beginning of the interrupt to here takes 6.36 uSec worst case (Rev01N+)
'xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
'this part of the code between xxxxxxxxx now takes 2.2 uSec
'Portb.1 = 0
'Look up into the proper sinus table and Update PWM registers
'FIRST CHANGE REFERENCE VOLTAGES BEFORE CHANGING SEQUENCE OUTPUTS
'SEQUENCE OUTPUTS SHOULD BE CHANGED ONLY DURING THE COIL CURRENT ZERO CROSSING
If Full_step = 1 Then
'use Full Step table
Ref1 = B(s + 1) 'the first element of the array is number 1
Ref2 = B(t + 1)
Else
If High_torque = 1 Then
' use high torque table
Ref1 = C(s + 1) 'the first element of the array is number 1
Ref2 = C(t + 1)
Else
'use microstep table
Ref1 = A(s + 1) 'the first element of the array is number 1
Ref2 = A(t + 1)
End If
End If
Compare1a = Ref1 'SET COMPARE REGISTES FOR OUTPUT
Compare1b = Ref2
'Portb.1 = 1
'xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
'-------------++++++++++++++++++++++++++----------------------------------------
'==============================================================================
' WHAT OUTPUT SEQUENCE?
'===============================================================================
' Sequence to output (DEPENDING ON CURRENT QUADRANT)
' from here to the end of the interrupt routine it takes 1.76 uSec worst case
' Portb.1 = 0
If Q = 0 Then 'sequence 1010 FIRST QUADRANT
Portb.7 = 0 'B\
Portb.5 = 0 'A\
Portb.6 = 1 'B
If Ref1 = 255 Then
Portb.0 = 0
Else
'NoP 'inclusion of this delay gives us 150 nSec dead time
Portb.0 = 1 'A
End If
I = 1 'sinus counter state to UP
Else
If Q = 64 Then ' sequence 1001 SECOND QUADRANT
Portb.5 = 0 ' A\
Portb.6 = 0 ' B
Portb.0 = 1 ' A
If Ref2 = 255 Then
Portb.7 = 0
Else
' NoP
Portb.7 = 1 ' B\
End If
I = 0 ' sinus counter state to Down
Else
If Q = 128 Then ' sequence 0101 THIRD QUADRANT
Portb.0 = 0 'A
Portb.6 = 0 'B
Portb.7 = 1 'B\
If Ref1 = 255 Then
Portb.5 = 0 'A\
Else
' NoP ' inclusion of this delay gives us 150 nSec dead time
Portb.5 = 1 'A\
End If
I = 1
Else ' sequence 0110 FORTH QUADRANT
Portb.0 = 0 'A
Portb.7 = 0 'B\
Portb.5 = 1 'A\
If Ref2 = 255 Then
Portb.6 = 0 'B
Else
' NoP ' inclusion of this delay gives us 150 nSec dead time
Portb.6 = 1 'B
End If
I = 0
End If
End If
End If
'------------------------------------------------------------------------------
Enable Interrupts
'Portb.1 = 1 'Used to measure time taken by Interrupt service routine with scope
Return
'
'+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
'+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
' WATCHDOG INTERRUPT ROUTINE
'+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
'+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
' ThisWatchdog interrupt routine will occur every 16 mSeconds.
' POWER_DOWN VARIABLE IS CLEARED IN EVERY STEP, IF IT'S VALUE IS INCREASING MEANS THAT NO STEPS ARE BEING RECEIVED
' If power_down counter is >= 187 (3 seconds with zero speed) then will reduce the coil current
' or If power_down counter is >= 63 (1 seconds with zero speed) then will reduce the coil current
' to the specified value (depends on the values of two resistors R39 and R40 on the schematics).
Watchdoggy: 'every 16 mSecond
' Portb.1 = 0
Time1 = Time1 + 1 'increment speed sensor timebase (48ms) ( modified in Rev01P)
Power_down = Power_down + 1 'increment power_down sensor
'----------------------------------
If Power_down > 62 Then '187 =>Current reduction after 3 seconds.(Value 63 => 1sec)
'------------------------------
Portd.5 = 1 'activate the current reduction hardware
Spd = 0 ' reset Spd accumulator, Modified in Rev01P+ (Thanks Tivoidehuong!)
Full_step = 0 ' reset tables. Added in Rev01P+
High_torque = 0
Power_down = 188 'LIMIT THE MAXIMUM VALUE OF THIS VARIABLE
Else
Portd.5 = 0 ' No Current reduction
End If
'this part of the watchdog interrupt routine will calculate the speed (proportional to rps).
'by counting the number of full-steps (ACTUALLY TWICE THE NUMBER OF FULLSTEPS) in 48 mSeconds ( modified in Rev01P).
If Time1 > 2 Then 'timebase=48mS (3 x 16mS) ( modified in Rev01P and Rev01P+)
Time1 = 0
'Since we are interested only in the low-medium speed range ( below 600 rpm)
'we can limit the Speed variable to 8 bits. Modified in Rev01N+
If Spd > 254 Then
Spd = 254
End If
Speed = Low(spd ) ' Latch accumulated Spd value on Speed variable (byte)
Spd = 0 ' reset Spd for the next 48ms period speed accumulator
End If
'Portb.1 = 1
'normally takes 1.60 uSec, worst case duration of this routine = 3.25 uSec
Return
'\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\
'TABLES------------------------------------------------------------------------
'\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\
'Table content will be loaded into three arrays, resident on RAM, during initialization
'data stored in tables might go into EEPROM in newer revisions in order to make space for more code
'Maximum Torque tables are calculated using Sin(2*Angle) from 0 to 45 degrees and value 255 the other
'45 degrees (90 degrees sector) Table values reflect the operation { value = 255 - calculated values on my
'excel spreadsheet (MICRO-STEPPING SEQUENCE AND TABLES_Rev03.xls ) } due to the inverted output of timer1 PWM
'---------------------------------------------------------------------------------------------------
' Sinus table for 1/16, 1/8, 1/4, Half and Full stepping
'---------------------------------------------------------------------------------------------------
S16:
Data 255 , 230 , 205 , 181 , 157 , 135 , 113 , 93 , 75 , 58 , 43 , 30 , 19 , 11 , 5 , 1 , 0
'---------------------------------------------------------------------------------------------------
' Sinus table for 1/10, 1/5
'---------------------------------------------------------------------------------------------------
S10:
Data 255 , 215 , 176 , 139 , 105 , 75 , 49 , 28 , 12 , 3 , 0
'--------------------------------------------------------------------------------------------------
' Maximum torque table for 1/16, 1/8, 1/4, etc Stepping
M16:
'Data 255 , 205 , 157 , 113 , 75 , 43 , 19 , 5 , 0 , 0 , 0 , 0 , 0 , 0 , 0 , 0 , 0 'non peak to rms compensation
Data 255 , 214 , 174 , 138 , 106 , 80 , 60 , 48 , 44 , 44 , 44 , 44 , 44 , 44 , 44 , 44 , 44
'--------------------------------------------------------------------------------------------------
' Maximum torque table for 1/10 and 1/5
M10:
' Data 255 , 176 , 105 , 49 , 12 , 0 , 0 , 0 , 0 , 0 , 0 'non peak compensation
Data 255 , 190 , 131 , 85 , 54 , 44 , 44 , 44 , 44 , 44 , 44
'-----------------------------------------------------------------------------
'FullStep torque Table for high RPS, will be shared between all microstep modes
F16:
'Data 0 , 0 , 0 , 0 , 0 , 0 , 0 , 0 , 0 , 0 , 0 , 0 , 0 , 0 , 0 , 0 , 0 'non peak to rms compensation
Data 75 , 75 , 75 , 75 , 75 , 75 , 75 , 75 , 75 , 75 , 75 , 75 , 75 , 75 , 75 , 75 , 75
