************************************************************************************
Electronic Speed Controller Program Version Beta 1.0 RELEASE NOTES
************************************************************************************
1)Addapted to more standard RC-set's
It turned out that for most RC-sets the receiver does not generate a 1 to 2 ms pulse
but actually more frequently they deliver a 1.1 to 1.9 ms pulse. This version of the
program expects 1.1 to 1.9 ms pulses from the RC receiver.
2)Larger ZERO means ZERO band
It turned out that you need a super stable transmitter to keep the neutral position
sufficiently well in place to realy stop the motor from making any movement. This
version has a ZERO band that is double the size it was in the Alpha 1.0 version.
3)Larger FULL-POWER-ZONE
It turned out that you need a super stable transmitter to make sure that full speed
on the transmitter translated indeed in full speed of the motor. In competition this
obviously is very important so I have now doubled the FULL-SPEED-ZONE.
4)Safer Failsave
The Failsave algorithm used in the Alpha 1.0 release was based on testing the
quality of the incoming pulses. No pulse would let the controller take the last
received pulse as the measure for the output signal. The result was that if the
controller was disconnected from the receiver the controller would continue to
output some power. Unacceptable for a failsave. In this release we also switch to
failsave when we do not get a pulse for more then 20 ms.
OPPORTUNITIES FOR IMPROVEMENT:
1)Current limiting capabilities
Link: http://www.avrfreaks.net/index.php?module=...ect&item_id=361
Electronic Speed Controller Program Version Beta 1.0 RELEASE NOTES
************************************************************************************
1)Addapted to more standard RC-set's
It turned out that for most RC-sets the receiver does not generate a 1 to 2 ms pulse
but actually more frequently they deliver a 1.1 to 1.9 ms pulse. This version of the
program expects 1.1 to 1.9 ms pulses from the RC receiver.
2)Larger ZERO means ZERO band
It turned out that you need a super stable transmitter to keep the neutral position
sufficiently well in place to realy stop the motor from making any movement. This
version has a ZERO band that is double the size it was in the Alpha 1.0 version.
3)Larger FULL-POWER-ZONE
It turned out that you need a super stable transmitter to make sure that full speed
on the transmitter translated indeed in full speed of the motor. In competition this
obviously is very important so I have now doubled the FULL-SPEED-ZONE.
4)Safer Failsave
The Failsave algorithm used in the Alpha 1.0 release was based on testing the
quality of the incoming pulses. No pulse would let the controller take the last
received pulse as the measure for the output signal. The result was that if the
controller was disconnected from the receiver the controller would continue to
output some power. Unacceptable for a failsave. In this release we also switch to
failsave when we do not get a pulse for more then 20 ms.
OPPORTUNITIES FOR IMPROVEMENT:
1)Current limiting capabilities
Link: http://www.avrfreaks.net/index.php?module=...ect&item_id=361
| CODE |
/************************************************************************************************************************ Electronic Speed Controller Program Version Beta 1.1 MAIN PROGRAM ------------------------------------------------------------------------------------------------------------------------- Written for : Atmel "AVR Assembler" and "AVR Assembler 2" when : 2006 April 20 where : BRUSSELS - BELGIUM - EUROPE who : Jan Huygh E-mail: jan.huygh@skynet.be *************************************************************************************************************************/ .Include "tn13def.inc" .Include "IO_Port_And_IO_Pin_Configuration.inc" .Include "Assign_Variables_To_Registers.inc" .Include "Macro_Definition_Initialize.inc" .Include "Macro_Definition_GeneratePWMSignal.inc" .Include "Macro_Definition_NewPulseCalculations.inc" .Include "Macro_Definition_FailSaveRoutine.inc" .Include "Macro_Definition_LookupTable.Inc" .ORG 0 Initialize ldi PWMStep,1 //PWMStep = 6 //With the main loop taking 12 CPU clock cycles we can calculate Fpwm as a function of PWMStep @9.6 MHz CPU clock. //Fpwm=(9.6*PWMStep*1000000)/(256*12) So: 1=>3125Hz 2=>6250Hz 3=>9375Hz 4=>12500Hz 5=>15750Hz 6=>18750Hz 7=>21875Hz FailSave: FailSaveRoutine MainLoop: //MainLoop: GeneratePWMSignal //Generate PWMSignal takes always exactly 6 cycles sbic PinB,RCPulseInputPin //If (RCPulseInputPin=0)then rjmp ML_Else1 // adiw RCPulseCounterH:RCPulseCounterL,1 // RCPulseCounter=RCPulseCounter+1 rjmp MainLoop // ML_Else1: //Else cpi RCPulseCounterH,0 // If (RCPulseCounter>0) then breq MainLoop // NewPulseCalculations // NewPulseCalculations takes always exactly 17 cycles // EndIf //EndIf rjmp MainLoop //GoTo MainLoop /*The main loop always takes 12 CPU clock cycles to execute except just after an RC pulse has been reveived. Just after an RC pulse has been received the main loop takes 13 cycles + cycles needed for NewPulseCalculations. This has no effect on the measurement accuracy of the incoming RC pulse. This has a minor effect on the stability of the generated PWM signal. Every 20ms the PWM signal will have one cycle that is 18 CPU cycles longer. With PWMStep = 6 we will have every 372 PWM cycles 1 cycle that takes 1.8 µs or 3.4% longer. With PWMStep = 1 we will have everey 62 PWM cycles 1 cycle that takes 1.8 µs or 0.5% longer. */ Data: LookupTable |