Fix/pwm new

[perpernorbi]

HardwarePWM was not working, probably since custom pwm implementation has been set to default. The Basic_HwPWM sample is not working. All GPIO pins are at Vcc, the voltage on GPIO2 does not change.

The custom pwm implementation contains a bug. Converting duties and periods is malfunctioning. See StefanBruens/ESP8266_new_pwm#24 for details.

These changes correct the conversion in pwm.c (using .patches/pwm.patch). Also, minor refactoring has been done in samples/Basic_HwPWM.

To test the changes, build and deploy Basic_HwPWM, and verify voltages using a multimeter. Voltages should approximately be the same, regardless ENABLE_CUSTOM_PWM is set to 0 or 1.

[slaff]

@perpernorbi Please, fix the coding style

!!! samples/Basic_HwPWM/app/application.cpp not compliant to coding style, here is the fix:
--- samples/Basic_HwPWM/app/application.cpp	2018-08-24 16:08:23.471101371 +0000
+++ /dev/fd/63	2018-08-24 16:08:29.427120477 +0000
@@ -27,7 +27,7 @@
 bool countUp = true;
 
 int maxDuty = HW_pwm.getMaxDuty();
-int32 inc = maxDuty/50;
+int32 inc = maxDuty / 50;
 
 void doPWM()
 {
@@ -58,12 +58,12 @@
 
 	// Setting PWM values on 8 different pins
 	HW_pwm.analogWrite(4, maxDuty);
-	HW_pwm.analogWrite(5, maxDuty/2);
+	HW_pwm.analogWrite(5, maxDuty / 2);
 	HW_pwm.analogWrite(0, maxDuty);
-	HW_pwm.analogWrite(2, maxDuty/2);
+	HW_pwm.analogWrite(2, maxDuty / 2);
 	HW_pwm.analogWrite(15, 0);
-	HW_pwm.analogWrite(13, maxDuty/3);
-	HW_pwm.analogWrite(12, 2*maxDuty/3);
+	HW_pwm.analogWrite(13, maxDuty / 3);
+	HW_pwm.analogWrite(12, 2 * maxDuty / 3);
 	HW_pwm.analogWrite(14, maxDuty);
 
 	debugf("PWM output set on all 8 Pins. Kindly check...");

Make sure to read the release notes related to coding style:
https://github.com/SmingHub/Sming/releases/tag/3.6.1

[slaff]

@MayaPosch can you check if that PR is working for you?

[MayaPosch]

@slaff I'll let you know tomorrow (Monday). Need to set up the infrastructure first before I can run a test :)

[MayaPosch]

@slaff Just tested the PR with my setup. Seems to have fixed the PWM issues. Also finally got the fine-grained control over the output now, with up to 0x6F (111) steps tested from 0 to nearly 3.3V with a DMM directly measuring the pin (GPIO14) in DC voltage mode.