Teensyduino 1.19 Release Candidate #1 Available

Status
Not open for further replies.
Paul

Paul

Administrator
Staff member
Here is a first release candidate for Teensyduino 1.19:


Old beta download links removed. Please use the latest version:
https://www.pjrc.com/teensy/td_download.html


Please give this a try and report any bugs. Try to include a sample program that reproduces the problem!


Here's a list of the changes since Teensyduino 1.18:

  • Add support for 72 MHz on Teensy 3.1
  • Fix relative positioning on Teensy 3.x USB mouse (now has hybrid relative+absolute)
  • Add more AVR emulation on Teensy 3.x: wdt.h, crc16.h, sleep.h, EMISK
  • Fix USB Serial break on Teensy3 (needed by some Labview applications)
  • Fix USB endpoint read with no data (thanks nox771)
  • Workaround for auto-reboot problems on Teensy 2.0 when using watchdog
  • Fix initialization of local static vars from non-const data on Teensy 3.x
  • Make all Teensy 3.1 HardwareSerial functions virtual (fixes issue with pointers to objects)
  • Add extra zero to USB serial number, works around a Macintosh CDC driver bug
  • Add Kinetis register definitions: DMA bits, VREF bit, ADC PGA
  • Fix IRQ_PRIORITY hard-coding in Serial3
  • Fix lockup in AVR emulation of SPI.setDataMode(SPI_MODE0)
  • Fix lockup in AVR emulation of TWBR when running at 24 MHz
  • Fix rare lockup in repeated Wire.begin() on Teensy 3.1
  • Update DS1307 library with workaround Arduino Due names.
  • Update Entropy library, now fully compatible with Teensy 3.x.
  • Add overclocking code (not enabled by default) for Teensy 3.x at 120, 144 & 168 MHz
  • Drop Teensy 1.0 support
 
Please be aware, using USB Serial on Windows will result in a new COM port number assigned.

The USB serial number code was changed to add an extra zero onto the end of the serial number, which works around a bug in all versions of Macintosh OS-X. Windows remembers the serial number and other info to keep the same COM port number for each board. Windows should reliably keep a COM port associated with each Teensy3 board, but it will be different between version 1.18 and 1.19.
 
Very nice!

I made a mistake in the post about the DMA registers (http://forum.pjrc.com/threads/25829-Bug-DMA-macros-only-work-for-Teensy-3-0) about the Channel n Priority Register DMA_DCHPRIn.
In the manual for Teesny 3.1 it says "Address: 4000_8000h base + 100h offset + (1d × i), where i=0d to 15d", so I assumed that i=0 corresponded to DMA_DCHPRI0 and i=x to DMA_DCHPRIx, I even pointed out that in Teensy 3.0 this was different.
Well, reading the DMA memory map I see that in fact the addresses are:
Code: 
4000_8100 Channel n Priority Register (DMA_DCHPRI3) 8 R/W See section 21.3.16/414
4000_8101 Channel n Priority Register (DMA_DCHPRI2) 8 R/W See section 21.3.16/414
4000_8102 Channel n Priority Register (DMA_DCHPRI1) 8 R/W See section 21.3.16/414
4000_8103 Channel n Priority Register (DMA_DCHPRI0) 8 R/W See section 21.3.16/414
4000_8104 Channel n Priority Register (DMA_DCHPRI7) 8 R/W See section 21.3.16/414
4000_8105 Channel n Priority Register (DMA_DCHPRI6) 8 R/W See section 21.3.16/414
4000_8106 Channel n Priority Register (DMA_DCHPRI5) 8 R/W See section 21.3.16/414
4000_8107 Channel n Priority Register (DMA_DCHPRI4) 8 R/W See section 21.3.16/414
4000_8108 Channel n Priority Register (DMA_DCHPRI11) 8 R/W See section 21.3.16/414
4000_8109 Channel n Priority Register (DMA_DCHPRI10) 8 R/W See section 21.3.16/414
4000_810A Channel n Priority Register (DMA_DCHPRI9) 8 R/W See section 21.3.16/414
4000_810B Channel n Priority Register (DMA_DCHPRI8) 8 R/W See section 21.3.16/414
4000_810C Channel n Priority Register (DMA_DCHPRI15) 8 R/W See section 21.3.16/414
4000_810D Channel n Priority Register (DMA_DCHPRI14) 8 R/W See section 21.3.16/414
4000_810E Channel n Priority Register (DMA_DCHPRI13) 8 R/W See section 21.3.16/414
4000_810F Channel n Priority Register (DMA_DCHPRI12) 8 R/W See section 21.3.16/414

So the definitions in mk20dx128.h should be:
Code: 
#define DMA_DCHPRI3		*(volatile uint8_t  *)0x40008100 // Channel n Priority Register
#define DMA_DCHPRI2		*(volatile uint8_t  *)0x40008101 // Channel n Priority Register
#define DMA_DCHPRI1		*(volatile uint8_t  *)0x40008102 // Channel n Priority Register
#define DMA_DCHPRI0		*(volatile uint8_t  *)0x40008103 // Channel n Priority Register
#define DMA_DCHPRI7		*(volatile uint8_t  *)0x40008104 // Channel n Priority Register
#define DMA_DCHPRI6		*(volatile uint8_t  *)0x40008105 // Channel n Priority Register
#define DMA_DCHPRI5		*(volatile uint8_t  *)0x40008106 // Channel n Priority Register
#define DMA_DCHPRI4		*(volatile uint8_t  *)0x40008107 // Channel n Priority Register
#define DMA_DCHPRI11		*(volatile uint8_t  *)0x40008108 // Channel n Priority Register
#define DMA_DCHPRI10		*(volatile uint8_t  *)0x40008109 // Channel n Priority Register
#define DMA_DCHPRI9		*(volatile uint8_t  *)0x4000810A // Channel n Priority Register
#define DMA_DCHPRI8		*(volatile uint8_t  *)0x4000810B // Channel n Priority Register
#define DMA_DCHPRI15		*(volatile uint8_t  *)0x4000810C // Channel n Priority Register
#define DMA_DCHPRI14		*(volatile uint8_t  *)0x4000810D // Channel n Priority Register
#define DMA_DCHPRI13		*(volatile uint8_t  *)0x4000810E // Channel n Priority Register
#define DMA_DCHPRI12		*(volatile uint8_t  *)0x4000810F // Channel n Priority Register
and
Code: 
#define DMA_DCHPRI_CHPRI(n)		((uint8_t)(n & 15)<<0)	// Channel Arbitration Priority

I see you haven't included the wrong ones, probably because you realized my mistake. Maybe you can include these in the next update.
 
Yes, I noticed after verify and upload, Windows 7 installed new drivers (which is slow as it checks windows update first).

Installed this rc on top of Arduino 1.0.5-r2/Teensyduino 1.18 without problems. Tested the 72Mhz, briefly:

Code: 
// benchmark code
// originally from http://arduino.cc/forum/index.php/topic,121568.60.html
// change integers to fixed sizes, and float to double, however:
// "Note, Arduino floating point uses 32-bit values for float, double, 
// and long double, while I believe the Arm boards use 64 bit.  This 
// means the mantissa and exponent ranges are higher.
// http://forum.arduino.cc/index.php?topic=128675.msg968187#msg968187

// 8 byte doubles
// ==============
// time elapsed = 18287 micros (Teensy 3.1, 96 Mhz = 1.755)
// time elapsed = 19763 micros (Teensy 3.1, 72 Mhz = 1.422)
// time elapsed = 23350 micros (Teensy 3.1, 48 Mhz = 1.121)
// time elapsed = 18713 micros (Arduino Due, 84 Mhz = 1.572)
// time elapsed = 29504 micros (Teensy 3.0, 96 Mhz = 2.832)
// time elapsed = 31417 micros (Teensy 3.0, 48 Mhz = 1.508)

// 4 byte doubles
// ==============
// time elapsed = 48480 micros (Teensy ++ 2.0, 16 Mhz = 0.776)
// time elapsed = 48484 micros (Teensy 2, 16 Mhz = 0.776)
// time elapsed = 48500 micros (Arduino Uno, 16MHz = 0.776)
// time elapsed = 50116 micros (Arduino Mega 2560, 16 Mhz = 0.802)


double sinanswers[400];

uint32_t time1 = 0;
uint32_t time2 = 0;
void setup()
{
  Serial.begin(57600); // USB is always 12 Mbit/sec
  delay(3000);
  Serial.println("Here");
  delay(1000);
  time1 = micros();
  for (int16_t i = 0; i < 400; i++)
  {
    sinanswers[i] = sin(i);
  }
  time2 = micros();

  uint32_t elapsed = time2 - time1;
  for (int16_t i = 0; i < 400; i++)
  {
    Serial.print(sinanswers[i],8);
    Serial.print(" ");
    Serial.println(i);
  }

  Serial.print("time elapsed = ");
  Serial.print(elapsed);
  Serial.print(" micros");
  Serial.println();
  Serial.print("Double size = ");
  Serial.print(sizeof(sinanswers[0]));
}

void loop()
{

}
 
Last edited:
168 MHz Seems to be working fine for me but if USB type is set to MIDI I get:
Is there something else I need to un-comment?

Code: 
C:\Users\nigel\Dropbox\arduino\hardware\teensy\cores\teensy3\usb_midi.c: In function 'usb_midi_write_packed':
C:\Users\nigel\Dropbox\arduino\hardware\teensy\cores\teensy3\usb_midi.c:89:37: error: 'TX_TIMEOUT' undeclared (first use in this function)
C:\Users\nigel\Dropbox\arduino\hardware\teensy\cores\teensy3\usb_midi.c:89:37: note: each undeclared identifier is reported only once for each function it appears in
 
Nantonos said:
// time elapsed = 18287 micros (Teensy 3.1, 96 Mhz) (8 bytes)
// time elapsed = 19763 micros (Teensy 3.1, 72 Mhz) (8 bytes)
// time elapsed = 23350 micros (Teensy 3.1, 48 Mhz) (8 bytes)
Click to expand...

Thanks for posting this!

I just tried it at 120, 144 & 168 MHz. At the clock gets higher, the slow flash memory really becomes the limiting factor.
 
Status
Not open for further replies.
Back
Top