The device achieves up to 16 MIPS throughput at 16 MHz. At least I'm thankful for the "red wire = pin 1" information being accurate, and sufficient to derive the right pinout.The low-power Atmel 8-bit AVR RISC-based microcontroller featuring 32KB self-programming flash program memory, 2.5KB SRAM, 1KB EEPROM, USB 2.0 full-speed/low speed device, 12-channel 10-bit A/D-converter, and JTAG interface for on-chip-debug. Debug Cable Pinout" - what I ended up with as working, looks very different to me, from what it says in the datasheet. I really don't get how they ended up writing that pinout to "Figure 3-2. The signals are GND, TCK, VTref, TDO, TMS, and TDI. With the AVR Dragon, you will require six wires at minimum to link it to the target board. You should choose based on the target board’s pinout as opposed to the target MCU type. These are the ARM Cortex and AVR JTAG headers. and FINALLY I could read the device signature from the ATmega128 board! However, the two adhere to two separate pinouts. Oh, well - then I tried with beeper to find how does the red wire map to the connector marked A (the top one) - and it turns out, it is not bottom right pin, as in the above image - it is top left! So, I tried to map the pins correspondingly: The red wire marks pin 1 of the 10-pin 50-mil connector. The connector markedĪ plugs into the debugger while the B side plugs into the target board. The figure below shows the pinout of the debug cable. The right of the keying when the connector is seen from the cable. Pin 1 of the 6-pin 100-mil connector is placed to Pin 9: TDI -> PF7 (ADC7/TDI), pin 54 of ATMEGA128īeeper says connections are fine, no shorts try again: "outside selected device's operating range"Īs I was about to lose my mind, I thought, I'll reread Atmel-ICE_UserGuide.pdf once more, and had to focus on this, as I'd already focused on everything else: Pin 5: TMS -> PF5 (ADC5/TMS), pin 56 of ATMEGA128 Pin 3: TDO -> PF6 (ADC6/TDO), pin 55 of ATMEGA128 Pin 1: TCK -> PF4 (ADC4/TCK), pin 57 of ATMEGA128 Rewired again as in above picture, measured with a multimeter beeper, that the pins in the above image are connected to pins of ATmega128: Repeat above some 10 times - "outside selected device's operating range"Ĭame up with some sort of a mirroring of the pin connections, try again - "outside selected device's operating range" Reread Atmel-ICE_UserGuide.pdf, rewire again, measure with beeper just to make sure - "outside selected device's operating range". OK, seemingly the VTG pin is supposed to sense power supply, I measure that with a scope, it tells me 3.3V constant. Read voltage 1.2V is outside selected device's operating range: 2.7 V to 5.5 V. The "squid" cable attached to the header on the bottom, and I confirmed I have the right connections (and no shorts) with a multimeter beeper.Īll good so far - but then, I tried to do Atmel Studio / Tools / Device Programming: choose Tool: Atmel-ICE Device ATmega128 Interface: JTAG -> Apply then can click Device signature: Read - and I got: So first, I looked at the Atmel-ICE_UserGuide.pdf looking at the pinout there, I derived the following pinout, shown on this image - the diagram piece is taked verbatim from the Atmel-ICE_UserGuide.pdf (apparently, version Atmel-42330C-Atmel-ICE_User Guide-10/2016), "Figure 3-2. Then the time came to connect it to the ATmega128 based board. So, as I had the Basic Kit, I had to basically build my own "mini squid cable", which was not too complicated in itself. Unfortunately, as it says in, this chip supports only ISP interface (so no PDI, thus no debug breakpoints), and JTAG as the only option that can both program and do breakpoints debugging. Then, I had to work with a ATmega128 based board, and so it happened, I needed to debug it with breakpoints. In these cases, I've always used the PDI interface, which allows both programming, and debugging with breakpoints.
So far, I've worked with XMEGA chips, CodeVisionAVR to compile, Atmel Studio 7 (Windows 10) to debug - and never had a problem. Well, I'm using an Atmel-ICE, however, I was given the Basic kit. Had quite a problem with this, so I thought I'd document it.
0 kommentar(er)
