Author Topic: Startup/JTAG poblem on TE0703+TE0720  (Read 3690 times)

he

  • Active Member
  • *
  • Posts: 1
Startup/JTAG poblem on TE0703+TE0720
« on: November 12, 2015, 11:04:14 AM »
The setup (TE0703+TE0720) worked fine for several months, e.g. programming the Zynq over JTAG. Yesterday I measured some GPIO signals on the VG96 connector pins with an oscilloscope. Now the board has some strange behaviour if it's powerd up. It could be possible that the adjacent GND and VCCIO35 pins were shortly connected. (VCCIO35 is connected with M3.3VOUT on the bottom of the carrier board.)

After power-up:
 - TE0720: LED1 and LED2 are ON and LED3 is very slightly ON
 - TE0703: D1 is shortly ON, D2 is ON, D3 is very slightly ON, D4 is OFF

Before this incident there was the correct behaviour where the zynq boots from SPI flash and the LED2 was blinking.

Based on the TE0720 LED status codes the LED3 should light up at least for short time at power-up otherwise there is a power supply problem. On the M3.3VOUT pin I measured 0.7V. I don't know if there should be 3.3V.
The status code where all three LEDs are ON says RESIN input is active low but I didn't change the switching buttons on the carrier board. Bank B34 is also supplied through the Jumper J5.
I tried also a other usb cable and a other power supply but there is no difference.

If the board is connected through USB to the PC it's not possible to connect through JTAG (with XMD console) or get a connection in the SDK terminal.
I found this topic http://forum.trenz-electronic.de/index.php/topic,317.0.html. Therefore, I tried the programs ToolZ and Lattice Design Programmer to detect a JTAG connection. In the Device Manager the usb connection works.

With Toolz  it is sometimes possible to connect on Channel A or Channel B but you only see in the message window the name (JTAG-ONB4), the ID (50800157) and vendor product (TE0703-0-0-0 or PP03-04). It is not possible to detect the chain or see informtion in the tree or in the properties. Otherise the message "FTDI with empty User EEPROM" is printed.
With Lattice  it is sometimes possible to see in the output that a board with FTDO USB Host Chip is detected and a HW-USBN-2B (FTDI) cable at port FTUSB-0 is connected. If the cable is selected the program doesn't respond anymore. Otherwise no cable is detected.

I also changed the S2.3 switch (default: S2.3 is ON) and tried both programs but there is the same behaviour.

There is also quiet monoton sound somewhere on the board.

So, I don't know if there is a hardware problem or the CLPD+Flash is erased.
Maybe someone could suggest some pins on the board where I can check the correct power supply or how to check a correct JTAG connection.

Thanks.
« Last Edit: November 12, 2015, 11:15:42 AM by he »

Antti Lukats

  • Hero Member
  • *****
  • Posts: 542
Re: Startup/JTAG poblem on TE0703+TE0720
« Reply #1 on: November 12, 2015, 11:28:58 AM »
The setup (TE0703+TE0720) worked fine for several months, e.g. programming the Zynq over JTAG. Yesterday I measured some GPIO signals on the VG96 connector pins with an oscilloscope. Now the board has some strange behaviour if it's powerd up. It could be possible that the adjacent GND and VCCIO35 pins were shortly connected. (VCCIO35 is connected with M3.3VOUT on the bottom of the carrier board.)

After power-up:
 - TE0720: LED1 and LED2 are ON and LED3 is very slightly ON
 - TE0703: D1 is shortly ON, D2 is ON, D3 is very slightly ON, D4 is OFF

Before this incident there was the correct behaviour where the zynq boots from SPI flash and the LED2 was blinking.

Based on the TE0720 LED status codes the LED3 should light up at least for short time at power-up otherwise there is a power supply problem. On the M3.3VOUT pin I measured 0.7V. I don't know if there should be 3.3V.
The status code where all three LEDs are ON says RESIN input is active low but I didn't change the switching buttons on the carrier board. Bank B34 is also supplied through the Jumper J5.
I tried also a other usb cable and a other power supply but there is no difference.

If the board is connected through USB to the PC it's not possible to connect through JTAG (with XMD console) or get a connection in the SDK terminal.
I found this topic http://forum.trenz-electronic.de/index.php/topic,317.0.html. Therefore, I tried the programs ToolZ and Lattice Design Programmer to detect a JTAG connection. In the Device Manager the usb connection works.

With Toolz  it is sometimes possible to connect on Channel A or Channel B but you only see in the message window the name (JTAG-ONB4), the ID (50800157) and vendor product (TE0703-0-0-0 or PP03-04). It is not possible to detect the chain or see informtion in the tree or in the properties. Otherise the message "FTDI with empty User EEPROM" is printed.
With Lattice  it is sometimes possible to see in the output that a board with FTDO USB Host Chip is detected and a HW-USBN-2B (FTDI) cable at port FTUSB-0 is connected. If the cable is selected the program doesn't respond anymore. Otherwise no cable is detected.

I also changed the S2.3 switch (default: S2.3 is ON) and tried both programs but there is the same behaviour.

There is also quiet monoton sound somewhere on the board.

So, I don't know if there is a hardware problem or the CLPD+Flash is erased.
Maybe someone could suggest some pins on the board where I can check the correct power supply or how to check a correct JTAG connection.

Thanks.

All the symptoms indicated fatal failure. Either base or both are dead.

A short circuit VCCx - GND is very dangerous, - there can be very large spikes due to the capacitance on the bus, this can and in many case will destroy some IC's

I am sorry to say, but one single one time HARD short any_supply to GND may destroy xx IC's on board.

You can try to talk to TE0703 when the module is removed, to see how much of the base is alive. You should be able to see the CPLD on base and that should be stable always, if not then base is damaged.

It looks that both are damaged, base and module. There is not much I can help here, try the base first, if it works (without module) then maybe only the module is damaged.