problem with JTAG configuration cyclone III

[Altera_Forum]

Hi, 

I am using cyclone III, epc3c10k. I have a problem of configuration from JTAG using usb blaster. Where I try to programming from quartus the JTAG can’t work. I connected the  ncomfig, nstatus, DCLK, DATA0,Conf_Done and Iit_Done to 3.3V with resistance of 10 kohm. I connected the: MSEL0, MSEL1, MSEL2, CLKUSR, NCE to the GND. nCEO unconnected. For JTAG connection: TMS and TDI to 2.5V (VCCA) with 10 kohm resistance TCK to GND with 1 kohm resistance. Pin 4 of JTAG Vtrgt to VCCA (2.5 V). Pin 2 and 10 of JTAG to GND. TDO to TDO of FPGA. When I powred the FPGA of 1.2 V (VCCINT) and 3.3V (VCCIO) and 2.5 V (VCCA) I have 1.06 v in the IO pin of FPGA, and 1.83v of TCK pin. I thing in normal case the voltage of IO pin should be about 3 V. I don’t know where is my mistake in the configuration or the schematic.  thanks for your help. i attached the configuration photo.

[Altera_Forum]

Have you assembled this board yourself? Could there be a build fault? 

 There's nothing obviously wrong with the parts of the schematic you've posted. Are you confident you've connected up all the power pins as required?  Perhaps you can post some more of the schematic.  Cheers, Alex

[Altera_Forum]

--- Quote Start ---  Have you assembled this board yourself? Could there be a build fault?  There's nothing obviously wrong with the parts of the schematic you've posted. Are you confident you've connected up all the power pins as required?  Perhaps you can post some more of the schematic.  Cheers, Alex  --- Quote End ---    Thanks for your reply.  yes i assembled this board myself.  I powred all the power pin as recommended in the guide connection of altera cyclone III family. i cheked all the connection of the pin 10 times. aall of theese are good connect. i tried to power the IO voltage about 2.5 V, it doesn't work also.  are you have any idea, about this problem ? thanks.

[Altera_Forum]

If you think all the power pins are connected correctly (as well all the ground pins) and you've done everything else you've stated then I would expect the device to appear in the JTAG chain when you scan it. 

 So, I can only recommend you check the resistor values you've actually fitted as the pull-up and pull-downs. If you've mistakenly fitted a value that's too low it certainly won't work.  I still suggest you post more of the schematic, just in case it reveals anything.  Beyond that, I'm afraid I suspect damage to the device :oops:  Cheers, Alex

[Altera_Forum]

I see it's TQFP144 package. Did you connect the exposed pad to ground?

[Altera_Forum]

thanks for your answers, 

 yes it's TQFP144 package. no the unused pad still in air (unconncted). more description : when i powred the card VCC I/O 2.5 V i have 1.1 V in all the I/O pin (i thing it should be between 2 and 2.5 V) and TCK voltage 1.83V (it's good value). But when i increase the voltage to 3.0 V or 3.3V i have 0 V in all the I/O pin and the TCK voltage 1.3 V. and the current of VCC I/O 3.3V about 20 mA, when i powred 2.5 V the current about 18 mA. (i thing the current should be about 5 mA).  Thanks for your help.

[Altera_Forum]

I'm having the same problem with a EP3C5E 144 pin package. No detection of JTAG. Its wired up the same way, only MSEL1 is pulled high to 3.3 volts which I think should be ok. 

 Our circuit board guy forgot to put the ground pad under the chip, so I'm thinking that's the problem with our board. The Altera documentation states: "The E144 pin package has an exposed ground pad at the bottom of the package. This ground pad is used for electrical connectivity . You must connect the ground pad to the ground plane of the PCB". However, when I poke a wire thru a over that ground pad it seems it is connected to ground already. Maybe thru other ground pins? In any case, I'm not sure if that is the problem with our board or not.  We have a 30 dollar Chinese eval board that we got here: http://www.wvshare.com/product/coreep3c5.htm and have pretty much copied the connections on our prototype. The eval board connects and works great, so I know our programmer and quartus works fine. They have grounded that ground pad, but I'm not sure that is the problem or not.  I guess I will keep looking for why our board wont communicate, the only thing I can come up with is that ground pad though.

[Altera_Forum]

I don't believe the lack of connection to the exposed pad will prevent a correctly powered device from being discoverable via JTAG. 

 I do base this on experience and a board we successfully had working, despite the lack of thermal pad on the PCB. This was using a Cyclone IV device. It could be that the thermal pad is connected differently internally to that of Cyclone III devices. However, I doubt it.  Some of the voltages you're reporting seem a little strange too based on how you are powering the device. Unfortunately, I think this is more likely to be damage to the device. Sorry.  Cheers, Alex

[Altera_Forum]

YES the lack of ground pad will prevent detection via JTAG! 

I had the same problem that this guy had, the circuitboard guy forgot to put a pad down and couldn't connect to jtag. I had troubles drilling a hole (went too deep thru the board into the chip!) , plus in a 6 layer board how do you not short some of the layers when you solder the wire in the hole? Cant get enough heat down there anyway. So I got another board and Solderd a wire to the tiny tiny corner of the chip where the leadframe is molded. (See leadframe diagram) Each corner has a tiny bit of metal that is connected to the bottom ground pad. I measured on an ohmmeter that the ground pad is NOT connected to any ground pins but is connected to the leadframe metal corners. Anyway its extremely fragile but now it connects via JTAG!!!!!!!!!!!! I don't get it but that ground pad must be different than any other ground on the chip......I might try to get a less fragile connection by dremelling out some of the epoxy corners to expose more of the leadframe so I have more to solder to.