Is chip fried? Can't access JTAG chain

[Altera_Forum]

I'm new to Cyclone IV design... I accidentally applied 3.3V to the core of a Cyclone IV today, so it's very possible that the chip is fried. When I try to configure it, I get the "Can't access JTAG chain" message. Would that be caused by a fried chip, or is there another problem? Thanks in advance

[Altera_Forum]

--- Quote Start ---  I'm new to Cyclone IV design... I accidentally applied 3.3V to the core of a Cyclone IV today, so it's very possible that the chip is fried. When I try to configure it, I get the "Can't access JTAG chain" message. Would that be caused by a fried chip, or is there another problem? Thanks in advance  --- Quote End ---  Its highly likely you have destroyed the device.  Did you try to measure the power supply voltages? Is the core at the correct voltage after this accident?  There is a JTAG debugger tool in Quartus. You could try toggling TCK, TMS, and TDI and then probe at the FPGA to see if the signals are toggling there. If they are, and TDO does not toggle, then your chip is dead.  If you have another version of this Cyclone IV board, then measure the supply rail impedance. If your toasted board has significantly different impedance than a good board, then thats also a bad sign.  Don't worry, I've made similar mistakes, generally with scope probes accidentally shorting traces together, eg., MAX II CPLDs don't like 12V on their I/O pins either :)  Cheers, Dave

[Altera_Forum]

Thanks Dave. That was good info. The supply voltages are good. And the supply rail impedance is practically 0. So maybe the chip's alright 

 When I step through the JTAG state machine in the debugger, I don't see TCK or TDI change. TDO and TMS do change, but they only fluctuate between ~0.7V and 0.9V. (I'm expecting 2.5V)   So I'm thinking the USB-Blaster that I bought on eBay is the problem. Cheap Chinese knockoff...

[Altera_Forum]

--- Quote Start ---  That was good info. The supply voltages are good. And the supply rail impedance is practically 0. So maybe the chip's alright    --- Quote End ---  This information is inconsistent. If an ohm-meter indicates 0-Ohms between the core voltage and ground, then its a short, and you would not see a voltage when you turn the supply on. (You be making the impedance measurement with the power off).  The I/O impedance can sometimes be close to 0-ohms if you have a lot of bus terminations, but for this test, you just want to look at the core.    --- Quote Start ---   When I step through the JTAG state machine in the debugger, I don't see TCK or TDI change. TDO and TMS do change, but they only fluctuate between ~0.7V and 0.9V. (I'm expecting 2.5V)   So I'm thinking the USB-Blaster that I bought on eBay is the problem. Cheap Chinese knockoff...  --- Quote End ---  That definitely sounds like an issue. The Terasic USB-Blasters are cheaper than the Altera ones, and I've never had a problem with them.  Cheers, Dave

[Altera_Forum]

--- Quote Start ---  This information is inconsistent. If an ohm-meter indicates 0-Ohms between the core voltage and ground, then its a short, and you would not see a voltage when you turn the supply on. (You be making the impedance measurement with the power off).  --- Quote End ---     What I meant about 0-ohms was just measuring the 3.3V-rail alone, from one end to the other. If I go 3.3V-to-ground, it's kind of weird. The reading starts at 0.5-Mohms and slowly climbs up, until displaying Overload. I guess that means I have a really weak short between 3.3 and ground? Probably another sign the chip is busted.   Thank you very much. I'll suck it up and buy a new one now :/

[Altera_Forum]

--- Quote Start ---  What I meant about 0-ohms was just measuring the 3.3V-rail alone, from one end to the other. If I go 3.3V-to-ground, it's kind of weird. The reading starts at 0.5-Mohms and slowly climbs up, until displaying Overload.   --- Quote End ---  That's actually a good sign; you don't have a short between power and ground. It just means that your voltmeter is charging up the bulk capacitance. Try turning the voltmeter leads around so that the ground lead (black) connects to VCCINT and the red lead connects to ground. That sometimes makes the reading settle down quicker. The key thing is that its not 0-ohm. It doesn't mean its ok, since it could have blown the circuit open, rather than shorted it.   By the way, you need to measure the impedance of the VCCINT core voltage. The 3.3V sounds a bit too high. Its more likely that that is an I/O voltage. Look at the circuit diagram, and measure the impedance of the core voltage from the output of the regulator; actually, check all the power supplies.    --- Quote Start ---   I'll suck it up and buy a new one now   --- Quote End ---  Buy yourself a new USB-Blaster while you are at it :)  Cheers, Dave