Zedboard power up failure, IC27

[bacif]

Hello all,

My Zedboard does not properly power up, power OK LED does not light up, but all the user LED's light up red. I debugged the problem that IC27 (Vadj generator) has no voltage at its output. Before the problem occurred, J18 was set to have 3.3V at Vadj. There was an Inverium FMC mezzanine card (TB-FMCL-PH) attached to the FMC connector, and there was a custom made ADC board connected to the PMOD B connector of Zedboard. In the attached photo you can see the whole experimental setup. In the photo, "Digilent" corresponds to Digilent Analog Discovery 2, "level shifter IC" on the breadboard (in the middle) and the receiver IC in the converter (right of the photo) are HEF4104B. The problem occurred after I used single ended digital inputs of Analog Discovery 2 to see the LVCMOS 3V3 outputs of Zedboard on the FMC mezzanine card.

As I understand from other posts in this forum and some others, either the IC27 or the whole Zedboard should be replaced. But before doing that, I need to understand the problem that lead to burn IC27 not to repeat it again. Can you please help me to find out?

Baris

[JColvin]

Hi @bacif,

If you just using the logic analyzer or the oscilloscope tools on the Analog Discovery 2, then I think it would be very unlikely that connecting those pins would cause an issue, especially considering that you already have the grounds connected. Do you happen to have a picture of the physical connection of the Analog Discovery 2 to your system?

Because you said that you had VADJ set to 3.3V, I presume that means you soldered on your own connector for it? Was there a particular reason you needed VADJ at 3.3V? The reason I ask is because the mezzanine card you linked to appears to just be a breakout board for a FMC LPC connector. Additionally, I am a little confused by your picture; where exactly is the mezzanine card? I realize that it would be attached to the FMC connector, but it looks like in your image that the Analog Discovery 2 is instead attached to the HEF4104B? Is this one way voltage translator connected to anything?

I realize it's a custom made ADC board so there likely isn't a datasheet or general schematic available for it, but what logic type and voltage levels does it use? I ask because the Pmod ports run at 3.3V CMOS and that is the same voltage rail that powers the "Power good" LED as well as the bank that controls the user LEDs.

In the interest of determining what all is happening on the power regulators, could you check the voltage on the following capacitors while the system is connected and powered?

Near IC27 on the bottom left on the underside of the board:
- C330 (should be 5V):
- C335 (should be 3.3V):
- C340 (should be whatever VADJ is set to):
Near IC28 on the bottom right on the underside of the board:
- C351 (should be 1.5V):
- C356 (should be 1.0V):
Near IC29 on the center right on the underside of the board:
- C368 (should be 5V):
Near IC30 on the top left on the topside of the board:
- C366 (should be 1.8V):
- C369 (should be 5V):
Near IC30 on the top right on the underside of the board:
- C388 (should be 12V):
- C392 (should be 5V):

Thank you,
JColvin

 

 

[bacif]

Hi @JColvin

Thank you for the response. Here are my replies:

Yes, the idea to use Analog Discovery 2 is to see the LVCMOS outputs driven by Zedboard over FMC mezzanine card (and yes, it's a breakout board for a FMC LPC connector) by its oscilloscope inputs or digital inputs. You can see the physical connection of the setup in the attached photo. On the breadboard, there is level shifter IC HEF4104BP. Zedboard outputs are connected to the primary side of HEF4104BP as well as Analog Discovery 2 oscilloscope or digital inputs. The secondary side of IC is connected to the D-sub connector on the right hand side which drives another HEF4104BP in a power converter. This secondary HEF4104BP is supplied by 5V in its primary side, that's why I needed to level up the 3.3V output of Zedboard FMC outputs with a previous HEF4104BP which you see on the breadboard.

Yes, J18 is soldered to set Vadj to 3.3V. (In the photo the jumper is set to have 2V5, just to check if the board was working with different Vadj setting, since another board with the same problem was able to turn on with 1V8 setting.)

The custom made ADC board contains an ADAS3023. The board is powered both by 3V3 over PMOD and by an external 5V (top right side of the board, lacks the ground cable on the photo). 

The power regulators are all good except C340 and C368. They are both 0V.

Do you have any more comments?

Baris

[JColvin]

Hi @bacif,

Based on your picture and the voltages you gave me for the capacitors, my best guess at the moment is that some ESD was applied to the FMC connector (which is much easier to do with the FMC breakout board you have attached with all of the exposed pins and the FMC connector in general is very sensitive to this, i.e. no protection, to maintain high speed data throughput), causing the damage you experienced. I'm not certain where the connections have all been made with regards to connecting to the 3.3V to 5V on the initial HEF4104BP, but from what I can tell they look reasonable to me (i.e. nothing jumping out at me with regards to power and ground shorting each other out).

The good news is that I'm also fairly sure that it is just IC27 that is damaged (as opposed to the entire Zedboard) as the two capacitors that were not at the expected voltage were C340 (measuring the VADJ voltage) and C368 which measures PG-ALL. PG-ALL (which controls the "power good" LED, hence why it is off), is controlled by a couple of voltage rails (both of which were fine based on the capacitor voltages I requested you check) and VADJ-FB, which if the VADJ supply is not working, would likely lead to the feedback for it not working properly as well.

For peace of mind though, it wouldn't hurt to load a different project onto the Zedboard to confirm that the other components are working as intended. One aspect you may not be able to readily test is the PS side of things as PG-ALL does also control Power-On-Reset button for the PS, so if that is not properly being initialized upon power up, it probably will not work as intended.

From your second post though, it seems that you have a second Zedboard with a similar set of issues?

Thanks,
JColvin

[xc6lx45]

Hi,

just a thought, looking at your diagram from a large distance.

Most likely you have some power supplies with two pins (non-grounded). The problem is that the output is floating at half the mains voltage, set by a high resistance (megaohms) voltage divider. Ironically, this is to protect the power supply against ESD / charge buildup on the secondary side that could break the transformer's insulation.

With such a supply, if you accidentally disconnect the ground connection to your circuit, you have half the AC voltage on the supply pin.

I'd double-check all involved power supplies and make sure your connection scheme has a well-established ground even if some random cable comes loose.

[bacif]

Hi @JColvin

Thanks for the reply.

You can see the connection to HEF4104BP on the schematic attached. The power supply has floating dc outputs, and I shorted the grounds on the supply side.

Unfortunately, I am not able to load any project onto Zedboard after this error. Please see below for the error message I got when I try to open hardware in Vivado:

open_hw_target
INFO: [Labtoolstcl 44-466] Opening hw_target localhost:3121/xilinx_tcf/Digilent/210299A1FCF3
ERROR: [Labtools 27-2269] No devices detected on target localhost:3121/xilinx_tcf/Digilent/210299A1FCF3.
Check cable connectivity and that the target board is powered up then
use the disconnect_hw_server and connect_hw_server to re-register this hardware target.
ERROR: [Labtools 27-55] Invalid index 0 passed to getDevice

Yes, I have another Zedboard with the same problem, IC27 failure. The process was like this: the setup is similar to the one I posted above except there is no breadboard, level shifter IC and D-sub connector and the attached power converter to D-sub. I run the project on Zedboard, supply some input signal to the ADC board over Analog Discovery 2 and measure FMC outputs by Analog Discovery 2. I used single ended digital inputs for measuring. It was OK with this desktop test but when I shifted to the experimental setup and even before anything is connected to the Zedboard, I got the same power up failure. However, for this board, during my desktop tests, J18 was set to have 2V5 at Vadj but I put LVCMOS33 on the constraint file for FMC output pins. So, maybe that was the problem leading to IC27 failure.

Baris

[bacif]

Hi @xc6lx45,

Thanks for your comments and advices. Yes, the power supply has floating dc outputs, and I shorted the grounds of the outputs on the supply side. The 5V output was feeding secondary (B) side of HEF4104BP and custom made ADC board, while 3V3 output was feeding primary (A) side of HEF4104BP. In such a case that you described, I think I would have some other components failed depending on where the ground has been lost as well as IC27 of Zedboard. I will give a check to the others.

Baris