FPGA, SoC, And CPLD Boards And Kits
FPGA Evaluation and Development Kits
6334 Discussions

Problem DevKit Agilex 5 - MK-A5E065BB32AES1

robertzab
Beginner
393 Views

Hi,

 

I have an issue with the DevKit - MK-A5E065BB32AES1.

Today, we received the product. The device was connected to the power supply – switch SW2 set to POS-3 – ATX. JTAG was connected. The "Board Test System" program was launched to check the system.

In the "Diagnostic" tab, many modules are unresponsive – indicators are red.

 

agilex5_1.png

Next, I go to the "Power Monitor" tab, and I also do not receive any feedback.

agilex5_2.png

The "Clock Controller" tab also shows no response.

agilex5_3.png

 

In the documentation, I found information regarding the power supply. In my case, the LED DS2 on the board is not lit.

 

agilex5_4.png

 

However, the FPGA itself programs successfully.

What should I do in this case? Do you have any suggestions?

 

Quartus version: 24.3.0

Labels (1)
0 Kudos
3 Replies
robertzab
Beginner
333 Views

I analyzed the schematic and concluded that there is no communication between the SOM and the mainboard. Specifically, the SoM_Prsnt# signal, which detects that the module is inserted, is in an open state – as shown in the schematic below at this location:

agilex5_6.png

 

Following this lead, I inspected the mainboard itself and, unfortunately, discovered some issues.

Firstly, the mainboard is bent – to be honest, this could happen during production:

 

PXL_20250123_094516794.jpg

 

I also checked the connectors that link the SoM to the mainboard (J9, J13). Unfortunately, the ends of these connectors are not fully mated (not fully locked) – as shown in the attached photo. However, the connectors in the middle are properly aligned.

Unfortunately, I am unable to fully press them in – it seems that the spacers between the SoM and the mainboard prevent the connectors from being fully inserted.

 

PXL_20250123_100923335.MP.jpg

PXL_20250123_100936757.MP.jpg

 

PXL_20250123_100948995.jpg


Theoretically, I could disconnect the SoM from the mainboard, remove the spacers, and reconnect it, but I’d prefer not to do this since it’s a new purchase, and removing components from this type of connector is quite difficult. Unfortunately, I’ll have to return it to the distributor under warranty.

 

 

 

0 Kudos
Farabi
Employee
258 Views

Hello,


If you found defect of the newly purchased board, please return it to distributor, we will replace another one for you.


regards,

Farabi


0 Kudos
robertzab
Beginner
132 Views

Hello,

I received a new module from the distributor.

However, after powering it on and running the diagnostic test (board_test_system), I’m still getting a message that two devices are not responding—see the picture below.

 

agilex_5_1.png

 

But since I’m not going to use those devices for my project, I moved on to connecting the Texas ADC (ADC3669) through the FMC connector. That’s where I encountered another issue: the pins PIN_BH62 and PIN_BH59, which are set up as a differential pair, aren’t functioning properly as inputs/outputs—they’re not configuring correctly. I created a simple program for data collection and added a SignalTap instance, using the converter’s signal as the trigger clock.

Here is what I tested:

  • I set PIN_BH62 and PIN_BH59 as a differential output—drove them with a PLL clock and checked the oscilloscope. Result: they showed a constant high level.
  • I set PIN_BH62 and PIN_BH59 as a differential input—applied a clock signal and observed them in SignalTap. Result: they remained high.
  • I disconnected the converter from the FMC connector—according to SignalTap, the pins then read low.
  • I reconnected the converter to the FMC connector—set PIN_BH62 as a single-ended input and PIN_BH59 as an output. On the converter side, I tied them together with a resistor. Observing SignalTap while driving the output high or low showed that the input correctly followed the output state.
  • I rerouted the clock on the converter to one of the data lines and checked the waveform. This change allowed me to see the clock in SignalTap.


In summary: what could be the reasons why PIN_BH62 and PIN_BH59, when configured as a differential pair, are not configuring correctly?

 

0 Kudos
Reply