Actions
Forum Thread Details
  • State Verified Answer
  • Replies 4 replies
  • Subscribers 354 subscribers
  • Views 586 views
  • Users 0 members are here
Related

Improving the 1 volt supply

Former Member

My design has a configurable number of compute engines.  When I try to use too many of them I exceed the limits of the 1 volt power supply.

I can load the FPGA, but when the PS commands the start of computation, the 1v supply drops to around 0.6v and, I think, the board resets.

Vivado says:
  vccint:  2.419 A
  vccbram: 0.013 A
  vccpint: 0.819 A

I'd like some help to replace or augment the power supply.
I thought I could use something like the
Murata OKR-T/6-W12-C
http://power.murata.com/en/products/dc-dc-converters/single-output-non-isolated/okr-t-6-w12-c.html

Some questions:
o  Is there a better approach to the problem?
o  Is this feasible with with fair soldering skills and no microscope?
o  Is there a better module that the Murata?
o  Do I have to disconnect the original MAX15021?
o  If so, what's the best way.
o  Will the enable signal be compatible?
o  Can you suggest mounting?

Thanks,
Michael

Parents
  • 0

    Hey Michael,

     

    While not ideal, since the MAX15021 is a synchronous device connecting another supply in parallel with it should be ok (as long as it is also a synchronous device).  The two supplies will essentially form an uneven current share. 

    The easiest way to do what you want to do would be to get an EVM that you could wire in to the output side of the output inductor L5.  Any EVM would work, here's an example - http://www.ti.com/tool/tps54425evm-608  You could also use a bench supply set to 1V.  The load will draw current from whichever voltage is higher, when it dips, the other supply will provide the power.  Again, this isn't ideal and with a bench supply you want to make sure sinking current won't damage it.

    Just be sure the output is set to 1V and if it isn't, modify the circuit so that it is correct.  After that, connect the input side of the EVM to the 12V on ZedBoard and make sure the enable is wired to Vin.  That will make the 1V supply come up right away (which will satisfy sequencing since it must come up first) and will increase the current capacity of that rail.  Ideally you would disable the 1V supply on the MAX15021, but to do so you would need to lift pin 25 on the device and tie it to GND. 

    Let me know if that makes sense and if you have any other questions.

    Thanks

    Chris

Reply
  • 0

    Hey Michael,

     

    While not ideal, since the MAX15021 is a synchronous device connecting another supply in parallel with it should be ok (as long as it is also a synchronous device).  The two supplies will essentially form an uneven current share. 

    The easiest way to do what you want to do would be to get an EVM that you could wire in to the output side of the output inductor L5.  Any EVM would work, here's an example - http://www.ti.com/tool/tps54425evm-608  You could also use a bench supply set to 1V.  The load will draw current from whichever voltage is higher, when it dips, the other supply will provide the power.  Again, this isn't ideal and with a bench supply you want to make sure sinking current won't damage it.

    Just be sure the output is set to 1V and if it isn't, modify the circuit so that it is correct.  After that, connect the input side of the EVM to the 12V on ZedBoard and make sure the enable is wired to Vin.  That will make the 1V supply come up right away (which will satisfy sequencing since it must come up first) and will increase the current capacity of that rail.  Ideally you would disable the 1V supply on the MAX15021, but to do so you would need to lift pin 25 on the device and tie it to GND. 

    Let me know if that makes sense and if you have any other questions.

    Thanks

    Chris

Children