Forum Discussion

Altera_Forum's avatar
Altera_Forum
Icon for Honored Contributor rankHonored Contributor
18 years ago

RIPPLES IN SUPPLY OF FPGA_plz help

HI everyone.....

Iam using a stratix II device. Before programming the PLL section of FPGA, supplies are observed

correct i.e. 1.21V and 3.29V.

As soon as I program the device, the supplies are showing high ripples in both 1.21V as well as 3.29V and voltage levels are coming down to 1.12V with ripple of 600mV

and 3.16V with ripple of 1.40V respectively.(appears like an AC signal!).

Why this fellow is misbehaving after programming:confused: ? And how can I rectify the problelm? plz help.

3 Replies

  • Altera_Forum's avatar
    Altera_Forum
    Icon for Honored Contributor rankHonored Contributor
    18 years ago

    Questions:

    Is it just the PLL section supply that ripples? If you don't use the the PLL section can you use the device ( with restricted clock capabilties )?

    If it is the first time you try to activate the whole device some outputs may be short circuit or the supply can't source enough current for the whole device.

    Do you use sperated supplies for the PLL section?

    Are the PLL supplies decoupled close to the device?

    Is the source impedance of the supplies low enough?

    Did you accidentally create a resonance circuit to filter the supply your PLL section: EG used a relative high Q inductor ( 10 uH || 100nf+10nF )? In that case you can just replace the inductor with a ferrite!

    Succes with debugging,

    Andries
  • Altera_Forum's avatar
    Altera_Forum
    Icon for Honored Contributor rankHonored Contributor
    18 years ago

    Hello, I also use a stratix 2 device (speed 3 bg1508).

    The voltage drop is quite high on this device, especially if the dc to dc are far away(15 cm)

    from the fpga.

    On my board, the drop is about 0.2V.

    To fix this, you can increase the voltage at the output of your dc to dc.

    Personnaly I have 1.4 V after the fuse of the dc to dc, and 1.2 V at the fpga side.

    Rippling is caused by a bad configuration of your dc to dc, we had the same issue.

    You probably forgot an extra rc at the dc to dc chip.

    Before we had a big rippling, and the fpga was far from stable (especially the clock after two pll in cascade).

    And before, when I had 1.0V at the fpga side, the entire device was not stable

    (because I'm just at the clock max f in the timing report).
  • Altera_Forum's avatar
    Altera_Forum
    Icon for Honored Contributor rankHonored Contributor
    18 years ago

    The ripple you are seeing could come from a number of contributors. Most often dc-dc supply is undersized (LC too small or freq too low). What frequeny is the ripple you see? Less than 1MHz your dc-dc convertor is the prime suspect.

    You should also make sure you don't have any shorted i/o. I always tell Quartus to set all unused io to tristated inputs rather than the default outputs driving ground.

    If the ripple/noise you see is much higher like 10's of MHz+ then you should examine your power supply filtering and decoupling.

    Hope this helps,

    Chris