Forum Discussion

SSenn's avatar
SSenn
Icon for New Contributor rankNew Contributor
6 years ago

EPCQL1024 doesn't response anymore, only with 0xFF

Hi,

I have a critical Problem with an Intel EPCQ-L Serial Configuration Device.

When I try to configure the EPCQL via the "JTAG indirect configure" feature

(JIC file), the EPCQL only returns 0xFF after a while. Therefore Quartus isn't

able to read e.g. the chip ID (0x21). This state is held until the power is

switched off. After a new power-on-reset everything works fine again until the same failure occurs again after some time.

This error happens only sporadically. I configure the EPCQL via the "JTAG indirect configure" feature every night and approximately once a month the described error occurs.

EPCQL: EPCQL1024 (Active Serial x4)

FPGA: Arria 10 GX (10AX115N2F40I1SG)

Quartus: 18.0.1 Build 261/28/2018

It seems to me that the EPCQL has some kind of 'dead lock'. This 'dead lock'

remains until the power is switched off but we cannot do that with our

customers. The good thing is that the data isn't corrupted after the error.

Everything works well after the power off.

Do you have any Ideas how we can fix this problem?

Is it possible to restart the EPCQL without to turning off the whole device?

Is this a known issue?

Many Thanks

5 Replies

  • YuanLi_S_Intel's avatar
    YuanLi_S_Intel
    Icon for Regular Contributor rankRegular Contributor
    6 years ago

    Hi Sebastian,

    This is not a known issue and is our first time hearing this. May i know how many unit is having the issue?

    Thank You.

  • SSenn's avatar
    SSenn
    Icon for New Contributor rankNew Contributor
    6 years ago

    Right now, we've only seen this behavior on one board. This board is part of our test setup.

    But I have to add that this is COST hardware (no in-house development) and this behavior is also new for the manufacturer.

  • ak6dn's avatar
    ak6dn
    Icon for Regular Contributor rankRegular Contributor
    6 years ago

    Typically to resolve an issue like this I would desolder the EPCQ part from the 'bad' board and swap it with an EPCQ part from a 'good' board.

    The error will follow the part if the EPCQ part is bad, or it the error stays on the board the FPGA it is attached to is likely bad.

    If the error follows the EPCQ part then I would return it to the manufacturer for F/A.

    If you are just a s/w development house (no h/w capability) I would return the bad board to the manufacturer as they would have to follow the above debug process.

  • SSenn's avatar
    SSenn
    Icon for New Contributor rankNew Contributor
    6 years ago

    First of all many thanks to YL_Intel and ak6dn.

    Today we saw the same error on a second board in a completely different test setup (but the same type of board). Therefore I would say that it isn't a broken IC. OK, maybe both boards are broken but this is very unlikely.

    I'm in touch with the manufacturer, but he is assuming there is a software problem.

    Is it possible to reset the EPCQL with a special command? I haven't found anything like this in the datasheet yet.

  • YuanLi_S_Intel's avatar
    YuanLi_S_Intel
    Icon for Regular Contributor rankRegular Contributor
    5 years ago

    Hi Sebastian,

    There is no way to reset it with special setup or command. You can perform DIE erase only if you are able to connect with the flash.

    Thank You.