Browse si-list: This Month's Archive • Main Archive Page • Related posts • Previous by Date • Next by Date
[SI-LIST] Re: Will ODT on DDR2 work?

From: Jory McKinley <jory_mckinley@xxxxxxxxx>
To: billw@xxxxxxxxxxx, si-list@xxxxxxxxxxxxx
Date: Wed, 7 Feb 2007 08:20:33 -0800 (PST)
I have seen (in simulations and not in all cases) where using the sstl18 with 
DCI on the Virtex-5 during a write cycle creates timing issues due to the shunt 
resistor network always "active".  If your sims show this then using the sstl18 
without DCI may be acceptable however in the read cycle you will see some 
overshot assuming no external resistors.
-Jory

Bill Wurst <billw@xxxxxxxxxxx> wrote: Ludvik,

You actually want to use both ODT at the memory device (enabled only 
during a write) and the SSTL18 I/O in the Virtex-5 with DCI.  It's 
unfortunate that the FPGA vendors didn't follow the memory folks and 
implement DCI such that the shunt terminations could be switched in only 
during a read, since during a write they only waste a lot of power, 
attenuate the signal, and create a mis-matched line for any reflections. 
  As far as I know, Virtex-5 is no different than Virtex-4 in this 
regard.  While it's true that switching the terminations in and out 
could aggravate SSN, if it's implemented properly SSN effects are 
minimized.  Just the same, generally speaking you will want to use DCI 
unless you have a special case and simulations reveal you don't need it.

Regards,
     -Bill


       /************************************
      /         billw@xxxxxxxxxxx         /
     /                                   /
    / Advanced Electronic Concepts, LLC /
   /           www.aec-lab.com         /
   ************************************
=========================================================
Ludvik Kindl wrote:
> Jeff & all, 
> 
> Thanks for help to you and to other people who responded, but I have one
> more question. You say "Yes of course ODT is enabled only on receive at the
> Micron", but I think that on Virtex-5 they have DCI which can also be used
> for terminating bidirectional lines. Why would you see problem with using
> DCI? My main interest is to free-up our extremely congested real estate area
> - it looks worst that Manhattan at 5:30 PM - and also I am hoping to get
> better SI performance, since resistors will be closer to receivers.
> 
> Sorry troubling, but I do not want to discover that digital guys cannot
> terminate my line for some reason.
> 
> Ludvik
> 
> +++++++++++++++++
> 
> 
>>-----Original Message-----
>>From: si-list-bounce@xxxxxxxxxxxxx [mailto:si-list-bounce@xxxxxxxxxxxxx]
>>On Behalf Of Jory McKinley
>>Sent: Tuesday, February 06, 2007 11:39 AM
>>To: jeffjohnson@xxxxxxxxxxxx; si-list@xxxxxxxxxxxxx
>>Subject: [SI-LIST] Re: Will ODT on DDR2 work?
>>
>>Jeff,
>>Yes of course ODT is enabled only on receive at the
>>Micron.  I think the point here is that can we get
>>away without using any on board resistors by taking
>>advantage of the ODT that the Micron part provides.
>>Why? probably not for the 30 cent or so savings (you
>>never know) but more than likely the space required to
>>place resistors for all the Data, Strobe, DM traces.
>>-Jory
>>
>>
>>--- Jeff Johnson  wrote:
>>
>>
>>>Jory, Ludvik,
>>>
>>>Remember that ODT only works when the I/O is in the
>>>"I" state.  I do not
>>>recommend saving 30 cents of resistors for the
>>>trouble-shooting headache you
>>>will create.  Refer to JESD79-2C for more
>>>information on how DDR2 works.
>>>
>>>
>>>Date: Mon, 5 Feb 2007 18:14:57 -0800 (PST)
>>>From: Jory McKinley 
>>>
>>>Subject: [SI-LIST] Re: Will ODT on DDR2 work?
>>>
>>>Hello Ludvik,
>>>
>>>In my experience it always a good idea to simulate
>>>this transfer but we have
>>>found that using Full ODT drivers on the Micron side
>>>tend to create some
>>>undesirable Signal Integrity issues at the
>>>Controller (in your case the
>>>Virtex-V). In most cases a Half strength ODT driver
>>>will work however you
>>>will need to simulate this while checking for
>>>overshot and ringback at the
>>>Virtex-V. Also make sure that you have adequate
>>>crosstalk rules in place.
>>>
>>>-Jory
>>>
>>>Jeffrey Johnson, C.I.D.
>>>
>>>Module R&D Engineer
>>>
>>>NanYa Technologies - Texas LLC
>>>
>>>Tel: (713) 774-4113 Ext. 135
>>>
>>>FAX:  (713) 774-3819
>>>
>>> Share your knowledge. It's a way to achieve
>>>immortality.
>>>
------------------------------------------------------------------
To unsubscribe from si-list:
si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field

or to administer your membership from a web page, go to:
http://www.freelists.org/webpage/si-list

For help:
si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field


List technical documents are available at:
                http://www.si-list.net

List archives are viewable at:     
  http://www.freelists.org/archives/si-list
or at our remote archives:
  http://groups.yahoo.com/group/si-list/messages
Old (prior to June 6, 2001) list archives are viewable at:
   http://www.qsl.net/wb6tpu
  



 
---------------------------------
Check out the all-new Yahoo! Mail beta - Fire up a more powerful email and get 
things done faster.

------------------------------------------------------------------
To unsubscribe from si-list:
si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field

or to administer your membership from a web page, go to:
http://www.freelists.org/webpage/si-list

For help:
si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field


List technical documents are available at:
                http://www.si-list.net

List archives are viewable at:     
                http://www.freelists.org/archives/si-list
or at our remote archives:
                http://groups.yahoo.com/group/si-list/messages
Old (prior to June 6, 2001) list archives are viewable at:
                http://www.qsl.net/wb6tpu
References:

[SI-LIST] Re: Will ODT on DDR2 work?
From: Bill Wurst
[SI-LIST] Re: Will ODT on DDR2 work?

Other related posts:
