[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: TDI and TDO signal
- To: Mailing List Recipients <pci-sig-request@znyx.com>
- Subject: Re: TDI and TDO signal
- From: Timothy Hellman/PicTel <Timothy_Hellman@smtpnotes.pictel.com>
- Date: 22 Aug 96 9:02:10 EDT
- Resent-Date: 22 Aug 96 9:02:10 EDT
- Resent-From: pci-sig-request@znyx.com
- Resent-Message-Id: <"E3Lvm.0.j7.3m57o"@dart>
- Resent-Sender: pci-sig-request@znyx.com
I'm convinced that the JTAG signals on the PCI connector are useless unless
you're designing an embedded system and know exactly what cards are
going into it. As pointed out below, you can't use the scan chain on a
motherboard unless all the slots are populated, and even then you'll have
to count on all your add-in boards implementing the JTAG pins properly.
On the other hand, add-in board manufacturers are strongly discouraged
from actually connecting the JTAG pins to any JTAG devices on their
boards, because not all motherboards implement these lines properly
(as was discussed recently in this forum), thereby allowing accidental
triggering due to noise coupling.
It was a nice idea, though.
Tim
=========================
Hi All,
This is one area of PCI that has confused me, perhaps someone can clarify. Is
there some type of specification, that mechanically, a PCI socket will keep
TDI and TDO connected when a PCI card is not present? If not, then
I'm not convinced that connecting TDI and TDO together is the proper
solution, either. It would appear to me that PCI slots that are not populated
would make TDI and TDO appear open, so even if you connect TDI and TDO together
in your card, there's still an incomplete scan path when the system is not
fully
populated with PCI cards.
However, by connecting TDI and TDO together on your card, you have made your
card
incompatible in systems supporting a 1149.1 multidrop architecture. In a
multidrop
architecture, the TDI, TDO, TMS and TCK are common to all daughter boards in
the system.
Each daughter board requires a multi-drop interface chip (see TI '8996
addressable
shadow port or similar) in order to address the daughter boards in the system
and select
which board will drive on TDO during the SHIFT-DR or SHIFT-IR states of 1149.1.
One advantage would be that not all slots need to be populated to complete the
scan
chain.
Regards,
CJ
IEEE 1149.1 working group member (however, these are my personal opinions and
not the opinion of the working group)
>>
>>In <199608210937.CAA12145@scruz.net> "John R Pierce"
><uunet!scruznet.com!pierce> writes:
>>>Read the PCI 2.1 SPEC. Those books are not the SPECIFICATION.
>>
>>>But, yes, you hook them up. otherwise other boundary scan boards would
>fail if
>>>the system supported boundary scans...
>>
>>>-jrp
>>>(btw, your email address is broken, '@@' don't work!)
>>
>>>----------
>>>> From: Carsten Bode <cb@@Diehl.DE>
>>>> To: Mailing List Recipients <pci-sig-request@znyx.com>
>>>> Subject: TDI and TDO signal
>>>> Date: Wednesday, August 21, 1996 4:56 AM
>>>>
>>>> Hello
>>>> I have a simple question regarding the design of a PCI card ( standard PC
>>>format ) that does not
>>>> support boundary scan.
>>>> Should I connect TDI and TDO or leave them open or pull them up or down
with
>>>a resistor?
>>>>
>>>> I found different statements in the Mindshare PCI book ( connect TDI and
TDO
>>>) and
>>>> the PCI book from Ed Solari ( recommends 5K pullups/pulldowns on various
>T...
>>>signals ).
>>>>
>>>> Regards,
>>>> Carsten Bode
>>>>
>>
>>
>>
>>--
>>My email: cjclark@intellitech.com
>>------------------------------------------------------------------------
>>+ Intellitech Corporation Scan Tools +
>>+ 66 Route 25, Meredith, NH 03253 for Design and Test. +
>>+ PH:603-279-6308 FX:603-279-5135 For info: info@intellitech.com +
>>------------------------------------------------------------------------
>>
>>
>--------------------------------
> Alan Deikman, ZNYX Corporation
> alan@znyx.com
>End of forwarded message.
>--
>My email: cjclark@intellitech.com
>------------------------------------------------------------------------
>+ Intellitech Corporation Scan Tools +
>+ 66 Route 25, Meredith, NH 03253 for Design and Test. +
>+ PH:603-279-6308 FX:603-279-5135 For info: info@intellitech.com +
>------------------------------------------------------------------------
--
My email: cjclark@intellitech.com
------------------------------------------------------------------------
+ Intellitech Corporation Scan Tools +
+ 66 Route 25, Meredith, NH 03253 for Design and Test. +
+ PH:603-279-6308 FX:603-279-5135 For info: info@intellitech.com +
------------------------------------------------------------------------
j œ Š