[Asterisk-Dev] TE405P *hardware* question

Paul Cadach paul at odt.east.telecom.kz
Tue Apr 13 18:32:35 MST 2004


Hi,

Check at the end of right column at page 2 of pointed manual:
"... Two forms of over-voltage protection are provided, one that permits 5V compliance, and one that does not. For 5V
compliance, a zener-like structure connected to ground turns on when the output rises to approximately 6.5V. When 5V
compliance is not required, a conventional clamp diode may be connected to the output supply voltage, Vcco. The type of
over-voltage protection can be selected independently for each pad".

As you pointed (and the same page 2 says too), both 3.3V and 5V PCI uses the same Vcco voltage, 3.3V, but as you can see
at pages 31-35, Xilinx have different symbols for input/output buffers for each pad:
- IBUF_PCI33_3 or IBUF_PCI33_5 or IBUF_PCI66_3;
- IBUFG_PCI33_3 or IBUFG_PCI33_5 or IBUFG_PCI66_3;
- OBUF_PCI33_3 or OBUF_PCI33_5 or OBUF_PCI66_3;
- OBUFT_PCI33_3 or OBUFT_PCI33_5 or OBUFT_PCI66_3;
- IOBUF_PCI33_3 or IOBUF_PCI33_5 or IOBUF_PCI66_3;

Page 43 specifies that for PCI33_3 and PCI66_3 maximum input voltage at pad is Vcco+0.5V, i.e. 3.8V (uses clamp diode to
Vcco), but for PCI33_5 there is 5.5V (uses zener diode to ground).

As for TTL specification, "0" voltage must be between 0 and 0.8V, and "1" voltage must be between 2.4 and 5V, so LVTTL
(3.3V)with their levels must comply with TTL spec. So, I don't see any is no problem if you connect zener-protected pins
(5V tolerant) to 3.3V system.


There was for Xilinx. But what about PEF? As I see for one used for TE410P (to remember - I don't have TE405P to check
which chips used for it), there was 3.3V-only chips, so IMHO TE405P must have any sort of 3.3V source (from onboard
regulator or  from PCI connector).


Any way, I don't see any problems to connect TE405P to 3.3V bus (after making second, 3.3V, key hole). But I'm not
Digium's hardware engineer... ;-)' I think there would be dual-voltage TE405P+ cards (TE405P with two key holes)
available when all TE410P "already-made" will be sold out. ;-)'


WBR,
Paul.

----- Original Message -----
From: "Andrew Kohlsmith" <akohlsmith-asterisk at benshaw.com>
To: <asterisk-dev at lists.digium.com>
Sent: Wednesday, April 14, 2004 2:05 AM
Subject: Re: [Asterisk-Dev] TE405P *hardware* question


> Thanks for the reply.
>
> > As I checked out Xilinx's datasheets CPLD used for TE410P's PCI bridge is
> > Spartan II-E family which is maximum 3.3V capable, while regular Spartan-II
> > family (which is, I think, used for TE405P) is supports both 3.3 and 5V for
> > I/O pins, but requires different termination for each voltage. I don't have
> > TE405P, but as I remember there was a sort of problems with some chipsets.
> > May be TE410P have ones too but they are not reproducible because "ugly"
> > chipset (or motherboard vendors) isn't support 3.3V for this chipset.
>
> http://direct.xilinx.com/bvdocs/publications/ds001_2.pdf
>
> Page 30 says that it can indeed use 3.3V or 5V for PCI, and that the clamps
> need to be different for 5V tolerant systems (which the TE405P is), so that
> much should work...
>
> The next page goes on to say that the standard doesn't require use of Vref of
> board term voltage, only 3.3V on Vcco.  Now it goes on to say that I/O
> configured for PCI 33MHz 5V standard are also 5V tolerant...  we know all
> that since the card is a 5V card.  :-)
>
> Page 43 gives the PCI 3V and 5V buffer Voltage specifications -- it looks like
> there should be no issues whatsoever, I believe that the Vih/il and Voh/ol
> specs are acceptable for both 3V and 5V systems.  Vcco does not change for 5V
> or 3V operation.
>
> The pinout tables are a little confusing for me, but with the information
> gathered so far there shouldn't be any obvious problem with sawing the key
> out for 3V operation.  I would still like to hear from Digium as to whether
> the IOREF is connected to the correct place on the PCI connector, and what
> the unstuffed voltage regulator was to be used for and its part number.
>
> Regards,
> Andrew
> >
> >
> > So, I think, the problem is different termination of I/O blocks at the
> > Xilinx for 3.3 and 5 V for Spartan-II family.
> >
> >
> > WBR,
> > Paul.
> >
> >
> > ----- Original Message -----
> > From: "Andrew Kohlsmith" <akohlsmith-asterisk at benshaw.com>
> > To: <asterisk-dev at lists.digium.com>
> > Sent: Tuesday, April 13, 2004 11:24 PM
> > Subject: [Asterisk-Dev] TE405P *hardware* question
> >
> > > Is there a specific design issue with making the TE405P a universal
> > > 3.3V/5V product?  I've done a cursory review of the Xilinx part and it
> > > seems that it should be just fine to plug in to a 3V or 5V PCI slot, if I
> > > take a hacksaw and create the appropriate key.
> > >
> > > I haven't yet traced out the VIO pins to make sure that the Spartan will
> > > honour the PCI bus voltage.
> > >
> > > Yes I am almost willing to sacrifice a $1500 card at this point in time.
> > > :-)
> > >
> > > Can anyone who is familliar with the hardware design of this product
> > > explain why there are separate 3V and 5V cards?  Hell even the X101P and
> > > TDM400P are dual-voltage capable, what is preventing it on the quad T1/E1
> > > cards?
> > >
> > > Regards,
> > > Andrew
> > > _______________________________________________
> > > Asterisk-Dev mailing list
> > > Asterisk-Dev at lists.digium.com
> > > http://lists.digium.com/mailman/listinfo/asterisk-dev
> > > To UNSUBSCRIBE or update options visit:
> > >    http://lists.digium.com/mailman/listinfo/asterisk-dev
> >
> > _______________________________________________
> > Asterisk-Dev mailing list
> > Asterisk-Dev at lists.digium.com
> > http://lists.digium.com/mailman/listinfo/asterisk-dev
> > To UNSUBSCRIBE or update options visit:
> >    http://lists.digium.com/mailman/listinfo/asterisk-dev
> _______________________________________________
> Asterisk-Dev mailing list
> Asterisk-Dev at lists.digium.com
> http://lists.digium.com/mailman/listinfo/asterisk-dev
> To UNSUBSCRIBE or update options visit:
>    http://lists.digium.com/mailman/listinfo/asterisk-dev
>




More information about the asterisk-dev mailing list