Walter
The Intel 486DX4 has a pin Vcc5 (J1 in the PGA package) that is the power
clamp diode cathode. The user ties it to either 3.3 or 5.0V.
Regards,
Tom Dagostino
971-279-5325
<mailto:tom@xxxxxxxxxxxxxxxxx> tom@xxxxxxxxxxxxxxxxx
LogoAddress
From: ibis-macro-bounce@xxxxxxxxxxxxx
[mailto:ibis-macro-bounce@xxxxxxxxxxxxx] On Behalf Of Walter Katz
Sent: Monday, March 14, 2016 9:58 AM
To: Arpad_Muranyi@xxxxxxxxxx; IBIS-ATM
Subject: [ibis-macro] Re: The next I/O Buffer Challenge
Arpad,
Actually, Tom has not one yet, since he has not sent a link to a Data Sheet
on the WEB. Tom is on the road so I am going to be a bit flexible here.
So I accept that there is a pullup_ref and a power_clamp_ref buffer (and
model) terminal, and they can be connected to different Data Book Names (aka
IBIS signal_names), and that the two Data Book Names can have different
voltages.
After the presentation tomorrow on Figure 16, we can try to determine how
one would measure the [Pullup] and [Power Clamp] IV curves.
The context of the reason for these I/O buffer challenges was to answer the
question:
Can I determine what the reference voltage is for an I/O pin from the syntax
of the existing IBIS specification?
My conclusion so far is that we can determine the reference "Pin" for all
devices as follows:
Legacy [Models]
Whichever pin is connected to pulldown_ref or gnd_clamp_ref
ECL
Whichever pin is connected to gnd_clamp_ref
Different pullup_ref and power_clamp_ref
Whichever pin is connected to pulldown_ref or gnd_clamp_ref
RS233
Problematic
Walter
From: ibis-macro-bounce@xxxxxxxxxxxxx
[mailto:ibis-macro-bounce@xxxxxxxxxxxxx] On Behalf Of Muranyi, Arpad
Sent: Monday, March 14, 2016 12:23 PM
To: IBIS-ATM <ibis-macro@xxxxxxxxxxxxx>
Subject: [ibis-macro] Re: The next I/O Buffer Challenge
Walter,
Tom beat me with an answer, so he is the winner of the lunch again,
but in the days when the 3.3V devices started to come out and they
were frequently used together with 5V devices, they started making
them "5V safe". In practice this meant that the pullup transistor's
parasitic diode had to be connected to a higher voltage than its
normal 3.3V pullup supply so that the diode would not clamp the 5V
high level signal to about 3.3 + 0.6 = 3.9V. If this clamping would
have been allowed to happen, the 5V pullup would have driven a lot
of current through the 3.3V clamping diodes. So in those days there
were a ton of 3.3V devices which had a 5V supply for the clamping
diodes of their pullup transistors. I don't think it would be too
hard to find such devices even today.
Thanks,
Arpad
======================================================================
From: ibis-macro-bounce@xxxxxxxxxxxxx
[mailto:ibis-macro-bounce@xxxxxxxxxxxxx] On Behalf Of Walter Katz
Sent: Sunday, March 13, 2016 12:00 PM
To: IBIS-ATM <ibis-macro@xxxxxxxxxxxxx>
Subject: [ibis-macro] The next I/O Buffer Challenge
All,
Congratulations to Tom Dagostino for winning the last I/O buffer challenge
and for receiving the challenge award of a lunch at a restaurant in
Portland.
The next challenge is to find a real buffer that has more than two rail
voltages. Specifically a non-ECL buffer that either:
[Pullup Reference] is not equal to [Power Clamp Reference]
Or
[Pulldown Reference] is not equal to [GND Clamp Reference]
Or for an ECL buffer that
[POWER Clamp Reference], [Pullup Reference] and [Pulldown Reference] are not
equal.
In order to win the challenge (and a free lunch) you must be the first
person to supply a link to a data sheet for a commercially available part
that requires more than two rail voltages for an IBIS buffer.
Walter
Walter Katz
<mailto:wkatz@xxxxxxxxxx> wkatz@xxxxxxxxxx
Phone 303.449-2308
Mobile 303.335-6156
Attachment:
IntelDX4.pdf
Description: Adobe PDF document