Analog Devices ee-70 Application Notes

Engineer To Engineer Note EE-70

Technical Notes on using Analog Devices’ DSP components and development tools

Phone: (800) ANALOG-D, FAX: [phone], EMAIL: [e-mail], FTP: ftp.analog.com, WEB: www.analog.com/dsp

Copyright 1999, Analog Devices, Inc. All rights reserved. Analog Devices assumes no responsibility for customer product design or the use or application of
customers’ products or for any infringements of patents or rights of others which may result from Analog Devices assistance. All trademarks and logos are
property of their respective holders. Information furnished by Analog Devices Applications and Development Tools Engineers is believed to be accurate and
reliable, however no responsibility is assumed by Analog Devices regarding the technical accuracy of the content provided in all Analog Devices’ Engineer-to­Engineer Notes.

ADSP-2106x SPORT DTx Pins:
Is There Potential MCM Data
Contention Between Different
SHARCs on Adjacent TDM
Timeslots?

Last Modified: 9/2/98

By: JT

Question:

Dear ADI SHARC Apps Engineer,

I plan to use multiple SHARCs (ADSP-21060s)
in my design, sending data to each other via the
SPORTs in multichannel mode. I think it's safe
to say they will all be within 25C of each other.

A question came up when I was reviewing the
SPORT timings in the data sheet. I am very
concerned about a possible contention on the
DTx pin of the SPORTs for multiple DSPs
transmitting in adjacent channels.

SHARC's channel to the first bit in the next
channel by another SHARC.

I am trying to get an understanding of how
much contention there really is. Is there really
7ns of contention possible? If this potential for
overlap exists, will it be small enough where
contention will be negligible? Also, Is there any
characterization data for a maximum overlap
specification?

Are there things we can control, perhaps in
software and how the port is used, that can
minimize the contention? Still, it seems that
limiting the environmental variables will lessen,
but not eliminate, the contention problem. Is this
correct? Our concern is that we are using our
design to test flight boards and that bus contention
will shorten the life of the devices contending.

Answer:

The apparent serial port contention issue is an
artifact of specifications that are meant to cover all
possible situations and are not intended to be
added or subtracted. i.e.:

Looking on page 36 in the 21060 Data Sheet in
(Figure 22), there are 2 timing parameters:

T

3.5 nanoseconds minimum

T

10.5 nanoseconds maximum

- Data Enable from External SCLK -

DDTEN

- Data Disable from External SCLK -

DDTTE

It appears that there can be a potential 7 ns of
overlap in transmitting for the last bit of one

• one specification comes from "Fast process"/ -

40C /"High VDD"

• and one comes from "Slow process"/ 85C /

"Low VDD"

This sets up an extreme worst case scenario. We
are already aware that this is a timing discrepancy in
the data sheet specification.

a

In reality this contention is unlikely to occur since
the parametric operating point of all devices
(voltage, temperature and process) are likely to be
similar and significantly more closely matched that
the intentionally skewed material that we used for
characterization.

not considered significant and is unlikely to result in
any adverse conditions. Given that the
measurement techniques on the automated tester
are conservative in their determination of release to
tri-state, this is likely not to exist.

With regards to the driver vulnerability to driven
conflict - assuming that there is a driven conflict at
each channel switch point, the SHARC drivers in
particular were designed to be more robust than
they need to be. You can easily show this by
intentionally overdriving an output pin to either
power rail continuously without damaging the
driver. If you did this to all driver
pins on the part simultaneously you may damage the
part, but a periodic driven conflict on a single pin
for a 7ns duration is very unlikely to
damage the driver.

The risk of long term damage due to the conflict is
significantly less than the standard risk of ESD
damage to the pin involved in normal handling.

Analog Devices Test Results

Information was gathered in response to your
question regarding an apparent overlap between
serial port data transmissions. The results are
shown in TABLE 1. The datasheet shows, for
externally clocked serial ports, an apparent overlap
of 7ns between devices enabling and disabling
SPORT data transmissions. Actual data however,
from production devices Device "A" (enabling
data transmission device) and Device "B"
(disabling data transmission), is shown in TABLE1
in nanoseconds.

SPORT data transmission overlap was only
evidenced in one of the above test cases ie: Both
parts at max temperature (85C) and opposite
voltage extremes. The overlap was 0.20nS. This is

TABLE1. SPORT MCM Mode Overlap Test

Temperature A=4.7V B=5.3V _
DevA DevB DevA DevB (B - A) ns
85C 85C 6.89 6.25 -0.64

85C 25C 6.89 5.26 -1.63
85C -40C 6.89 4.63 -2.26
25C -40C 6.02 4.63 -1.39

-40C -40C 5.23 4.63 -0.60

Temperature A=5.3V B=4.7V _
DevA DevB DevA DevB (B - A) ns
85C 85C 6.41 6.61 +0.20

85C 25C 6.41 5.71 -0.70
85C -40C 6.41 4.81 -1.60
25C -40C 5.53 4.81 -0.72

-40C -40C 4.94 4.81 -0.13

Temperature A=5.3V B=5.3V _
DevA DevB DevA DevB (B - A) ns
85C 85C 6.41 6.25 -0.16

85C 25C 6.41 5.26 -1.15
85C -40C 6.41 4.63 -1.78
25C -40C 5.53 4.63 -0.90

-40C -40C 4.94 4.63 -0.31

NOTE:

PartA always the device enabling a data transmission
PartB always the device disabling a data transmission

A "-" in the "B-A" category indicates no overlap
A "+" in the "B-A" category indicates overlap

Additional Information:

1) Chapter 10: Serial Ports, ADSP-2106x SHARC
User's Manual, Second Edition, Analog Devices Inc,

Norwood, MA, (1996)

2) ADSP-21062/ADSP-21062L Data Sheet, ,Rev A,
Analog Devices Inc, Norwood, MA (1998)

EE-70 Page 2

Notes on using Analog Devices’ DSP components and development tools

Phone: (800) ANALOG-D, FAX: [phone], EMAIL: [e-mail], FTP: ftp.analog.com