Timing measurements

RMoore at corp.adaptec.com RMoore at corp.adaptec.com
Thu Nov 7 12:06:00 PST 1996

* From the SCSI Reflector (scsi at symbios.com), posted by:
* RMoore at corp.adaptec.com
Form: Internet
Text: (56 lines follow)
In Figure 24 of SPI-2 Rev. 11, I previously raised a question about how the 
setup and hold time measurements are specified. Here I would like to expand 
the question further.

My initial problem with this diagram is that the 60 mV dynamic receiver 
threshold is not being properly applied. 60 mV was requested by Frank 
Gasparic of Symbios but I think the intention of his request was to ensure 
that the applied signal level was large enough to ensure an accurate timing 
measurement, not that the actual timing measurements be made at the +60 mV 
falling edge and -60 mV rising edge levels. Here I may be wrong but I would 
appreciate any comment that Frank might have on the matter.

Another issue which may be significant is the fact that the setup and hold 
measurements are specified as an inside measurement (latest crossing of the 
earlier signal to earliest crossing of the latter signal). It was thought 
that this arrangement would be a worst-case condition, but in reality an 
outside measurement (earliest to latest) is required if a worst case test 
condition is to be applied. A diagram should make this point clear:

   CASE 1:
               | \____________________
               |         \
               |          \
               |          |\__________
               |<-- tm -->|
   CASE 2:     |          |
   __________  |          |
             \ |          |
              \|          |
   ______________________ |

In Case 1, the timing tm is an outside measurement (if tc1 is the "real" 
timing according to the device's "real" threshold, then tc1 < tm). In Case 
2, tm is an inside measurement (tc2 > tm). If tm is the same in both cases, 
then in Case 2 the waveforms are further apart (tc2 > tc1). It is easier for 
a receiving device to handle Case 2, so Case 1, the outside measurement, is 
the worst case.

My recommendation: Modify Figure 24 so that an outside measurement is called 
for (this renders the 30 mV vs. 60 mV debate moot), specify minimum values 
of Va and |Vn| to be used in the test, and specify a minimum edge rate to be 
used for the test.

On a purely editorial note, the figure is missing a line segment at the 
lower right of the data/parity signal.

	Richard Moore
	Adaptec Irvine Technology Center
Use Proportional Font: true
Attachment Count: 0

X-NVL-Content-Type: UNKNOWN
X-NVL-Content-Typename: UNKNOWN
X-NVL-Content-Charset: X-IBM-437
X-NVL-Content-Filename: ATTRIBS.BND
X-NVL-Content-Transfer-Encoding: X-UUENCODE

begin 777 ATTRIBS.BND


* For SCSI Reflector information, send a message with
* 'info scsi' (no quotes) in the message body to majordomo at symbios.com

More information about the T10 mailing list