FW: SAS 3.3V LED driver characteristics - Feedback by 8/2/02 please!

Brown, Russ Russ_Brown at maxtor.com
Mon Jul 29 14:36:08 PDT 2002


* From the T10 Reflector (t10 at t10.org), posted by:
* "Brown, Russ" <Russ_Brown at maxtor.com>
*
t10 folk:

I'd like to propose a programmable current sink driver solution for the
READY LED.

Some of the LED driver issues are:

1. Drive manufacturers want the 5.5V tolerant requirement reduced to 3.6V
max so that the LED driver can be implemented on a high speed CMOS ASIC, and
with no external components (added cost).

2. For system designers, with LED forward voltages ranging from about 1.6V
to 2.5V, the lower 3.3V +/-10% supply voltage doesn't leave enough voltage
margin to adequately control LED brightness in a configuration where the
driver is a switch to GND and a series resistor is used to define the
current.

3. A single current source value is not adequate to cover the range of LED
applications.

SAS systems will likely end up having mode pages to allow access to some
drive parameters, and that could include 2 bits to select the LED drive
current. The LED pin would then act as a current sink to GND when ON. The
series current-defining resistor on the back-plane would not be used
(although back-plane manufacturers may still want some lower value current
limiting resistor to protect the LED against accidental short to GND on the
READY LED pin.)

A programmable (2 bits) current sink solution could support 4 LED drive
current settings, for example OFF, and nominal current values of 5mA, 10mA
and 20mA.

Current source tolerance could be held to +/-25% over a voltage compliance
range of 0.4V to 2.2V, with an absolute max voltage of 3.6V.

The drive ASIC would dissipate a bit more power than if an external resistor
were used to set current, but worst case for Vdd=3.6V and Vled = 1.6V would
be 2V at 25mA, or 50 mW, and that should be tolerable.

This idea was discussed in the PHY working group conference call on 29 July,
but requires feedback from a much broader audience.


Russ Brown
Maxtor Corp



-----Original Message-----
From: Dave Davis [mailto:Dave_Davis at uk.xyratex.com]
Sent: Monday, July 22, 2002 2:04 PM
To: John Lohmeyer; Alvin.Cox at seagate.com
Cc: T10 Reflector
Subject: RE: SAS 3.3V LED driver characteristics - Feedback by 8/2/02
please!


* From the T10 Reflector (t10 at t10.org), posted by:
* "Dave Davis" <Dave_Davis at uk.xyratex.com>
*
Alvin et al,
	I want to express a concern here on the brightness of LEDs that this
spec might support. Most manufacturers use some type of lightpipe to
transfer the light from the LED to the front of the enclosure. All of these
suffer loss, some due to alignment and most due to material characteristics.
Whilst the LEDs are improving there is still a challenge in the materials
that will push the use of brighter, higher current LEDs. Whilst these could
be driven using additional transistors, all of us are striving for passive
backplanes for increased reliability and hence are not desirable. 

Has this been considered in the recommendation ? I believe that we currently
run high brightness LEDs at around 19mA and that a reduction to 10mA would
be too difficult to meet.

Regards , Dave Davis , Product Architect,
Storage Solutions Development, Dept 459 ,
Locn 1000/61, XYRATEX ,UK, mailto:dave_davis at uk.xyratex.com
TEL:Int 6466 Ext +44 (0)2392 496466 FAX: +44 (0)2392 492284
Http://www.xyratex.com

-----Original Message-----
From: John Lohmeyer [mailto:lohmeyer at t10.org] 
Sent: 22 July 2002 16:57
To: Alvin.Cox at seagate.com
Cc: T10 Reflector
Subject: Re: SAS 3.3V LED driver characteristics - Feedback by 8/2/02
please!

* From the T10 Reflector (t10 at t10.org), posted by:
* John Lohmeyer <lohmeyer at t10.org>
*
Alvin,

I wholeheartedly agree with moving to 3.3-volt LED driver specifications.

However, I have two issues with the Drive LED on specification:

1. The sign in the IOL inequality for the Drive LED on specification is 
backwards to normal practice for chip specs. This inequality should be less 
than or equal to 10 mA, not greater than 10 mA. (This says the chip can 
sink up to 10 mA over the VOL range.)

2. The 0.15 VOL spec is aggressive. Do you really want your chip vendors to 
scrape chips that can only sink 10 mA at 0.16 Volts on the LED driver pin, 
but are otherwise perfectly good parts?

John




At 7/19/2002 03:23 PM, you wrote:
>* From the T10 Reflector (t10 at t10.org), posted by:
>* Alvin.Cox at seagate.com
>*
>The majority of the SAS PHY working group desires to change the LED driver
>characteristics to be 3.3V compliant rather than 5V compliant. As a new
>interface, this seems to be the best opportunity to change. IC supply
>voltages are lower and LED efficiencies are increasing. One complication is
>the old specification allowed a rather high output-low voltage. At 0.5V
>maximum, it is a significant proportion of the nominal 3.3V supply. The
>current was also rather high when considering the LED's available today.
>The SAS PHY working group makes the following recommendation for
>consideration in the SAS specification for the READY LED electrical
>characteristics. We realize that this may be an aggressive specification at
>this time, but feel that it most likely can be achieved with today's
>commonly used technologies. Check the SAS specification (Section 5) for the
>complete set of READY LED implementation requirements.
>
>Output characteristics of the READY LED signal (tested through the device
>mated connector)
>
>     State              Output current             Output voltage
>
>Drive LED off      -100 µA < IOH < 100 µA         0 < VOH < 3.6 V
>Drive LED on             IOL > 10 mA              0 < VOL < 0.15V
>
>
>Alvin Cox
>Seagate Technology, LLC
>
>*
>* For T10 Reflector information, send a message with
>* 'info t10' (no quotes) in the message body to majordomo at t10.org

--
John Lohmeyer                  Email: lohmeyer at t10.org
LSI Logic Corp.                Voice: +1-719-533-7560
4420 ArrowsWest Dr.              Fax: +1-719-533-7183
Colo Spgs, CO 80907

*
* For T10 Reflector information, send a message with
* 'info t10' (no quotes) in the message body to majordomo at t10.org
*
* For T10 Reflector information, send a message with
* 'info t10' (no quotes) in the message body to majordomo at t10.org
*
* For T10 Reflector information, send a message with
* 'info t10' (no quotes) in the message body to majordomo at t10.org




More information about the T10 mailing list