April 23, 1998

FYI

I enjoyed Bill Travis' article "Tiny connectors squeeze into handheld wireless devices" (EDN, Feb 16, 1998, pg 85), and I would like to point out some additional data.

Sumitomo (www.sel-rtp.com) makes flexible coaxial cable (originally for medical devices) that has a larger market share and is substantially smaller (0.015 in.) and more flexible than that made by Amp (www.amp.com) and Precision Tube (www.precisiontube.com). I suggest contacting Sumitomo, which makes cable assemblies for many customers. You can also look at a DEC (www.dec.com)  laptop computer that uses these coaxes to interface to the company's LCD panel.

The Elco (www.avxcorp.com) 1.6-mm stacking height is not exactly leading-edge. JAE (www.jae.com) and other companies have had 1.5×0.5-mm centers for several years. JAE now has stacking connectors with contacts on 0.3-mm centers precisely for the reasons Travis points out (for example, the fact that these devices are smaller, lighter, and faster).

Don Chambers
JAE Electronics, via the Internet

Corrections and updates

In "Point-to-point ring turbocharges PCI bus" (EDN, Nov 6, 1997, pg 13), the 144-nsec-latency value should have been called the ring-transport delay. This delay is the average round-trip delay that would occur for the ring portion of a transaction on a 16-node Sebring Ring. The total latency that a PCI master sees is the transport delay plus the delays through the SRC (Sebring Ring Controller) chips at both the initiator and the target nodes and any wait states on the target bus. These delays are roughly comparable with the delays through a PCI-to-PCI bridge chip.

An SRC chip does not have to wait more than 56 nsec to begin transmission on the ring, whereas one may have to wait for more than one full PCI burst to complete before receiving a bus grant in a conventional PCI environment. The net result is an average latency on the order of 500 nsec for a single word read, even when multiple gigabytes/sec of data are being pushed through the network.

Jack Regula
Sebring Systems, Los Gatos, CA

In EDN's March 26, 1998, issue (pg 18), we accidentally printed some information about HP's 16600/16700 series logic-analysis systems earlier than we should have. We apologize for the error.

Sound off

Send your letters to Signals and Noise Editor, EDN, 275 Washington St, Newton, MA 02158 or e-mail us at kase@edn.cahners.com.   Our fax is 1-617-558-4470. EDN reserves the right to edit letters for clarity and length.

| EDN Access | Feedback | Table of Contents |