rlg@STYX.DESKTALK.COM (Richard L. Gralnik) (04/19/91)
Hi. A recent discussion of the comparative merits of slide latches versus screw/thumbscrew as a means of attaching a cable to a system port has prompted this unofficial opinion poll. Which do you prefer and why? or put another way - Do you like slide latches (the connector used for ethernet cables)? Why or why not? Do you like screw on connectors? Why or why not? Feel free to forward this note to friends. Also, please reply directly, not to the list. I'll publish the results if there are any. Thanks, Richard Gralnik (rlg@desktalk.com)
ckollars@deitrick.East.Sun.COM (Chuck Kollars - Sun Technical Marketing - Boston) (04/20/91)
In article <9104182050.AA07838@desktalk.com> rlg@STYX.DESKTALK.COM (Richard L. Gralnik) writes: >Hi. A recent discussion of the comparative merits of slide latches versus >screw/thumbscrew as a means of attaching a cable to a system port has >prompted this unofficial opinion poll. > >Which do you prefer and why? > >or put another way - > >Do you like slide latches (the connector used for ethernet cables)? Why >or why not? The results of your survey are likely to be seriously skewed. Ethernet transceiver cables often don't work, and most folks blame it on the slide latch. So I'd guess you're going to get a lot of responses denigrating the slide latch. But the full story is that, although the Ethernet transceiver cable connector on many systems in fact doesn't work very well, it's _not_ the fault of the slide lock. The spec (see drawing on page 94, just above section 7.6.2) says that the connector should go on the _outside_ of the backpanel. But in order for printed circuit boards to be stuffed and soldered by automatic machinery then married to the system later, the connector is often mounted on the _inside_ of the backpanel. Result -- the connector wobbles and makes poor electrical connection even though the cable is inserted "all the way" and the slide lock is "closed". The Ethernet spec was apparently written by someone who either was not a mechanical engineer, or did not have any experience with automated manufaturing. The reputation of the slide lock will probably never recover from that oversight. --- chuck kollars <ckollars@East.Sun.COM> Sun Technical Marketing, located in Sun's Boston Development Center
rlstewart@eng.xyplex.com (Bob Stewart) (04/23/91)
> The Ethernet spec was apparently written by someone who either was not > a mechanical engineer, or did not have any experience with automated > manufaturing. The reputation of the slide lock will probably never > recover from that oversight. Yup. I know the guy. At the Ethernet 10 year reunion last fall, he asked what the designers would do differently. He (an electrical engineer) perpetrated the slide lock. He regrets it, and he wouldn't do it again. For what it's worth, the idea of the slide lock comes from the late 70s, when automated manufacturing techniques were probably a bit different. Bob
henry@zoo.toronto.edu (Henry Spencer) (04/23/91)
In article <5634@eastapps.East.Sun.COM> ckollars@east.sun.com (Chuck Kollars - Sun Technical Marketing - Boston) writes: >... The spec (see drawing on page 94, just >above section 7.6.2) says that the connector should go on the _outside_ >of the backpanel. But in order for printed circuit boards to be >stuffed and soldered by automatic machinery then married to the system >later, the connector is often mounted on the _inside_ of the >backpanel. Result -- the connector wobbles and makes poor electrical >connection... There is absolutely no problem with having the connector on the inside of the panel provided the slide lock is mounted there too. Manufacturers have been known to get this right. Of course, it's harder, and we all know that most manufacturers would rather do things the easy way than the right way. Especially a certain workstation vendor whose name starts with S... The *intentions* of the slide-lock advocates were good. All too often, screw-equipped connectors aren't screwed in because the relevant screwdriver is not handy. A poor locking system which gets used is better than a good one that doesn't. Unfortunately, they didn't realize that (a) transceiver cables are sufficiently heavier than rs232 cables that use of locking is effectively mandatory anyway, and (b) the slide locks are flimsy and easily damaged. -- And the bean-counter replied, | Henry Spencer @ U of Toronto Zoology "beans are more important". | henry@zoo.toronto.edu utzoo!henry
jerry@olivey.ATC.Olivetti.Com (Jerry Aguirre) (05/07/91)
In article <5634@eastapps.East.Sun.COM> ckollars@east.sun.com (Chuck Kollars - Sun Technical Marketing - Boston) writes: >The Ethernet spec was apparently written by someone who either was not >a mechanical engineer, or did not have any experience with automated >manufaturing. The reputation of the slide lock will probably never >recover from that oversight. The use of the slide lock also predates the "pizza box" and personal computer designs. All the old systems that I have seen use an internal cable to go from the board to the chassis and the connector is mounted on the outside of the mounting plate. Under those conditions it is OK though I always try to strain relief it where I can. Even the case where the connector is soldered to the board it would be easy to correct the problem. Just mount the connector to (the outside of) a small metal plate before soldering it to the board. That metal plate can then be screwed to the chassis where the connector would have been. The DB15 and slide lock assembly can then protrude tru a larger hole in the chasis. A 10 cent fix. A smart connector manufacturer would start selling the PC mount versions with a mounting bracket already attached. Having the slide lock come pre-assembled would please those "automated manufacturing" types. Are those "mechanical engineer"s deliberately violating the design spec to cut costs or are they just ignorant of it? That should be the real question being asked here! Screws are not perfect either. When was the last time you removed a cable and had the barrel nut come out with it? Maybe we need a new spec. Anyone for a modular transceiver cable? Jerry Aguirre
roy@phri.nyu.edu (Roy Smith) (05/07/91)
jerry@olivey.ATC.Olivetti.Com (Jerry Aguirre) writes: > Screws are not perfect either. When was the last time you removed a > cable and had the barrel nut come out with it? Some screws are better than others. The problem you are referring to is again the fault of stupid designers who use screws to attach the mounting sockets for the connector screws; i.e. an implementation problem, not a concept problem. I remember the lowly ADM-3/3a/5 terminals which had moulded-in screw sockets. I've never had one of those come off with the cable when I unscrewed the hold-down screws. The after-the-fact fix to connectors which have screwed-on-sockets is a dab of epoxy on the inner nut to keep it from comming off. We've got lots of terminals/printers/etc with dabs of epoxy inside, but people shouldn't have to resort to such antics to keep a cable connected. And, the answer to not having a screwdriver handy, is to make cables with thumbscrews, like Apple used to use for the old-style LocalTalk connectors (Mac-512 and LaserWriter). -- Roy Smith, Public Health Research Institute 455 First Avenue, New York, NY 10016 roy@alanine.phri.nyu.edu -OR- {att,cmcl2,rutgers,hombre}!phri!roy "Arcane? Did you say arcane? It wouldn't be Unix if it wasn't arcane!"