Path: utzoo!censor!geac!torsqnt!news-server.csri.toronto.edu!cs.utexas.edu!uunet!isis!mnemosyne.cs.du.edu!aburt From: aburt@mnemosyne.cs.du.edu (Andrew Burt) Newsgroups: comp.dcom.modems Subject: Practical Peripherals 9600 Message-ID: <1990Dec9.054531.20417@mnemosyne.cs.du.edu> Date: 9 Dec 90 05:45:31 GMT Organization: University of Denver, Dept. of Math & Comp. Sci. Lines: 72 Well, I thought I'd add my experiences (so far) with a PPI 9600 bps modem. Having read the earlier posting about problems with slower speeds, I was very curious to try this out... and found it to be true. I also spent some time with PPI tech. support and think the problems will be solved soon. Without changing any settings, the modem has had major problems with (non-mnp) 2400 or 1200 bps modems. MNP modems or 9600 bps -- no problems. The problems with 2400/1200 have been varied. Starting with not connecting. It'll just hang. Tech. support says it can take up to 45 seconds, but a few times it just never happened. When it has connected, the negotiations for protocol, compression, etc., have greatly disturbed the system at the other end. All the above problems were solved by using AT&Q0 to force the modem to make a straight async connection, no frills. So, looking good... except that the connections seem *awfully* noisy. Like, a steady trickle of garbage data, one or two bytes per second. This never happens with any of the other 2400/1200 modems, and happens *always* with the PPI. At this point I called tech support. I found them to be very willing to help. He (Phil) didn't insult my intelligence with "say, moron, is the modem plugged in?" type questions. I felt that if I really didn't know anything, they'd be very helpful, but since I did, they retargeted the level of help. So we spent a couple hours (at their expense) calling modems back and forth. It finally turned out we could reproduce the problem when he changed his firmware to the same rev. as mine (1.05). They're up to 1.18. At 1.18, no problem. (The only "problem" that still existed was that withOUT at&q0, the feature negotiations confused our autobauding getty; understandable. With at&q0, of course, even this went away. I can live with that.) Apparently this noise problem only shows up on mediocre lines, since when we connected directly we didn't have any problems; it was only connecting to the campus phone system (historically noisy, though recently not much problem) that I had major problems. So, the problems appear to lie in their firmware. Apparently they go through revs of firmware quickly, upping the # every time a bug is discovered during testing (they didn't release 1.06...1.1?, they were all internal). 1.18 still hasn't been released. But will be soon. I volunteered to be a guinea pig for 1.18, so I'll repost after a while. Another note: The power switch is a contact switch, thus for BBSs, not much good (it comes up "off" after being plugged in, or after a power outage). Phil reports that the next rev. of the board itself will have a jumper for initial state: power on/off. Due out in January, he said. My hunch is that they wanted market share, and went to market a wee bit early; classical management "mistake" (or is it really a mistake in the long run?). But that's just my gut feeling. Despite the 1.05 problems, I'm very positive about this unit. I like the price, the 5 year warranty, and of course the v.32/v.42bis/etc. I felt their tech support was very... supportive. After recent bouts with Wordperfect (a couple antagonistic "you don't know what you're talking about" types) and Borland (very unhelpful -- said I'd have to wait 4 months before they'd even LOOK AT a bug that was 100% provably their problem, and was delaying release of a product; and they wouldn't even supply me with the code so I could fix it!) -- after that & a couple other minor nuisances with assorted tech support, I'm really happy with PPI. Gee, tech support that actually tries to help... now there's a novel idea! Anyway, I would recommend the modem, though I would wait a leeetle while for them to get the firmware solidified. Or buy now & swap chips. -- Andrew Burt uunet!isis!aburt or aburt@du.edu "Kwyjibo on the loose!"