russell@ucb-vax.ARPA (06/01/85)
In order to get connected to the CSNET PhoneNet, we bought a Racal-Vadic VA3451 Auto-Dial modem. The modem was delivered in early 1985, and the manual that came along with it was entitled "VA3451 Modem Installation/ Operation Manual", "Initial Issue", "November, 1984". There appears to be an problem with the modem and an error in the manual. After we put the modem into service, it occasionally got itself into a funny state. The modem was off-hook even though there was no connection, and the modem was generating a carrier. The lights that were lit were: HS, DSR, and DTR. Further, disconnecting the modem from its DTE by removing the RS232 plug (thereby dropping DTR) did not cause the modem to go on-hook. Neither did unplugging the modem from the phone line. It was possible to force the modem to go on-hook either by toggling power or by flipping the DA/VO/MA switch on the front panel away from VO and then back. I telephoned the Racal-Vadic Service Hotline, and we tried a several months worth of different things, but nothing would make the problem go away. On Wednesday, 5/29/85, Diane Dustman, CSNET-CIC Technical Staff, phoned me to tell me that our modem was off-hook. She asked me to reset the modem so that CSNET-RELAY could begin to call our host again. I told her that I knew about the problem and that I was working on it. I asked her if she often had to telephone site liaisons asking for a modem to be reset. She said that she sometimes did, but that it didn't strike her as a widespread problem. In any case, she mentioned my problem to Dan Long, CSNET Technical Liaison, and it jogged his memory. He sent on to me something he had received in early 1984. It was an article, signed by "--Lauren--", that had been distributed to the INFO-MICRO, INFO-CPM, UNIX-WIZARDS, and TELECOM interest groups. Charles Lindahl at "ti-csl" (Texas Instruments Central Research Labs in Dallas) had forwarded the article to Dan Long. Briefly, the article described exactly the problem that we were experiencing with our modem. The article said that the fix involved changing the strapping so that the A1 switch was OFF. The setting of the A1 switch is ignored, and A1 is considered as ON, if the modem is set to its "Standard Configuration" via the A6 switch. The article said that A6 should be set to ON in order to turn off the "Standard Configuration", i.e., to enable all the other switches. I had set A6 to the non-"Standard Configuration" setting, but according to my manual, that setting was achieved with A6 OFF, not ON as was indicated in the article. The fact that the manual I have is wrong was verified by a Racal-Vadic Field Service representative. I will forward a copy of this to him, in order to be sure that the error in the manual is corrected. Unfortunately, I cannot tell you yet that, by setting A6 to ON and A1 to OFF, our problem has been solved. I had mailed the modem back to Racal-Vadic for repair just before I got the information from Dan Long. In any case, I would like to express my thanks to Diane, Dan, Charles, and Lauren for getting this information to me. Michael Russell russell@ucd.csnet ...!ucbvax!ucdavis!bluebell!russell ucdavis!bluebell!russell@berkeley.arpa russell%bluebell%ucdavis.uucp@berkeley.arpa