Aucbvax.1616 fa.info-micro utzoo!duke!decvax!ucbvax!CSTACY@MIT-AI Thu Jun 11 02:35:19 1981 INFO-MICRO Digest V3 #48 INFO-MICRO AM Digest Thursday, 11 June 1981 Volume 3 : Issue 48 Todays's Topics: Desoldering Chips, Modems - DC Hayes & 212 Speed Selecting & Autodialers, Memory - 16K Hybrids & ExpandoRam modifications ---------------------------------------------------------------------- Date: 6/10/81 2350 EDT From: CSTACY at AI Subject: Desoldering chips Here are the responses to DAN's query about ways to desolder many chips from a pc board without damaging them. Thanks to all who replied, and to DAN@ML. Except for the fumes (use of open area and fan highly recommended), torching ICs works very well. It is not bad on the chips; apparently they tend to be far more sensitive to length of overheating than to temperature, and they get heated for a far shorter time than with irons since the torch makes it possible to melt all the pads at once. Fumes ar STEP FARTHER OUT. See you on the ski slopes of Encaladeus, Ken Haase ------------------------------ Date: 1 Jun 1981 10:59:16-EDT From: cjh at CCA-UNIX (Chip Hitchcock) Subject: mining on Io As an acquaintance of the person who deduced Io's structure and dynamics (and got a couple of publications in SCIENCE for it) I would agree with all of REDFORD's points about the untenantability of Io. HOWEVER, the end of the msg again forgets ?'s point that we can't expect continuous linear advances over the course of history. In particular, he makes a false analogy with mining in Colorado. After all, Colorado today is practically civilized (with the exception of Coors)---they're even going to hold a World SF convention there this summer! I would argue that there are a couple of believable analogies between \\1880's// Colorado and Io-as-it-could-be: both are a substantial amount of travel time removed from what is called c high heat is offset by the fact that it only takes about 5 seconds to unsolder a chip. The best idea to avoid fumes is to do the whole process outside. At worst, you need a relatively sturdy table with a vice on it. Also be careful to avoid splattering yourself with hot solder. Unfortunately, I have seen the boards you want to unsolder, and it doesnt look like you'll have room to get the vice grips around the chips. You might consider getting 7 or 8 of those spring loaded clips designed to be used with the H-bars, and unsolder an entire row of chips at once. Dave tried the H-bar, and it was a total lose. It has slits that are supposed to fit over the pins, but the pins didnt tend to be straight enough to fit into the slits. Another idea might be to cut apart the board and uses the busses that are already there (eg soldering wire to the PC card at appropriate places). Good luck Bill W -- WESTFW at WHARTON Having removed many a socket and chip myself, and learning alot in the process, may I recommend solder wick. I have never tried torching, but as you point out, this is rather destructive. I have tried and lost when using the H-brander and associated extraction tools, it could have been because of the low (Rat-Shack) quality or perhaps low tip temperature. In any case, I have found solder wick the most relibable method, when it HAS to come off. -- NEAL at MIT-MC ONE METHOD I'VE SEEN WORK VERY WELL IS TO FIND A CONTAINER LARGE ENOUGH TO HOLD THE PC BOARD IN FLAT. HEAT UP PEANUT OIL IN IT, PLACE THE BOARD IN FOR AROUND 10-15SECS AND THEN TAKE IT OUT AND SMACK IT ON A TABLE. ALL THE IC'S WILL FALL OUT AND SINCE THERE HASN'T REALLY BEEN ANY TEMPERATURE DIFFERENCE ON THE PINS OF THE IC, IT WILL WORK JUST FINE. IN FACT THIS METHOD CLEANS THE BOARD OF ALL ITS PARTS. THE ONLY THINGS THAT DON'T SURVIVE ARE SWITCHES AND SOCKETS AND OTHER PLASTICS. -- JEFFH at MIT-MC A technique used by the MIT Electronic Research Society (MITERS) is to use something that heats up a surface to the melting point of the solder and set the board down and lift out the chips. I don't know where MITERS got this thing, but it doesn't have to have much thermal capacity only thermal conductivity to work. This means it doesn't have to be a massive amount of metal to heat up (although the one at MITERS does). -- MAP at MIT-AI ------------------------------ Date: 3 May 1981 02:25-EDT From: Barry A. Dobyns Re: DC Hays modem query Does anyone know if there is any difference between the 1977 DC Hayes modem board and the current one that i should be aware of. I have been offered one for about $150 and i am very tempted at that price. -thanks --barry ------------------------------ Date: 21 May 1981 17:34:33-PDT From: ihnss!karn@berkeley Re: 212's and speed select I just went through the hassles with automatic speed selection on our 212's, and have our '70's DH/DM-11's selecting the originate speed on cu and uucp calls. In order to control the originate speed from the interface, you must have the 212L1A (earlier models don't have it) and set the internal switch to allow speed selection via connector pin 23 (instead of the HS button). I had to jumper pin 14, Secondary Transmit Data, to pin 23 inside the modem so the unix driver could control the line through the DM-11. I don't make use yet of the ANSWER mode speed indication from the modem (remember, the originating 212 always controls the speed mode). The 212 manual shows this as being on pin 12, which just happens to be conveniently available as secondary receive data in the DM-11. I don't see any conflict in these two signals; they are distinct and on separate lines. Another point: You'll have to modify the conn.c routine in uucp to set the mode BEFORE the call is placed, as the 212 requires the originate mode to be set before going into data mode. Hope this helps. Phil ------------------------------ Date: 6 May 1981 01:41:29-PDT From: sdcsvax!dbw at Berkeley Re: auto-dialers I am interested in purchasing an auto-dialer but I am not sure where to start. As usual cost is an issue but I am looking for good quality as well. Is there some mechanism for obtaining this information, or if not does anyone out there have any suggestions? If it makes any difference, this will be going on a 16 bit uprocessor. dbw ------------------------------ Date: 6 May 1981 1850-PDT From: Geoffrey C Mulligan (at The Pentagon) Vadic has just announced a new OEM board. It is vadic 300 baud auto-originate and auto-answer modem. I don't remember right off hand the Vadic number of the board. It has connections for the new modular plugs. There are also a couple of s-100 based auto-dialer modems available. geoff ------------------------------ Date: 3 May 1981 (Sunday) 0439-EDT From: WESTFW at WHARTON-10 (William Westfield) Re: 16K * 8 hybird Ram chip ? Does anyone make a hybird chip containing 8 4116 type chips to yield a 16K * 8 ram ? I know Harris semiconductor has a hybird with 16 4K*1 CMOS static chips on it, yielding 8K*8 on a 40pin DIP, but it would be nice to have the 16K*8 if you're willing to put up with dynamicism (?). Bill W ------------------------------ Date: 05/27/81 03:35:47 From: W8SDZ at MIT-MC Re: Modifications for ExpandoRam I and II Postal-Address: Chuck Weingart, 2152 W Iowa, Chicago IL 60622 EXPANDORAM I WITH Z80 4MHZ Here's how to use an Expandoram I board in a 4Mhz Z80 system: 1. You need 200ns rams 2. Change R1 to 150 ohm, R2 to 2.7K ohms, R3 to 4.7K ohms 3. If you have an early board, lift pins 9 and 11 on U19 out of the socket. Follow directions if you have a late board with a Z80. 4. Break the connection to U25 pin 5. Connect U25 p5 to U13 pin 3. If you have only 250ns chips, fiddle with the values of R2 and R3, and make sure the CPU has an M1 wait state enabled. R3 controls the length of the refresh cycle, and in a 4Mhz system with an M1 wait, you have about 700 ns to do the refresh. Without the M1 wait, you only have 450 ns or so, so R3 must be reduced. Do whatever works. Feedback I have gotten has indicated the resistor values are OK for 250ns, sometimes. EXPANDORAM II FIXES These are the fixes to SD Sales Expandoram II for CPUs such as Cromemco, Vector Graphics, etc. 1. X U11p4. J U11p4 to U10p3 2. X U10p4. J U10p4 to U10p10 3. Rmv C15, R7, CR2, C12. 4. x U18p6. J U6p13 to U6p1 5. X U22p13. J U22p13 to U22p7 6. X U18p11. J U18p11 to S100 pin 75 7. Take CR2 (step 3) connect across R5 - cathode to C3. 8. Replace C3 w/ 270pf. X U6p10. J U6p10 to U6p2 9. X U6p12. J U6p12 to U16p15 10. X U22p4. J U6p11 to U22p5 or 10 11. X U11 p14. J U11p14 to U21p6 12. X U11p15. J U11p15 to U20p1 13. Ground + side of C14 if REV A board. (they forgot to) Abbreviations: X = "break connection to" J = "connect Jumper between" Rmv = Remove. These changes are my own! They have been given to SD sales, and my reply was "Oh, we'll look at them." They are now being given out as their own changes. Most of the changes are for timing problems, one change is to use PRESET instead of a power on clear for resetting the board, and two bugs are fixed: the notorious lock-up if anything funny happens (like a wait state), and the M1 wait generator enabled whether the board is enabled or not. I use all of these modifications in my system. I put on the changes suggested by SD Sales, and there is no conflict with mine, but their changes dont affect anything, and I took theirs off again. EXPANDORAM II BANKING Here is how to change the banking port address on the SD sales Expandoram II to 40H: 1. If you havent modified the board (above) break the connections to U18 pins 3 & 4. Connect U6 p15 to U6 p1. This frees up an inverter, and fixes a bug at the same time. 2. Break the connections to U19 pin 3 and connect a jumper around the breaks, because the signal doesnt stop at U19. 3. Connect a jumper between U18 pin 4 and U19 pin 3. Connect a jumper between U18 pin 3 and U23 pin 16. 4. Replace U23 with a 74LS240. The banking port will be at 40H. It is not exactly compatible with Cromemco, however. because the latter has eight bits and eight banks, and any board can be in multiple banks. The Expandoram can be in only one bank at a time, and uses only three bits on the port to specify that. You may wish to alter which bank port bits are latched. If you want, do not replace U23, and the address will be at 0BFH. ------------------------------ End of INFO-MICRO Digest ************************ ----------------------------------------------------------------- gopher://quux.org/ conversion by John Goerzen of http://communication.ucsd.edu/A-News/ This Usenet Oldnews Archive article may be copied and distributed freely, provided: 1. There is no money collected for the text(s) of the articles. 2. The following notice remains appended to each copy: The Usenet Oldnews Archive: Compilation Copyright (C) 1981, 1996 Bruce Jones, Henry Spencer, David Wiseman.