INMC 80 News

  

February–April 1981, Issue 3











Page 17 of 55











– P99 38 6 ( )
38 4 39 2 ()
– P99 38 8 ( )
3 13 MREQ 38 9 ( )
38 10 38 11 ( )
32 13 MREQ. RFSH 38 12 ( )
38 13 39 9 ( )
38 13 39 10 ( )
39 8 RAS3 ( )

4) Connect the new address select pad P99 to the appropriate decode pin(s). ( )

5) Perform the clock modification as described in INMC newsletter No 7. (This delays the column address strobe to the 4116s by delaying the clock waveform from NASBUS Pin 5 to IC 31 Pin 3 through 2 inverters).

6) Perform the ‘gridding’ operation described in the NASCOM construction article.

7) These may be necessary but I have managed to do without them:
Increase the power supply decoupling in the areas of ICs 40-47 and at the address multiplexers ICs 20 & 21.
Increase the values of the series resistors R7-14 to allow for the extra capacitance of the added RAM chips.

The resulting memory should be fast enough to run with the processor at 4MHz without a ‘WAIT’ state. It was in my case. Unfortunately the author (and the INMC) does not hold himself responsible for any damage caused by inaccuracies in these notes, and the constructor should check that what he is attempting is safe.

==ooOoo==

Following on from the above, we can’t recommend soldering sockets to the tops of very expensive MOS chips. So only attempt. this mod. if you are prepared to accept the (remote) possibility of a handful of dead 4116s.

Having warned you off, now on to how to do it with the least risk:
There are two dangers, first static, and the usual precautions should be observed. Secondly, soldering so close to the package allows heat to travel through the lead frame to the gold wire bonding that connects the lead frame to the chip itself, and damage to the bond can follow.

So: cover a piece of polystyrene tile with cooking foil, and plug the IC into it. The foil connects all pins together reducing risks from static, and also conducts the heat away quickly. Plug the socket through another piece of foil, and fold it up (like wrapping a toffee). This keeps heat away from the socket. Solder two diagonally opposed pins to line everthing up, and if in line, solder the remaining pins. Then carefully remove all traces of the cooking foil between the chips with tweezers.

This method has been used very successfully for piggy-backing 2708s for use as selectable NASBUG/NAS-SYS monitors on Nascom 1s, and Tandy suggest it as a method of mounting the lower case character generator chip on the TRS 80.


This is an OCR’d version of the scanned page and likely contains recognition errors.











Page 17 of 55