Micro­power

  

Volume 2 · Number 2 · April 1982

Page 30 of 37

64K RAM ON A Nascom RAM B BOARD

by Douglas M. Barr

This article is aimed at the owner of a Nascom RAM-B board who wants to increase the available RAM to the full 64K that the Z-80 microprocessor can address. There may well be something of interest for readers who are considering the modification of other boards populated with 4116 dynamic RAM chips; be warned, however, that although the modified board will work on a Nascom 2 or a Gemini multiboard system, it will NOT work on a Nascom 1 as described. The modification uses the MEXT line, whih is essential to the correct operation of a Nascom 1. That said, there is no reason why some enterprising Nascom 1 owner should not circumvent this problem and find some suitable alternative way of routing what we shall call the ‘RAM BLK 3’ signal on the board

One of the very attractive features of Nascom products so far has been the literature that the firm has published with their hardware, and I am assuming that you have the circuit diagrams which came with the board when you bought it. In particular, I shall be referring to figures 10, 11 and 12, which are the circuit diagrams issued with the manufacturer’s instructions for the board, but it should be possible to follow the gist of the theory behind the modification from this article alone. If you are interested only in the practical aspects of ‘how to do it’, then the article should be self explanatory.

The modification should not be beyond the skill of the average Nascom owner, but I am well aware that for some who are new to the game, what I am proposing may sound a bit daunting. For their benefit I shall take things rather slowly, and I ask any old hands to bear with me if some of what I say is ‘old hat’ to them. After you have completed the modifications you will have a board which will support 64K of RAM at 4 Mhz without wait states, will retain the ability to ‘write protect’ the banks of 4116s, and will also retain the Nascom page mode of operation. For several months I have been running such a modified board on my Nascom 2, and more recently I have used it also on a Gemini Multiboard system. What is interesting is that I used 200 ns RAMs, apparently without any adverse effects, but if you add up all the gate and delay-line times, you will realise that there is probably a certain element of luck in getting the board to run at 4 Mhz, and that the RAMs may be running a bit faster than spec. Those of a weak disposition may consider the use of 150 ns RAMs if they cannot bear the thought of the memory dumping at an inopportune moment.

The modification involves soldering sockets or chips on top of other delicate chips – with all the risks that entails. Many argue that the best way to do this is to wrap a couple of layers of cooking foil around a spare piece of polystyrene tile and plug the chip into the foil. This does two things; it ‘commons’ all the pins of the chip, and it helps conduct away heat and thereby reduces the risk of damage to delicate connectors inside the DIL package. If you are short of space inside your Nascom and there is not too much space between the boards, you may have to solder the

Page 30 of 37