INMC News

all of five minutes. I now have an 8 AMP bridge which, although bigger, can be easily fitted with the aid of a piece of aluminium which also serves as a heat sink.

Since the RAM board draws about one and a half amps a healthy earth is required – a thick piece of wire from the power supply to the expansion board. However, this creates an earth loop, which perhaps is not desirable. (Suggest that all earths then taken from expansion board Ed.)

Getting It Working

With the kit now assembled and 8K of RAM plugged in that should have been the end of the story, but there were the inevitable problems. The memory could be read from and written to , but very soon the machine would lock up. On power-up everything looked fine but the system would not initialise properly on pressing the reset key. Many hours were spent looking for non-existant faults. Eventually it became apparent that the MEMEXT signal was at fault but not why. With the internal memory select used there was no problem. With external memory select used the problem reappeared, but on monitoring MEMEXT with an AVO it disappeared again. In an attempt to cure the fault the chips at each end of the MEMEXT signal were changed but to no avail. Since the AVO cured the problem a resistor was soldered between the signal line and earth, this made a minor improvement. Next a capacitor was tried between the signal and earth at the Nascom 1 end, this fixed it. Decoupling the expansion board end had no effect.

Conclusion

This board is not easy to build and may require a small decoupling capacitor to be added to the main N1 board on the MEMEXT line. Having this signal so close to the earth and the +5v lines would appear not to be a good idea. A criticism of the design is that tri-state buffers are used unnecessarily on the control and address lines, surely ordinary buffers would have done the job. Since getting this expansion working I have experienced no other problems and am very happy with it.

Mr. Curtiss neglected to mention the manufacturers of this product, but we believe it is the one supplied by Comp Components of Barnet.

There were a couple of other points in the letter attached to Mr. Curtiss’ article which ought to be mentioned. One of the reasons he gave for choosing the S-100 mother board was that it allowed greater choice of ‘goodies’ that may be fitted. Whilst generally true, this particular board is not truly S-100 compatible, and may give rise to problems under special circumstances. Another reason was that in his opinion the Nas-bus was responsible for some of the trouble experienced with Nascom expansion. In our opinion the trouble is not concerned with the Nas-bus but with the series 1 memory card. Happily these troubles can all be cured. Finally Mr. Curtiss asks why a non ASCII keyboard was chosen for Nascom and says, “Surely it would only cost a few pounds

(continued on page 18)