Previous column ComputerAnswers    Next column ComputerAnswers


ComputerAnswers Column 15


Copyright 1984-1989 Simon N Goodwin


After reading your article 'Trundle along' published earlier this year, I purchased a Trundle kit from Maplin Electronic Supplies Ltd, for my son's 16th birthday. The kit arrived in good order, and during his summer vacation my son assembled the Trundle from the kit and his old ZX-81. No problems were encountered and all seemed well.

However the Trundle does not work correctly because the Infra-red sensors do not respond. We have checked exhaustively and can find no wiring or other faults. Your article gives no fault-testing procedures but we have found that the correct response is obtained if the infra-red sensors are shorted or a strong infra-red source (a 60W light bulb) is held within an inch of the sensor. But no response is produced, even if the infra-red emitters are placed in direct contact with the sensor.

We wrote to Maplin and asked them to comment and to check the infra-red emitters. Maplin did not provide any comments but they said the emitters were tested and found OK. Subsequently we also sent the sensors which Maplin also declared to be OK, although again we could not get a word of comment on what the fault could be.

I should be most grateful if you could suggest what is wrong or a test procedure to find out.
Name withheld by request.

Mr. Atfield, of Maplin's Technical Department, says that technical queries are always answered if they are addressed to his department, at PO Box 3, Rayleigh, Essex, SS6 8LR. You have tried all of the fault-finding steps that I'd recommend over and above the eight pages of instructions and advice in the March 1985 PCW. It might be a good idea to make quite certain that you have wired up the sensors and emitters the right way round. In each case the longest lead should go to (or nearest to) the ground rail.

Mr Atfield says that Maplin will 'service' the machine for you if you send it back to them. They ask for a £10 deposit which will be refunded if they turn out to have supplied you with faulty components. If the mistake is yours they charge £10 per hour for 'salvage' work.


I have a Sharp MZ-711 computer - one of the Sharp MZ-700 series. I cannot find any software and cassette tapes for the computer.
Nigel Chappelle, Woodley, Berkshire.

Your nearest Sharp software suppliers are Kuma Computers Ltd, at Unit 12, Horseshoe Park, Horseshoe Road, Pangbourne, Berkshire (07 357 4335). Next closest are Sharpsoft in London, at 82 Great East Street EC2 (01 739 8559). You should be able to get suitable blank cassettes from any computer shop or department. The MZ-700 is an obsolescent machine now, but it is still quite well supported by mail-order - several Sharp software publishers advertise regularly in PCW.


I use an Acorn Atom computer and I've been getting very interested in Prestel and other 'on-line' databases recently, while using a friend's BBC micro. I am now trying to obtain a communications package for the Atom. I contacted my local Acorn dealers and they asked me what an 'Atom' was! I would appreciate any advice from you or other PCW readers.
Keith Maton, Harlow, Essex.

I couldn't think of anyone who produces an Atom comms. package, so I called Acorn (hope springs eternal...). They told me that 'the Atom is more or less deceased', which perhaps explains your dealer's question. The Atom was launched in 1980, before micro communications really caught on in the UK, so you may have trouble finding appropriate software. Acorn used to advertise a 'Teletext VDU card', designed to display Teletext and Ceefax pages, but this was based upon their even older 1K RAM 'System 1' ('Start with System 1 and continue as and when you like...') which has not been available for years.

You can probably communicate with any standard modem via the Atom's optional 6522 VIA (Versatile Interface Adaptor), but you may well have to write the software yourself and that's not a trivial task. If any readers can help, will they please write in care of PCW, or contact Keith's friend's BBC micro on Prestel Mailbox number 992 552 188.


I have an Apricot F1E printer and the superb Smith- Corona D200 dot-matrix printer. Although I can manage the whole range of printing styles of which this printer is capable I have been unable to print out a simple picture produced by Sketch, the graphics program which is supplied with the printer. Can you please help?

In addition can you suggest a reasonably priced program for artistic sketching which can use this printer's full graphic capabilities - bit image, dot addressable and block graphics?
W.T Moore, Glasbury-on-Wye, Herefordshire.

The Sketch program only supports the C-Itoh printers (including the colour model) which Apricot themselves supply. Apricot do not produce an alternative program themselves, but the list of third-party graphics programs for their computers runs to five pages, so there ought to be something to suit you. The Apricot dealer who sold you the system will have a copy of the latest issue of the 'Apricot Gold Catalogue', which lists such programs. Most of them are likely to be quite expensive, though, since the F1E is the 'baby' of the Apricot range.

This story has a double moral. Watch out for a hidden price when you're offered a 'free gift' - in this case the price of an Apricot printer if you want to use Sketch seriously. Secondly, remember that hardware is useless without appropriate software - it's no good having the latest whizzo printer with fabulous new features if you can't find any software to drive it.


I have been very impressed with the Trace/Step/Walk utility in the November PCW, which is better even than the 'Trace' option on Simon's Basic. Unfortunately I have just one problem. I have saved a machine-code version as explained in the REM statements but I cannot load this using lines 1 and 2 when I have a program in memory, since I have saved to disk and the lines want tape input.

I have tried juggling with lines such as LOAD "Trace*",8 instead of just LOAD but this crashes horribly. I do not mind having to type NEW, load Trace and then load my program again but it would be useful to be able to by-pass this kerfuffle. Does the author have any comments?
Peter Bilbrough, Yarm, Cleveland.

The author, Alexander Sassoon, tested the program from disk so it should work as you would wish it to do. You should make certain that there is an END command at the end of line 1. The command you specify in your letter will load anything on the disk with a name starting with the word 'TRACE', so you might cure the problem by typing the name in full:


From looking at the loading routine it seems that the computer would be certain to crash if the variable K was set to any value other than zero before the lines 1 and 2 were invoked. If your program sets K you should invoke line 1 with RUN rather than GOTO, as RUN re-initialises all variables. If you don't want to clear the variables you could use K=0:GOTO 1 instead of RUN, but don't put the K=0 command in either line - it must be typed as a direct command.


I have an Apple II+ and would like to know if it is possible to write Applesoft programs for an RS-232 interface card to control printers, modems and RTTY (in conjunction with Amateur Radio). Please can you suggest ways of doing this and advise me if the same program can control different modems or RTTY units.

I have seen a TTL to RS-232 converter in kit form, available from Maplin. Will this be suitable if it is used with the game paddle socket and programmed accordingly?

Is it possible to use the cassette input of the Apple to receive Morse (continuous wave) signals, finding the length of each dot or dash and the gap between letters and words and printing the message on the screen?
D Williams, Scunthorpe, South Humberside.

You can certainly control printers and modems from your Apple, using any of the plug-in RS-232 interface cards that are available. Most of these are based on a standard chip and offer a range of speeds from 75 up to 19200 baud. The most important speeds for modem communication are 75, 300 and 1200 baud. These are usually selectable by throwing small switches on the card, although more expensive cards allow you to change the speed with software - this is obviously more convenient, but it complicates the electronics.

A modem will convert the groups of pulses sent from the interface card into audio tones which can be transmitted by telephone or radio. You could try transmitting the tones from the Apple's cassette port directly, but quite a wide bandwidth is required - at least 9KHz for reliable operation. The signal is very sensitive to distortion and you'll only be able to pass messages to other Apple users, so this is not really a very practical option, especially as there are strict legal limits on the bands Radio Amateurs can use for data transmission.

RTTY is more difficult to cope with since the tones used are not the same as those generated by common modems. Three different sets of tones are in common use - Amateurs often use 1275 and 1445 Hertz for 0 and 1 respectively.

Another difference between modem communication and RTTY is the encoding system used for characters. ASCII (American Standard Code for Information Interchange) is the seven or eight bit code used by most computers and modem links, whereas Murray or Baudot code (two names for the same thing) is used for RTTY. This is a five-bit code which uses two 'sets' of codes to communicate capital letters, digits and punctuation. The most common RTTY data-rate is 50 baud, although other rates are sometimes used.

It should be clear that, in general, quite different software is needed for RTTY and modem communication. In the case of RTTY you might just as well take a signal directly from the Apple's cassette port or game paddle socket, generating and de-coding the tones in software. The data-rate is slow enough to allow you to write the majority of the code in Applesoft BASIC, though Integer BASIC might be a better choice if you've got it. A few short machine- code routines will probably be needed.

The game paddle socket only provides four single-bit digital outputs and three inputs, so it is probably not compatible with the Maplin kit. The socket also has a 'strobe' output and four slow analogue inputs.

The original Apple II reference manual (January 1978 - the one with a red cover and handwritten pages) contains the circuit of a mini RS-232 interface. This is made from two transistors and three resistors, so I don't think you'll find anything much simpler. It is good enough to drive a printer, slowly, but it wouldn't be much use with a modem since it doesn't cater for character input.

You should be able to read Morse code through the Apple's cassette port. I've done this on my Video Genie computer, although I found the technique rather sensitive to fluctuations in keying speed and phrasing. I don't know of any commercial software to do the job, but the US amateur radio fraternity may be able to guide you in this regard.

Applesoft Basic is fast enough to cope with most Morse speeds, so long as you write your program carefully. It is best to try to make the system 'self-clocking', i.e. capable of adjusting automatically to the speed of the input.

One way to do this is to measure the length of received tones and wait till you receive a tone at least twice as long as its predecessor. Work out the duration midway between the two and call anything shorter than that a dot and and anything longer a dash. Dashes are supposed to be three times longer than dots, so this simple rule divides the two quite effectively. Build up the pattern of dots and dashes in a string as you go along. When you encounter a pause longer than a dash, compare the pattern you've found with an array of valid patterns. If you don't find a match, try to re-assess the data-rate; otherwise, print the corresponding character.

It is a good idea to keep the valid patterns in two parts - those starting with a dot and those starting with a dash - since you then only need search one group when each character is received. A little trial and error will probably be needed to get things working, and a system like this can be improved almost indefinitely, to cope with wider speed changes, background noise and so on.

On the hardware side, a signal level of about 4 volts peak-to-peak at the cassette port can be sensed by PEEKing address -16288. You'll probably have to derive this signal from a speaker output, since most headphone outputs will lack sufficient output power. If the value you PEEK is intermittently greater than 127, a tone is being received - if it is consistently less than 128, there is no tone. The pitch doesn't matter much, so long as it is audible; you'll need a Beat Frequency Oscillator to detect a Continuous Wave signal in any case, so you should be able to select the pitch manually.


We are specialists in structural and environmental engineering, working as consultants. Presently we have access to DEC 10 and IBM 300 computers, and we are using the Fortran language for our problems.

We are interested in purchasing a personal micro with necessary acessories and software related to our field. It is difficult to decide what type of PC we should acquire. We are not much conversant with the electronic side of the machines and their specifications, so we request your advice in this regard.
Dr P Gopal Krishnan & Dr N.K Pareek, Tripoli, Libya.

It sounds as if your main need is for a machine which can run a full implementation of Fortran at a respectable speed. I assume that you want to run your existing software, and develop it further. If you intend to use a commercial computer aided design or engineering package you should choose the software first, and then buy whatever computer the software suppliers recommend.

If hardware is your main concern, there are two types of machine which you might consider: the new, fast MSDOS machines, and the computers based on the 68000 processor family. Both types of machine can run Fortran and are capable of processing floating-point numbers at speed.

The Research Machines Nimbus and the host of IBM PC-AT compatible machines are worth considering; the Nimbus is very fast and has good graphics, which may be significant in your applications. It runs ProFortran and RM Fortran, both of which resemble Fortran 77 (the latest widely-used standard, and broadly speaking a superset of Fortran 66 and Fortran 4. Other MSDOS machines will run DR Fortran, which is slightly cheaper than the other two. You should look for a machine which runs the MSDOS operating system and uses one of the more recent Intel processors - either the 80186 or the souped-up 80286.

The 68000-based machines range from Sinclair's QL through Atari's ST and the Commodore Amiga, to designs aimed at the CAE market such as the Wicat and Apollo Domain. Several Fortran compilers are available for the last two, but those machines are expensive - they are really work-stations rather than personal computers. The other 68000 machines are much cheaper than the IBM- compatibles; they run Fortran at least as fast, but they tend not to be so well constructed.

The Sinclair QL, despite a bad press and haphazard marketing, is a very fast number-crunching machine at a bargain price, even when you've added the disk drives needed to process large files. The Fortran compiler, from Sinclair, is a cheap but professional product, and there's an excellent QL compiler for the C language as well. You should think about moving your applications into C since it is a much better language for fast, reliable scientific programming. Good Fortran and C compilers for the Amiga and Atari ST are promised, but I have yet to see these in action.

Few suppliers I spoke to had links with Libya, so you should put facilities for local maintenance high on your list of requirements; this will probably help to make your decision for you. There is no shortage of computers that will do what you want. Ultimately, support is likely to be much more important than technical specifications.

Link to the top of this document    Link to the main index
Previous column ComputerAnswers    Next column ComputerAnswers