Slashdot Mirror
Current Radio Rules Mean Sinclair ZX Spectrum Wouldn't Fly Today
First time accepted submitter wisewellies writes "Ben clearly has way too much spare time on his hands, but he decided to see just how well an antiquated ZX Spectrum would hold up to modern EMC requirements. His blog is a good read if you're looking for something to do while pretending to work! From the blog: 'This year is the 30th anniversary of one of my favourite inventions of all time, the Sinclair ZX Spectrum. A few weeks ago, I finally bought one: a non-working one on eBay that I nursed back to health. Fortunately there was very little wrong with it. Unfortunately it's a 16K model, and a fairly early one at that, which won't run much software in its native state. This probably accounts for its unusually pristine condition. We took half an hour in the chamber to perform an approximate series of EN55022 measurements, to check its radiated emissions against today's standard. The question is, what have we learned as an industry since 1982?'"
Sorry, I couldn't help myself .....
Hey Hey 16k
upgrade the motherboard.
"I bless every day that I continue to live, for every day is pure profit."
"It’s not just a failure; it’s an abject one" Really? Now I admit the situation could be a ALOT worse with the accessories and cables, and until you've ran the test you don't know. But it's only about 6dB above the line, I've seen a lot worse problems [try 20dB!]. There is a good chance this would be a relatively easy fix when you start looking at the problem.
A ferrite bead on the power supply cable would probably fix the "bad power" supply if indeed that's what it is. And some judicious copper taping would likely fix the other problems. Worse case you do a board spin and add ferrite beads to the I/O and possibly move suspect traces into internal layers. Worse WORSE case you change the clocking to use spread spectrum which would likely not require any changes except in the clocking circuits. None of those would prevent a "modern" version of the product from going to market.. And a good engineer could probably implement them in less than 6 weeks in a production environment...
Plus it doesn't even manner, if you were going to bring a sinclair back to market it would draw about 20mA, run on USB power and be completely implemented on a single chip.... Because it has roughly the same processing power as a PIC uC.
The ZX didn't fly back in the day either. /rimshot
the preceding comment is my own and in no way reflects the opinion of the Joint Chiefs of Staff
EMC - Electro-Magnetic Compatibility
I have a ZX Spectrum in the loft I often see when I go up there. As far as RFI is concerned our regulations back then were non-existent. I once saw a BBC Micro for the German market, it was encased in metal and built like a tank. The ones on the UK market were plastic and caused havoc with my Amateur Radio gear until I quietened it a fair bit by coating the inside of the case with graphite spray and grounding it. TV's were another problem as they were susceptible to interference from Amateur radio transmissions operating within the legal limits and specifications and we had inspectors who audited our stations for compliance. It was all down to the manufacturers saving may be a penny or 2 by using a cheaper front-end transistor for TV's sold into the UK.
I wrote my first real program on a Sinclair. It was for TV troubleshooting and it took you down to the section. Storage was a cassette tape and the output was composite video for black & white TV.
Then I bought the memory expansion, took it to work and made a program for it to do cost estimate calculations. It was the 2nd computer anywhere in the company. I got promoted from cost estimating to Systems Administrator all in one go. I stayed with that company almost 30 years, then I left to start my own software company.
A few years ago I was telling that story to a client. He pulled a mint condition Sinclair -- still in the original box -- out of his desk and gave it to me. He said it bought it to learn computers and never used it. It was like giving me the keys to my first car.
Unfortunately it's a 16K model, and a fairly early one at that, which won't run much software
Hey! I was still supporting the 16k version with a game released in NINETY-two.
The original TRS-80 was a wideband RF jammer. Cheap PCB design, plastic (unshielded) case, lots of ribbon cable external interconnects operating at megahertz frequencies.
One of the better ways to see whether the machine was frozen or just processing a long-running (but productive) internal loop was turn on an AM radio in the same room. Within about 3 feet, the RF noise would override all but the strongest stations and allow you to monitor the CPU's execution by the hums and burbles of the RF noise.
It's why the original TRS-80 became the Model I, rapidly superseded by the all-in-one Model III (with lots of internal shielding).
Welcome to the Panopticon. Used to be a prison, now it's your home.
Here's the actual blog instead of some stupid Register article: http://focusritedevelopmentteam.wordpress.com/
>The question is, what have we learned as an industry since 1982?
Quite a bit.
I have three books on electromagnetic compatibility. The most recent, Electromagnetic Compatibility Engineering (2009), is comprehensive and thick enough to stun an ox.
I once made a QBasic program named "NOISE.BAS" on a 386 computer.
When ran, it made the radio which was playing, produce noise for 7 seconds.
All the program contained was:
SLEEP 7
We're really lucky that the FCC clamped down on RF emissions from electronics. Otherwise, we'd all be looking at big electromagnetic compatibility charts before buying anything, trying to find combinations known to work well together. Offices would need RF spectrum analyzers to figure out who brought in something that was messing up other gear. I mentioned in another post that you couldn't operate a Milton Bradley Big Trak and an TRS-80 near each other. The other side of stopping RF emissions is that the shielding makes electronics much less sensitive to RF interference.
The development of really good RF noise management technology made modern cell phones possible. The concept of a handheld device with four radios (GSM, WiFi, GPS, and Bluetooth), all operating simultaneously within a few inches of each other, was totally beyond the RF technology of a generation ago. Two generations ago, it was so bad that marine radio stations had miles of separation between the receivers and the transmitters.
You may both be thinking of the ZX80 or ZX81, which used this hack to drive the screen. The ZX Spectrum (released 1982) had proper display circuitry and did not suffer from this issue.
The ZX80's display hack was all a cheap way to get the data streamed out of RAM. To do this they placed the cpu's program counter(!) at the start of display memory, every time it tried to execute an instruction it would read a byte from memory - which was picked up to generate the display - but the data wasn't returned to the CPU and instead it was fed a byte of zero bits. 0x00 is the NOP instruction on Z80, so it would just increment the program counter and read the next byte. This means successive bytes appear in sequence on the data bus, without having to include a second agent that was capable of making accesses into the RAM - a considerable design simplification. Later, for the Spectrum, a separate circuit which generated and incremented its own address was able to access memory, and it had to be arranged that it would get priority over the CPU (which meant the CPU was considerably slower when running programs in the bottom 16k of memory).