Cold Water

R Tape loading error, Part 2

In the first part I showed how the Sinclair ZX Spectrum stored data on tape. This second part explains what is stored, and what causes a tape loading error.

The ZX Spectrum BASIC offers a SAVE command for saving all kind of data. It can be used to save a BASIC program, variable arrays, but also arbitrary parts of memory. These files are always saved in two separate blocks. The first block is called header. It contains the file name, data type, and other meta information. The second block follows about a second later and contains the data itself.

The internal structure of each block is identical. The first byte distinguishes between header ($00) and data blocks ($FF). The final byte is a parity checksum. Everything between these two bytes is the payload.

A header block always contains a payload of 17 bytes. The first byte identifies the file type, followed by the file name (10 characters), followed by the length of the data block, and closed by two optional parameters that have different meanings depending on the file type. The length and the two parameters consume two bytes each, with the lower byte coming first because the Z80 CPU is little endian.

This is an example header block of a screenshot:

00$00 = Header
0003$03 = Binary file (Code or SCREEN$)
11001BLength: 6912 bytes ($1B00)
130040Parameter 1, here: starting address ($4000)
150000Parameter 2, here: unused

A screenshot is actually just a memory dump that starts at address $4000 (which is the starting address of the screen buffer) and is exactly 6912 bytes long (the ZX Spectrum has a resolution of 256×192 monochrome pixels plus 32×24 bytes color attributes, giving a screen buffer size of 6912 bytes).

For other file types, the two optional parameters have different meanings. For example, a BASIC program file stores the line number to start at after loading.

The final byte is the parity. It is used for error detection, and computed just by XOR-ing all the bytes that have been read. The result must be $00, otherwise a "R Tape loading error" is reported.

This kind of error detection is rather weak. Due to the nature of the XOR operation, two wrongs give a right. This means that when the block contains an even number of bad bits at the same position, they will be undetected. It is also not possible to correct reading errors, as the XOR operation only allows to identify the position of the bad bit, but not the actual byte that contained the error. More sophisticated error correction algorithms would have slowed down the loading process, though.

The parity is computed as a final step, after all the bytes have been read from the block on tape. For that reason, the loader can only decide at the end of the recording whether the loading was successful or not.

But then, why does the tape loading error sometimes appear while the block is still loading? Well, in the first part I have explained that the loading routine just reads an unknown number of bytes. It ends when waiting for a pulse change took to long. Now, if there is an audio gap on tape, the signal seems to end just in the middle of the block. It is then very likely that the parity checksum is wrong because there are still bytes missing.

Some simple copy protections made use of the way the Spectrum loads data from tape. A very common way were “headerless” files, where the header block was left out and only the data block was recorded on tape. The BASIC LOAD command was unable to read those files because of the missing header.

R Tape loading error

In the early time of home computers, at the beginning of the 1980's, hard disks and even floppy disks were too expensive for home use. The cheapest way for storing large amounts of data was the cassette tape. Cassettes and tape recorders were affordable and available in almost any household.

In this blog article, I'm going to explain how the Sinclair ZX Spectrum stored programs on cassette tapes. Other home computers of that time, like the Commodore 64 or Amstrad CPC, worked in a similar fashion.

Cassette tapes were designed to store audio signals like voice or music, so the inventors of the home computers had to find a way to convert data to audio signals. The easiest way is to serialize the data to a bit stream of 1's and 0's, and generate a long rectangular wave cycle for "1" and a short rectangular wave cycle for "0". This is what the ZX Spectrum actually does!

A short wave cycle is generated by giving power to the audio output for 855 so called T-states, and then turning off the power for another 855 T-states. A "T-state" is the time of a single clock pulse of the Z80-A CPU. As the CPU of a classic ZX Spectrum is clocked with 3.5 MHz, a T-state has a duration of 286 ns. The duration of a short wave cycle is thus 489 µs, giving an audio frequency of about 2,045 Hz. The long wave cycle is just twice as long.

Due to all kind of filters in the analog audio path, the rectangular signal is smoothed to a sinusoidal signal when played back. A Schmitt trigger inside the ZX Spectrum's hardware converts the audio signal back to a rectangular shape. Since the audio signal can have different amplitudes or could even be inverted, the hardware only cares for signal edges, not for levels. All that the loader routine now has to do is to measure the duration of the pulses, regenerate the bit stream, and put the bytes back together.

If you think that things cannot be that easy, you are right. 😄 The most difficult part for the loader is to find the start of the bit stream. If it is off by only one cycle (or even just a pulse), all the bytes are shifted by one bit, and the result is useless. All kind of noise on the tape makes it impossible to just wait for the signal to start, though.

For this reason, the recording starts with a leader signal, followed by a sync wave cycle, followed by the bit stream itself. The leader signal is just a continuous wave with a pulse length of 2,168 T-states, giving an 806 Hz tone that is displayed by red and cyan border colors on the TV. The sync wave cycle is a pulse of 667 T-States "on", followed by 735 T-states "off". After that, the actual data stream begins, which is displayed in blue and yellow border colors. When the last bit was transmitted, the data stream just ends.

So when the ZX Spectrum loads a file from tape, it first waits for the 806 Hz leader signal. If it was detected for at least 317 ms, it waits for the sync pulses, then it starts reading the bit sequence until there is a timeout while waiting for the next pulse.

It is a very simple way to store data on tape. And still, it is surprisingly reliable. After 30 years, I could recover almost all files from my old cassette tapes. Some of them were of the cheapest brands I could get my hands on back in 1987.

The only disadvantage is that this method is very slow. With 489 µs for a "0" and 978 µs for a "1", saving just 48 KBytes of data can take up to 6 minutes, giving an average bit rate of 1,363 bps (yes, bits per second). If we were to save a single 3 MBytes mp3 file that way, it would take almost 5 hours (and 5 cassettes with 60 minutes recording time each).

Some commercial games used speed loaders and copy protections. Speed loaders just reduced the number of T-states for the pulses, which increased the bit rate. Some copy protections used a "clicking" leader tone, where the leader signal was interrupted before the minimal detection time of 317 ms was reached. The original loader routine could not synchronize to these kind of signals, so it was impossible to read those files into copy programs. Those protection measures could still be circumvented by copying directly from tape to tape, but this only worked a few times due to increasing audio noise.

In the next article, I will take a deeper look at the bit stream contents, and I will also explain where the dreaded "R Tape loading error" comes from.

Premium Wall Bias Lighting, Part 3

I haven't forgotten about you. Some private stuff kept me from completing this project for a while. To make it up, I have added OpenSCAD files for a 3D printed case.

The controller was a little tricky to complete, mostly because of the very different component heights. I decided to use two circuit boards that are stacked onto each other by headers.

On the upper board, there are only the two buttons and the LCD, as well as the transistor and resistor for the LCD backlight. As I only used one-layer TriPad strip boards, I had to use this one upside down for the male headers to point downward. This rather unconventional use made it a little tricky to solder the buttons and LCD headers on the actual bottom side of the board.

The soldered controller boards. The lower board contains all the other components, as well as the wiring. The rotary encoder also made it to the lower board, because it is much taller than the other buttons. This way, the top of the button caps are almost level and nice to look at.

The result is surprisingly compact for a DIY solution. The button caps and the LCD are just perfectly positioned for a case.

With plastic feet attached, you can use the controller as it is. You can also get a plastic case with transparent top, drill three holes in it for the button caps, and mount the sandwich with spacers. But if you have the chance, you should definitely go for a 3D printed case.

I have set up a project at GitHub. It contains the circuit diagram, the bill of materials, the firmware source code, and OpenSCAD files for a printed case. There is no firmware binary yet, as you need to adapt the source code to the length of your LED strip anyway.

You will find the OpenSCAD files for the case in the GitHub project. There are bonus OpenSCAD files in the project, for printing a customized case. Due to the absence of properly layouted PCBs, I am aware that each controller is going to look differently when finished. In the parameter.scad file, you can change all kind of parameters, so you should be able to make your individual case in, well, almost any case (silly pun intended). 😄

The SPI flash memory of the Feather M0 Express is not used yet. In a future release, I may add a settings menu for the LED strip size. The controller is also forgetting all its settings when disconnected from the power. This needs to be addressed in a future release as well.

But after all, this is a start for your own DIY wall bias lighting. Feel free to send pull requests for enhancements!

Again, remember that you must remove the jumper before connecting the Feather to an USB port, otherwise your computer will be damaged.

Premium Wall Bias Lighting, Part 2

The completed prototype on a breadboard In the first part, I have assembled a working proof-of-concept for my premium wall bias lighting. Thanks to CircuitPython, it just took a couple of minutes to program a light effect once the hardware was working.

Now it's time to extend the hardware to its final stage. I'd like to have a LC display that shows the current settings. A button and a rotary encoder allows to browse through different menus and change the parameters. And finally, the strip shall be switched on and off by an illuminated power button.

Thanks to the bread board, the components were quickly added and connected to the Feather with some wires. Polling the buttons is a basic functionality of CircuitPython. It was also incredibly easy to poll the rotary encoder, because CircuitPython already comes with a library for that.

It took a lot more time to set up the LC display. CircuitPython supports SPI out of the box, but the SSD1803A controller of the display uses a weird protocol. Each command byte must be split up into two nibbles (4 bits), which are packed into bytes again, with the bit order reversed. The SPI library does not offer support for it, so I had to do all this bit mangling in Python, which turned out to become a rather ugly piece of code.

But then, finally, a minimal version of the firmware was working. I could turn the light on and off, select between two light effects, and I could also control the brightness.

However the Feather often took long breaks, where it did not react on key presses for multiple seconds. I guess the reason for that is Python's garbage collector, which stops the world while it is collecting unused objects and freeing some memory. This was actually a pretty annoying behavior that rendered the controller unusable.

After I added a third light effect, I also started to run into frequent out of memory errors. It seems that I have reached the limits of what is technically possible with CircuitPython on a Feather.

Was my approach too ambitious?

Luckily it wasn't. The Feather can also be programmed in C++, using the well known Arduino IDE. It comes with a lot of libraries that are ready to use. It's all very lightweight and is looking very promising. So why did I use Python in first place? Well, it is because I wrote my last lines of C++ code about 20 years ago. 😅

Porting the existing Python code to C++ was easier than I had expected. The SPI library now even supports reversed bit order, so it was much easier to address the LC display. On the down side, I had to test several libraries until I found a reliable one for the rotary encoder.

The C++ code consumes a fraction of the Python code's memory, so there is a lot left for extensions. The garbage collection breaks are also gone now, so the controller instantly responds to key presses. And I haven't even used the Feather's SPI flash memory yet. 😀

I have added some more light effects, and menus for adjusting brightness, saturation, and color temperature. Everything is working as expected now. It's time to finish the prototype phase and draw a circuit diagram.

R2 is the series resistor for the power button LED. A green LED would need an 68 Ω resistor at 3.3 V. However the LED is directly connected to the Feather, so the current should not exceed 7 mA (maximum rating is 10 mA). A 500 Ω resistor limits the current to a safe value. If you need more current for a fancy power LED, you can use one of the three 74HCT125 drivers left, or add a transistor.

R3 is the series resistor for the LCD backlight. The manufacturer specifies a 27 Ω resistor when the backlight LEDs are connected in series and powered with 5 V. If you use a different backlight, change the resistor accordingly. The BC 548 transistor permits up to 100 mA in this configuration.

Remember: You must remove the jumper JP1 before connecting the Feather to an USB port, or your computer will be damaged.

In the next part, I'm going to grab my soldering iron and build a final version. It's high time. The many wires on the breadboard prototype are annoying when operating the rotary encoder. Also its pins are too short and are often disconnecting from the breadboard when I use it.

Premium Wall Bias Lighting, Part 1

A good way to relieve the strain from your eyes while working on a PC, is to illuminate the wall behind your monitor. Jason Fitzpatrick wrote an interesting article about what bias lighting is and why you should be using it.

Many light sources can be used as bias lighting. I have used an old bedside lamp for a while. But what about something more stylish? What about a LED strip on an aluminum profile?

In this project, I am going to make a Wall Bias Lighting myself, and write a controller software for it. The source code will be released on my GitHub profile eventually, so you will be able to customize it.

Proof of Concept

To make it a true premium lighting, I use a LED strip that consists of SK6812 RGBW LEDs. It can produce colors, but it also has separate white LEDs for a clean neutral white. Even better: Each LED can be addressed and the color changed individually. It would be possible to illuminate the wall behind the monitor in a bright white, while the visible parts of the strip are in a soft blue that won't dazzle the eyes.

AdaFruit sells these LED strips under their brand name NeoPixel, but there are also no-name strips on the market that are fully compatible and considerably cheaper. The strips are usually sold on reels of up to 5 meters length. They can be shortened to the desired size with scissors, and have an adhesive tape on the back so they can be glued to aluminum profiles.

This is the bill of material for the first proof-of-concept phase of the project:

  • An SK6812 RGBW LED strip with 60 LEDs per meter
  • An aluminum wall profile for LED strips
  • 1x AdaFruit Feather M0 Express
  • 1x Level converter (read below)
  • 1x 1000 µF/16 V capacitor, 1x 500 Ω resistor (read here why they are needed)
  • 1x 5 V power brick. Each LED is said to consume up to 60 mA (I couldn't find concrete figures), so you will need 18 W per strip meter if you want to set all four colors of all LEDs to maximum brightness.

The assembly was rather simple. First I cut the profile and the strip to the desired length and glued them together. Then I connected the strip to the power supply, and the strip's data line to the Feather via the level converter.

The next thing on the to-do list was a quick test drive, to check if some of the LEDs are defective. So I installed CircuitPython on the Feather, and wrote a tiny test program that just cycles through the colors red, green, blue, and white. With this pattern, even a single defective LED would immediately catch one's eye.

I turned on the power supply, aaaand... Nothing! 😲 All the LEDs stayed black.

I checked and double checked the wiring, but everything seemed to be correct. I tested my test program on the single NeoPixel that is mounted on the Feather, and it worked there.

Puzzled, I connected my scope to the data line of the LED strip. It immediately revealed the culprit.

The Feather runs on 3.3 V, and so the signal on the data line has an amplitude of 3.3 V.

The LED strip runs on 5 V though, and also expects a signal amplitude of 5 V. The logic converter between the Feather and the strip is supposed to convert the 3.3 V signal to 5 V. However, the BSS138 based bi-directional logic level converter from my spare part box turned out to be too slow for this purpose. The output level starts at 3 V and then ramps up to 4 V.

This is not sufficient for the SK6812, which needs a 5 V signal and a very precise timing with clean signal edges. Both was not given, so the LEDs stayed black.

I replaced the logic level converter by a standard 74HCT125 buffer IC, and tried again. The LED strip immediately came to life and cycled through the colors. The scope now shows a clean (well, more or less clean) 5 V signal.

My proof-of-concept is working. 🎉 This is what the circuit looks like:

While the LED strip is powered by the power brick, the Feather is going to be powered by USB as long as I am developing the software. Later I will also supply the Feather with LED power, so it runs stand-alone.

Never connect the Feather to an USB port while it is supplied by an external power source. It could damage your computer.

What next? I'm going to add a power button, so I can turn the light on and off. For controlling the brightness and light effects, I am also going to add a display, a rotary switch, and another button. Stay tuned…