Niche Knowledge: Z80 parallel port SD card on Zeta2

green circuit board with secure digital card slot on left and 40-pin parallel interface connectors on the right
Mini PPISD board: a slow SD card mass-storage system for 8-bit computers

Almost no-one will need this knowledge, but I might need to remember it. In order to add Mini PPISD support to a RomWBW 3.01-supported system, you need to create a file called something like Source/HBIOS/Config/ZETA2_ppisd.asm (for yes, I’m using a Zeta SBC V2) containing:

#include "cfg_zeta2.asm"

Running make from the top-level directory should create a ROM image called Binary/ZETA2_ppisd.rom for you to write to flash. Since my floppy drive isn’t feeling too happy, I had to resort to buying a TL866II Plus programmer to write the chip.

And it worked!

 RomWBW HBIOS v3.0.1, 2021-03-12

 ZETA V2 Z80 @ 8.000MHz
 0 MEM W/S, 1 I/O W/S, INT MODE 2
 512KB ROM, 512KB RAM

 CTC: MODE=Z2 IO=0x20
 UART0: IO=0x68 16550A MODE=38400,8,N,1
 DSRTC: MODE=STD IO=0x70 Sun 2021-03-14 17:47:13 CHARGE=OFF
 FD: IO=0x30 UNITS=2
 SD0: SDSC NAME=SD    BLOCKS=0x003C7800 SIZE=1935MB

 Unit        Device      Type              Capacity/Mode
 ----------  ----------  ----------------  --------------------
 Char 0      UART0:      RS-232            38400,8,N,1
 Disk 0      MD1:        RAM Disk          384KB,LBA
 Disk 1      MD0:        ROM Disk          384KB,LBA
 Disk 2      FD0:        Floppy Disk       3.5",DS/HD,CHS
 Disk 3      FD1:        Floppy Disk       3.5",DS/HD,CHS
 Disk 4      SD0:        SD Card           1935MB,LBA

 ZETA V2 Boot Loader

 ROM: (M)onitor (C)P/M (Z)-System (F)orth (B)ASIC (T)-BASIC (P)LAY (U)SER ROM  
 Disk: (0)MD1 (1)MD0 (2)FD0 (3)FD1 (4)SD0 

 Boot Selection? 

I was pleasantly surprised how easy it is to use a TL866 programmer under Linux. minipro does all the work, though. To write and verify the whole 512K Flash ROM, it’s:

minipro -p SST39SF040 -w ZETA2_ppisd.rom

The programmer supports over 16000 devices, of which around 10000 are variants (form factor, programming voltage, speed, OTP, etc). It’ll also verify over 100 different 74-series logic chips. It’s not a super cheap device (mine was a little over $80, from Simcoe Diy) but it does a lot for that price.

Next stop: try rebuilding BBC BASIC with RomWBW’s timer support included ..

Seeeduino XIAO simple USB volume control with CircuitPython

round computer device with USB cable exiting at left. Small microcontroller at centreshowing wiring to LED ring and rotary encoder
Slightly blurry image of the underside of the device, showing the Seeeduino XIAO and the glow from the NeoPixel ring. And yes, the XIAO is really that small

Tod Kurt’s QTPy-knob: Simple USB knob w/ CircuitPython is a fairly simple USB input project that relies on the pin spacing of an Adafruit QT Py development board being the same as that on a Bourns Rotary Encoder. If you want to get fancy (and who wouldn’t?) you can add a NeoPixel Ring to get an RGB glow.

The QT Py is based on the Seeeduino XIAO, which is a slightly simpler device than the Adafruit derivative. It still runs CircuitPython, though, and is about the least expensive way of doing so. The XIAO is drop-in replacement for the Qt Py in this project, and it works really well! Everything you need for the project is described here: todbot/qtpy-knob: QT Py Media Knob using rotary encoder & neopixel ring

I found a couple of tiny glitches in the 3d printed parts, though:

  1. The diffuser ring for the LED ring is too thick for the encoder lock nut to fasten. It’s 2 mm thick, and there’s exactly 2 mm of thread left on the encoder.
  2. The D-shaft cutout in the top is too deep to allow the encoder shaft switch to trigger.

I bodged these by putting an indent in the middle of the diffuser, and filling the top D-shaft cutout with just enough Blu Tack.

Tod’s got a bunch of other projects for the Qt Py that I’m sure would work well with the XIAO: QT Py Tricks. And yes, there’s an “Output Farty Noises to DAC” one that, regrettably, does just that.

Maybe I’ll add some mass to the dial to make it scroll more smoothly like those buttery shuttle dials from old video editing consoles. The base could use a bit more weight to stop it skiting about the desk, so maybe I’ll use Vik’s trick of embedding BB gun shot into hot glue. For now, I’ve put some rubber feet on it, and it mostly stays put.

Hey! Unlike my last Seeed Studio device post, I paid for all the bits mentioned here.

Nyan Cat, except it gets faster — RTTTL on the Raspberry Pi Pico

Raspberry Pi Pico with small piezo speaker connected to pins 23 and 26
piezo between pins 26 and 23. Sounds a little like this:

It was inevitable:

  1. A Raspberry Pi Pico; plus
  2. a tiny piezo PC beeper; plus
  3. MicroPython; plus
  4. dhylands / upy-rtttl; plus
  5. nyancat.rtttl; plus
  6. my unfailing sense of knowing when to stop, then ignoring it

brings you this wonderful creation, which plays the Nyan Cat theme forever, except it gets 20% faster each time. This is weapons-grade annoying (thank’ee kindly), so I’m not going to include a recording here. If you must hear an approximation, paste the RTTTL into Play RTTTL Online and enjoy.

# Raspberry Pi Pico RTTTL example
# nyan cat, but it gets faster
# scruss - 2021-02: sorry, not sorry ...

# Uses from
# Nyan Cat RTTTL from

from rtttl import RTTTL
from time import sleep_ms
from machine import Pin, PWM

b = 90    # bpm variable

# pin 26 - GP20; just the right distance from GND at pin 23
#  to use one of those PC beepers with the 4-pin headers
pwm = PWM(Pin(20))
led = Pin(25, Pin.OUT)

def play_tone(freq, msec):
    # play RTTL notes, also flash onboard LED
    print('freq = {:6.1f} msec = {:6.1f}'.format(freq, msec))
    if freq > 0:
        pwm.freq(int(freq))       # Set frequency
        pwm.duty_u16(32767)       # 50% duty cycle
    sleep_ms(int(0.9 * msec))     # Play for a number of msec
    pwm.duty_u16(0)               # Stop playing for gap between notes
    sleep_ms(int(0.1 * msec))     # Pause for a number of msec

while True:
    nyan = 'nyancat:d=4,o=5,b=' + str(b) + ':16d#6,16e6,8f#6,8b6,16d#6,16e6,16f#6,16b6,16c#7,16d#7,16c#7,16a#6,8b6,8f#6,16d#6,16e6,8f#6,8b6,16c#7,16a#6,16b6,16c#7,16e7,16d#7,16e7,16c#7,8f#6,8g#6,16d#6,16d#6,16p,16b,16d6,16c#6,16b,16p,8b,8c#6,8d6,16d6,16c#6,16b,16c#6,16d#6,16f#6,16g#6,16d#6,16f#6,16c#6,16d#6,16b,16c#6,16b,8d#6,8f#6,16g#6,16d#6,16f#6,16c#6,16d#6,16b,16d6,16d#6,16d6,16c#6,16b,16c#6,8d6,16b,16c#6,16d#6,16f#6,16c#6,16d#6,16c#6,16b,8c#6,8b,8c#6,8f#6,8g#6,16d#6,16d#6,16p,16b,16d6,16c#6,16b,16p,8b,8c#6,8d6,16d6,16c#6,16b,16c#6,16d#6,16f#6,16g#6,16d#6,16f#6,16c#6,16d#6,16b,16c#6,16b,8d#6,8f#6,16g#6,16d#6,16f#6,16c#6,16d#6,16b,16d6,16d#6,16d6,16c#6,16b,16c#6,8d6,16b,16c#6,16d#6,16f#6,16c#6,16d#6,16c#6,16b,8c#6,8b,8c#6,8b,16f#,16g#,8b,16f#,16g#,16b,16c#6,16d#6,16b,16e6,16d#6,16e6,16f#6,8b,8b,16f#,16g#,16b,16f#,16e6,16d#6,16c#6,16b,16f#,16d#,16e,16f#,8b,16f#,16g#,8b,16f#,16g#,16b,16b,16c#6,16d#6,16b,16f#,16g#,16f#,8b,16b,16a#,16b,16f#,16g#,16b,16e6,16d#6,16e6,16f#6,8b,8a#,8b,16f#,16g#,8b,16f#,16g#,16b,16c#6,16d#6,16b,16e6,16d#6,16e6,16f#6,8b,8b,16f#,16g#,16b,16f#,16e6,16d#6,16c#6,16b,16f#,16d#,16e,16f#,8b,16f#,16g#,8b,16f#,16g#,16b,16b,16c#6,16d#6,16b,16f#,16g#,16f#,8b,16b,16a#,16b,16f#,16g#,16b,16e6,16d#6,16e6,16f#6,8b,8c#6'
    tune = RTTTL(nyan)
    print('bpm: ', b)
    for freq, msec in tune.notes():
        play_tone(freq, msec)
    b = int(b * 1.2)

Visualizing PWM

my desk is not usually this tidy

I built a DS0150 — successfully, on my second try — and wanted to measure something. My demo MicroPython program from MicroPython on the terrible old ESP8266-12 Development Board has been running since May 2020, and the RGB LED’s red channel is conveniently broken out to a header. Since the DSO150 has precisely one channel, it let me see the how the PWM duty cycle affects the voltage (and brightness) of the LED.

I can’t think of oscilloscopes without being reminded of a scene from one of my favourite sci-fi books, The Reproductive System by John Sladek. Cal, the hapless new hire in the Wompler toy factory-turned-military research lab, is showing the boss some equipment while making up more and more extravagant names for them for the clueless owner:

At each exhibit, Grandison [Wompler] would pause while Cal named the piece of equipment. Then he would repeat the name softly, with a kind of wonder, nod sagely, and move on. Cal was strongly reminded of the way some people look at modern art exhibitions, where the labels become more important to them than the objects. He found himself making up elaborate names.
“And this, you’ll note, is the Mondriaan Modular Mnemonicon.”
“—onicon, yes.”
“And the Empyrean diffractosphere.”
“—sphere. Mn. I see.”
Nothing surprised Grandison, for he was looking at nothing. Cal became wilder. Pointing to Hita’s desk, he said, “The chiarascuro thermocouple.”
“Couple? Looks like only one, to me. Interesting, though.”
A briar pipe became a “zygotic pipette,” the glass ashtray a “Piltdown retort,” and the lamp a “phase-conditioned Aeolian.” Paperclips became “nuances.”
“Nuances, I see. Very fine. What’s that thing, now?”
He pointed to an oscilloscope. Cal took a deep breath.
“Its full name,” he said, “is the Praetorian eschatalogical morphomorphic tangram, Endymion-type, but we usually just call it a ramification.”
The old man fixed him with a stern black eye. “Are you trying to be funny or something? I mean, I may not be a smart-aleck scientist, but I sure as hell know a television when I see one.”
Cal assured him it was not a television, and proved it by switching it on. “See,” he said, pointing to a pattern of square waves, “there are the little anapests.”

— The Reproductive system, by John Sladek (text copypasta from Grey Goo in the 1960s)

So we were displaying roughly 500 anapests/s there. Not bad, not bad at all …

Raspberry Pi Pico: DS18x20 in MicroPython

Hidden away in the Pico MicroPython guide is a hint that there may be more modules installed in the system than they let on. In the section about picotool, the guide has a seemingly innocuous couple of lines:

frozen modules: _boot, rp2, ds18x20, onewire, uasyncio, uasyncio/core, uasyncio/event, uasyncio/funcs, uasyncio/lock, uasyncio/stream

The third and fourth ‘frozen modules’ are a giveaway: it shows that support for the popular Dallas/Maxim DS18x20 1-Wire temperature sensors is built in. Nowhere else in the guide are they mentioned. I guess someone needs to write them up …

DS18x20 digital temperature sensors — usually sold as DS18B20 by Maxim and the many clone/knock-off suppliers — are handy. They can report temperatures from -55 to 125 °C, although not every sensor will withstand that range. They come in a variety of packages, including immersible sealed units. They give a reliable result, free from ADC noise. They’re fairly cheap, the wiring’s absurdly simple, and you can chain long strings of them together from the same input pin and they’ll all work. What they aren’t, though, is fast: 1-Wire is a slow serial protocol that takes a while to query all of its attached devices and ferry the results back to the controller. But when we’re talking about environmental temperature, querying more often than a few times a minute is unnecessary.

So this is the most complex way you can wire up a DS18x20 sensor:

breadboard with raspberry Pi Pico, DS18x20 sensor with 47 kΩ pull-up resistor between 3V3 power and sensor data line
Raspberry Pi Pico connected to a single DS18x20 sensor

and this is how it’s wired:

   DS18X20    Pico
   =========  =========
   VDD      → 3V3
     --47 kΩ--
   DQ       → GP22
   GND      → GND

 (47 kΩ resistor between DQ and 3V3 as pull-up)

Adding another sensor is no more complicated: connect it exactly as the first, chaining the sensors together —

breadboard with raspberry Pi Pico, two DS18x20 sensors with 47 kΩ pull-up resistor between 3V3 power and sensor data line
Two DS18x20 sensors, though quite why you’d want two temperature sensors less than 8 mm apart, I’ll never know. Imagine one is a fancy immersible one on a long cable …

The code is not complex, either:

# Raspberry Pi Pico - MicroPython DS18X20 Sensor demo
# scruss - 2021-02
# -*- coding: utf-8 -*-

from machine import Pin
from onewire import OneWire
from ds18x20 import DS18X20
from time import sleep_ms
from ubinascii import hexlify    # for sensor ID nice display

ds = DS18X20(OneWire(Pin(22)))
sensors = ds.scan()

while True:
    sleep_ms(750)     # mandatory pause to collect results
    for s in sensors:
        print(hexlify(s).decode(), ":", "%6.1f" % (ds.read_temp(s)))

This generic code will read any number of attached sensors and return their readings along with the sensor ID. The sensor ID is a big ugly hex string (the one I’m using right now has an ID of 284c907997070344, but its friends call it ThreeFourFour) that’s unique across all of the sensors that are out there.

If you’re reading a single sensor, the code can be much simpler:

# Raspberry Pi Pico - MicroPython 1x DS18X20 Sensor demo
# scruss - 2021-02
# -*- coding: utf-8 -*-

from machine import Pin
from onewire import OneWire
from ds18x20 import DS18X20
from time import sleep_ms

ds = DS18X20(OneWire(Pin(22)))
sensor_id = ds.scan()[0]  # the one and only sensor

while True:
    sleep_ms(750)         # wait for results
    print(ds.read_temp(sensor_id), " °C")

The important bits of the program:

  1. Tell your Pico you have a DS18x20 on pin GP22:
    ds = DS18X20(OneWire(Pin(22)))
  2. Get the first (and only) sensor ID:
    sensor_id = ds.scan()[0]
  3. Every time you need a reading:
    1. Request a temperature reading:
    2. Wait for results to come back:
    3. Get the reading back as a floating-point value in °C:

That’s it. No faffing about with analogue conversion factors and mystery multipliers. No “will it feel like returning a result this time?” like the DHT sensors. While the 1-Wire protocol is immensely complicated (Trevor Woerner has a really clear summary: Device Enumeration on a 1-Wire Bus) it’s not something you need to understand to make them work.

Quick labelled Fritzing Raspberry Pi Pico layout

half-size breadboard with Raspberry Pi Pico mounted on top. Labels for each of the pin functions are on the left and right
and now, with labels!

Nothing particularly new or innovative here, but if you’re making simple Raspberry Pi Pico circuits and need to explain them to folks, this little Fritzing template might help. It’s mashed up from:

  1. the pinout diagram (chopped and scaled);
  2. the Raspberry Pi Pico Fritzing part.

Since both of these components are from the Raspberry Pi Foundation’s Getting Started documentation, it’s supplied under the same licence.

Presentation: Getting Started with MicroPython on the Raspberry Pi Pico

I just gave this talk to the Toronto Raspberry Pi Meetup group: Getting Started with MicroPython on the Raspberry Pi Pico. Slides are here:

or, if you must, pptx:

If I were to do this again, I’d drop the messy thermistor code as an example and use a DS18x20, like here: Raspberry Pi Pico: DS18x20 in MicroPython

also, simple potentiometer demo I wrote during the talk:

Book: Fortran techniques with special reference to non-numerical applications (1972)

Programming flow diagram, with the flow of a program using subroutines on the left ("closed coding") and the same structure on the right written as a series of GOTO-controlled sections ("open coding") to save computer memory and execution time
Subroutines do, however, bring with them considerable
overheads in both space and execution time

Imagine you have a programming task that involves parsing and analyzing text. Nothing complicated: maybe just breaking it into tokens. Now imagine the only programming language you had available:

  • has no text handling functions at all: you can pack characters into numeric types, but how they are packed and how many you get per type are system dependent;
  • allows integers in variables starting with the letters I→N, with A→H and O→Z floating point;
  • has IF … THEN but no ELSE, with the preferred form being
    IF (expr) neg, zero, pos
    where expr is the expression to evaluate, and neg, zero and pos are statement labels to jump to if the evaluation is negative, zero or positive, respectively;
  • has only enough memory for (linear, non-associative) arrays of a couple of thousand entries;
  • disallows recursion completely;
  • charges for computing time such that a solo researcher’s work might cost many times their salary in a few weeks.

Sounds impossible, right? But that’s the world described in Colin Day’s book from 1972, Fortran techniques with special reference to non-numerical applications.

The programming language used is USA Standard FORTRAN X3.9 1966, commonly known as Fortran IV after IBM’s naming convention. For all it looks crude today, Fortran was an efficient, sod-the-theory-just-get-the-job-done language that allowed numerical problems to be described as a text program and solved with previously impossible speed. Every computer shipped with some form of Fortran compiler at the time. Day wasn’t alone working within Fortran IV’s text limitations in the early 1970s: the first Unix tools at Bell Labs were written in Fortran IV — that was before they built themselves their own toolchain and invented the segmentation fault.

The book is a small (~ 90 page) delight, and is a window into system limitations we might almost find unimaginable. Wanna create a lookup table of a thousand entries? Today it’s a fraction of a thought and microseconds of program time. But nearly fifty years ago, Colin Day described methods of manually creating two small index and target arrays and rolling your own hash functions to store and retrieve stuff. Text? Hollerith constants, mate; that’s yer lot — 6HOH HAI might fit in one computer word if you were running on big iron. Sorting and searching (especially without recursion) are revealed to be the immensely complex subjects they are, all hidden behind today’s one-liner methods. Day shows methods to simulate recursion with arrays standing in for pointer stacks of GO TO targets (:coding_horror_face:). And if it’s graphics you want, that’s what the line printer’s for:

Damped cosine 2d function density plot rendered as mono-spaced characters, approximately 60 colums across, made up of only X, 0, *, +, - and space characters
… the most serious drawback to a density plot of the type shown above is the limited number of characters used to represent the height above the page.”
(This image was deemed impressive enough by Cambridge University Press that they used it as the cover of the book. The same function became a bit of a visual cliché, with home computers being able to render it in colour and isometric 3D less than a decade later.)

Why do I like this book enough to track down a used copy, import it, scan it, correct it and upload it to the Internet Archive? To me, it shows the layers we now take for granted, and the privilege we have with these hard problems of half a century ago being trivially soluble on a $10 computer the size of a stick of gum. When we run today’s massive AI models with little interest in the underlying assumptions but a sharp focus on getting the results we want, we do a disservice to the years of R&D that got us here.

The ‘charges for computing time’ comment above is from Colin’s website. Early central computing facilities had the SaaS billing down solid, partly because many mainframes were rented from the vendor and system usage was accounted for in minute detail. Apparently the system Colin used (when a new lecturer) was at another college, and it was the custom to send periodic invoices for CPU time and storage used back to the user’s department. Nowhere on these invoices did it say that these accounts were for information only and were not payable. Not the best way to greet your users.

(Incidentally, if you hate yourself and everyone else around you, you can get a feel of system billing on any Linux system by enabling user quotas. You’ll very likely stop doing this almost immediately as the restrictions and reporting burden seem utterly alien to us today.)

While the book is still very much in copyright, the copy I have sat unread at Lakehead University Library since June 1995; the due date slip’s still pasted in the back. It’s been out of print at Cambridge University Press since May 1987, even if they do have a plaintive/passive aggressive “hey we could totally make an ebook of this if you really want it” link on their site. I — and the lovely folks hosting it at the Internet Archive — have saved them from what’s evidently too much trouble. I won’t even raise an eyebrow if they pull a Nintendo and start selling this scan.

Colossal thanks to Internet Archive for making the book uploading process much easier than I thought it was. They’ve completely revamped the processing behind it, and the fully open-source engine gives great results. As ever, if you assumed you knew how to do it, think again and read the How to upload scanned images to make a book guide. Uploading a zip file of images is much easier than mucking about with weird command-line TIFF and PDF tools. The resulting PDF is about half the size of the optimized scans I uploaded, and it’s nicely tagged with metadata and contains (mostly) searchable text. It took more than an hour to process on the archive’s spectacularly powerful servers, though, so I hate to think what Colin Day’s bill would have been in 1972 for that many CPU cycles … or if even a computer of that time, given enough storage, could complete the project by now.


One of the earlier acknowledgements of the inevitability of TIMTOWTDI in programming:

In computing there is always more than one correct way of approaching a given problem. Generally a standard mathematical method for solution can be found, or a method developed. Programs using the same method can still be written in more than one correct way.
from Digital Equipment Corporation, PDP-8 Handbook Series: Programming Languages (May 1970), p.12-6

Admittedly, it’s talking about BASIC — and by BASIC, PDP-8 BASIC was very basic¹ indeed — but there’s always more than one correct way to implement a solution.

¹: no text string handling, variable names limited to two characters in [A-Z][0-9] format, IF…THEN can only take a line number as argument (as with Dartmouth BASIC), one statement per line, max 350 lines or so. I’d heard that DEC thought that BASIC was going to be a passing fad and that their own FOCAL language was going to “win”, so their BASIC offerings were deliberately given less attention than FOCAL. Hmm …

bench64: a new BASIC benchmark index for 8-bit computers

Nobody asked for this. Nobody needs this. But here we are …

commodore 64 screen shot showing benchmark results:
basic bench index
>i good. ntsc c64=100

1/8 - for:
 60 s; 674.5 /s; i= 100
2/8 - goto:
 60 s; 442.3 /s; i= 100
3/8 - gosub:
 60 s; 350.8 /s; i= 100
4/8 - if:
 60 s; 242.9 /s; i= 100
5/8 - fn:
 60 s; 60.7 /s; i= 100
6/8 - maths:
 60 s; 6.4 /s; i= 100
7/8 - string:
 60 s; 82.2 /s; i= 100
8/8 - array:
 60 s; 27.9 /s; i= 100

overall index= 100

bench64 running on the reference system, an NTSC Commodore 64c

Inspired by J. G. Harston’s clever but domain-specific ClockSp benchmark, I set out to write a BASIC benchmark suite that was:

  1. more portable;
  2. based on a benchmark system that more people might own;
  3. and a bunch of other less important ideas.

Since I already had a Commodore 64, and seemingly several million other people did too, it seemed like a fair choice to use as the reference system. But the details, so many details …

basic bench index
>i good. ntsc c64=100

1/8 - for:
 309.5 s; 130.8 /s; i= 19 
2/8 - goto:
 367.8 s; 72.1 /s; i= 16 
3/8 - gosub:
 340.9 s; 61.7 /s; i= 18 
4/8 - if:
 181.8 s; 80.1 /s; i= 33 
5/8 - fn:
 135.3 s; 26.9 /s; i= 44 
6/8 - maths:
 110.1 s; 3.5 /s; i= 54 
7/8 - string:
 125.8 s; 39.2 /s; i= 48 
8/8 - array:
 103 s; 16.3 /s; i= 58 

overall index= 29
It was entirely painful running the same code on a real ZX Spectrum at under ⅓ the speed of a C64

(I mean: who knew that Commodore PET BASIC could run faster or slower depending on how your numbered your lines? Not me — until today, that is.)

While the benchmark doesn’t scale well for BASIC running on modern computers — the comparisons between a simple 8-bit processor at a few MHz and a multi-core wildly complex modern CPU at many GHz just aren’t applicable — it turns out I may have one of the fastest 8-bit BASIC computers around in the matchbox-sized shape of the MinZ v1.1 (36.864 Z180, CP/M 2.2, BBC BASIC [Z80] v3):

>I GOOD. NTSC C64=100

1/8 - FOR:
 3.2 S; 12778 /S; I= 1895 
2/8 - GOTO:
 6.1 S; 4324.5 /S; I= 978 
3/8 - GOSUB:
 3.1 S; 6789 /S; I= 1935 
4/8 - IF:
 2.9 S; 4966.9 /S; I= 2046 
5/8 - FN:
 3.5 S; 1030.6 /S; I= 1698 
6/8 - MATHS:
 1.5 S; 255.3 /S; I= 4000 
7/8 - STRING:
 2.6 S; 1871.6 /S; I= 2279 
8/8 - ARRAY:
 3.1 S; 540.3 /S; I= 1935 


That’s more than 9× the speed of a BBC Micro Model B.

Github link: bench64 – a new BASIC benchmark index for 8-bit computers.

Archive download:

Raspberry Pi Meetup tonight: the SeedStudio Wio Terminal

Anthopomorphized line drawing of the Wio terminal, with a simple smiling face on the screen, waving arms and legs with feet underneath
Wio Terminal-chan, the mascot for SeedStudio’s Wio Terminal

Hey – the Toronto Raspberry Pi Meetup Group is meeting online tonight! All welcome: you don’t have to be in/near Toronto to attend.

I’ll be introducing the SeeedStudio Wio Terminal: a flexible, small input and display device. The Wio Terminal has many interesting uses — including as an adjunct to or even alternative to the Raspberry Pi

Thursday, December 10, 2020
7:00 PM to 8:30 PM EST

Signup link:
or directly on Google Meet:


Stand for PROTODOME’s 4000AD chiptune album

4000AD album PCB in stand, with Print+ 3D printed headphones in foreground
4000AD album PCB in stand, with Print+ 3D printed headphones in foreground

So Dr. Blake “PROTODOME” Troise (previously) made a chiptune album that’s entirely synthesized by an Atmel/Microchip ATmega328P microcontroller in realtime. And every chip needs a PCB, right? So Blake released the album as a physical device you can solder up for yourself.

Of course, having the PCB lying flat doesn’t allow you to see Marianne Thompson’s great pixel cover art, or read the liner notes on the back — and risks having the circuit short out on random tinny things on your desk. (Maybe that’s just my desk, though.)

This stand allows you to display the board at a convenient 75° angle, but also allows the PCB to be flipped forward so you can read the liner notes comfortably. Yeah, I may have been a crate-digger at one time.


(or Thingiverse link:

4000AD album PCB in stand, front view
4000AD album PCB in stand, rear view (messy Kapton taping optional)
4000AD album PCB stand in black low-sheen PLA

calculator for engineering nerds

For something to do with my head, I’m taking the RAC Advanced Ham Radio course. The exam uses a non-programmable scientific calculator. I thought that all my calculators were programmable, but we found this one lurking in the basement and it’s just perfect:

Casio fx-115MS scientific calculator
SI prefixes above 1…9 on the fx-115MS: f, p, n µ, m, k, M, G, T

This is one of the few calculators I’ve seen that both displays and takes inputs in SI units. How to put it into SI engineering display mode is explained in this delightful (archived) site Casio fx-115MS.

Entering numbers with SI prefixes is simple: type the number, then Shift and hit the prefix. So to enter 300000, you’d type 300 Shift 4 to get 300 k.

You do have to be a little careful reading the display in this mode, though. The display above reads 221 × 10-3 (from the m at right), or 0.221.

I don’t see any calculator in Casio’s current range that offers this handy feature. Guess I’m lucky I found it before the exam!

speech on Raspberry Pi: espeak-ng

Audio can be a bit dismal on a Raspberry Pi. Once you get a configuration that works, sometimes you’re not sure how you got there and you’ll do anything to keep that arcane setup going. It’s better than it was.

Speech synthesis or TTS adds an extra layer for potential failure. One of the popular Linux TTS systems, eSpeak, hasn’t seen much development in almost a decade and seems to only work through workarounds and hand-waving.

Thankfully, there’s a fork of eSpeak that is maintained: espeak-ng. Better yet, it’s packaged with Raspberry Pi OS and can be installed quite easily:

sudo apt install espeak-ng espeak-ng-data libespeak-ng-dev

In my simple tests, it output everything I expected of it.

eSpeak had a Python module that kinda worked, but espeak-ng’s is much more ambitious, and (mostly) does what it sets out to do. You can install it like this:

sudo pip3 install py-espeak-ng

py-espeak-ng has some documentation, but it’s still got some trial and error in getting it to work. The biggest issue that held me up was that the module needs to be initialized with a voice that espeak-ng already knows about. If you don’t specify a voice, or specify one that the system doesn’t know about, you won’t get any errors — but you won’t get any output, either.

Here’s a small Python example that you’ll probably want to try with no-one else within earshot. It repeats the same English phrase (a favourite of elocution teachers) in every English regional language that espeak-ng knows about. In addition, since I’m a dictionary nerd, it outputs phonetics too.

# -*- coding: utf-8 -*-
# an espeakng elocution lesson from scruss, 2020-07
#     I suffered this at school, now you get to as well!
# You will need to:
#     sudo apt install espeak-ng espeak-ng-data libespeak-ng-dev
#     sudo pip3 install py-espeak-ng

from espeakng import ESpeakNG
from time import sleep

# you have to initialize with a voice that exists
#   `espeak-ng --voices=en` will list English ones
esng = ESpeakNG(voice='en-gb')
esng.pitch = 32
esng.speed = 150

phrase = "Father's car is a Jaguar and pa drives rather fast. "\
    "Castles, farms and draughty barns, all go charging past."

for voice in esng.voices:
    if voice['language'].startswith('en-'):
        print('Using voice:', voice['language'],
              'for', voice['voice_name'], '-')
        esng.voice = voice['language']
        ipa = esng.g2p(phrase, ipa=2)
        print(voice['language'], 'phonetics:', ipa)
        esng.say(phrase, sync=True)

Be thankful you can’t hear the output. The IPA output, however, is a thing of beauty:

Father's car is a Jaguar and pa drives rather fast. Castles, farms and draughty barns, all go charging past.

Using voice: en-029 for English_(Caribbean) -
en-029 phonetics: fˈɑːdaz kˈɑ͡əɹ ɪz a d͡ʒˈaɡwɑ͡ə and pˈɑː dɹˈa͡ɪvz ɹˈɑːda fˈa͡astkˈa͡asɛlzfˈɑ͡əmz and dɹˈa͡afti bˈɑ͡ənzˈɔːl ɡˌo͡ʊ t͡ʃˈɑ͡əd͡ʒɪn pˈa͡ast

Using voice: en-gb for English_(Great_Britain) -
en-gb phonetics: fˈɑːðəz kˈɑːɹ ɪz ɐ d͡ʒˈaɡwɑː and pˈɑː dɹˈa͡ɪvz ɹˈɑːðə fˈastkˈasə͡lzfˈɑːmz and dɹˈafti bˈɑːnzˈɔːl ɡˌə͡ʊ t͡ʃˈɑːd͡ʒɪŋ pˈast

Using voice: en-gb-scotland for English_(Scotland) -
en-gb-scotland phonetics: fˈa:ðɜz kˈaːr ɪz ɐ d͡ʒˈaɡwaːr and pˈa: drˈa͡ɪvz rˈa:ðɜ fˈa:stkˈa:sə͡lzfˈaːrmz and drˈa:fte bˈaːrnzˈɔːl ɡˌoː t͡ʃˈaːrd͡ʒɪŋ pˈa:st

WeAct F411 + MicroPython + NeoPixels

Further to the Canaduino STM32 boards with MicroPython writeup, I thought I’d start showing how you’d interface common electronics to the WeAct F411 boards. First off, NeoPixels!

Rather than use the Adafruit trade name, these are more properly called WS2812 LEDs. Each one contains a tiny microcontroller and it only takes three connections to drive a long chain of addressable colour LEDs. The downside is that the protocol to drive these is a bit of a bear, and really needs an accurate, fast clock signal to be reliable.

The STM32F411 chip does have just such a clock, and the generic micropython-ws2812 library slightly misuses the SPI bus to handle the signalling. The wiring’s simple:

  • F411 GND to WS2812 GND;
  • F411 3V3 to WS2812 5V;
  • F411 PA7 (SPI1_MOSI) PB15 (SPI2_MOSI) to WS2812 DIn

Next, copy into the WeAct F411’s flash. Now create a script to drive the LEDs. Here’s one to drive 8 LEDs, modified from the library’s advanced example:

# -*- coding: utf-8 -*-

import time
import math

from ws2812 import WS2812

ring = WS2812(spi_bus=2, led_count=8, intensity=0.1)

def data_generator(led_count):
    data = [(0, 0, 0) for i in range(led_count)]
    step = 0
    while True:
        red = int((1 + math.sin(step * 0.1324)) * 127)
        green = int((1 + math.sin(step * 0.1654)) * 127)
        blue = int((1 + math.sin(step * 0.1)) * 127)
        data[step % led_count] = (red, green, blue)
        yield data
        step += 1

for data in data_generator(ring.led_count):

Previously I said you’d see your WS2812s flicker and shimmer from the SPI bus noise. I thought it was cool, but I suspect it was also why the external flash on my F411 board just died. By pumping data into PA7, I was also hammering the flash chip’s DI line

Canaduino STM32 boards with MicroPython

Volker Forster at Universal Solder was kind enough to send me a couple of these boards for free when I asked about availability. By way of thanks, I’m writing this article about what’s neat about these micro-controller boards.

always neat packaging from Universal Solder

Can I just say how nicely packaged Universal Solder’s own or customized products are? They want it to get to you, and they want it to work.

I’d previously played around with Blue Pill and Black Pill boards with limited success. Yes, they’re cheap and powerful, but getting the toolchain to work reliably was so much work. So when I read about the WeAct STM32F411CEU6 board on the MicroPython forum, I knew they’d be a much better bet.

Canaduino Black Pill Carrier Board with STM32F411 (and battery) installed

Volker sent me two different things:

Let’s start with the STM32 Screw Terminal Adapter:

Canaduino Black Pill Carrier Board (front)

It’s a neat, solid board built on a black 1.6 mm thick PCB. Apart from the obvious screw terminals — essential for long-term industrial installations — it adds three handy features:

  • a real-time clock battery. If you’re using a micro-controller for data logging, an RTC battery helps you keep timestamps accurate even if the device loses power.
  • mounting holes! This may seem a small thing, but if you can mount your micro-controller solidly, your project will look much more professional and last longer too.
  • A 6–30 V DC regulator. Connect this voltage between Vin and GND and the regulator will keep the board happy. From the helpful graph on the back of the board, it doesn’t look as if things start getting efficient until around 12 V, but it’s really nice to have a choice.
Canaduino Black Pill Carrier Board (back)

I made a little slip-case for this board so it wouldn’t short out on the workbench. The project is here: Canaduino STM32 Screw Terminal board tray and you can download a snapshot here:

The boards themselves are pretty neat:

two STM32F411 Black Pill boards from Canaduino

Gone are the lumpy pin headers of the earlier Blue and Black Pill boards, replaced by tactile switches. The iffy micro USB connectors are replaced by much more solid USB C connectors. According to STM32-base, the STM32F411 has:

  • 100 MHz ARM Cortex-M4 core. This brings fast (single-precision) floating point so you don’t have to fret over integer maths
  • 512 K Flash, 128 K RAM. MicroPython runs in this, but more flash is always helpful
  • Lots of digital and analogue I/O, including a 12-bit ADC
  • A user LED and user input switch.

About the only advanced features it’s missing are a true RNG, a DAC for analogue outputs, and WiFi. But on top of all this, Volker added:

the all-important 128 Mbit flash chip (and capacitor) fitted by Universal Solder

128 Mbit of Flash! This gives the board roughly 16 MB of storage that, when used with MicroPython, appears as a small USB drive for your programs and data. I found I was able to read the ADC more than 22,000 times/second under MicroPython, so who needs slow-to-deploy compiled code?

STM32F411 board pinout
board pinout from STM32F4x1 MiniF4 / WeAct Studio 微行工作室 出品.
Avoid A4-A7 if you’re using a flash chip.

Building and Installing MicroPython

This is surprisingly easy. You’ll need to install the gcc-arm-none-eabi compiler set before you start, but following the instructions at mcauser/WEACT_F411CEU6: MicroPython board definition for the WeAct STM32F411CEU6 board will get you there.

I had to run make a couple of times before it would build, but it built and installed quickly. This board doesn’t take UF2 image files that other boards use, so the installation is a little more complicated than other. But it works!

Once flashed, you should have a USB device with two important MicroPython files on it: and is best left alone, but can be used for your program. I’m going into more details in a later article, but how about replacing the program with the fanciest version if Blink you ever saw:

# -- fancy Blink (scruss, 2020-05)

from pyb import LED
from machine import Timer
tim = Timer(-1)
tim.init(period=1000, mode=Timer.PERIODIC,
         callback=lambda t: LED(1).toggle())

None of that blocking delay() nonsense: we’re using a periodic timer to toggle the user LED every second!

debugging the mystery huge potentiometer using two ADC channels

I’m really impressed with the Universal Solder-modified board as an experimentation/discovery platform. MicroPython makes development and testing really quick and easy.

[and about the mystery huge potentiometer: it’s a Computer Instruments Corporation Model 206-IG multi-turn, multi-track potentiometer I picked up from the free table at a nerd event. I think it’s a 1950s (so Servo-control/Cybernetics age) analogue equivalent of a shaft encoder, looking at the patent. Best I can tell is that each pot (there are two, stacked, with precision bearings) appears to have two 120° 10k ohm sweep tracks offset 90° to one another. The four wipers are labelled -COS, -SIN, +COS and +SIN. If anyone knows more about the thing, let me know!]

MicroPython on the terrible old ESP8266-12 Development Board

… + 1 + 1 + 1 …

I just found my first ESP8266 dev board. This was from way back before Arduino support, and long before MicroPython

esp8266-dev-boards from ESP8266 Support WIKI

It’s not really in a useful form factor, but it’s got some sensors and outputs:

  • an LDR on the ADC channel
  • RGB LED for PWM on pins 15, 12 & 13
  • red LEDs pins 16, 14, 5, 4, 0, 2 with inverted logic: set them low to light them.

My board can’t quite be the earliest of the early, as it has 1 MB of flash. This is enough to install MicroPython, so I wrote a tiny test program for the outputs:

  • run a binary counter every second on the six red LEDs;
  • cycle through a colour wheel on the RGB LED while this is happening.

Here’s the code:

# esp8266 old explorer board
# see

from time import sleep
from machine import Pin, PWM
# LEDs are 16, 14, 5, 4, 0, 2 - L to R
# inverted logic: 1 = off
leds = [Pin(2, Pin.OUT, value=1), Pin(0, Pin.OUT, value=1), Pin(4, Pin.OUT, value=1), Pin(
    5, Pin.OUT, value=1), Pin(14, Pin.OUT, value=1), Pin(16, Pin.OUT, value=1)]

# RGB for PWM on [15, 12, 13]
rgb = (PWM(Pin(15)), PWM(Pin(12)), PWM(Pin(13)))
# LDR on ADC

def cos_wheel(pos):
    # Input a value 0 to 255 to get a colour value.
    # scruss (Stewart Russell) - 2019-03 - CC-BY-SA
    from math import cos, pi
    if pos < 0:
        return (0, 0, 0)
    pos %= 256
    pos /= 255.0
    return (int(255 * (1 + cos(pos * 2 * pi)) / 2),
            int(255 * (1 + cos((pos - 1 / 3.0) * 2 * pi)) / 2),
            int(255 * (1 + cos((pos - 2 / 3.0) * 2 * pi)) / 2))

i = 1
while True:
    i = i + 1
    i = i % 64
    w = cos_wheel(4 * i)
    for j in range(3):
        rgb[j].duty(4 * w[j])

    for k in range(6):
        if i & (1 << k):

super-special serial port standards

The PC I put together a few years ago (well, Scott Sullivan told me which bits to get, I bought them and assembled it) is still working really well. It was quite spiffy in its day — i7-4790K, 32 GB DDR3, Asus H97M-E — and is quite fast enough for me.

One thing, though, has never worked. The hardware serial port (the old kind, not the USB kind) refused to do anything. Only in the last day or so did I work out why and managed to fix it.

PC serial ports for roughly the last 25 years connected to the motherboard like this:

motherboard pin 1 → RS232 pin 1; motherboard pin 2 → RS232 pin 6; motherboard pin 3 → RS232 pin 2; motherboard pin 4 → RS232 pin 7; motherboard pin 5 → RS232 pin 3; 
motherboard pin 6 → RS232 pin 8; 
motherboard pin 7 → RS232 pin 4; motherboard pin 8 → RS232 pin 9; motherboard pin 9 → RS232 pin 5; motherboard pin 10 not connected
ZF SystemCard – Data Book (1998)

This rather strange mapping makes sense as soon as you see an IDC ribbon-cable DB-9 connector:

serial cable for the SBC6120-RBC, unhelpfully the wrong way up

Going along the cable from left to right (reversed in the photo above), we have:

    1 2 3 4 5 6 7 8 9

    1   2   3   4   5
      6   7   8   9

This was good enough for everyone except ASUS, who decided that they needed their own way of arranging cables. Because of course they would:

ASUS wiring: motherboard pin 1 → RS232 pin 1; motherboard pin 2 → RS232 pin 2; motherboard pin 3 → RS232 pin 3; etc.
Oh ASUS, how could you?

With a bit of resoldering, I’ve got a working serial port. You can never have too many.