Here you’ll find a few direct links to important content of this page:

Electronic projects:

Deutsch Lieber auf Deutsch? Hier lang.

Today I learned a cool new trick, made possible by the Item Preprocessing that is available since Zabbix 3.4. This is not a Zabbix blog, but I consider this so useful and so not-intuitive that I just have to write it down.

For the occasional visitor: Zabbix is an open source enterprise class monitoring solution. I use it not only at work, but also at home — not only to monitor the meat on my BBQ. :-)

In my opinion, Zabbix is really great when it comes to anything that can be expressed in numerics. It can also handle string data, but having had minor problems with that I try to avoid strings whenever it is possible. States can be expressed numerically and stored in integer values, using value mapping it is possible to show the states in a meaningful way.

Today I came across a problem that I wasn’t able to implement at first. A web service told me the state of an application, which could be either RUNNING, STOPPING or STOPPED. It’s easy to create a string-item to handle this, and to implement a trigger that notifies if the service is STOPPED. However, in this case I had to create a trigger that tells if the service is in the state STOPPING for more than a certain time, indicating that it might have a problem shutting down. I couldn’t find a solution to this, since the trigger function str("STOPPING",15m) would trigger if at least one of the values during the last 15 minutes was “STOPPING”.

Item Preprocessing to the rescue!

Three steps

Three steps

Zabbix 3.4 brought a feature called Item Preprocessing. A value can be fetched, and before it gets stored it can be dissected or converted in several ways. One of these is, to apply regular expressions to the fetched value, in a “find and replace” kind of way.

It turned out that I needed a fairly complex expression, but in the end I was able to convert the states from the webservice to simple integers. Searching the web, i found something about “conditional replacement”, and using this great online regex tester I was able to come up with this beauty:

Using this, I can convert the string that is extracted from the web service’s output using a JSON path in two further steps:

  • First, attach a “dictionary” to the value: replace the full value (.*) with with itself, followed by the replacement values: \1:STOPPED=0:STOPPING=1:RUNNING=2.
  • Then, replace the regex (STOPPED|STOPPING|RUNNING)(?=.*:\1=(\d)) by the value of the second capturing group \2.

In this way the item can be configured as an unsigned integer, since there are only the numbers 0, 1 or 2 that have to be stored. And I can apply the usual trigger-function-magic to notify if the value is 1 for a certain length of time. Added benefit: I can use the graph view of this item to see if there were any deviations from the wanted “RUNNING” state, when they occured and how long they lasted.

Better ideas?

I’m pretty enthusiastic about this way of processing the values. But I’m also very interested in opinions: is there a better way to deal with this problem? Something obvious that I haven’t seen?

Deutsch Lieber auf Deutsch? Hier lang.

BBQ temperatures in Zabbix

BBQ temperatures in Zabbix

Unfortunately, this project is still work in progress. But I want to tell of the first test run anyways. The electronics didn’t do exactly as it was supposed, but you see where this is going.

About two years ago, I built a prototype of a BBQ thermometer, which I showed here (german only, sorry). It really was just a quick hack, not really intended for repeated usage. I have learned a lot since then, and now I want to build something like that, but this time as a ‘real project’.

Being able to measure core temperatures of multiple pieces of meat, I didn’t have to think about the project’s name for long: Multimeater! :-D

The Hardware

The Hardware

Like some of my last projects, this is based on an ESP8266-module, with software developed in the Arduino IDE. So it already runs WiFi. My firmware is — as in some of my other recent projects — based on the Homie-framework. It sends collected values via MQTT to my broker, from there they are delivered to Zabbix. Values are measured with two high temperature sensors each connected to a MAX6675, and four common meat thermometers on a MCP3208.

Six temperatures? Yes. The plan is, to acquire the inner temperature of the pit on two positions. Under certain circumstances, these temperatures can be too high for the meat sensors. So I take two separate sensors for this, one on each end of the pit (which is a bit lengthy on a smoker, so there really is a relevant bit of a temperature drop in between). And with four meat thermometers I can monitor core temperatures of — you guessed it — up to four pieces of meat.

Zabbix is usually used as a monitoring software for server systems. But it doesn’t just work in enterprise environments, I am collecting some… let’s say unusual values at home. ;-)

In this case, I just used it to see graphs of the temperatures. But if I’m really starting to smoke me a pulled pork over night, I can tell Zabbix to wake me up if temperatures are out of a predefined range.

As mentioned, today’s run was just a test. And of course, several things didn’t work out as they were planned. I had a problem with the high temperature measurements in the pit. Those values turned out to be completely useless. But since I just bought three pieces of meat for today, I was able to use the fourth meat sensor for the pit temperature. Then I misplaced the sensor in my piece of ham, so the graph looks a bit strange. And to top all this off, heavy rain poured down during the test. Currently, there is no roof over the smoker, so that didn’t make it easier, either.

All in all, everything looked quite promising. I keep bringing this forward, and as soon as I really get it to work I’m going to publish it here, of course with circuit and firmware.

Oh, and since I’m sure that I’ll get that question: technical problems aside, I had some awesome pieces of meat from this test. Everybody really enjoyed it. :-D

The circuit as it ran for two years

The circuit as it ran for two years

Normally, it would make much more sense to publish a project when it is done. In this case, I was sure that I had done exactly that. It’s just a few days since I realized that I didn’t. Funny, it’s been exactly three days before I replaced the project by something else. However, even if it’s a bit late, I’ll deliver…

In early 2014, we renovated our living room. One wall got a rocky structure, and I thought that was the perfect place for a grazing light from above. This screamed for LED-strips. Fooling around with my daughters, I mentioned a plan of installing a pink light. The plan was to go with warm-white, but they insisted on a pink light… two against one, I was in defense position. :-)

Theoretically, a simple RGB-strip can be made to give white light. Practically, most times that’s a really ugly tone of white. So I installed a warm-white strip next to the RGB one. The controller that was delivered with the RGB-strip wasn’t able to power this combination, so I needed a RGBW controller. I couldn’t find an affordable one, so I had to build it myself.

That's how it works

That’s how it works

If I remember correctly, that was my first Arduino project. And it was surprisingly easy to do. I built a first prototype with an Arduino Uno, and with the great infrared receiver library by Ken Shirriff I was able to control a RGB and a white LED after just two hours. Using my trusty Logitech Harmony, which also controls the rest of my media system.

The second prototype I built using MOSFETs, to test how I would be able to control the LED-strips with an Arduino. They are powered with 12V, after all.

The final build is pictured on the photo above: an Arduino Pro Mini (the one without USB interface), and a simple PCB that is wired for controlling the strips. There’s an infrared receiver wired to the Arduino. Initially, that was a TSOP31238, but I fried it just before finishing the project by wiring it wrong. Wanting to get this done, I gutted an old DVD player. I don’t have the slightest idea what type of receiver this is, but it works. :-D

That means: it did work. Almost exactly for two years. Till today. Now it’s obsolete — as I said. Sources and documentation go by the name of IRlicht, I published today.

Oh, and if that’s of anybodys concern: until today, the lamp was used almost only in the white mode. The colors — and especially the color changing modes — were used only for testing and for showing how it works. It was used rarely enough that I had to look up the used keys in the source code…

H801 Wifi in case

H801 Wifi in case

I don’t remember how I found this, but a while ago I ordered one of these modules from china. It’s called H801Wifi, and it’s sold for nine Euros. There’s also an app that controls the lamps from a phone, but I didn’t even bother to test that.

Two years ago I built a project that runs by the name of IRlicht (I was sure that I published it at that time — seems that I still have to). It does almost the same: it controls the brightness of LED strips. My own projekt does that on four channels: RGBW. Red, green, blue and white. This device controls RGBWW, meaning that it would be possible to attach for example a warm and a cold white strip, in addition to the RGB one.

My DIY-one is controlled via infrared remote control. This module is driven by an ESP8266, so it works in a WiFi network. I’m fooling around with a firmware for small WiFi-devices for a while, that communicates with the MQTT protocol. This module would be a great platform to use it on. It’s just questionable if it’s possible to get my own firmware to this off-the-shelf-device…

Turns out: yes. It’s possible. :-)

Flash-jumper set, connected to serial

Flash-jumper set, connected to serial

I’m not the first one to try that. Andreas Hölldorfer did it, and he wrote about it. I didn’t expect it to be that simple. Almost disappointing… ;-)

On the PCB — and thanks to great photos in most of the offering shops I knew that ahead of the purchase — is a well labeled serial interface. And right next to that is a connector that literally wants a jumper to be set on it. When the jumper is set, it’s possible to flash a new firmware over the serial interface. Right from the Arduino IDE. I just connected my USB-serial-interface between computer and the module (without connecting external power to it, I don’t know if that would have damaged anything), and I configured the module in the IDE like this:

  • Board: “Generic ESP8266 Module”
  • Flash Size: “1M (64k SPIFFS)
  • Upload Speed: “1152200”

After connecting RX and TX in the right way, I was able to upload my firmware.

I mentioned to have an almost fully working firmware for my use, based on the excellent Homie for ESP8266 framework. ‘Almost’, because till now it just does RGB, not RGBWW. But the only thing I had to adapt for using RGB on this device instead of my usual hardware were the IO-Pins. The “Generic ESP8266 Module” header file doesn’t know any readable names for the pins, so I had to use the numbers. Andreas Hölldorfer already found out the mapping, though it seems that he’s got a different revision of the hardware. This worked for me:

Pin Function
15 Output red
13 Output green
12 Output blue
14 Output white 1
4 Output white 2
1 Internal LED green / Signal
5 Internal LED red / Power

I like that, after initial flashing the OTA-Update (Over The Air) works. This way, I could already close the box again. All further updates will be uploaded over the air. :-D

Most pins are in use

Most pins are in use

The ESP8266 is an interesting chip, I mentioned it here several times (sorry: mostly german). In a short description, it’s a freely programmable microcontroller. Compared to Arduino & Co., it’s a real number cruncher. The ESP is faster, has more memory, and the best thing: it has builtin WiFi. You’ll find more information all over the network, a basic overview is contained in Wikipedia.

I tested several boards when playing with this chip. Today, I want to describe one that goes with the name Witty Cloud.

Basically, it’s a board with an ESP-12-F, a USB connector that delivers power, a little pushbutton, a LDR (Photoresistor) and a RGB-LED. So there’s plenty of hardware to play with. When you buy the module, you receive a stack of two PCBs. The lower one has a second USB connector, which is equipped with a serial converter. So it’s not only used as a power source, but also as a programming and debugging interface. Furthermore, there’s a reset- and a flash-button on the lower board. After programming, you just need the upper board, and you can even send newer firmware versions to it over the air (OTA).

The  lower board is only used for programming

The lower board is only used for programming

Stacked like this, the module costs less than three Euro, you just need a USB cord and a compiler to start programming. I suggest using the Arduino IDE, it’s very easy to use, even for beginners. After installation of the ESP8266 extensions, it’s best to select WeMos D1-Mini in the board manager, this way everything works fine.

Unfortunately, it’s not easy to find proper documentation for the Witty. So i scribbled the picture above, at first to have some kind of reminder for myself. So the pin labeled GPIO13 is connected to the blue channel of the RGB-LED, in the Arduino environment it’s called D7.

Label Pin (Arduino) Purpose
REST Reset
ADC A0 Analog input, connected to LDR
CH_PD Chip Power-Down
GPIO16 D0 GPIO, freely usable
GPIO14 D5 GPIO, freely usable
GPIO12 D6 GPIO, green channel of RGB-LED
GPIO13 D7 GPIO, blue channel of RGB-LED
VCC +5V power
TXD TX Serial interface
RXD RX Serial interface
GPIO5 D1 GPIO, freely usable
GPIO4 D2 GPIO, connected to pushbutton
GPIO0 D3 GPIO, connected to flash-button, not really freely usable
GPIO2 D4 GPIO, connected to blue LED on the ESP-Module
GPIO15 D8 GPIO, red channel of RGB-LED
GND Ground

I would be highly interested in a circuit of the board, and if you have any corrections or suggestions: just let me know.

Mein Fazit: ein echt interessantes Board. Wer mehr GPIO braucht sucht vielleicht lieber nach einem NodeMCU, wer sowieso einen LDR oder eine RGB-LED braucht sollte zugreifen. Ich habe mittlerweile einige davon hier, und eine Firmware mit der ich die Dinger hier im Haus verteilen möchte ist auch fast fertig.

Oh, das Bild habe ich übrigens mit einer Grafik aus diesem Projekt gemacht, das ist die Witty Cloud für Fritzing.

The completed controller

The completed controller

This is a project from 2004. I had it covered extensively on my old homepage, including some more pictures, circuit diagrams and a description on how I analyzed the keyboard controller.

Nowadays, PCs don’t have a PS/2-interface anymore. I’d base a project like this on microcontrollers if I would do it again, and interface the computer via USB. So I leave the details in the past and give just a short description of the project.

At that time, I was interested in MAME, the Multi Arcade Machine Emulator. You can play all the old arcade game classics with that. A PC keyboard is not the ideal interface for thos games, so I started building my own controllers.

Ratsnest

Ratsnest

Originally, the joysticks were sold to fit a Sega Dreamcast console. I threw out the electronics and implanted a PC keyboard controller instead. I had to find out the keyboard matrix layout and interface all the joysticks microswitches to the right contacts on the controller board. Not too complicated.

The second player’s contacts were connected to a second unit with an old serial cable. Additionally, I built a Y-adapter for the PS/2-interface, so I was able to control the PC outside of MAME.

The result surely wasn’t a pretty build, but: it worked. :-D

Nowadays, I don’t have a PC with a PS/2-interface anymore. I already dismantled one of the joysticks and built something else. But that’s a different story…

See also…

What's wrong with this?

What’s wrong with this?

This is another old project, dating back to about 2004. I had it covered on the old Schatenseite, and I want to give a short presentation on the blog to preserve it.

I was given an old Commodore C64 that somebody decided to dispose of. The first version, in germany this case is dubbed ‘Brotkiste’ (bread box). It didn’t turn on anymore, and lacking any personal Commodore history I didn’t hesitate long before taking it apart.

Unfortunately, I didn’t know about microcontrollers at that time, otherwise I would have been able to save the functionality of the keyboard. I wasn’t able to, so parts of the underside of the keyboard had to go, to make space for CPU cooler and RAM. :-(

I inserted a VIA EPIA M10000b, so the C64 ran at 1GHz. The system was completed by 256MB RAM, a 2,5″-20GB-harddisk, a CD-ROM from a notebook and an external power brick.

The C64-PC

The C64-PC

Built like this I could attach the unit to my TV set, mainly for playing some games on it. But to be honest: this was one of the projects that made more fun during the build than afterwards…

Nowadays I would build this in a different way. A Raspberry would look great in the case. Needing much less space it would even be possible to keep the keyboard in a working condition. And running the Retropie distribution you could even have this certain 80’s feeling… :-D

See also…

I mentioned this project in an article (german only) on this blog, it dates back from November 2012. Somehow I never showed it, I don’t really know why.

Those who read this blog for a while know that I have a deep faible for computer keyboards. You don’t get the full picture from my english language articles, but in german you can find many keyboard related articles on this site.

Amongst others, my collection contains two vintage Apple keyboards. There’s a — really not too shabby — Apple Extended Keyboard II, which was built between 1990 and 1994. Many enthusiasts claim that this is the best keyboard ever produced by Apple, partly of course because of it’s mechanical nature with the Alps switches.

Even more original — and in my eyes it’s a typical Apple product — is the Apple Adjustable Keyboard, 1993 vintage. Typical Apple? As I said, in my eyes: it was expensive for its time, and above all the visual appearance of the product is more important than functionality. Even though there are arrow-keys on the main keyboard, I personally wouldn’t want to miss function- and navigation-keys (Home, End, PgUp, PgDown). And if you connect the additional keypad, you really have to have long arms to reach your mouse… :-/

However, I wanted to give both a try. And since I don’t have access to a computer with ADB interface — Apple killed it in 1999 — I needed a converter. You can buy this stuff, but you don’t have to.

On geekhack.org, there’s a japanese developer called hasu, who is tmk on Github. He created and published the TMK Keyboard Firmware Collection. That’s a firmware that can be run on several AVR-based platforms, notably the Teensy 2.0 which is well known in keyboard communities. The firmware is able to ‘talk’ several different keyboard protocols. Of course, USB in one direction, to be connected to any modern computer. In the other direction there are protocols like PS/2, Sun or Apple Desktop Bus. The custom firmware is loaded with features even expensive modern high end boards fail to deliver. It’s possible to program macros, use media keys or control the mouse pointer with the keyboard.

Unfortunately, I didn’t have a Teensy at home. But I had all the parts to build something compatible, and all the tools to etch a fitting PCB. So I fired up KiCad and designed a minimal circuit, from that it wasn’t too difficult to create a PCB layout:

(The KiCad files need a bit of a workover, so I won’t publish them here. But it should be easy to recreate the circuit following the two screenshots.)

ADB USB Converter

ADB USB Converter

The finished converter measures in at 2x5cm, that’s really tiny. But the important thing is: it works!

I could probably use the ADB mouse with an off-the-shelf converter. But honestly: even a hardcore fanboy would prefer a modern pointing device without the ball in it. That is: a while ago I had the opportunity to play a bit with System 7.5 (the Apple operating system from the same era as the two keyboards), and I have to confess that the Adjustable Keyboard did its part for the authentic user experience… :-D

In Star Trek TNG, Captain Picard is shown as the epitomy of wisdom. I just recognized that yesterday, watching the episode Devil’s Due:

Data, from your own experience of performing Ebenezer Scrooge, you’re aware how fear can be a very powerful motivator. […] And in the hands of a con artist, fear can be used to motivate obedience, capitulation, the exploitation of innocent people. And that is what I believe has happened here.

The episode is from 1991, and everybody can decide for themselves if this quote fits into our current reality. For me, TNG always has been the best part of the Star Trek franchise.

A look back

A look back

Just until a few hours ago, the Schatenseite — my main homepage — was running with the Typo3 content management system. I installed it about twelve years ago, and I’m still certain that people can build awesome sites with it. However, Typo3 maintenance takes more engagement than I’m willing to invest. And Typo3 as a CMS was too big for my little page, anyway — which was clear from the start… :-D

I’m running my blog since 2004, and I think most of my content fits better into a blog than into a classic ‘homepage’ format. So old projects are ‘fading out’, instead of me taking them offline. So I moved my blog to the main page. I preserved parts of my old homepage by ordering some articles into the blog. I kept chronoligical order with this, so what I wrote in 2006 will be found in 2006, even if I copied the content in 2015. I put a sticky article onto the front page, so it’s easy to find the main project pages.

Ages ago, I heard a very wise sentence. Unfortunately, I don’t remember who said that:

Important URLs never change.

I tried to honor that with my relaunch. So on one hand, the most important pages of the old Typo3-site should lead to the right articles in this blog. And on the other hand all old blog URLs should lead to the new location, too. So most links and feeds should continue working properly. People with links to my page, especially those who subscribed the feed, should check if they know about the new locations. For the time being, I’ll keep the rewrite rules intact. But who knows what’s coming…

There’s one more thing I introduced to the blog: parts of the old homepage were already translated into — my very broken — english. I want to expand the english part in the future, so from now on this is a bilingual blog. However, not for the whole content. I like how WordPress keeps the page with the Polylang plugin. I don’t have to translate every single article. But rest assured: the main parts will be available in english, too.

I’d really like to hear what you think about the relaunch. Especially of course, if there still are errors to fix. Does the page look OK? Do the feeds still work for you…?