Binary Clock Version 3 (or 14 Years of Binary Clock Building)
This project is so much fun and makes me smile each time I'm working on it. It's related to a lot of memories and feels a bit like timetraveling. Binary Clocks have some sort of history in my life. This is the third Binary Clock I've build in my life. Many years have passed between each one (each time about seven I just noticed) and looking through my project pictures tells a lot of stories and stirs up memories. Not only related to this specific sort of project but also about my life back then in general. What do they say? It's amazing how fast the boring now becomes the awesome old time. It also reminds me of HOW long I know good old Phelan by now. He also inspired me to build a binary Clock first hand. When thinking about it I come to the conclusion that we both inspire each other a lot. Which is very grateful for.
2000 - First version
I had to laugh so hard when realising that the reason for building my current Binary Clock is the same like back then in around 2000 when I was building the first one: Damn! I need a clock in my room! Back then it was my room in the house where I grew up. Now, this one is for my living room. In 2000 I told Phelan about and he suggested building a Binary Clock. And so I did. But I didn't keep it. Phelan was absolutely fascinated by it that I just gave it to him as a Christmas present some months later. And as a result: lacked of a clock in my room some more months to come.
On the pictures below you can see the prototype and the final first version. It puts some things into spotlight. Primary what I've learned from building those projects. For a start: how I'd build my boards back then. Without any CAD tool, routed and transfered by hand entirely. Just look at all those holes aslant and ugly. It worked but was a pain. Especially creating the layout on a sheet of paper took a lot of time. It also required a copy machine to mirror the layout so it could be transfer properly to the PCB. That also explains why the all components are mounted like SMD parts: I forgot to mirror the layout. But - eh, screw it - it was just a prototype.
After that I got myself EAGLE from CADSoft my mother supplied me with an (expensive) professional license for the Pro version. It took me a whole weekend to study the user manual and figure out how each function of the software works the right way. Up to the present day it's still my CAD tool of choice for designing PCBs. My first project that got entirely designed with EAGLE was the final Version of my first Binary Clock. It was so much fun to finally have a clean layout and lined up holes for all the components on the board. EAGLE made things so much easier. Open, change, save, print. Done. I could easily change the layout or re-locate components. Also routing was way more easy now. And - hell, yes - I was able to build double sided layouts!
The first version was everything but perfect and I learned a lot from the prototype and the not-so-final version:
From the final version: Ground planes are so important when working with digital and HF circuits! The tracks on the prototype were short enough for signals to cause significant inductions. But on the large version (larger than a A4 sized sheet of paper) the tracks carrying the clock signal were prone to pick up noises. When switching off the light in the room the Minutes and Seconds jumped to random values. Hello inductive coupling. Adding pull-up resistors fixed the problem on a way I could live with it. For a while. Still, every now and then a large signal caused the Minutes and Seconds to change. The clock went all postal when I started to work on HF Transceivers or Switch Mode Power Supplies - or everything else that created strong (electro-) magnetic fields. I eventually fixed it by adding a sheet of grounded aluminium foil between insulating self-adhesive tape to the backside of the panel. It shielded everything. Almost. I also learned that: it's important to run burn-in and long-time tests. Just because something works on the bench doesn't mean it also works in the field.
From the final version: 20PPM frequency shift/tolerance doesn't sound like its much but had a visible effect. To name it: the clock ran ahead or fell back about several Seconds just after hours of operation. I read literature on "frequency shift compensation" and decided to add a variable capacitor to the clock source unit if I ever happen to build another clock. Which I did.
From the prototype version: Reading the clock at night becomes a big problem. Often it was not possible to tell which LED was which and therefore it's was not possible to calculate the proper time. Because of that I added a small additional resistor to each LED that feed a bypass-current to it, making it glow a bit in the dark by default. After that it was possible to tell which LED was on and which not during night from the difference of the brightness alone.
From the prototype version: It's important to hold and reset the clock source when entering the programming mode. Otherwise it will happen that one of thw (H:M:S) block will jump one number up when leaving the programming mode due to a present signal transition.
From the final version: God damnit! Never EVER clean acrylic plastic with alcohol. I have no idea about the reaction that takes place but it somehow causes the plastic to grow cracks to the point where it won't just look ugly but become very fragile and easy to break. I had to build the Front panel once more. Which was expensive, time consuming and frustrating. Lesson learned.
Photos of the first prototype...
... and final version
2007 - Second version
Version two was build around 2007. Lions_ had asked me to build one for him. It had to fit into a chassis for EU-sized laboratory cards. There we go.
Setting the clock has become more easy. One button (start programming) locks the gate for the 1Hz Clock source and clears the divider. Another button supplies a 2Hz clock to the Seconds as long as it is held down. A third button (locks out the 2Hz clock signal) and supplies a 100Hz Signal to the Seconds, allowing to fast forward and set the Minutes an Hours. It takes some attempts until one gets used to it but then it works like a charm. After all buttons are released the 1Hz clock signal is enabled again after a delay of three Seconds. If an exact time is required the clock must be set three Seconds ahead of time so it's in sync when the button is released.
Photos of the second version
2014 - Third version
Now I'm building my third Binary clock. Again, the reason is, again: good damnit! I need a clock in my room! This time in the living room. I found my old "DCF 77 controlled 7 Segment Clock" project in one of the project boxes. But the motivation to finish it wasn't that huge. Several year ago I got a fancy DCF77 receiver (heterodyne receiver with two IF stages) that motivated me. I'm going to finish that another time. Let see of long it'll take. :)
Anyway. The third version is controlled by a small Mikrocontroller. I settled on using a ATTINY26. The LED panel operates in Matrix mode with a refesh rate of 1,2Khz. Flicker free. This also allows me to dim the Clock by using PWM. During summer with a lot of light in the room the clock needs to be bright. But during winter or in the evening hours I want to dim it. Or turn the panel off entirely (i.e. when watching a movie).
I'm wondering what kind of Binary Clock I'll build in 2021 . Seven years from now on. And also: How will my boards look then?
Oh my. Look at the clock. Now I spend another two hours on writing small documentation. All I wanted to do was to post some photos and say "Look, that's the Binary Clock I'm currently building". :)
Photos of the current version
2000 - First version
I had to laugh so hard when realising that the reason for building my current Binary Clock is the same like back then in around 2000 when I was building the first one: Damn! I need a clock in my room! Back then it was my room in the house where I grew up. Now, this one is for my living room. In 2000 I told Phelan about and he suggested building a Binary Clock. And so I did. But I didn't keep it. Phelan was absolutely fascinated by it that I just gave it to him as a Christmas present some months later. And as a result: lacked of a clock in my room some more months to come.
On the pictures below you can see the prototype and the final first version. It puts some things into spotlight. Primary what I've learned from building those projects. For a start: how I'd build my boards back then. Without any CAD tool, routed and transfered by hand entirely. Just look at all those holes aslant and ugly. It worked but was a pain. Especially creating the layout on a sheet of paper took a lot of time. It also required a copy machine to mirror the layout so it could be transfer properly to the PCB. That also explains why the all components are mounted like SMD parts: I forgot to mirror the layout. But - eh, screw it - it was just a prototype.
After that I got myself EAGLE from CADSoft my mother supplied me with an (expensive) professional license for the Pro version. It took me a whole weekend to study the user manual and figure out how each function of the software works the right way. Up to the present day it's still my CAD tool of choice for designing PCBs. My first project that got entirely designed with EAGLE was the final Version of my first Binary Clock. It was so much fun to finally have a clean layout and lined up holes for all the components on the board. EAGLE made things so much easier. Open, change, save, print. Done. I could easily change the layout or re-locate components. Also routing was way more easy now. And - hell, yes - I was able to build double sided layouts!
The first version was everything but perfect and I learned a lot from the prototype and the not-so-final version:
- Features
- OFF-Brightness for H:M:S, adjustable
Photos of the first prototype...
... and final version
2007 - Second version
Version two was build around 2007. Lions_ had asked me to build one for him. It had to fit into a chassis for EU-sized laboratory cards. There we go.
Setting the clock has become more easy. One button (start programming) locks the gate for the 1Hz Clock source and clears the divider. Another button supplies a 2Hz clock to the Seconds as long as it is held down. A third button (locks out the 2Hz clock signal) and supplies a 100Hz Signal to the Seconds, allowing to fast forward and set the Minutes an Hours. It takes some attempts until one gets used to it but then it works like a charm. After all buttons are released the 1Hz clock signal is enabled again after a delay of three Seconds. If an exact time is required the clock must be set three Seconds ahead of time so it's in sync when the button is released.
- Features
- ON-Brightness for H:M:S, seperately adjustable
- OFF-Brightness for H:M:S, seperately adjustable
- Adjustable crystal frequency to compensate PPM tolerance drift
- 100Hz Clock output. Helps when compensating the PPM tolerance drift
Photos of the second version
2014 - Third version
Now I'm building my third Binary clock. Again, the reason is, again: good damnit! I need a clock in my room! This time in the living room. I found my old "DCF 77 controlled 7 Segment Clock" project in one of the project boxes. But the motivation to finish it wasn't that huge. Several year ago I got a fancy DCF77 receiver (heterodyne receiver with two IF stages) that motivated me. I'm going to finish that another time. Let see of long it'll take. :)
Anyway. The third version is controlled by a small Mikrocontroller. I settled on using a ATTINY26. The LED panel operates in Matrix mode with a refesh rate of 1,2Khz. Flicker free. This also allows me to dim the Clock by using PWM. During summer with a lot of light in the room the clock needs to be bright. But during winter or in the evening hours I want to dim it. Or turn the panel off entirely (i.e. when watching a movie).
- Features
- Dual Color LEDs. RED = background light, GREEN = active
- OFF-Brightness for H:M:S, manually adjustable
- ON-Brightness for H:M:S, adjustable in 5 steps
- Adjustable crystal frequency to compensate PPM tolerance drift
- 2kHz Clock output. Helps when compensating the PPM tolerance drift
- The menue structure
- BTN 1: PRGM SECs, PGRM MINs, PGRM HRs, PGRM END
- BTN 2: (regular mode): LEDs OFF, LEDs DIM 1 LEDs DIM 2, LEDs DIM 3 LEDs DIM 4
- BTN 2 (program mode): Feeding 1Hz to the selected block (H:M:S) as long as it is held down
I'm wondering what kind of Binary Clock I'll build in 2021 . Seven years from now on. And also: How will my boards look then?
Oh my. Look at the clock. Now I spend another two hours on writing small documentation. All I wanted to do was to post some photos and say "Look, that's the Binary Clock I'm currently building". :)
Photos of the current version
