Dieter's CD47 Nixie Clock Project
last update: March-22-2005 (stay tuned!!)

Back to Dieter's Nixie Tube Page

Email to Dieter

It was in May 2003 when I at last get some of the very nice Rodan CD47 Tubes.
You can find the technical data and some nice pictures of that tube here:
CD47/GR-414- Pictures and Technical Data
CD47/GR-411- Pictures and Technical Data

You'll find the matching datasheets in the datasheet archive.
Search for this type and its substitutes there.

A.J. Franzman also have a nice gallery of that tube, which you can find here: A.J.'s Gallery

I decided to build a CD47 Nixie Clock. I thought that it would be interesting for some Nixie enthusiasts to see the building process. So I statred this site, which I will update step by step.

Step1: The Power Supply
For the Power supply I made my thoughts about the features first. They were:

• Quasi- resonant flyback SMPS (Switch mode Power Supply)
• Power: approx. 100 Watts (some might say "Isn't it over- engineered", but you'll see at the end of that project, that it's ok. Moreover, I like over- engineering! ;-)
• Plate Voltage: +170V/0.5A
• Voltage for CPU: +5V/1A
• Voltage for other electronic: +12V/1A (this will be for the radar sensor, the mic- amplifier and the light sensor for example)
• Galvanic seperation
• Primary Voltage: 100VAC - 230VAC / 50Hz - 60Hz
• Cycle by cycle current protected
• Under- voltage lock out
• Latched Over- voltage protection
• Thermal shutdown
• Low- Power output stand- by mode (5V only)
• Off- line switching- regulated
• Switched on/off and regulated by the CPU of the clock.
• Plate voltage adjustable by the CPU in a menu (Nixie Tubes will display the plate voltage!)
• Output voltage accuracy over all conditons: <1% (I know this is crazy, but I like it accurate!)
• High quality line filter
• Stand by mode: Nixies will be switched off when nobody is around- this will keep them alive as long as possible. The sensors for the detection will be a ECM (microphone) and a radar sensor.

I built up everything around a STR-S6707 switching regulator. You can download it's datasheet.
As you can see there is nothing to adjust, but the output voltage 170V. This can be done by a pot for the test- phase, and later by the CPU by PWM- modulation. I started this website arter already made the experiments with the supply, so I unfortunately have no pictures of the experiments.

The Schematic:

After I built the developing circuit on a experimental board, I immediately created the real board (double- sided)
Here you can see
the Layout:

Click on image ot enlarge!

Here you can see the top side of the finished board:

and here the bottom side:

You can see the primary and secondary separation very well. It remembers a SMPS in a TV-set or PC-monitor.

The SMPS- Partlist:
CD47 Nixie Tube Clock SMPS- Partlist (klick to display the Partlist)

SMPS- board- equipment:

First I soldered all lower parts (resistors, diodes, ferrite beads.....)

Here you can see the opto - coupler TLP733:

Then I soldered the capacitors, transistors and other parts.

The line filters:

The heatsink for the v-regulators was an old used one (I don't know where I pulled it out):

Here you can see the transformer which I pulled out of an old PC-monitor and equipped it so, that it works with the switching regulator:

The heatsink for the switching- regulator I took out of an old TV-set (vertical deflection circuit :-):

It also works as a screen for the transformer (switching- pulses!!!):
Peter Stonard @ neonixies asked me to put the eagle files online for download.
Here they are

there's little space at the board.

SMPS- Test- Run:
I made the test- run with a 230V/60W light bulb:

It works like cream.
The output voltage is adjustable from 160 to 180 Volts by the pot (later by the CPU). The Stand- By- Mode also works perfect. When switched to stand by, the 5V- CPU voltage shows no voltage drop, not even any move of the meter. You can short any output, you can leave the output open, you can do what you want. It works.

September-21-2003

Step2: Developement of the ECM- Sonic- Sensor:
The CD47 tubes may be switched on, when a noise occurs (door opened, TV turned on or something like this). I developed a simple sonic- sensor with an ECM which will be connected to the controller (analog input).
I first planned to let it work on the 5V- stand- by- power. But as I use a radar sensor, which needs 12V to work, I planned a small DC-DC (5 to 12V) for the Radar- Sensor and for the Acoustic- Sensor.
The sensor should have the following features:

• adustable gain
• high gain
• output => DC- voltage (range 0 to 5 volts). The voltage depends to the ambient noise- level.
• 12VDC single supply
• simple schematic

According to this features I developed the following schematic:

M1 is a standard 2-pin ECM. It's powered by R65.
IC9 is a 741 OP. It works in single supply mode, while the output is not set to UB/2, but to 7V.
This is for to get only 5V max. at the output.
C1 and C2 are for the AC- coupling.
R83 is the Gain- Pot. (OP is set up to an inverted amplifier, R78 and R83 do the feedback)
R70 is necessary to charge/ decharge C2.
D2 rectifies the pulses.

This schematic may be used for every Nixie- Clock. You don't need a processor. You also can use the output to trigger a longtime- monoflop.

Here you can see how I do my experiments:

This is not the board I'll use in my clock of course- it's only an experimental- board. Not nice, but very quick!

October-13-2003:

Step 3: Developement of the Radar- Sensor- Circuit:
The CD47 tubes may be switched on, when a movement occurs (door opened, people moving around or something like this). This I'll solve with a Radar- sensor which will be connected to the controller (digital input).
The sensor should have the following features:

• adjustable sensitivity (by a pot)
• 6m detecting range.
• 180° Radiation Pattern. (should recognizes movements 180° from side to side)
• Output power: approx 6-8 dBm
• Small dimensions
• output => Digital output (0/5 volts).
• simple schematic

I've choosen a Siemens KMY24 to meet this conditions.
According to it's appliction notes I used the following schematic:

The KMY24 is a radar- sensor which uses the doppler effect. It's simple to use and provides 2 phase- shifted doppler- signals. The signals run through some filters. As you can see, the first OP Stage is a typical differential amplifier. The second stage also a filter, and the third stage is just an impedance-transformer. R82 adjusts the sensitivity. The output provides the 0/5V rectangular signal for the Microcontroller.

This schematic may be used for every Nixie- Clock. You don't need a processor. You also can use the output to trigger a longtime- monoflop.
Here you can see the third experimental board (made in just 20 minutes or so- it also will be repalced as soon as the mainboard is ready, of course):

It works very well, and the max. detecting range is about 7m.
J.C. Wren @ neonixies asked me for the datasheets for the KMY24. Here it is.

October-14-2003:
Step 5: Developement of the Driver- Circuit:
The CD47 can not be driven by 74141 or 7441 nixie drivers, because the operating current is to high. So I had to do it in another way.
The driver circuit should have the following features:

• small dimensions
• direct drive (no multiplexing)
• fast drive possible (for dimming the tubes)
• high efficiency
• simple schematic

I'll take 8 octal D-type transparent latches type 74HC373 and connect a smd- resistor and a high voltage SMD- transistor to every output- so I get 64 outputs. 60 I use for the cathodes an 4 for the colon- tubes. The colon- tubes are not developed, yet. I made a quick 'rats-nest' experiment, which worked excellent:

So I came to the following simple schematic:

I will get one 8 bid data and one 8 bid adress- bus... easy to drive with every controller.
When everything would be as easy as this........ ;-)))))

May-30-2004:
Step 6: Developement of the Boards:

After a longer rest I finally found a little time to develope the Mainboard and the connection- board.
I decided to put the microphone and the ECM- amplifier onto the connection- board.
I also decided to put a RS232 interface in the clock, so I can use a DFC77 module or a GPS- mouse or any other time- module or equipment for the clock.

Here you see the finished schematic of the mainboard.CLICK ON THE IMAGES TO ENLARGE!

Here you se the finished layout of the mainboard.
It's large. It measures: 600x92 mm!

Here you see the finished schematic of the connection- board.

Here you se the finished layout of the connection- board.

March-22-2004:
Step 7: The mainboard an the connector board is done!

Finally I had the time to finish the bare boards.....
Here you see the result.
Check the measurements: 600x92 mm!

The Mainboard:

The Peripheral Board:

Next step will be the bulding of the very special Colon- Tubes and the board- solderings.

last update: March-22-2005 (stay tuned!!)

TO BE CONTINUED SOOOOOON!!!!!
Thanks for reading.

Back to Dieter's Nixie Tube Page

Go to Tube-Tester.com

Email to Dieter
Impressum & Datenschutz

Schlüsselwörter: nixie nixieröhre nixieröhren röhre röhren tube tubes buizen indicator numerical vfd vacuum clock nixieclock uhr readout