Nixie clock with the CT7001/FCM7001 chip?
Dekatron42
Tomasz Kowalczyk
W dniu środa, 17 kwietnia 2019 20:11:41 UTC+2 użytkownik Dekatron42 napisał:
I've come across a few of the CT7001/FCM7001 clock chips which were designed to be used with LED displays and I've been looking for a Nixie clock design with this chip but I haven't found any yet - does anyone here know about such a circuit diagram?/Martin
Dekatron42
Bill Notfaded
Bill Notfaded
Bill Notfaded
Bill Notfaded
Bill
jf...@my-deja.com
I built three six-digit clocks using that chip and B7971 nixies. I was inspired by the article “Behold the Giant Nixie Clock - using a minimum of new parts” that appeared on page 70 of the July 1976 issue of 73 magazine. This can be found on the internet. That was more than 40 years ago, and all three clocks have continued working with zero nixie failures, one failed power transformer, two failed electrolytic capacitors, and one failed CT7001 chip.
From memory, here are some important differences from the article:
The magazine article used a MM5314 clock chip. The polarities and levels for the digit and segment drives are different in the CT7001 (common cathode versus common anode configurations), but it was straightforward to do the necessary level shifting and polarity inversion. You need to study the data sheets for the two chips. I vaguely recall that my design involved discrete transistors and cascode amplifiers, but that TI later came out with suitable HV drivers (75468).
I did document my circuit, but after 40+ years and several moves, I doubt that I could find it. Since the clocks all continued to work with only easy and obvious repairs, I was never motivated to refer to it.
The MM5314 derives its timing from the mains. Using the circuit in the data sheet, it tends to run fast; I think that is because it is susceptible to counting any noise spikes on the power line (this might be reduced by adding a low-pass filter onto the line reference input). The CT7001 can run slightly fast or slightly slow, depending on the MUX frequency; I believe that it uses a PLL and the free‑running frequency of the display MUX pulls it. If I were to do it over again, I would use a variable resistor to set the MUX rate and adjust it to minimize the drift.
Both of these chips “forget” after a power outage, but the CT07001 will try to keep time if there is a battery backup by using its internal MUX oscillator. As far as I know, there is no shortcut in either chip to set the time directly, so you have to go through the whole procedure of incrementing each digit from its initial state.
One additional feature of the CT7001 is the 365-day calendar. It does not automatically do February 29, but setting it manually makes you good for the next four years.
The nixies in all of my clocks produce acoustic noise and RFI at the MUX rate. Someone suggested it was from my switching power supply, but this explanation was unlikely since I used linear power supplies and the noise was clearly emanating from the tubes.
I was roundly criticized for using the B7971 as “only” seven-segment displays since they were so expensive and capable of so much more, while the critics considered seven-segment displays to be ugly. When I bought mine, they were $1 each (including sockets) and were cheaper than Monsanto MAN-1 displays.
An interesting (to me) alternative would be to find a CT7002 and use a 74141 to drive regular nixies with formed characters. A more ambitious approach would be to use a CT7001 and build a seven-segment-to-BCD decoder (this would be an interesting exercise for the student, and I can think of at least two different approaches).