Here's how it works -
NOTE : In the following schematics, TTL devices are shown as 7400 series. These are used as generic 7400 series part numbers. During construction, 74LS, 74ALS or 74HCT should be used because of the low power requirements and availability. (I know most of you know this already.)
The power supply is not shown in the schematics. It can't be seen in the picture either because it is under the digital circuit board. A simple 5 volt regulated supply of 500mA to 1A will work well. I used a 6.3V, 500 mA power transformer, a full wave rectifier and a 220 uf capacitor as a filter. Regulation is done with a LM340-5 - a 7805 would work just as well. The power transformer is a NOS (new old stock) 6.3V filament transformer I acquired from an electronics surplus outlet at a low cost. The transformer was originally meant to power filaments (heaters) in vacuum tubes (thermionic valves). Although this transformer is probably 40 years old, it's 6.3V output works very well for 5 Volt regulators.
The first schematic -
The oscillator is a LM311 comparator along with a 32768Hz tuning fork type clock resonator. The schematic shows a 7pf capacitor in series with the tuning fork. This value will get you close to 32768Hz. However, if you have a 6 digit frequency counter, a 10pf trim cap should be used to adjust the frequency. I chose the LM311 circuit because the 311 is commonly available and inexpensive. There probably are better time base ICs available.
The 32768Hz signal is divided by 16384 to get a 2Hz signal. This is done with a 74HCT4020 counter. The signal is then divided bt 12 with a 7492 and then by 10 with a 7490. This gives a signal of one cycle per minute. The MUX0 and MUX1 outputs from the 74HCT4020 are used to multiplex the seven segement displays.
The circuitry to set the clock is done with the 7414 Schmidt trigger inverters connected to the set buttons. These in turn connect to a cluster of AND and OR gates to select between a 2Hz signal and the 1 minute and 1 hour signals. The 2Hz signals being used to quickly cycle the counters while holding down the set buttons.
The 1 minute signal is then connected to the 7490 decade counter, which counts the minutes from 0 to 9. The D output of the 7490 is then connected to the 7492 which counts from 0 to 5, or 0 to 50 minutes. The counter is reset from it's own D output and the hour stage is clocked from the C output. This may seem incorrect (and may be if I made an error in my schematic) but remember that the 7492 is not a straight binary ripple counter and is a modulo 12 counter.
The second schematic -
The one hour signal is connected to a 7493. This is where the design gets convoluted. The hours digit must count from 1 to 9 (for 1 to 9 o'clock), then 0 to 2 (for 10 to 12 o'clock). The tens of hours digit must count from 0 to 0 (for 1 to 9 o'clock), then 0 to 1 (for 10 to 12 o'clock). The 7493 re-loads to 1 at the 13th count (12) reguardless of the hour. The 7485, a magnitude comparator, selects between the 7493 and a 7490, depending on whether it is greater that 9 o'clock. The 7490 resets on the 3rd count. The 7485 and 74157 selects between the 7493 (0 to 9 o'clock) and the 7490 (10 to 12 o'clock). The 7410 triple input NAND gate is used to decode the 13th hour from the 7493 and the remaining gates in the IC are used as inverters. I hope this makes sense, but this is the best I can do translating digital logic to the English language.
The 74153s are used to multiplex the seven segment displays. Minutes, tens of minutes, hours and tens of hours are selected at a high rate of speed from MUX0 and MUX1 outputs of the 74HCT4020 counter.
The third schematic -
A 7448 is used to drive the common cathode, seven segement displays. The 75156 is used to select the display during multiplexing. The two LEDs are the colon between the hours and minutes.
Note about construction :
I used wire wrap connection techniques with machine pin sockets, not wire wrap sockets. I ran the wiring on the component side of the circuit board which makes the job a lot easier. Since only a few wraps can be put on the relatively short posts of machine pin sockets, the connections have to be soldered. I put wire loops on the board in which to route the thin wire wrap wire through. I used this technique because I have a lot of salvaged machine pin sockets and the machine pin sockets give the board a lower overall profile. Wire wrap sockets would be easier to work with.
The display circuit board with the time setting buttons was constructed separately from the board with the ribbon connector to the main digital circuit board. The reason I did this is that the displays I used did not lend themselves to IC sockets and required some relatively high density soldering.
Send questions and comments to WireHead@GalacticElectronics.com
(c) Jon Qualey, August 2006
Galactic Electronics Home