Jameco Digital Clock Kit turns off then back on later

[CoopedUp]

I got this Clock kit already assembled and it works great when I plug it into my variable transformer but when I plug it into my 12vac transformer it works about 70% of the time. It keeps track of the time when it turns off but you cant read it when it turns off. Anyone have experience with these or have an idea of what could be wrong with it? Thanks

[floobydust]

I recognize that as Jameco JE725 105507  2-1/2" high 6-digit clock kit. I built mine 25 years ago, did a bunch of mods and it's still running. It's actually really nice, you can read it 50 ft. away.
The clock has no dimmer or brightness control, it's pretty simple circuit.
The power system is a mains 12VAC 50/60Hz 500mA-1A wall adapter, which is then rectified and giving around 15VDC to the MM5314 clock IC. As a test, you could power the clock from around 15VDC and it should light up all 6 digits- although the time will not advance.

Your AC adapter I would confirm it outputs around 12VAC. Just a bit confused where the variac is.
If it doesn't want to run from AC and the voltage is reasonable (not too low), I would suspect the filter capacitors C3, C4 axial 1,000uF 16V may have gone bad. That can make the display flicker or just count up one digit with the set pushbuttons. The caps are too fat and hit the display window, so the kit builder might have done something different with them. They need to be small so are only 16V rated which is close to what they run at.

[CoopedUp]

I checked the transformer and it is outputting 14v without load and just over 13v with load. I also checked with 14vdc and the lights turned on and stayed on. I will look into swapping the caps, Idk when this thing was built or even made but I do know they don't make them anymore. Thanks for the tip

[floobydust]

I have this schematic for the JE725.

[CoopedUp]

I finally got around to swapping out the capacitors and it still turns off and on randomly

[greenpossum]

I finally got around to swapping out the capacitors and it still turns off and on randomly

They were due for replacement anyway. Leave a voltmeter connected to the DC output of the rectifier and see what happens when the display goes out.

If no joy there the other thing worth checking is the RC circuit the clock chip uses for generating the multiplexing frequency. Change the capacitor and resistor.

[CoopedUp]

When I unplug it, The time shows back up before going dark. If that means something let me know, I'm going to check the RC circuit for the clock chip

[timenutgoblin]

If I was working on this clock, I would check diodes CR1 - CR4 (bridge rectifier) and I would check their solder joints. Personally, I would consider replacing the diodes (one by one). Interesting that the clock maintains time, but the display is intermittent. Perhaps U2 has solder joint problems.

[CoopedUp]

well, I swapped out the capacitor in the RC circuit and broke it in the process so I went back to the old capacitor and broke a trace in the process so I fixed that and low and behold I fixed it somehow... still not entirely sure what was wrong or how it kept its time but I have yet to observe it turning off. Thanks for the help, if there is any more problems I will let you guys know

[greenpossum]

The RC circuit is for generating the multiplexing of the digits. The time is counted using the mains frequency injected via a different pin.

[CoopedUp]

Well that explanation of the RC circuits use makes sense of what was wrong, the multiplexing was broken slightly

[floobydust]

There's not much left that would make the entire display go blank intermittently.
Try a light tap here and there to see if it's a bad solder joint. The IC's are sensitive to static electricity so don't go for a walk and then touch the PCB.
I would ensure the MM5314 clock IC's /output enable pin 1 is grounded (i.e. connected to pin 2 VDD=GND). If that pin floated then it could cut in and out. You can check with an ohmmeter. U2 should also have power on its pin 1 +15VDC, or ohmmeter pin 1 to the new caps. Old IC sockets can oxidize too.
Note the Jameco JE725 schematic is missing pins 1,2 and 10,11 (12/24hr, 50/60Hz) jumper.

If the clock IC's multiplex oscillator stops (C2/R9), the display usually stays stuck with only one digit lit, it stops scanning but doesn't go completely dark as I recall. I'm not 100% sure though as the IC has a 100Hz min. mux frequency, like it's synchronous.

The clock IC's max is 20VDC and U2 CD4049 is 18VDC, so a power line surge might have happened which can damage an IC. It's kind of a last guess. Some 12VAC adapters have high output, especially at lowest load 1:11:11 for fewest LEDs lit compared to 20:08:08 for max. current drain.
MM5314 can still be bought from scalpers, but it is expensive.

edit: I just tried stopping the mux oscillator by grounding pin 23 through a 10k. The display stops scanning and lights up only one digit, but the time keeps advancing. Could not make the display go blank that way.

[greenpossum]

as the IC has a 100Hz min. mux frequency, like it's synchronous.

The mux frequency is more like 1kHz. 100Hz would be too slow and you would see flicker as it would be 25Hz per digit. You can verify this, as the RC time constant is about 1ms. At that frequency it would not be synchronous with the timebase nor would it need to be anyway as the least significant digit is is seconds, far longer than the mux period.

It is possible they put in an anti-stuck feature in later chips to prevent burnout of one digit if the oscillator fails.

[floobydust]

My question is whether the clock IC is all combinational logic or a synchronous state machine inside. So the mux could be the master logic clock, explains why it can't go to DC and has a 100Hz low limit.

I have MM5314 from 1972 which behaves differently than the late 1970's and 1980's parts.
An old 1972 gold pins part will boot up with a few blank digits which fixes after doing a slow or fast set. It does not debounce the buttons at all, so pressing "hold" causes the seconds to jump ahead, and it has inter-digit ghosting at fast mux frequencies.
The newer IC's boot to the same time usually, no switch bounce issues and are much better behaved. I think there are revisions to the silicon.

[greenpossum]

My question is whether the clock IC is all combinational logic or a synchronous state machine inside. So the mux could be the master logic clock, explains why it can't go to DC and has a 100Hz low limit.

As The Saint used to say, your guess is as good as mine.