Digital clocks

Regarding digital clocks…why can they only be set forward? If my power goes off at 11:55PM, I’ve got to cycle through almost the entire time cycle to get it set correctly.

In the same vein, don’t you hate it when someone refers to 12:00AM or 12:00PM?


My (digital) clock radio has 4 buttons that can be used to set both the time and the alarm time: Forward Fast, Forward Slow, Reverse Fast and Reverse Slow. If you want to change the time forward or back a few minutes, you use the appropriate Slow button; if you want to set it forward or back several house, you would use the appropriate Fast button (and then the Slow button if you overshoot).

Is this related to a column?

Your Official Cat Goddess since 10/20/99.

“I get along well with everybody.” --I.M.F.

Reverse buttons are certainly uncommon, at least in cheaper models, but boy, they are a good idea! And it’s not just overshooting while setting the time:

I have often noticed that if I use the buttons carelessly, adjusting the alarm time will also increase the current time by a minute or two. I have no idea how this happens; I’m really not pressing the ‘set time’ button, but it happens quite often if I slip off the ‘set alarm’ button (presumably). I guess I routinely buy really cheap clocks, because I’ve had the experience with several of them. Does anyone know what quirk in the circuitry may be behind this?

At any rate, I hate having to forward the clock by 23 hrs 58 mins!

Switches are probably the most expensive part of a cheap clock. Extra buttons cost more.

John W. Kennedy
“Compact is becoming contract; man only earns and pays.”
– Charles Williams

John W. Kennedy is correct about switches being an expensive part of a cheap clock. I bet many cheap clocks have the circuitry on board for forward/backwards setting, but no switch. This lets the manufacturer only have to build one type of circuit board, and just stick it in different types of housings (some with extra switches). The overall cost is cheaper than if they had to build two different types of boards.

Is this in reference to a Straight Dope column?

Weeell, okay – somebody’s gotta do it. Here’s the link:

Thank you! You really are incredible! Okay, what I’m raving about is that I’ve found the same thing going on - without ever touching the “Time Set” button, if I press the “Hour” button and the “Alarm Set” button wrong, like if I let them go on and off together (?.. it’s hard to explain) the clock will “flicker” and the time (not alarm time) will be set forward substantially.
I am just glad someone else noticed this, since it means either I am not going insane, or I am going insane with company.

Ah, thank you too, Boris! I’m also glad that someone shares my observation. Plus, you’ve described the circumstances more clearly: It has indeed to do with pressing/releasing two buttons at the same time, namely ‘set alarm’ and one of the fast/slow or hour/minute buttons (depending on the model).

But for a moment there I thought you were a bit overentusiastic about my providing the link…

Clocks probably have a problem with multiple buttons being pressed because there isn’t a seperate IO line to the CPU for each button.

A computer keyboard is a grid of switches, the D key could (for instance) set of the A, S, D, F, G, etc, and 3, E, D, C keys. Then the processor on the keyboard picks this up, sees the letter in common is a D, and sends ‘D’ to the computer.

The actual layout of the switch isn’t defined, as long as the keyboard sends the right codes to the computer. But if you find a program that’ll show you keypresses, not the letters they generate, and try pressing combinations of keys, likely you’ll find certain combinations don’t work. For me, shift-alt-LeftArrow doesn’t work, but shift-alt-RightArrow does. Was hell playing Doom. :slight_smile:

Similarly, alarm clocks with eight buttons (Forward Fast, F. Slow, Reverse Fast, R. Slow, Time Set, Alarm Set, Sleep Timer, Snooze) could either have eight wires directly into the CPU that correspond to these buttons, or it could have three, that being the number of digits you need to represent eight in binary.

(Actually, it would have five, that being the next power of two above 8 * 7, which is the number of combinations of keys, assuming you allow for any two keys at once.)

They might try to get by with less, coding the Fast/Slow Forward/Reverse buttons on one set of two, and the other buttons on the other set of two, allowing for the CPU to distinguish one keypress in each bank of four.

But, the problem comes when you press two keys in the same bank at once. This can result in the CPU thinking a third, different button was pressed, or both being lost, or only the first one being registered.

Anyways, there are reasons why accidentally pressing two buttons they didn’t intend to be pressed together could register as a third different button.

A thorough explanation, WhiteNight, but it doesn’t quite answer the question. The phenomenon that Boris and I have described occurs when we press a perfectly ‘legal’ combination of keys, such as ‘set alarm’ and ‘fast forward’. The problem seems to be pressing and/or releasing them simultaneously rather than in a slow sequence of ‘press A – press B – release B – release A’. (At any rate, such a slow and controlled sequence seems to solve the problem in my case.)

My personal guess is a fluctuation in voltage the moment a curcuit is closed (opened) by pressing (releasing) a key. One such fluctuation at a time is safe, but two at once are too much and a momentary short-circuit occurs somewhere. The fact that this leads to an increase in current time may be due to the fact that the keys are interconnected the way you explained.

But that’s just my guess.


I’ve noticed the same thing also, but I think the explanation is probably a lot simpler- what seems simultaneous to us is probably a many nanosecond differential to an electronic circuit. Thus, while we think we’re releasing “Set Alarm” and “Fast Forward” at the same time, to the circuit it sometimes seems we are holding down “Fast Forward” alone for sufficient time for it to respond to that command. Just a WAG, but one that seems very likely to me.

Only that ‘fast forward’ alone is not supposed to do anything. You normally have to press it together with either ‘set time’ or ‘set alarm’ for it to have any effect (at least in my clock).


Oops! Sorry for the misunderstanding. On my clock “Fast Forward” alone does advance the time.

Al, if your power returns at 11:55 pm, perhaps you could wait a few minutes before setting your digital clock.

Also, I am one of those people who refers to 12:00 noon as 12:00pm. I won’t presume to ask that you forgive me for this sin. But maybe an explanation will moderate your hatred somewhat.

My clocks all display time only to the second. This means that noon is displayed as 12:00:00. I try my best to set my clocks so that 11:59:59 is displayed until the exact instant of noon. Then the display switches to 12:00:00. I agree that the exact instant of noon is neither am nor pm. But now let time advance a picosecond and consider my dilemma. My display to read 12:00:00, because, alas, I do not have clocks that display picoseconds. At this point, 12:00:00 pm seems to me to be more correct that 12:00:00 noon. This continues to be true at the picoseconds tick by. For most of that second, 12:00:00pm is correct.

Shouldn’t you people be at work or in school?

WhiteNight- that isn’t usually how keypads are handled. The cheapest way to decode 8 keys is to use six lines. The keys are arranged in a 3x3 grid (you could actually read 9 keys with this arrangement). Because of the way the key multiplexing is done, you need to satisfy X*Y >= N, where N is the number of keys you have to read, and X+Y = number of I/O lines. Of course you want X+Y to be as small as possible to minimize cost. This scheme can read more than one key pressed at the same time. is a datasheet on a chip that does all this, with an explanation of how it works.

I think nebuli’s theory is correct- just because we think we hit two keys simultaneously doesn’t mean the actual switch contacts happened at the same time. Physical switch contacts actually bounce several times over a few milliseconds.

I had a clock where the SET ALARM button wasn’t reliable- sometimes it would become “unpressed” while I setting the alarm time. VERY irritating. I had to press extremely hard to get it to stay pressed, which probably hastened it’s degeneration.


I am at work. This is my… umm… coffee break? (Hi, boss!)

It would be nice if Holg and Boris post the make
and model of their clock. It’d be interesting to see if they both have the same clock!

WhiteKnight is talking about CPU’s. I don’t think cheap clocks have CPU’s. It’s overkill. I’d imagine they use combinational

As to why the problem occurs, I’d vote for either bounce on the contacts is confusing
the circuitry, or the extra pressure required
to press 2 buttons simultaneously is causing
a flex in the circuit board and the “time set” button is getting pressed briefly.