My Radio Shack clock has an 8-minute snooze time.
See JillGat about why my links don’t work. They worked before she messed with them. Sheese!!! You’d figure the moderators around here would know how to make links work! If you just simply put the full URL in the form as is, it works just fine – no UBB, no HTML, no butchered-up versions of them. . .like this (picture of JillGat 
):
http://www.tsoft.net/~raych/Devil.jpg
Ray
I measured the snooze interval on my two clock radios, one from Sony and one from GE, both with LED digits (if that matters). Both snooze between eight and nine minutes, coming back on when the minutes digit changes.
If I hit the snooze button at 59 seconds past the minute, then I get an 8:01 snooze. Hit it at 1 second past the minute, and snooze for 8:59.
My guess is that this is the common standard, but most observers would see a nine-minute difference in the displayed time because the clocks don’t display the seconds - if I hit snooze at 7:15:59 and it comes back on at 7:24:00, someone would think it was nine minutes (24 minus 15) although it was actually barely eight.
This would invalidate the hypothesis that it goes on a 15-bit binary counter of 60 Hz power line cycles, because that is six seconds MORE than nine minutes.
It seems this topic is being argued without enough supporting data, so I thought we could use some more.
[[See JillGat about why my links don’t work. They worked before she messed with them. Sheese!!!]]
I didn’t mess with em at all. I simply removed the text under them.
Jill
Sheese?
Having been awakened by the same LED digital display clock radio for the past 20 years, I have more or less intuitively understood the logic of the 9-minute snooze cycle. Considering the basic simplicity of the microprocessor in the average cheap clock radio and its ability to handle events, this arrangement makes sense.
On my alarm clock (and I’m sure it’s not unique in this regard) I can view the total snooze time left. When the alarm goes off, the timer reads “00” (because there is no 6 in the tens place). The moment the snooze button is pressed, the timer jumps to 59. It will then count down each minute until it reaches 50 and sound the alarm again. It immediately goes to 49 when the snooze button is pressed. Thus, the 9 minute cycle.
Pressing the snooze ALWAYS takes it to the closest minute ending in 9. (Actually, the closest minute ending in 0, from which it then subtracts 1.) This holds true regardless of whether the alarm is actively sounding at the moment. If the timer is at 36 minutes and I press the snooze, the timer goes back up to 39 minutes. If I press the snooze when the timer is at 35 minutes, it drops down to 29.
Therefore, nine minutes is the maximum snooze time available if you’re using a simple microprocessor that’s reacting to events in a base 10 counting system (unless you wanted to go to a 60 minute snooze cycle). Pretty simple really if you think about it as a very basic computer program… if the counter time ends in 0, trigger the alarm. When the snooze button is pressed, go to the nearest 10 and subtract one. Nine minutes.
My old “flip-number” type clock had 9 minute snooze alarm, but my 8 year old LCD clock has a snooze alarm that you can set for anytime you want, from 1 minute (one lousy minute!) up to 9 minutes. Set mine at 7 minutes so I can hit it twice and get in 14 more minutes before I feel guilty enough to get up.
My old “flip-number” type clock had 9 minute snooze alarm, but my 8 year old LED clock has a snooze alarm that you can set for anytime you want, from 1 minute (one lousy minute!) up to 9 minutes. Set mine at 7 minutes so I can hit it twice and get in 14 more minutes before I feel guilty enough to get up.
“If people were meant to pop out of bed in the morning, we’d all sleep in toasters.”
-Garfield
No, JillGat, the URLs, as I had them originally in my post, were simply that, URLs. . .with nothing else related to them except what the MB software does to make them links. The result of your change to them shows “URL=”, which, of course, won’t work, because it is none of:
- A simple URL,
- A UBB URL link, or
- An HTML URL link.
I don’t know what you did to cause the defective result, but click on the below URL to verify that it works as is:
http://www.straightdope.com
. Then put this URL in a test post and do whatever you did to my post before and see that it changes to something that doesn’t link.
Ray
I do not understand why this is such a mystery. Why could Cecil not get a straight answer from the engineers at clock manufacturers? I used to have a DIGITAL 10-minute snooze, alarm clock. I used it all through college and for several years thereafter.
When it died I simply went to the store and bought another one that LOOKED exactly like the old one. Imagine my surprise on that first morning to find a 9-minute snooze? I was furious! No mention on the box. Up until THAT point, I had assumed that all clocks came with 10 minute snoozes.
My point is this: I know from experience that GE made a clock with a 10-minute snooze feature. Then they changed it to 9-minutes several years later.
There can be no technical limitation or reason for this. It had to be a choice that SOMEBODY made for some supposed logical reason. I just want to know what that reason WAS???
Cecil, I wish you could actually get tot the bottom of this.
There are those of us who don’t need the alarm to go off. About 95% of the time my wife and I wake up naturally, a minute or so before the alarm goes off. However, our clock and all three of the kids clocks are at 9 minutes for the snooze cycle.
Cecil’s answer # 8 suggested that 9 min was the longest interval attainable if using the “ones” column to advance the snooze. This is actually the correct answer. Cecil discounted this theory by suggesting that they could easily use the “tens” column and advance the snooze by one. The problem there is that if the alarm is originally set for 6:48, the first snooze would occur after only 2 minutes, when the “tens” column changed from 4 to 5. Nine minutes is the longest interval that can be used if the engineers want it to be a consistent interval regardless of the original time of alarm.
The reason a snooze alarm is 9 minutes is because it’s a snooze alarm – the person looking at the clock is going to be half asleep! If the interval is 10 minutes, the groggy person is going to be a lot less likely to be able to tell if it’s 7:10, 7:20, 7:30 or 8:40 (and now you’re really late!) The 9 minute interval makes each bleary glance at the clock register a definitely different time (hey, idiot, time is advancing! Get yourself out of bed!)
At least, that’s the way it works with my snooze alarm.
I agree with Cbaddest - the simplest and therefore most likely explanation is that a snooze interval between eight and nine minutes is the longest possible if the logic of the clock is saying “make the radio come back on when the ones digit reads X.” If the time is 7:14:45 when you hit snooze, it substitutes the digit “3” for “X”, so as soon as the ones digit reads 3, the radio will come back on (at 7:23:00). This logic would be fairly easy to implement, and would use fewer bits of memory (four) than keeping track of a number of minutes and seconds counting down a separate snooze timer. Recall that this design came from a time when real estate on a chip was much more expensive than it is today.
This will make for an interval between eight and nine minutes, depending on the number of seconds past the minute when snooze was hit. In this example the snooze would be for eight minutes, fifteen seconds. This is exactly how my two clock radios behave.
Someone stated that a GE engineer should know why their clock model changed from a ten minute interval to a nine minute one, but maybe not. Maybe they just started buying the clock logic chip from a different vendor, which implements the snooze with what seems to have become the industry standard.
So, basically, I think Cecil blew this answer. He didn’t have hard data supporting any of the possible explanations, and discarded the simplest because of his misunderstanding. (Count the tens column? Huh?)
-Curtis
I have to disagree on the comment that no digital clock uses line frequency for keeping time. I have a 20-year digital alarm clock that defintiely uses line frequency. I know because it has a 50Hz/60Hz selector switch on the bottom, and I can make it go slower by flipping the switch. It uses LED (or some other light emitting device) display and has no moving parts - no synchronous motors inside.
As I understand it, using the line frequency is a very cost-effective and accurate method of keeping time. A 555 timer is out of the question - it relies on a current source (resistor) and capacitor, neither of which are very stable. You need to use a quartz oscillator to match (exceed?) the line frequency accuracy.
I did get to the bottom of it, you idiot. If, having heard the answer, people want to continue inventing chuckleheaded theories, that’s their business.
I agree with Nanobyte, Cecil just blindly assumed all clocks have a snooze of 9 minutes. This just isn’t so. I have one
with an 8 minute interval. But I do notice
that 9 minutes seems to generally be the max
for snooze.
So, how about THIS for a theory: Clocks tend
to have a lot of BCD counters in them. This means that in 4 binary bits, you count up to
9, then reset to zero and bump up the next
significant digit by one. Hence, a 1 digit
BCD counter can count up to 9 max, instead of
15 like a normal 4 bit counter. This means
that in order to count off more than 9 minutes you’d need 2 BCD counters. To cheapen things, they use just 1.
In response to JimB. Any mechanical clock that is plugged into an electrical outlet counts cycles.
Regarding #6, the Engineers comment that “clocks don’t count that way”, he means they
don’t count in binary they count in BCD.
(Binary Coded Decimal)
However, the designers could certainly throw in a binary counter if they wanted to.
I stand by my theory that snooze is 9 minutes max because they use a 1 digit BCD counter for it. If they want to use a binary counter they would be able to go up to 15 minutes, and indeed a few select manufacturers might.
Cheap digital clocks don’t have CPU’s, it’s all combinational logic and state machines.
Yes, I’m an Engineer. :’)
The following post by Tralfaz (12-03-1999 09:20 PM) went totally unappreciated:
I checked one of my own clocks again (the other has a defective ‘set alarm’ button and can’t be tested) and – lo and behold! – it has the same feature. I have to press the button for the 59-minute sleep timer (which also appears to be an essential feature for clocks of this type) to view the remaining time left, so I assume it’s really the same timer used for two different purposes. This also explains why the alarm is usually restricted to about an hour (give or take a few minutes depending on when you press the snooze button, as Tralfaz explained), which has not yet been discussed in this thread at all.
The pieces finally seem to fit together. There is not one-digit counter, as most of us thought. There is actually a (multi-purpose) two-digit counter with a trigger on the final digit. The circuitry must be very simple. The 9-minute snooze is still a design choice, but, as Tralfaz says, the simplest one besides a one-hour snooze.
Cecil, you had not gotten to the bottom of this! As to why only 9-minute snooze was discussed in the column, Hyxtryx: It is still the dominant feature for (cheap) clocks. No one denies that other models have been built.
this is exactly what I wanted to say
(20year experience with 9min snooze)