Emerson Clock Radios
 

For years, my family's go-to were Emerson clock radios/alarm clocks and they always kept good time. The last two I've bought over the last 15 years lose about 3 minutes a month. The worst culprit is the CKD3630 Sound Cube CD Player Clock Radio. :dammit: Why is this and is there a way to fix it?

The old motor driven clocks followed the precisely controlled 60 cycle power. They could not respond to transients on the power line.
Later electronic clocks followed the 60 cycle power (many still do), but were (are) subject to counting occasional transient spikes as additional cycles, and therfore run fast.
You may have a clock that depends on an internal reference, not the 60 Hz power frequency. However, three minutes a month is terrible accuracy for a crystal controlled clock,if that's what it is. (Think of a digital wristwatch for comparison.)

I have a little crystal controlled clock radio (runs from batteries or a wall wart DC supply, so no line frequency reference) with much better accuracy than you are seeing.

.


You should take it apart and see how it gets it's time base, is it from the AC line, or
does it have a on board osc. Open it up & post a pic of the clock part, IC Chip board
etc. you may even find a trim pot of some kind to get it running closer to the
right time..... Find the Clock Chip & get the sec sheet & lets see what makes
it tick.....


.

He's had two Emerson units show the same problem (running slow). Assuming they're synced to the powerline*, and it's 'dirty' with transients, would the spikes necessarily show as additive (positive-going), or could they induce negative-going dropouts?:saywhat: I dunno, just wonderin'.:headscrat

* EDIT. That's assuming both units to be digital.

I suppose it's possible for a circuit to be filtered heavily to avoid spikes and then miss a cycle if there is an occasional reduced amplitude one. I can only say I haven't seen slow running, only fast, on the cheap units I have owned.

In either case it seems to be a basic failing of something that counts pulses rather than something with inertia (a motor) that coasts through transients.

I wonder if they're made by the same place that made the radios in my parents Hondas...The clock on Mom's runs 12 minutes fast between daylight savings time resets, and dad's runs IIRC 4 minutes slow...Any clock that drifts more than 2 minutes annoys me especially when my family members don't want it changed because they somehow like keeping track of 4 different clocks that are all wrong relative to each other...and they wonder why I have no sense of time.

My children (Son & Daughter in law) who live with me keep the bathroom clock set about 3 minutes ahead on purpose, they seem to think it gives them a jump on the morning somehow :dunno:

I still haven't figured out what time has been "saved" by the three minute time zone between the bathroom & the rest of the house, but I guess it works for them :lmao:

I have a new Sony that picks up a minute or two a month.

I can sometimes understand that. If you're bad estimating duration and need to compensate. When I had cable I used to set my work bench clock ~3 minutes fast so when my shows came on I would have extra time to run upstairs and start watching if in the middle of something.

This is more being too dumb to figure out how to change it without the manual, being too lazy to find the manual, getting used to the unacceptable, then during time changes relying on me to fix the hour then wanting to keep the minutes wrong....on one clock and 2-3 minutes off it's not the end of the world, but when it's 4 or more clocks each off by different larger amounts for no meaningful/real reason it starts to get excessively difficult to manage and confusing. On the car clocks I generally cut the drift in half and try to set it by that amount in the opposite of the drift so for the half of the year between adjustments it's at least closer to correct on average than it otherwise would be.

Years ago I had one of these mains-based clock... sometimes it become crazy and doubled the seconds count (it blinked at 1/2 second interval) for a minute or so...
In fact, mais based clock is in reality a pity, since in the reality, the mean frequency in practice is slight > 60Hz (or 50 for the 50Hz countries) in the long run, even with national operators dictating near 60Hz for entire interconnected system (for example, here in Brazil) and power plants need to sync with it perfectly or they are automatically disconnected (of course, due to electrical phasing mismatch/overload etc). Here also is common to a mains based clock to having 1 to 4 minutes/month added.
I have a DIY clock with 6dB passive plus 24dB/oct active filtered clock followed a carefully calibrated squarer to pick the mais freq, and even so, this time difference occurs.
I also have a crystal clock in the car that advanced 5 minutes in a month... :thumbsdn: when I have some time I will sub the internal circuit for one from these home wall clocks that have far more accuracy.
Recently I built this vector CRT clock: http://www.sgitheach.org.uk/scope2.html having a promising RTC; will be very happy if the RTC is precise as described in it's datasheet (about the time issue)...

What is the source of this claim? Power systems (in the U.S., at least) have historically been regulated to produce correct time of day with motor driven clocks, as far as I know. I cannot think of a reason to run fast on average.

Here are some actual measurements from a place in the Netherlands:
https://wwwhome.ewi.utwente.nl/~ptde...20using%20them.

However, the freedom from transients, dropped cycles, etc., has not necessarily improved over time.

Good observation.

Too many coincidences with too many observations led me to believe that (along with some other technicians frinds of mine), even after talking with a colleague that is from a sector near the operators here in the Itaipu DAM (where I work, but at other sector not related to this). He is not a authority in this subject, but he explained at that time the same explanation in this page you mention here, that the ONS here (National Electric System Operator; Operador Nacional do Sistema Elétrico in Brazilian Portuguese) enforces exactly mean 60Hz in long run, but I didn'd see a graph for the long run, and I wondered if in a month the national grid deviated slight above 60Hz mean.
In short run, frequency is subject to some variations due to various demands.
And I talked with him about this same subject of clocks at that time, but he believe that all clocks being bad filtered etc.

Perhaps I need to ask to an actual operator about this subject.

Of course, is not possible to cathegorize my clock being ultra scientific accurate (more professional measuring is needed). And, of course, most other ones have paltry filtering.

That's correct, is a belief, to be yet checked with some more professional work (or to find one already made); maybe the more correct sentence here is: "I strongly believe that in practice this"...

It's weird, My clock is an old Radio Shack digital with NOAA Radio and it keeps perfect time. Her clock loses the three minutes a month and the Emerson it replaced does the same thing

I have an el cheapo Timex watch (quartz based, analog dial) bought in 2007. It's dead-on to the minute between DS resets. Also have a quartz wall clock that loses about 2 minutes between DS resets.