Annoying Clock on Microwave

[John Tserkezis]

Bob Villa wrote:

Donate it...and buy a new one. I've never had an appliance with a
digital clock that kept poor time. Coffee maker, oven/range, MW, or
VCR.

You mean you had a VCR that *DIDN'T* flash 12:00 all day?

That could only mean one of two things:
You're either a Rocket Scientist, or you had a six-year-old around.
--
If it's not worth doing well, it's not worth doing.

 

[William Sommerwerck]

"John Tserkezis" <jt@techniciansyndrome.org.invalid> wrote in message
news:4db34c2b$0$13392$afc38c87@news.optusnet.com.au...

You mean you had a VCR that *DIDN'T* flash 12:00 all day?

Most Public TV stations broadcast time code. The "later" VHS VCRs searched
for a PBS station and used the time code to set the clock.

 

[John Tserkezis]

William Sommerwerck wrote:

"John Tserkezis" <jt@techniciansyndrome.org.invalid> wrote in message
news:4db34c2b$0$13392$afc38c87@news.optusnet.com.au...

You mean you had a VCR that *DIDN'T* flash 12:00 all day?

Most Public TV stations broadcast time code. The "later" VHS VCRs searched
for a PBS station and used the time code to set the clock.

Sorry, my email addy doesn't show it, but I'm in Australia. I've
verified that timing codes are NOT sent by ANY networks here.
More so, of the few that knew what I was talking about, said they had
heard of it, but no, they did not, and never will support it.

That was about a decade ago while digital terrestrial TV was still a
bit experimental here, but even then, I don't know of any digital PVRs
today that maintain time that way.
There's the odd one out that does NTP if they're 'net connected, but
everything else relies on standard crystal clocked circuits.
Probably a bad example though. At that that level of complexity,
no-one in their right mind would go to the trouble of mains-locking a
clock, when implementing a single-chip Real Time Clock is so much
easier. Especially so when RTC functionality is built into other multi
purpose chips.


I suppose this comes down to the question of is mains locking "good
enough"?
Since I can't get a clear answer from anyone, I'm staying with crystals
or NTP where applicable.
--
OUCH! Got my floppy caught in my PKZipper!

 

[John Tserkezis]

Jeff Liebermann wrote:

... Find the frequency, identify the
CPU, and install a replacement resonator.
http://www.tdk.com/crdata.php

At this price-point, there is no "CPU" to look for. They're all Chip
On Board designs (the black blob on the PCB itself).

The resonator would be a separate component, and I wouldn't think that
an RC oscillator would be good enough to keep mere minutes within a day
anyway.

Besides simply replacing the resonator probably wouldn't show any
significant improvement. You'd need to go all out and use an
upper-market crystal, associated oscillator (with fine tuning) and drive
the clock that way. Would likely need some divider in there too if you
can't match the frequency otherwise.

And now you're starting to lean into the "not worth it" territory.
--
No matter where you go, there you are.

 

[Jeff Liebermann]

On Sun, 24 Apr 2011 11:10:13 +1000, John Tserkezis
<jt@techniciansyndrome.org.invalid> wrote:

Jeff Liebermann wrote:

... Find the frequency, identify the
CPU, and install a replacement resonator.
http://www.tdk.com/crdata.php

At this price-point, there is no "CPU" to look for. They're all Chip
On Board designs (the black blob on the PCB itself).

Are you sure? I haven't repaired enough modern microwave ovens to
even have an opinion, but I have yet to see such a PCB. The problem
is that there are usually a mess of wires that need to connect to
various components in the microwave oven. Handling a black glob PCB
usually breaks the wire bonds. Black glob boards are usually more
suitable for enclosed devices and robotic assembly such as
calculators.

I also gave myself a Google Images tour of "microwave oven PCB" and
variations. There were only about 50 assorted images of such PCB's
and none of them showed a black glob.

The resonator would be a separate component, and I wouldn't think that
an RC oscillator would be good enough to keep mere minutes within a day
anyway.

They're available in surface mount packages and are difficult to
distinguish from a large capacitor.

Besides simply replacing the resonator probably wouldn't show any
significant improvement.

I beg to differ. It's probably off frequency for some reason, and
would bring it closer to acceptable. 120 seconds per day is lousy. 40
seconds per day might be a little less lousy.

You'd need to go all out and use an
upper-market crystal, associated oscillator (with fine tuning) and drive
the clock that way. Would likely need some divider in there too if you
can't match the frequency otherwise.

Yep. However, if the CPU can be identified, the data sheet usually
has instructions for using the internal clock oscillator with either
crystal, ceramic resonator, or external input. At worst, it would
involve also changing 2 capacitors.

And now you're starting to lean into the "not worth it" territory.

I've been there many times. I'm one of those compulsive repairmen
that insists on fixing things, even if they're not profitable.
Personally, I like the ceramic resonator type real time clocks because
when the power dies, and if they have a backup battery, they keep
running. Unfortunately, most microwave ovens don't even have a big
capacitor to keep the time during minor power interruptions, so it's
kinda a wasted preference.

--
Jeff Liebermann jeffl@cruzio.com
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558

 

[Phil Allison]

"John Tserkezis is a MASSIVE JERK"

So you live in a place _not_ on the grid?

Sigh.

I do live on the grid.

** He lives in Sydney - same as me.

I, along with everyone else was promised that
while short term timing errors on the grid are expected, longer term
errors are tweaked and by far less.
I could have lived with this if they actually meant it.

** The do actually - fuckhead.

However, in real life, it's STILL not good enough for clocks. Like I
said, I have many non-temperature crystal locked clocks that are by FAR
better.

** That is complete bollocks.

I have cheap Chinese digital watches (NOT on my wrist) at the whim of
the seasonal temperature environment that have minor seasonal shifts,
but over the year are better than mains-locked clocks.

** The problem is not with the frequency of the AC supply.

I can't even get a straight answer googling. Lots of nice graphs that
really don't answer the question too. The BEST was along the lines of
"is very precise" and "it shouldn't vary too far".

In engineering terms, that's just plain bullshit.

** The supply grid has absolutely precise time keeping over the long term.

The PROBLEM is that any given premises may suffer losses of power for short
periods from time to time - over a year these all add up of course. And
they all add in the same direction !!!!

Most mains derived clocks react to a power outage by going into a mode where
the display blinks and the time resets to midnight or some such. OTHERS
( with battery back up) however revert to an internal oscillator that is
only based on a cheap R and C and give no warning that there has been an
outage.

The most common outages in homes are due to ELCBs and breakers tripping.


..... Phil

 

[Jeff Liebermann]

On Sun, 24 Apr 2011 10:58:06 +1000, John Tserkezis
<jt@techniciansyndrome.org.invalid> wrote:

William Sommerwerck wrote:

"John Tserkezis" <jt@techniciansyndrome.org.invalid> wrote in message
news:4db34c2b$0$13392$afc38c87@news.optusnet.com.au...

You mean you had a VCR that *DIDN'T* flash 12:00 all day?

Most Public TV stations broadcast time code. The "later" VHS VCRs searched
for a PBS station and used the time code to set the clock.

Sorry, my email addy doesn't show it, but I'm in Australia. I've
verified that timing codes are NOT sent by ANY networks here.

Are you talking about analog TV or digital TV? With analog, it was at
the discretion of the station. With digital, it's inscribed into the
ATSC (US) and DVB-T (everyone else) standards. Unfortunately, I don't
know enough about either of these to explain how to get to these time
codes.

More so, of the few that knew what I was talking about, said they had
heard of it, but no, they did not, and never will support it.

In the US, the local PBS station (KTEH) went the other way. They
claim to be sending time code signals, but none of my 3 assorted VCR
boxes would find it.

That was about a decade ago while digital terrestrial TV was still a
bit experimental here, but even then, I don't know of any digital PVRs
today that maintain time that way.

Nope. I have a Philips HDRW720. It can use a PBS station for real
time setting. See the owners manual at:
<http://download.p4c.philips.com/files/h/hdrw720_17b/hdrw720_17b_dfu_aen.pdf>
on page 22 under "automatic time setting". The problem is that there
has been no firmware update to deal with the 2007 change in daylight
savings time.

My Tivo 2 can also use a PBS station for date and time, when connected
to an antenna.

Most digital DVR's are now integrated with either a cable TV modem or
a satellite receiver. Time sync comes from the network. My DirecTV
R16-300 DVR is about 3 seconds early, for no obvious reason, which
makes time shifted recordings start 3 seconds early, and chop off 3
seconds from the end. Grrr....

There's the odd one out that does NTP if they're 'net connected, but
everything else relies on standard crystal clocked circuits.

Samsung combined Blue-Ray players and DVR work that way. Buried in
the setup is the option to specify an NTP server. I like to use:
us.pool.ntp.org

Probably a bad example though. At that that level of complexity,
no-one in their right mind would go to the trouble of mains-locking a
clock, when implementing a single-chip Real Time Clock is so much
easier. Especially so when RTC functionality is built into other multi
purpose chips.

Yep. However, if the functionality is there as in a network connected
device, and the computer horsepower is available, then network time is
just a matter of adding some code.

I suppose this comes down to the question of is mains locking "good
enough"?

It's marginal. Loss of power means loss of clock, with no automatic
method of restoring the real time. No thanks.

Since I can't get a clear answer from anyone, I'm staying with crystals
or NTP where applicable.

Perhaps a wind up clock would suffice?

--
Jeff Liebermann jeffl@cruzio.com
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558

 

[William Sommerwerck]

"KR" <kenreed1999@gmail.com> wrote in message
news:19295b80-cb87-416b-930b-05ae71bd9761@x37g2000prb.googlegroups.com...
On Apr 23, 5:36 am, "William Sommerwerck" <grizzledgee...@comcast.net>
wrote:

The only reason I can guess for the poor timekeeping is that the
circuitry/board running the clock is used on their 220 volt 50 Hz
versions without modification, so the makers decided not to use
line sync for the clock, and probably didn't want to spring for
a crystal oscillator either, so used an RC one instead.

A chip would likely be designed to work at either frequency,
with a jumper making the selection.

IF it is set to 50HZ and you are on a 60HZ supply -- then it
would run faster, not slower.

That has nothing whatever to do with the point I was making.

Open it up and see if there is some sort of trimmer capacitor
near the clock chip, adjust it and see what happens. Might
have to fiddle with it over some time to get it right.

I find it impossible to believe that /any/ microwave oven would use a
crystal or resonator -- which costs money -- when the power frequency is
/right there/, and free.

 

[KR]

On Apr 23, 5:36 am, "William Sommerwerck" <grizzledgee...@comcast.net>
wrote:

The only reason I can guess for the poor timekeeping is that the
circuitry/board running the clock is used on their 220 volt 50 Hz
versions without modification, so the makers decided not to use
line sync for the clock, and probably didn't want to spring for
a crystal oscillator either, so used an RC one instead.

A chip would likely be designed to work at either frequency, with a jumper
making the selection.

IF it is set to 50HZ and you are on a 60HZ supply - then it would run
faster, not slower.

Open it up and see if there is some sort of trimmer capacitor there
near the clock chip, adjust it and see what happens. Might have to
fiddle with it
over some time to get it right

 

[Michael A. Terrell]

William Sommerwerck wrote:

"KR" <kenreed1999@gmail.com> wrote in message
news:19295b80-cb87-416b-930b-05ae71bd9761@x37g2000prb.googlegroups.com...
On Apr 23, 5:36 am, "William Sommerwerck" <grizzledgee...@comcast.net
wrote:

The only reason I can guess for the poor timekeeping is that the
circuitry/board running the clock is used on their 220 volt 50 Hz
versions without modification, so the makers decided not to use
line sync for the clock, and probably didn't want to spring for
a crystal oscillator either, so used an RC one instead.

A chip would likely be designed to work at either frequency,
with a jumper making the selection.

IF it is set to 50HZ and you are on a 60HZ supply -- then it
would run faster, not slower.

That has nothing whatever to do with the point I was making.

Open it up and see if there is some sort of trimmer capacitor
near the clock chip, adjust it and see what happens. Might
have to fiddle with it over some time to get it right.

I find it impossible to believe that /any/ microwave oven would use a
crystal or resonator -- which costs money -- when the power frequency is
/right there/, and free.

They would use the same crystal as the microprocessor, and run the
clock in software.


--
You can't fix stupid. You can't even put a Band-Aid™ on it, because it's
Teflon coated.

 

[John Tserkezis]

Jeff Liebermann wrote:

Sorry, my email addy doesn't show it, but I'm in Australia. I've
verified that timing codes are NOT sent by ANY networks here.

Are you talking about analog TV or digital TV? With analog, it was at
the discretion of the station. With digital, it's inscribed into the
ATSC (US) and DVB-T (everyone else) standards. Unfortunately, I don't
know enough about either of these to explain how to get to these time
codes.

I was talking about analogue. Digital here has only appeared in the
last decade or so, and even then, early on, there were no boxes you
could actually buy to do that. (funny though, the politicians could get
wide screen TVs and set top boxes donated quite easily to them - for
"testing purposes" you understand).

It's only the last few years the prices have come down to something
that's actually reasonable.

On the analogue side, we had *electronics* here that was compatible
with a variety of systems around the world for a LONG time, non of which
were compatible with our TV stations.

In the US, the local PBS station (KTEH) went the other way. They
claim to be sending time code signals, but none of my 3 assorted VCR
boxes would find it.

At least you had TV stations that WERE transmitting codes, we didn't.
Not till digital anyway.

on page 22 under "automatic time setting". The problem is that there
has been no firmware update to deal with the 2007 change in daylight
savings time.

Arrg. Don't start me on daylight savings. I'm actively unimpressed
unless the device supports USER DEFINABLE dst rules.

My Tivo 2 can also use a PBS station for date and time, when connected
to an antenna.

TiVo is another thing that failed miserably in Australia. The group in
charge of keeping things running in Australia and New Zealand is down to
7 people (from 100+?) and will only keep the current licence holders
going till that runs out.
Shame really. If it actually worked like it does in the US, it would
be quite viable. Except they knobbled a lot of the functionality, and
charged more for the privileged.

I suppose this comes down to the question of is mains locking "good
enough"?

It's marginal. Loss of power means loss of clock, with no automatic
method of restoring the real time. No thanks.

That's pretty much my attitude. Mains locking is cheaper to implement
than crystal locking, and if mains locking was better than crystal then
no-one would implement cyrstal locking in the first place.
But it IS better, AND you get very viable long term battery backed
timekeeping with crystals too. That leaves no reason for mains locked
clocks to be implemented at all, outside of "novelty" clocks. Really,
is saving two cents really worth it when you're pissing off your
customers that much?

Since I can't get a clear answer from anyone, I'm staying with crystals
or NTP where applicable.

Perhaps a wind up clock would suffice?

I don't do windups. Outside of antiques that is.
But an antique isn't a timepiece, it's an example of a combination of
the limited technology of the day, doing a job that a two dollar Chinese
wrist-strapped device can today.
--
None of you exist, my Sysop types all this in.

 

[William Sommerwerck]

I find it impossible to believe that /any/ microwave oven
would use a crystal or resonator -- which costs money --
when the power frequency is /right there/, and free.

They would use the same crystal as the microprocessor,
and run the clock in software.

I was, of course, referring to a separate crystal (he said, retro-conning
the issue).

 

[Michael A. Terrell]

William Sommerwerck wrote:

I find it impossible to believe that /any/ microwave oven
would use a crystal or resonator -- which costs money --
when the power frequency is /right there/, and free.

They would use the same crystal as the microprocessor,
and run the clock in software.

I was, of course, referring to a separate crystal (he said, retro-conning
the issue).

William, I've never seen a microwave where the clock wasn't just
software in the MPU and I've scrapped a lot of them over the years.
Just one 20 cent crystal, and a MPU that cost less than a dollar in
production quantities.


--
You can't fix stupid. You can't even put a Band-Aid™ on it, because it's
Teflon coated.

 

[KR]

On Apr 25, 9:05 pm, "William Sommerwerck" <grizzledgee...@comcast.net>
wrote:

I find it impossible to believe that /any/ microwave oven
would use a crystal or resonator -- which costs money --
when the power frequency is /right there/, and free.
They would use the same crystal as the microprocessor,
and run the clock in software.

I was, of course, referring to a separate crystal (he said, retro-conning
the issue).

It is possible to have a second crystal, as I have seen a custom IC
that included an RTC inside have its own separate 3.579mhz
crystal in addition to the crystal for the CPU function. Both
connected to the same IC on separate pins.
In this case though, this second crystal would be just for the RTC
function,
as the low speed would save power, and also the rest of the chip would
likely be shut down also to save power hence no software clock.
This was an early 1990s design



As microwaves don't have a RTC in them (that I have seen) and simply
reset to 12:00 or whatever after a power failure,
there is no need for the more sophisticated system described above.



I agree with Michael A Terrell,

There wouldn't be a separate crystal for the clock in a modern
microwave,
The "clock" and processor would be incorporated into a single chip,
and the timekeeping would
be surely a software function. The chip would very likely have a
single crystal
or resonator.

In my original comment, I suggested looking for a trimmer cap
across the crystal and trimming it to fine adjust the clock if the
inaccuracy is a real issue for the OP .

You also can check the value of the crystal, there is also a very very
remote possibility that there was the wrong value, but very close
installed at the factory - something like a 3.562 MHZ instead of a
3.579545 MHZ
for example, but without sitting and working out the inaccuracy this
would cause, this might not be right and might make too big a
difference to be the problem.


It will not hurt the rest of the system to run a tiny fraction faster/
slower.

 

[josephkk]

On Wed, 27 Apr 2011 06:50:05 -0700 (PDT), KR <kenreed1999@gmail.com>
wrote:

On Apr 25, 9:05 pm, "William Sommerwerck" <grizzledgee...@comcast.net
wrote:
I find it impossible to believe that /any/ microwave oven
would use a crystal or resonator -- which costs money --
when the power frequency is /right there/, and free.
They would use the same crystal as the microprocessor,
and run the clock in software.

I was, of course, referring to a separate crystal (he said, retro-conning
the issue).


It is possible to have a second crystal, as I have seen a custom IC
that included an RTC inside have its own separate 3.579mhz
crystal in addition to the crystal for the CPU function. Both
connected to the same IC on separate pins.
In this case though, this second crystal would be just for the RTC
function,
as the low speed would save power, and also the rest of the chip would
likely be shut down also to save power hence no software clock.
This was an early 1990s design



As microwaves don't have a RTC in them (that I have seen) and simply
reset to 12:00 or whatever after a power failure,
there is no need for the more sophisticated system described above.

That IS a RTC. It just has no battery backup to ride through power

outages.

I agree with Michael A Terrell,

There wouldn't be a separate crystal for the clock in a modern
microwave,
The "clock" and processor would be incorporated into a single chip,
and the timekeeping would
be surely a software function. The chip would very likely have a
single crystal
or resonator.

In my original comment, I suggested looking for a trimmer cap
across the crystal and trimming it to fine adjust the clock if the
inaccuracy is a real issue for the OP .

You also can check the value of the crystal, there is also a very very
remote possibility that there was the wrong value, but very close
installed at the factory - something like a 3.562 MHZ instead of a
3.579545 MHZ

It would save a lot more power at 32,786 Hz, which is also standard for
the application.

for example, but without sitting and working out the inaccuracy this
would cause, this might not be right and might make too big a
difference to be the problem.


It will not hurt the rest of the system to run a tiny fraction faster/
slower.

 

[KR]

On Apr 29, 3:28 am, josephkk <joseph_barr...@sbcglobal.net> wrote:

On Wed, 27 Apr 2011 06:50:05 -0700 (PDT), KR <kenreed1...@gmail.com
wrote:



On Apr 25, 9:05 pm, "William Sommerwerck" <grizzledgee...@comcast.net
wrote:
I find it impossible to believe that /any/ microwave oven
would use a crystal or resonator -- which costs money --
when the power frequency is /right there/, and free.
They would use the same crystal as the microprocessor,
and run the clock in software.

I was, of course, referring to a separate crystal (he said, retro-conning
the issue).

It is possible to have a second crystal, as I have seen a custom IC
that included an RTC inside have its own separate 3.579mhz
crystal in addition to the crystal for the CPU function. Both
connected to the same IC on separate pins.
In this case though, this second crystal would be just for the RTC
function,
as the low speed would save power, and also the rest of the chip would
likely be shut down also to save power hence no software clock.
This was an early 1990s design

As microwaves don't have a RTC in them (that I have seen) and simply
reset to 12:00 or whatever after a power failure,
there is no need for the more sophisticated system described above.

That IS a RTC.  It just has no battery backup to ride through power
outages.

Sorry - I was referring to a RTC as a dedicated clock IC either in or
out of a custom IC.
battery backed up

I would think that a microwave clock would be done in software though,
and not have
a dedicated clock section in the custom IC, not that it matters
though.

 

[josephkk]

On Fri, 29 Apr 2011 22:15:18 -0700 (PDT), KR <kenreed1999@gmail.com>
wrote:

On Apr 29, 3:28 am, josephkk <joseph_barr...@sbcglobal.net> wrote:
On Wed, 27 Apr 2011 06:50:05 -0700 (PDT), KR <kenreed1...@gmail.com
wrote:



On Apr 25, 9:05 pm, "William Sommerwerck" <grizzledgee...@comcast.net
wrote:
I find it impossible to believe that /any/ microwave oven
would use a crystal or resonator -- which costs money --
when the power frequency is /right there/, and free.
They would use the same crystal as the microprocessor,
and run the clock in software.

I was, of course, referring to a separate crystal (he said, retro-conning
the issue).

It is possible to have a second crystal, as I have seen a custom IC
that included an RTC inside have its own separate 3.579mhz
crystal in addition to the crystal for the CPU function. Both
connected to the same IC on separate pins.
In this case though, this second crystal would be just for the RTC
function,
as the low speed would save power, and also the rest of the chip would
likely be shut down also to save power hence no software clock.
This was an early 1990s design

As microwaves don't have a RTC in them (that I have seen) and simply
reset to 12:00 or whatever after a power failure,
there is no need for the more sophisticated system described above.

That IS a RTC.  It just has no battery backup to ride through power
outages.



Sorry - I was referring to a RTC as a dedicated clock IC either in or
out of a custom IC.
battery backed up

I would think that a microwave clock would be done in software though,
and not have
a dedicated clock section in the custom IC, not that it matters
though.


I'm cool with this. No worries.

Plus there are plenty of uC that could wake up once per second just enough
to keep good time, but a crystal and battery are more parts and cost.

 

[Abi Normal]

On Apr 22, 12:35 pm, jeff_wisnia <jwisniaDumpThisP...@conversent.net>
wrote:

 From my "I can't believe they'd make it this way" list.

Last year I purchased a new Westinghouse Model WST-3501 microwave oven
for our office lunchroom.

It heats and cooks things fine, but has one lousy feature which I can't
avoid noticing and which irks me very time I see it.

The TOD, which is displayed whenever the unit is on standby, loses about
two minutes a day, and has to be reset every few days to keep from
falling ridiculously behind.

I'm shocked, shocked, that the clock isn't synced to the 120 volt 60 Hz
power the unit runs on, like every other line powered appliance with a
digital TOD display I own.

The only reason I can guess for the poor timekeeping is that the
circuitry/board running the clock is used on their 220 volt 50 Hz
versions without modification, so the makers decided not to use line
sync for the clock, and probably didn't want to spring for a crystal
oscillator either, so used an RC one instead.

I tried emailing Westinghouse about this and they fobbed me off to a
subsidiary who couldn't even bother to respond to an email.

Thanks for reading. Anyone got any other ideas about why the stupid
clock in this microwave loses time?

Jeff

--
Jeffry Wisnia
(W1BSV + Brass Rat '57 EE)
The speed of light is 1.8*10e12 furlongs per fortnight.

Why not donate the SOB and get a new one for $40 and STFU!