info
Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss
Electromechanical Appliance Timers
41 views
Skip to first unread message
Steve Kraus
Feb 19, 2017, 8:39:06 AM2/19/17
to
I've gotten curious about the accuracy or not of these. I am speaking of
the motor-driven timers with a 24-hour dial on the front.
Obviously not something someone would use where precision is required so
this is purely an academic question.
I was under the impression that these were driven by tiny sync motors and
therefore as accurate as the power line frequency just like a traditional
electric wall clock. Not easy to set precisely but should repeat nearly to
the second day after day. Or so I thought.
Recently, using a timer on Christmas lights, I noticed variation as much as
a minute. But that was a newer timer and I don't know what kind of motor
it has. I know on the older ones, the motor rotor could randomly start in
either direction but there was device in gear train that would kick it to
reverse if it happened to start by running in reverse. Doesn't that sound
like a sync motor?
So for a few days I've been running a little experiment. I have a number
of timers of various vintages. I ran each to where it just kicked on then
unplugged it. I put the group on a couple of outlet strips and powered
them at the same time, each hooked to an indicator lamp. At the expected
time each night I put a smartphone running an NTP-connected clock app
nearby and video the scene with another phone so as to produce a record of
when each turned on. I was expecting them to switch on a few seconds
before the designated time (based on how many seconds elapsed during the
setup adjustment between the timer triggering and my unplugging it) but the
variation has been much greater, nearly a minute early on one but generally
all over the map and different each night and not in unison as though the
power line frequency* had varied. Maybe this is from mechanical variations
from the switch part of the timer snapping over. Or maybe they are not
sync motors after all.
* I think the utilities used to make up for lost or gained time keeping it
to ą5 seconds. Now I think it's ą25 seconds. That doesn't account for the
variation I see though it could be some of it.
Again, there is no practical point to this; a digital timer can be used
where precision is needed. I am just curious.
the motor-driven timers with a 24-hour dial on the front.
Obviously not something someone would use where precision is required so
this is purely an academic question.
I was under the impression that these were driven by tiny sync motors and
therefore as accurate as the power line frequency just like a traditional
electric wall clock. Not easy to set precisely but should repeat nearly to
the second day after day. Or so I thought.
Recently, using a timer on Christmas lights, I noticed variation as much as
a minute. But that was a newer timer and I don't know what kind of motor
it has. I know on the older ones, the motor rotor could randomly start in
either direction but there was device in gear train that would kick it to
reverse if it happened to start by running in reverse. Doesn't that sound
like a sync motor?
So for a few days I've been running a little experiment. I have a number
of timers of various vintages. I ran each to where it just kicked on then
unplugged it. I put the group on a couple of outlet strips and powered
them at the same time, each hooked to an indicator lamp. At the expected
time each night I put a smartphone running an NTP-connected clock app
nearby and video the scene with another phone so as to produce a record of
when each turned on. I was expecting them to switch on a few seconds
before the designated time (based on how many seconds elapsed during the
setup adjustment between the timer triggering and my unplugging it) but the
variation has been much greater, nearly a minute early on one but generally
all over the map and different each night and not in unison as though the
power line frequency* had varied. Maybe this is from mechanical variations
from the switch part of the timer snapping over. Or maybe they are not
sync motors after all.
* I think the utilities used to make up for lost or gained time keeping it
to ą5 seconds. Now I think it's ą25 seconds. That doesn't account for the
variation I see though it could be some of it.
Again, there is no practical point to this; a digital timer can be used
where precision is needed. I am just curious.
pf...@aol.com
Feb 19, 2017, 5:01:04 PM2/19/17
to
On Sunday, February 19, 2017 at 8:39:06 AM UTC-5, Steve Kraus wrote:
These devices usually run on a cartridge motor, often paired with a non-synchronized motor if any torque is really needed. The synchron more-or-less governs the standard motor - within very broad limits. As you note, these are not designed to be precision devices as there is really nothing to be gained on a cheap timer that gets reset every so often. But any load on the motor for these devices will definitely slow things down until it re-aligns.
If it is a purely cartridge-driven device - a regular electric clock for instance, there is a very short gear train of lightweight gears and a soft-switch for the alarm function, if any.
More-so, tariffs now are far less stringent on AC frequency, as you noted. Keep in mind that it is no longer a small, happy family of large utilities, but a much broader family of small generators, many using very small solid-state grid-tie inverters that may be very precise in frequency regulation - but may not be instantly synchronized to the grid when they come on-line. And solar micro-converters are nightmares as not only do they need to align to the grid, but to each other as each comes and goes on and off-line. The process is very nearly instantaneous, and the are not supposed to tie until properly synched - except...
Peter Wieck
Melrose Park, PA
These devices usually run on a cartridge motor, often paired with a non-synchronized motor if any torque is really needed. The synchron more-or-less governs the standard motor - within very broad limits. As you note, these are not designed to be precision devices as there is really nothing to be gained on a cheap timer that gets reset every so often. But any load on the motor for these devices will definitely slow things down until it re-aligns.
If it is a purely cartridge-driven device - a regular electric clock for instance, there is a very short gear train of lightweight gears and a soft-switch for the alarm function, if any.
More-so, tariffs now are far less stringent on AC frequency, as you noted. Keep in mind that it is no longer a small, happy family of large utilities, but a much broader family of small generators, many using very small solid-state grid-tie inverters that may be very precise in frequency regulation - but may not be instantly synchronized to the grid when they come on-line. And solar micro-converters are nightmares as not only do they need to align to the grid, but to each other as each comes and goes on and off-line. The process is very nearly instantaneous, and the are not supposed to tie until properly synched - except...
Peter Wieck
Melrose Park, PA
olds...@tubes.com
Feb 19, 2017, 5:57:40 PM2/19/17
to
They are not meant for precision.
Steve Kraus
Feb 22, 2017, 12:50:18 AM2/22/17
to
I'm not familiar with the term "cartridge motor."
Here is a photo of one type of timer motor. I was under the impression,
perhaps wrongly, that this is a sync motor. This one does start either way
and there's a thing in the gear train to jam it and cause it to reverse if
it starts up in reverse.
https://imgur.com/gallery/y5hNb
If this is in fact a sync motor then the variation I see must just be
something with the actual switch mechanism and how it snaps over each time.
Here is a photo of one type of timer motor. I was under the impression,
perhaps wrongly, that this is a sync motor. This one does start either way
and there's a thing in the gear train to jam it and cause it to reverse if
it starts up in reverse.
https://imgur.com/gallery/y5hNb
If this is in fact a sync motor then the variation I see must just be
something with the actual switch mechanism and how it snaps over each time.
pf...@aol.com
Feb 22, 2017, 7:31:47 AM2/22/17
to
On Wednesday, February 22, 2017 at 12:50:18 AM UTC-5, Steve Kraus wrote:
> I'm not familiar with the term "cartridge motor."
http://image2.cccme.org.cn/i_supply/2011-03-05/20110305120000000920290.jpg
> I'm not familiar with the term "cartridge motor."
This is a cartridge motor with the gear-train included. They come in many variation, including a double-layer device for additional torque. They are usually about 40mm across, but can be smaller/larger.
Steve Kraus
Feb 22, 2017, 2:41:32 PM2/22/17
to
"pf...@aol.com" <pf...@aol.com> wrote in
I need to go to Goodwill or someplace and find a sync motor clock with a
second hand and see how much variation it has (no switch mechanism friction
to overcome) against an accurate reference like an NTP server as that would
tell me how the power line is varying these days. That may be part of it.
> This is a cartridge motor with the gear-train included.
So that *is* a sync motor?
I need to go to Goodwill or someplace and find a sync motor clock with a
second hand and see how much variation it has (no switch mechanism friction
to overcome) against an accurate reference like an NTP server as that would
tell me how the power line is varying these days. That may be part of it.
Steve & Lynn
Feb 22, 2017, 9:23:19 PM2/22/17
to
> If they are only a minute off, or even 5 minutes, you're doing good.
> They are not meant for precision.
>
IT TOOK SITRE MAGANA ABOUT FIVE MINUTES TO CUT HIS HAMSTER UP WITH A
> They are not meant for precision.
>
PAIR OF SCISSORS.
Steve Kraus
Feb 23, 2017, 1:59:58 PM2/23/17
to
Four of five timers were trending the same way suggesting power line
frequency.
From 2/20 to 2/21 to 2/22:
Timer B1 got 47 seconds later then 38 seconds earlier.
Timer B2 got 55 seconds later then 58 seconds earlier.
Timer C1 got 41 seconds later then 32 seconds earlier.
Timer C2 got 82 seconds later then 39 seconds earlier.
Timer D1 got 10 seconds earlier then 14 seconds later.
The B's are nearly identical to the one whose motor I showed. The C's I've
never had open. The D's I've looked at but have no clue if it's a sync
motor or not. If not, it's kind of funny as that one seems to hew most
closely to the expected turn-on time.
Getting a vintage electric wall clock off eBay. Will set it accurately to
an NTP server and then watch it for subsequent variations. Inexpensive way
to look at this and then ultimately a useful device.
frequency.
From 2/20 to 2/21 to 2/22:
Timer B1 got 47 seconds later then 38 seconds earlier.
Timer B2 got 55 seconds later then 58 seconds earlier.
Timer C1 got 41 seconds later then 32 seconds earlier.
Timer C2 got 82 seconds later then 39 seconds earlier.
Timer D1 got 10 seconds earlier then 14 seconds later.
The B's are nearly identical to the one whose motor I showed. The C's I've
never had open. The D's I've looked at but have no clue if it's a sync
motor or not. If not, it's kind of funny as that one seems to hew most
closely to the expected turn-on time.
Getting a vintage electric wall clock off eBay. Will set it accurately to
an NTP server and then watch it for subsequent variations. Inexpensive way
to look at this and then ultimately a useful device.
Jeroni Paul
Feb 23, 2017, 6:46:49 PM2/23/17
to
Steve Kraus wrote:
> Getting a vintage electric wall clock off eBay. Will set it accurately to
> an NTP server and then watch it for subsequent variations. Inexpensive way
> to look at this and then ultimately a useful device.
Years ago I designed and built a nixie clock with digital TTL counters using mains frequency as timing source. Here in Spain the typical variation I see is between 1 and 10 seconds per day and within some days I've seen it go as far as 30-40 seconds, then slowly correct back. That has been always the behaviour of that clock from 2007 to now. At long term time keeping is good.
> Getting a vintage electric wall clock off eBay. Will set it accurately to
> an NTP server and then watch it for subsequent variations. Inexpensive way
> to look at this and then ultimately a useful device.
I've found this website clock https://time.is/ a good time reference, I successfully calibrated a quartz clock with adjustable trimmer in a few days checking once per day and it kept perfectly aligned for a month. My previous attempts with some cheap DCF77 radio controlled clocks failed, apparently they randomly added or removed a few seconds on each sync.
Steve Kraus
Feb 23, 2017, 10:55:43 PM2/23/17
to
I just had an idea. I have a UPS sitting around out of service as the
batteries went dead and even with good batteries I don't know how long it
would run even at very low load. I don't know how precise it is though
when I looked at it on a scope it was a nice sine wave and looked stable in
frequency compared to the mains. I could rig a 24VDC supply and use the
output of the UPS to run my group of timers that I am testing. If the UPS
has an accurate output then the timers should be accurate as well. If the
UPS is fast or slow then at least the timers will get ahead or behind the
same amount each day, not varying like I have observed. Hopefully it would
be stable at whatever frequency it is and not go up and down.
batteries went dead and even with good batteries I don't know how long it
would run even at very low load. I don't know how precise it is though
when I looked at it on a scope it was a nice sine wave and looked stable in
frequency compared to the mains. I could rig a 24VDC supply and use the
output of the UPS to run my group of timers that I am testing. If the UPS
has an accurate output then the timers should be accurate as well. If the
UPS is fast or slow then at least the timers will get ahead or behind the
same amount each day, not varying like I have observed. Hopefully it would
be stable at whatever frequency it is and not go up and down.
0 new messages