Circuit to drive an old railway clock

Thread Starter

Doctor_Ed

Joined Feb 10, 2022
9
First post here. Dabbler in electronics but seriously behind the times now.

I need a circuit to deliver 24 VDC pulses once a minute to an old railway clock. The pulses advance the minute hand. Each pulse should be one or two seconds long and preferably damped. Power is around 20-25 mA.

I already have a circuit that will do that, almost. It works most of the time. That is not very good for a clock. It uses a reed switch tripped by an AC synchronous motor which controls an FET. The diagram is attached to this post.

There are problems with the existing design. The three sections of interest are

- Power supply - the voltage divider is not providing the power needed
- A timer chip would be much better than the mechanical 1-minute switching section
- It would be nice to damp the power going to the clock so the armature works quietly.

I would appreciate suggestions for a better circuit.
 

Attachments

Thread Starter

Doctor_Ed

Joined Feb 10, 2022
9
Thanks for the link to the power supply. Looks neat.

Yes, it would be great to have an all electronic setup. Years ago I looked at frequency counters and wasn't able to figure them out, so I would need guidance. I have no experience with them.
 

Thread Starter

Doctor_Ed

Joined Feb 10, 2022
9
Thanks, boostbuck. I am wondering if 0.16% is satisfactory for a clock. A counter working off the 60 Hz might be better.

Buying a made-up module might be best but I'm flexible.
 

AlbertHall

Joined Jun 4, 2014
12,344
Thanks, boostbuck. I am wondering if 0.16% is satisfactory for a clock. A counter working off the 60 Hz might be better.

Buying a made-up module might be best but I'm flexible.
In UK using the 50Hz supply will keep the clock accurate in the long term but there will be short term deviations.
With any crystal derived frequency the clock will gradually drift.

If my maths is right 0.16% is about one minute per day.
 

crutschow

Joined Mar 14, 2008
34,280
f my maths is right 0.16% is about one minute per day.
I calculate over 2 minutes per day but, either way, that's a large drift.
Most cheap battery clocks seem to be good for a minute or two per month.
The board does have a calibration trimmer, so you perhaps could tweak it to be much better.
 

Thread Starter

Doctor_Ed

Joined Feb 10, 2022
9
Okay, the inherent accuracy of the Cebek is probably similar to those cheap battery clocks. So after adjusting the trimmer, it might be pretty good. However, adjusting a clock like this forward or back a minute or two every month is going to be awkward. It would involve unplugging it for a short while, or rigging an advance button to it, or opening the case and doing a mechanical adjustment.

So a counter off the mains would still be preferred. I'm on 60 Hz here.
 

LowQCab

Joined Nov 6, 2012
4,022
Here's a simple to build version with your requested "Soft-Switching".
This feature may have to be turned-off and abandoned,
depending upon how the Escapement-Mechanism works,
because it might "need" a hard hit to operate reliably.
.
.
.
Clock Driver 60hz  .png
 

Tonyr1084

Joined Sep 24, 2015
7,852
A few clocks I have run off the 60Hz. They tend to be fast - about a minute a month. After a few months I set them back about 2 to 3 minutes so I don't have to mess with them more often than about three times a year. Unplugging them isn't an option because they will need to be fully reset. One clock has battery backup. Time keeping is even worse then. The clock can game a few minutes over night.

My programmable thermostat has a quartz crystal oscillator. Since I only run heat I don't bother with it in the warm months. But at the beginning of the heating season I'll set the time accurately. Within two months it's gained about 10 minutes. It's horrible at time keeping. So since I heat for about five months I'm resetting it too often. Over night while I'm in bed it drops the temperature. Before the time I usually get up it boosts the temperature back to 72˚F. The time keeping isn't too much of a problem because in the morning I like getting up to a warm house. So if it starts 10 or 15 minutes early - oh well.
 

Alec_t

Joined Sep 17, 2013
14,280
Radio-clock modules are extremely accurate, shouldn't need adjusting, and aren't too expensive. Access to a 1sec 'tick' signal is probably possible, for driving a power pulse circuit.
 

AlbertHall

Joined Jun 4, 2014
12,344
A few clocks I have run off the 60Hz. They tend to be fast - about a minute a month.
In the UK the long term total number of cycles is kept correct. I have a 'flip clock' powered and timed by our 50Hz supply. With no resetting of time (except after a power cut) it has kept time to within a minute for years.
 

marcf

Joined Dec 29, 2014
288
Just curious, but does your clock have any resemblance to this this?

I maintained these classroom clocks when I worked for a large west coast university in the 90's.

If you need any 'gory details' on how they worked, let me know.

One feature that they had was they would synchronize every hour when the hour solenoid was activated by the controller. You could also set them to 6AM /PM with a 12 hour reset pulse, and the controller was programed to do this at 6AM. This was handy to deal with momentary power glitches and set them all for daylight savings time. (activate the hour advance 11 times to set them back 1 hour) Also at times the clocks in one building would become 'scattered' due to a power loss, and rather than having to reset all of them, you could give the building a 12m hour correction with the master controller , advance them to the last hour and they would all be at the same time when the minute hand crossed 12.
 

Attachments

Thread Starter

Doctor_Ed

Joined Feb 10, 2022
9
Here's a simple to build version with your requested "Soft-Switching".
This feature may have to be turned-off and abandoned,
depending upon how the Escapement-Mechanism works,
because it might "need" a hard hit to operate reliably.
.
.
.
View attachment 260431
Thanks for the amazing detailed circuit. It will take me a while to absorb it all. I noted the 5V box and its description. Fascinating gadget that is new to me. (I'm of the resistor-capacitor-tube school.) Why the parallel capacitors in the 5V section? It looks like the LEDs are there just as visual confirmation of current in the respective lines.

Where is the compartitor in the circuit?

I guess 390R means 390 Ω and 4K7 means 4.7 KΩ.:)
The soft switching worked with my old circuitry, so should work with this one too.

I think I can build this! Appreciated.
 

Thread Starter

Doctor_Ed

Joined Feb 10, 2022
9
Just curious, but does your clock have any resemblance to this this?

I maintained these classroom clocks when I worked for a large west coast university in the 90's.

If you need any 'gory details' on how they worked, let me know.

One feature that they had was they would synchronize every hour when the hour solenoid was activated by the controller. You could also set them to 6AM /PM with a 12 hour reset pulse, and the controller was programed to do this at 6AM. This was handy to deal with momentary power glitches and set them all for daylight savings time. (activate the hour advance 11 times to set them back 1 hour) Also at times the clocks in one building would become 'scattered' due to a power loss, and rather than having to reset all of them, you could give the building a 12m hour correction with the master controller , advance them to the last hour and they would all be at the same time when the minute hand crossed 12.
Mine has a horseshoe-shaped core with a coil. There is an armature in the gap, and it turns about 45 degrees when the coil is energized. The armature is connected to a mechanism that drives the minute hand. The hour hand follows with gears. There is provision for resetting the clock, but I don't know how that works. I think the idea is to reset it to 12:00, pause it, and then resume timekeeping when it is actually 12:00.
 
Top