Kundo Kundo electronic question

watchwldr940

Registered User
Nov 26, 2008
401
20
18
Hello All
I have a "Kundo Electronic" clock that I'm trying to bring back to life. The old, large battery was discarded long ago and now only two fine wires remain that go up to the coil. I see that the various parts houses sell adapters to let you use "AA" batteries. Is this all I would need? Seems to me there should be some kind of electronics there as well, but I don't see any. Is this a salvagable project? If I do need some type of circuit board, where can I get one? Thanks. George
 

Tinker Dwight

Registered User
Oct 11, 2010
13,666
78
0
Calif. USA
Hi
Some have the electronics in the coils and some
in the base.
If it is the type with the electronics in the coil,
you need to be careful about polarity.
Tinker Dwight
 

watchwldr940

Registered User
Nov 26, 2008
401
20
18
Thanks for the reply. Just for kicks I held the ends of the wires one on the + side of a "D" cell and one on the - side, and it runs! So I guess I have the type with the electronics in the coil, and they must work. So I should be able to buy a battery holder and install it and should be good to go. George
 

Tom Kloss

NAWCC Member
Donor
Dec 5, 2003
1,882
6
38
N.E. Pennsylvania
Country
Region
Hi George,

These Kundo electronic clocks do operate on 1 ½ volt batteries. If it's the kind that has a little circuit board in the bottom underneath alongside the battery as Tinker said, it's pretty straight forward repair. If it has no circuit board then all the electronics are in the coil. For the battery in mine, I went to radio shack and got a C cell battery holder and mounted it underneath the case there is just enough room to get it in but you have to remove the origional battery holder. Mine is running well on one C cell.

T.J. Kloss :cool:
 

Scottie-TX

Deceased
Apr 6, 2004
936
65
0
81
Mesquite, TX
Country
Region
What is the battery life for a "C" cell in one of these. Reason I ask is that you don't want a battery that is so large that the battery begins rotting before it fails to provide sufficient current. I suspect that the current necessary to drive an interval coil may be relatively high and battery life may not be a concern. .
I have here a Toshiba wall hanger radio that uses 4, "D" cells. It is so overpowered that the batteries will last several years and will actually work fine on half voltage.
 

dAz57

Registered User
Dec 7, 2011
2,018
32
48
sydney Australia
Country
Region
What is the battery life for a "C" cell in one of these. Reason I ask is that you don't want a battery that is so large that the battery begins rotting before it fails to provide sufficient current. I suspect that the current necessary to drive an interval coil may be relatively high and battery life may not be a concern. .
I have here a Toshiba wall hanger radio that uses 4, "D" cells. It is so overpowered that the batteries will last several years and will actually work fine on half voltage.
on a "C" two years or so, you would have to measure the current draw and calculate on the amp hour of the battery to figure on the life of the battery

on those clocks that have the tall square battery at the back of the movement, I gut the battery case and fit a "D" holder inside leaving the back open so the customer can change the battery and leaves the clock looking original.

those with the battery under the base there is just enough room to mount a "C" battery holder

I just tell people to stick to Duracells, nice fresh one from the shop, but people still winge on the cost or they use a battery that has been rattling around the kitchen drawer 3 years out of date.
 

Tinker Dwight

Registered User
Oct 11, 2010
13,666
78
0
Calif. USA
Hi
You'll not be able to determine it with a current meter.
It will pulse the current on each swing.
You'll need to use a shunt and a storage oscilloscope
to get the actual peak and average that over the
time of a swing.
Tinker Dwight
 

Scottie-TX

Deceased
Apr 6, 2004
936
65
0
81
Mesquite, TX
Country
Region
Yeah, my thoughts also: This'd be a pretty involved calculation given duration between pulses and average voltage over the duration of the pulses.
 

Tinker Dwight

Registered User
Oct 11, 2010
13,666
78
0
Calif. USA
At least tough to do without an oscilloscope. There are ways to do it with some
pots, capacitors and extra batteries. Then use a voltmeter.
Tinker Dwight
 

rlwindle

Registered User
Mar 18, 2011
396
21
18
70
Dallas, TX
Country
Region
Just for kicks I held the ends of the wires one on the + side of a "D" cell and one on the - side said:
I have a Junghans ATO with one AA battery (see below), I have mine connected and held in place by a rubber band. Works great. I also have a Kundo that uses a D battery.
 

Attachments

lmester

NAWCC Member
Dec 30, 2009
456
12
18
West Virginia
mesterhome.com
Country
Region
Scottie's question about battery life got me thinking. I'm running one from a C cell. I was thinking that it's been over a year. I checked the battery tonight. It's been 1 year 10 months. I wrote the date on the battery when I installed it. The battery voltage is down to 1.174 volts. It's probably not going to run too much longer.

Also, I only have an older analog scope but I do have a fancy multimeter that will measure average current. With a fresh battery the average current was .42 mA. It's now down to .29 mA. Looking up the info for a Duracell C cell they list 7.8 AH capacity. If you assume a .42 mA drain you get a life of a little over 2 years. That's going to be only approximate for many reasons. The current draw has dropped as the battery voltage decreased, AH rating varies depending on the load on the battery, Battery self discharge rate etc... Also, Duracell lists the cutoff (end of life) voltage for the battery as .8V. When it finally stops I'll know what the actual minimum voltage is for the clock.

It's going to be interesting to see how low the voltage can go before the clock stops. The germanium transistor in the circuit will run on a fairly low voltage. I don't think I'm going to reach the cutoff point for the transistor. Watching the clock I can see that there is now only slightly more pendulum swing than is needed for the pawl to advance the ratchet wheel. The pendulum may still be swinging when it stops running.

I'll post an update when it finally stops.
 

praezis

Registered User
Feb 11, 2008
528
52
28
Germany
Country
Region
Also, I only have an older analog scope but I do have a fancy multimeter that will measure average current. With a fresh battery the average current was .42 mA. It's now down to .29 mA.
Your values appear to be quite high. A Kundo with single coil & electronic board (the newer version) consumes 0.11 mA average @ 1.5V. The similar ATO consumes about 0.035 mA average current.
Regards,
Frank
 

eskmill

NAWCC Fellow
NAWCC Member
Aug 24, 2000
7,135
40
0
Region
Other factors affect the battery current of these and similar "weak current" battery clocks.
1.Ambient temperature and the cell chemistry. 2. The opposing current in the drive coil generated as the permanent field magnet passes through the drive coil prior to current cut off by either the semiconductor switch or the dry contacts switch in the ATO and Bulle battery clocks.

I am unsure if the Germanium junction is able to instantaniously block some of the induced inverse current back into the dry cell. We know that semiconductor junctions turn-on in the forward direction very rapidly but known to be slow to stop conduction. (only the Shottky junction shuts off fast) Some critical analysis of the current using an analog oscilloscope and a sensitive current probe might provide more insight into how and why these "weak-current" battery clocks often rival quartz oscillator digital clock dry cell life.

Somethings to consider.
 
Last edited:

Tinker Dwight

Registered User
Oct 11, 2010
13,666
78
0
Calif. USA
Hi
One of those super caps and a resistor, tied to a
power supply would work well to get the average,
I'd think.
I've not tried this because I do have an oscilloscope
and a bench supply.
Tinker Dwight
 

lmester

NAWCC Member
Dec 30, 2009
456
12
18
West Virginia
mesterhome.com
Country
Region
Your values appear to be quite high. A Kundo with single coil & electronic board (the newer version) consumes 0.11 mA average @ 1.5V. The similar ATO consumes about 0.035 mA average current.
Regards,
Frank

It's likely that my average current readings are incorrect. I used the easy method to measure average current. My fancy multimeter. I looked through the manual for the meter. I can't find any useful specifications on it's limitations. Frequency response, minimum pulse width etc. It's average current reading could be very inaccurate with the short duration current pulses generated by this clock.

Anyway, calculating the battery life With the average current that my meter indicated and using the typical Ampere-Hour capacity of a C cell gives about two years of battery life.

A reality check is that I'm getting close to two years and the battery is nearly dead:)


With your current readings the battery should last several years longer. possibly the self discharge rate of the battery would now be significant. At just a few percent discharge rate per year I didn't bother with that in my calculations.

How did you measure the average current drain for your clocks?

I hope that you don't say "With my digital storage oscilloscope". I only have an old analog scope. My scope can do the job but it'll be a pain. A low frequency waveform with a short pulse width, I'll probably have to stare at my scope screen for a long time to get accurate measurements!

Finally, measuring transient signals and battery capacity is not easy or exact. We may just have to wait and see how long our battery really lasts. The next battery might just run for a lot longer or shorter since the capacity varies even when buying the same brand of battery. Also, I have no idea what the normal tolerances are with the Kundo Electronic clocks. What is the typical current draw & battery life? They might have decided that the worst clock leaving the factory should get at least a year of operation from a zinc-carbon battery. Some of us may have a clock that uses much less power. And we now have alkaline batteries that perform much better than the zinc-carbon cells available when these clocks were made.
 

Tinker Dwight

Registered User
Oct 11, 2010
13,666
78
0
Calif. USA
Hi
Do you have an adjustable bench supply?
Find a large electrolytic capacitor. One in
the 20,000 uF is a good value to start
with.
Use a 15K resistor in series and the capacitor
across the battery terminals ( observe
the capacitors terminals polarity ).
Test it without the clock to see what the
long term leakage is. If it has been sitting
for some time it may need a little forming.
Adjust the supply to maintain 1.5 volts
across the capacitor. Measure the voltage
across the resistor with a meter that is
a digital of vtvm.
Once you have a stable voltage on the cap
and the voltage across the resistor
doesn't change.
Attach the clock and then watch how
much the capacitor voltage drops.
Increase the voltage of the supply
until the capacitor maintains 1.5V.
Measure the drop across the resistor
and use Ohms law to determine the
average current.
When checking the voltage across
the capacitor, do this quickly. Leave
the meter on the resistor to allow
it to be stable.
Tinker Dwight
 

praezis

Registered User
Feb 11, 2008
528
52
28
Germany
Country
Region
Hi Luke,

I agree with you that a multimeter may not be able to work well, with these ultra low frequencies (1...2Hz)

Les' remarks reminded me, that I forgot to take in account the generated AC, so my Kundo value would be even lower.

I am very sorry, but in fact I use a digital storage scope :) Older models (like my Hameg) are very cheap today.
Usually I calculate the current from the oscillogram of the voltage across the coil:
((Battery voltage) - (peak value of generated AC)) : (coil resistance) x (ON-time) : (time of 1 period)
See the picture: it gives 49µA (coil 740 Ohm)
But I do not know if a Germanium transistor will drain an additional permanent low current off-state.

My ATO wall clock runs in the 12th year with 2 C cells.

Rgds
Frank
 

Attachments

Typ1-2-3

Registered User
Oct 13, 2010
44
0
6
Germany
Country
And a Junghans ATO with mechanical contact runs a little more than 7 years. And that's a clock of the 40's!!!

Frank
 

Tinker Dwight

Registered User
Oct 11, 2010
13,666
78
0
Calif. USA
Hi
You can't measure the current by looking at the
voltage across the coil. The current is the difference
in voltage between the back emf and the 1.5 volts.
You have to measure the source current with
a meter shunt resistor.
Imagine, continuing that saw tooth waveform
under the pulse and using that voltage between
that waveform and the 1.5v.
Remember it is the difference in voltage that causes
the current, not the absolute voltage.
The waveform does not have enough information
to determine the current.
Tinker Dwight
 

praezis

Registered User
Feb 11, 2008
528
52
28
Germany
Country
Region
Hi
You can't measure the current by looking at the
voltage across the coil.
Hi,
I am sure you can.

The current is the difference
in voltage between the back emf and the 1.5 volts.
Exactly (at least if you add "divided by the coil's resistance"). The same is stated in my above formula and can be taken from the oscillogram.
You need no extra shunt: there is only one current path and current is switched on/off.
Of course only the portion during the sqare pulse is regarded.
Rgds
Frank
 

Tinker Dwight

Registered User
Oct 11, 2010
13,666
78
0
Calif. USA
How do you know what the back EMF waveform looks
like under the pulse? The current that is flowing is
only 1.5 minus that voltage across the coils resistance.
Under that pulse is the miller feedback from the transistor
plus any emitter resistance coupled through the tickler
and the back EMF. Do you have a magic way of figuring
these?
Tinker Dwight
 
Last edited:

lmester

NAWCC Member
Dec 30, 2009
456
12
18
West Virginia
mesterhome.com
Country
Region
Praezis,

12 years on a battery! It looks like I'll only be getting about two. Either I got unlucky and have a poor quality battery or I might have some excessive leakage current. Very possible with a germanium transistor. I need to check the quiescent current. Also, you mentioned two C cells powering your clock. Are they in series or parallel? About the storage scope. They were just becoming available when I was in college. You could buy a nice car for what one cost! I'd never even thought of buying one. I'll definitely take a look on the internet and see what's available used. My analog scope is pretty worthless for these type of measurements and since it doesn't have a lighted reticle I can't use a scope camera with it.

Tinker,

Yes, I have a bench supply. Using a resistor & cap sounds like a good possibility. Let the cap do the math for you. If I'm thinking correctly the resistor and cap are basically acting as an integrator. I'll need to pick a cap large enough so that there is little voltage change when the coil is triggered.

I'm wondering how significant the back EMF would be? You would have to energize the coil with the magnet stationary and then compare with the pulse current when the clock is running. My Kundo has the electronics in with the coil. I'm not going to risk pulling the coil cover off just for testing. This would be an interesting test to do if you have a clock with the electronics in the base.
 

Tinker Dwight

Registered User
Oct 11, 2010
13,666
78
0
Calif. USA
Hi
You can see from the picture that back EMF is most of
the voltage across the coil. Back EMF is not the current
causing part, only the difference in the voltage of the 1.5
volts. There are some hidden parts of the signal as well.
The current flow in the base lead changes as the transistor
turns on. This is couple through the tickler coil ( that if
I recall was about a 2:1 ratio ).
Tinker Dwight
 

praezis

Registered User
Feb 11, 2008
528
52
28
Germany
Country
Region
Luke,
my 2 cells are in series. But it is a special case as I run my ATO with an electronic device that keeps it radio-clock synchronized. The current drain is the same as with one cell in usual set-up.
I run the pendulum with low but secure amplitude. This will double or more the lifetime of the cell.

Tinker,
I suppose the back EMF looks like the bottom peak. Even if you approximate with a straight line you will land in the middle of the ballpark, good enough for calculating the cell life.

The clock in my oscillogram
a) is contact operated as are most of my clocks
b) has an unusual high back EMF. Most clocks have 0.3 - 0.8 Vp rather.

Rgds
Frank
 

Tinker Dwight

Registered User
Oct 11, 2010
13,666
78
0
Calif. USA
Hi
A high back EMF means an efficient clock. The smaller
the difference the between the back EMF and the
battery voltage the better.
A perfect motor would have exactly the same voltage
if there was no load.
It is true that for the purposes of battery life, it is fine
to estimate the delta. Around 50 ua looks to be a fine
value.
It is interesting that in a normal motor, the back EMF
can be used for speed control since it is propotional
to the speed of the motor.
Tinker Dwight
 

Wynen

Registered User
Jan 21, 2009
59
0
6
Country
Region
@Tinker Dwight:
for a rough estimation of power consumption and efficiency, let's use the simplified equivalent circuit in the diagram below (sorry in German)
Leistungsbilanz.GIF
The real coil is replaced by an ideal L coil and the DC resistance R of the wire. U(in) = induced voltage by the moving magnet = back EMF, U(batt) = batterie voltage
The power consumption P(auf) is of course:
P(auf) = U(batt) * i
The power dissipation (in the resistance of the coil) is
P(verl) = ( U(batt) - U(in) ) * i
The power output P(ab) is the difference between the consumption and the dissipation (very simplified)
P(ab) = P(auf) - P(verl)
with a maximum output at U(in) = U(batt) / 2
The efficiency n at maximum output is
n = 0,5 (50%)
The power consumption at U(in) =U(batt) and also the power output tends toward zero.

I know, this is very simplified but it shows some relationships in the electrical clock.
For example, a well constructed electrical clock (back EMF U(in) at least U(batt/2) is able to keep the amplitude of the pendulum constant. An increase in the amplitude will increase the back EMF U(in) and therefor decrease the power output. To increase the back EMF it is not a good idea to increase the coil. This would result in a higher resistance of the coil i.e. more power dissipation. It is better to increase the magnetic field or speed of the balance.

Best regards from Germany
Hartmut
 

Tinker Dwight

Registered User
Oct 11, 2010
13,666
78
0
Calif. USA
Hi Hatmut
One of the more interesting things is that if you increase the magnetic field,
that would increase the efficiency, it would also result in a decreased swing
of pendulum.
I know this is counter intuitive.
This is because one can not get more back EMF than the value of the
battery.
The back EMF is a multiplication of the velocity of the coil ( result of the amplitude of the swing )
and the strength of the magnetic field.
On a lightly loaded motor ( such as the clock pendulum ), increasing the magnetic
field has no choice but to decrease the amplitude to keep the back EMF from
exceeding the source voltage.
This means, a really strong magnet might not give enough swing to drive
the ratchet wheel of the clock. A really weak magnetic field might not generate
enough torque to make up for the loss of driving the ratchet wheel.
The problem of not enough magnetic field is easy to see on an oscilloscope
as the back EMF will be near or less than 50% of the source.
It is an interesting problem that there needs to be a balance between the
magnetic field and the battery voltage to get the optimal swing.
This was done by the designer by adjusting the number of turns in the coil
to achieve the right swing amplitude.
The magnetic field was set by the most field that they could get into the
iron bar they were using( a safe way to control that variable ).
Tinker Dwight
 

Wynen

Registered User
Jan 21, 2009
59
0
6
Country
Region
Hi Tinker,

you are completely right, concerning the strength of the magnetic field.
One small addition: The back EMF is a proportional to the velocity of the coil (i.e. the amplitude) and the gradient(!) of the magnetic field (dB/dx).
So to increase the back EMF you also have the opportunity to increase the gradient of the magnetic field for example by specially shaped magnets as you can see the shape of the horse shoe magnet in the Brillie clock.

What I just want to show is, that you can discuss the performance of these kind of electric clocks by analyzing this simple model.
E.g. by measuring the back EMF (should be in worst case half of the battery), you can decide, if the magnet is poor or not.


Hartmut
 

Tinker Dwight

Registered User
Oct 11, 2010
13,666
78
0
Calif. USA
Hi
Yes, gradient(!) or number of line of the field. More lines
more voltage.
I have an old 2090 Nicolet oscilloscope ( payed $50 for ) to
display such things.
I just wanted to note that it is possible to have too much
magnetic field and that is a problem as well.
About the best way to think of it is that it tightens up the swing
and makes it less.
Another thing that will increase efficiency is to reduce the resistance
of the wire. When I was younger, we used to rewind motors
with silver plated wire to get the most in a small package.
Tinker Dwight
 

Wynen

Registered User
Jan 21, 2009
59
0
6
Country
Region
@Twinker Dwight
It's a good idea to examine some of the clocks. I should have a look for my old Scope.
Thank you for the discussion.

Hartmut
 

lmester

NAWCC Member
Dec 30, 2009
456
12
18
West Virginia
mesterhome.com
Country
Region
I'd never though of the back EMF as being an indicator of the efficiency of the clock & strength of the magnet. When I had my clock apart to check the battery voltage I noticed that I've never got around to putting in a proper battery holder. I've got some C holders ordered.

When I replace the battery holder I'm going to play around a little with the magnet. I have some neodymium magnets that should fit in place of the original magnet. This could be interesting. I may be able to have it running with very little pendulum amplitude. Maybe even so small that it won't drive the ratchet wheel.

I'm also going to scope the waveform. if back EMF is low and I don't find any problems with the circuit I'll probably re-magnetize the original magnet.
 

lmester

NAWCC Member
Dec 30, 2009
456
12
18
West Virginia
mesterhome.com
Country
Region
My clock finally stopped. The battery was down to .9v. As I expected, the pendulum was still swinging but was not moving far enough to engage the pawl. It ran for about 2-1/2 years. I tried Tinker's suggestion and used a resistor and capacitor to measure the current. With 1.5 volts applied it was drawing .436 mA. I got .42 mA using the average current function on my multimeter. It looks like the meter is fairly accurate. This means that my clock is taking much more current than others have measured.

It was mentioned on this thread that the strength of the magnet will affect the efficiency. I was planning on trying a neodymium magnet. Unfortunately, the smallest ones I have were a little too large in diameter to fit. I decided to remagnetize the existing magnet. I had my home made remagnetizer cranked up to an insane amount of power and couldn't get the magnet very much stronger than it already was. When my wife asked what those bangs were I decided I'd pushed it far enough:whistle: Anyway, with the magnet strengthened I only had a small drop in current.

Also, the quiescent current (the current when pendulum is stopped and the circuit is idle) is nearly the same as the operating current. I'm guessing that I have a leaky component in the circuit. Probably the transistor.

I'm not going to attempt any repairs on it. This clock has the transistor mounted on the coil bobbin. Not easy to get it apart without messing up the brass coil cover. The coil on this clock has never been apart. The brass cover is in perfect shape.

I have a Schatz Lectronic that has the coil and electronics mounted in the base. I may do some tests on it.
 

enigmadan

Registered User
Apr 12, 2020
10
0
1
58
Country
An interesting thread here, much of which is beyond my understanding.
I bought a Kundo dome style at least ten years ago on ebay, put a battery in, and after many attempts to regulate it, it kept gaining time, so I left it alone, and there it sits, still running on the same C battery!
I live in the desert, and my house goes through wide swings in temperature through the days and nights, and seasons.
I wanted to see just how long it would run on the same battery, but it didn't dawn on me it may well be corroded and doing damage under there now, so I'm going to check it out.
Have to admit though, that's some impressive engineering to still get that kind of performance on a decades old device with moving parts.
 

Schatznut

NAWCC Member
Sep 26, 2020
506
214
43
SoCal
Country
Region
What is the battery life for a "C" cell in one of these. Reason I ask is that you don't want a battery that is so large that the battery begins rotting before it fails to provide sufficient current. I suspect that the current necessary to drive an interval coil may be relatively high and battery life may not be a concern. .
I have here a Toshiba wall hanger radio that uses 4, "D" cells. It is so overpowered that the batteries will last several years and will actually work fine on half voltage.
These movements consume on the order of 300 microwatts, so it is highly likely the battery will begin to leak before it can no longer provide the current to run the clock.
 

S_Owsley

NAWCC Member
Jan 24, 2011
291
74
28
67
San Francisco
Country
Region
I'm running my Kundo on a single 1.5v AA cell. It runs a bit slow with the pendulum adjustment weight all the way at the top. I have a guess that perhaps a single AA cell isn't providing enough amperage for the circuitry design? The electronic measurement talk I will admit is above my head, but my basic question is: would adding an additional AA battery in series potentially solve my slow running condition? (It loses a minute or two per day). Otherwise, it might be more fine tuning at the top spring lever that catches the "escape wheel"?
 

Schatznut

NAWCC Member
Sep 26, 2020
506
214
43
SoCal
Country
Region
I'm running my Kundo on a single 1.5v AA cell. It runs a bit slow with the pendulum adjustment weight all the way at the top. I have a guess that perhaps a single AA cell isn't providing enough amperage for the circuitry design? The electronic measurement talk I will admit is above my head, but my basic question is: would adding an additional AA battery in series potentially solve my slow running condition? (It loses a minute or two per day). Otherwise, it might be more fine tuning at the top spring lever that catches the "escape wheel"?
I'm running one of mine on a single AA also and it runs fine and keeps time well.

In an ideal pendulum, T = 2(pi)*sqrt(L/g) where T is the period, L is the length from the pivot to the center of mass of the pendulum and g is gravitational force. You can see the speed is totally a function of the length of the pendulum and not the impulse to the pendulum. These clocks are not ideal, however, in that there is a spring connecting the pendulum to the pivot point. The spring may be incorrect - too long or not quite rigid enough. If there isn't enough energy in the circuit, it will stop tripping the escapement. The pendulum will continue to oscillate but the hands will not advance. Adding a battery in series will not help. The adjustment on the top of the escapement lever is necessary to set the correct depth such that it doesn't hang up on the escapement wheel.
 
  • Like
Reactions: S_Owsley

Forum statistics

Threads
168,096
Messages
1,465,832
Members
48,075
Latest member
Derek
Encyclopedia Pages
1,060
Total wiki contributions
2,954
Last book
400 day pendulums by happycusa