# Transformer coil orientation

Status
Not open for further replies.

#### polashd

##### Member
In the attachments there are two circuits (circuit 01 & circuit 02) which I intent to charge a photo-flash capacitor for zapping Nicad batteries.

I’m confused about the transformer connection to other parts.

In circuit 01 all three coils are winded in the same direction (using right hand technique).
-L1 starts from pin 3 – ends at pin 2
- L2 starts from pin 4 – ends at pin 1
- L3 starts from pin 6 – ends at pin 5

In the diagram I labeled transformer points as A B C D & E
I need help to connect the transformer (ie. Which pin to connect to which point, like pin 2 = point C ….).

In circuit 02 there are 2 coils.
-L1 starts from pin 2 – ends at pin 1
- L2 starts from pin 4 – ends at pin 3
I connected the pins like the diagram.
Not yet tested the circuit to be sure that the transformer coil orientation is correct!

I got both the circuits from internet and added the feedback / control system myself.
Is the feedback system correct?

#### MrAl

##### Well-Known Member
Hi,

Schematically, the dot shows the orientation of the polarity. So if you have two coils L1 and L2 and you apply a positive signal to L1 at the dot terminal you will get a positive signal at the dot of L2.

Physically, if you wind two coils that both have the same length of wire on the same magnetic core by laying one wire next to the other and winding in the same direction so you end up with two windings with close spaced wires, if you connect the start of one of those wires to the finish of the second wire the voltage across the two loose ends will be double of what the voltage is across each individual coil. That means that the voltages added and that means that both starts are positive in polarity so they would both be indicated with dots.

So schematically the dots have the same polarity, and physically the starts have the same polarity if the coils are wound in the same direction on the core, so the translation from the physical to the schematic is such that the two starts both have dots and the two finishes have no dots.

It's always a good idea though to test unknown structures with a small test signal to make sure how the windings are orientated. If you connect the start of one winding to the finish of the other winding and you get a reduction in voltage then they are not 'dotted' correctly. If you see an increase in voltage then they are dotted correctly. That's because if you connect two starts together the voltages subtract, but one start to one finish and the voltages add.
Again that assumes they are both wound in the same direction. If not, then the start of one will actually act like a finish.

BTW the end with the first turn laid is usually called the 'start' and the end with the last turn laid is called the 'finish'. I suppose there could be variations on this though.

Last edited:

#### audioguru

##### Well-Known Member
I have not used Ni-Cad battery cells for many years. The cells that were discharged for a long time or over-charged for a long time became shorted and needed to be zapped but then they soon shorted again and needed zapping frequently. Ni-MH cells don't do that and do not need zapping.

You do not need a high voltage, you need high current. I used 12VDC and charged a 1000uF capacitor. The capacitor produces the high current that burns away the short in the Ni-Cad cell.

#### dr pepper

##### Well-Known Member
I concur.
Even if I wind a trans myself I check it with the sig gen, place a signal on the primary and trace the primary and secondary with a 'scope, you can see if the 2 waveforms are in phase or not, you can then put the dots directly on the trans.
Certainly for my projects the dot doesnt have to be the start of the winding, so long as all the windings are marked up the same way, phase is more important than actual winding direction, unless your building a dc electromagnet or a linearity choke.

#### spec

##### Well-Known Member
In the attachments there are two circuits (circuit 01 & circuit 02) which I intent to charge a photo-flash capacitor for zapping Nicad batteries.

I’m confused about the transformer connection to other parts.

In circuit 01 all three coils are winded in the same direction (using right hand technique).
-L1 starts from pin 3 – ends at pin 2
- L2 starts from pin 4 – ends at pin 1
- L3 starts from pin 6 – ends at pin 5
View attachment 100839
In the diagram I labeled transformer points as A B C D & E
I need help to connect the transformer (ie. Which pin to connect to which point, like pin 2 = point C ….).

In circuit 02 there are 2 coils.
-L1 starts from pin 2 – ends at pin 1
- L2 starts from pin 4 – ends at pin 3
I connected the pins like the diagram.View attachment 100840
Not yet tested the circuit to be sure that the transformer coil orientation is correct!

I got both the circuits from internet and added the feedback / control system myself.
Is the feedback system correct?

Where are you from: care to put it in your user page beside 'Location' so that it shows in the window on the left of your posts.

This is a widely used relaxation oscillator which is very tolerant of component values.

I assume that 1000μ and 30μ refer to the number of turns on the transformer.

The connections that you have shown for the transformer are correct, providing pin 2 and pin 4 are the same phase. It does not matter if they are both starts or finishes. But the ratio of the collector turns to the feedback turns should be around 1:3, not 30:1000 (1:33), which means that the inverter will probably not start with a 1.5V battery.

The inverter that you show, if it started, would probably produce about 15V across the capacitor, but it is impossible to be definitive because the characteristics of the transformer are not given. Can you say what voltage you require across the capacitor?

Normally, for a high voltage output, a separate winding of about 15 turns to 30 turns would be required for the collector of the driver transistor and the rectified output would be taken from the 1,000 turn winding.

The feedback circuit you have designed in quite clever, but the ratio of the feed back resistors, R2, R3, in not correct; it is way too high and would stabilize at 257 volts. R5 serves no purpose and can be removed.

Suggest make R2, 5.6K then the output voltage can be calculated by, 0.6V + (100uA * R3). For example if you wanted an output voltage of 10.6V, R2 would need to be 100K.

By the way, the feedback circuit would not be very accurate, but probably good enough for your job.

In addition the feedback circuit may simply stop the oscillator from working once the target voltage on the capacitor has been reached. Some form of hysteresis may be required to jolt the oscillator back to life. But maybe you would want the oscillator to stop once the target voltage has been reached.

One final point: I do not think that a 470uF capacitor would be able to provide sufficient current to reliably blast away the crystals from a dead NiCad cell: 10,000 uF capacitor charged up to 12V, 24V, and 36V (selectable) would be nearer the mark. But, of course, the amount of current required to blast the crystals depends on the size of the NiCad cell. And the voltage required would depend on the degree of cystalization.

In my days of recovering old NiCad batteries, I used to flash them across 12V automobile batteries in series, but I would not recommend that you do this because, if you are not careful, the NiCad battery can explode.

spec

http://batteryuniversity.com/learn/article/how_to_restore_nickel_based_batteries

Last edited:

#### polashd

##### Member
Can you say what voltage you require across the capacitor?

One final point: I do not think that a 470uF capacitor would be able to provide sufficient current to reliably blast away the crystals from on a dead NiCad cell: 10,000 uF would be nearer the mark. But, of course, the amount of current required to blast the crystals depends on the size of the NiCad cell.

In my days of recovering old NiCad batteries, I used to flash them across a 12V automobile battery, but I would not recommend that you do this because, if you are not careful, the NiCad battery can explode.

spec

http://batteryuniversity.com/learn/article/how_to_restore_nickel_based_batteries
I want to charge the cap at around 200v. Instead of high capacitance(10,000 uF) and low voltage(12v) I'm using low capacitance(450 uF photo flash) and high voltage(~200v) as both the methods are found in the net.

I've tested circuit 02 and its not working. when I connect to the battery I get few hundred mv (around 400mv) at the cap, then it starts to drop. It seems that the circuit is not oscillating, the initial current flow gives that output or something else is happening. I don't know why?

#### audioguru

##### Well-Known Member
In my days of recovering old NiCad batteries, I used to flash them across a 12V automobile battery, but I would not recommend that you do this because, if you are not careful, the NiCad battery can explode.
OMG! You want a short duration high current to burn out the crystal that has penetrated the internal insulation, not to blow up the battery cell. The high current might weld the batteries together then the results will be spectacular.

#### dr pepper

##### Well-Known Member
Have you tried reversing the connections on L1?

#### spec

##### Well-Known Member
OMG! You want a short duration high current to burn out the crystal that has penetrated the internal insulation, not to blow up the battery cell. The high current might weld the batteries together then the results will be spectacular.
A shot burst of high current is exactly what you get. Perhaps you are mislead by the term flashing : just swipe the battery terminal with the end of a wire. It works a treat.

spec

#### spec

##### Well-Known Member
I want to charge the cap at around 200v. Instead of high capacitance(10,000 uF) and low voltage(12v) I'm using low capacitance(450 uF photo flash) and high voltage(~200v) as both the methods are found in the net.
OK, but that is not optimum.

I've tested circuit 02 and its not working. when I connect to the battery I get few hundred mv (around 400mv) at the cap, then it starts to drop. It seems that the circuit is not oscillating, the initial current flow gives that output or something else is happening. I don't know why?
As I said in post #5, I would not expect your circuit 02 to work because of the 30:1000 feedback turns ratio: should be more like 30:15.

I have to go and do some real work now, but will probably get back to this topic in about five hours.

spec

#### polashd

##### Member
Have you tried reversing the connections on L1?
I did this with L2, the output was even lower.

#### spec

##### Well-Known Member
Hy polashd,

The circuit below should work. It uses the transformer of your circuit 01. The transformer pins shown in the schematic do not necessarily match the pin numbering of your transformer. What counts is the relationship between the starts or finishes of the windings.

spec

NOTES
(1) As ChrisP58 states in post #14, μ in the original post refers to μ Henrys rather than turns as shown on the above schematic

Last edited:

#### MrAl

##### Well-Known Member
I concur.
Even if I wind a trans myself I check it with the sig gen, place a signal on the primary and trace the primary and secondary with a 'scope, you can see if the 2 waveforms are in phase or not, you can then put the dots directly on the trans.
Certainly for my projects the dot doesnt have to be the start of the winding, so long as all the windings are marked up the same way, phase is more important than actual winding direction, unless your building a dc electromagnet or a linearity choke.
Hi,

Yes you can dot the two finishes or if you wind one the other way then you have to dot one start and one finish. It's just easier to explain with dots on the starts.

#### ChrisP58

##### Well-Known Member

I assume that 1000μ and 30μ refer to the number of turns on the transformer.

spec
It looks like he's using LTSPICE as a schematic capture/simulator.

LTSPICE uses the inductance values of a coupled inductors to model a transformer. So the numbers 1000u and 30u will be 1000 μHeneries and 30 μHeneries.

#### spec

##### Well-Known Member
It looks like he's using LTSPICE as a schematic capture/simulator.

LTSPICE uses the inductance values of a coupled inductors to model a transformer. So the numbers 1000u and 30u will be 1000 μHeneries and 30 μHeneries.
Thx ChrisP58: you learn something new every day.

spec

Status
Not open for further replies.