Center Tapped and Regular Transformer

Hi,

Could you please explain the following points regarding a center tapped
transformer (some questions apply to a non-center tapped "regular"
transformer as well):

a. Is the center tapped transformer wound differently than a
non-center tapped transformer? Or is it just a regular transformer for
which the center point of the secondary winding is "brought outside".

b. Considering the secondary voltage of a transformer is Vs, the two
terminals of the secondary are at +Vs/2 and -Vs/2. This implies a
voltage gradient across the secondary. The gradient passes through a
zero point which we "tap". What causes this voltage gradient?

c. Can we say that all the turns in the secondary winding of a
transformer have the same amount of flux passing through them at a given
instant or do they have a different amount of flux (with the flux
depending on the position of the turn)?

Thanks,
Anand

 

I don't know the answer to these, and I'm interested to know, but they
are sound very much like homework questions and I wouldn't be surprised
if the answers you get are along the lines of "do your own homework".

Now, I would guess that Wikipedia and Google are going to help you in
your quest for information. Good luck.

 

"John Larkin"

** That would need to be a rather large ( multi kVA) transformer.

The "rule of thumb" for a laminated iron E-core is 5 to 6 turns per volt,
for 1 square inch core cross section.

( 5 turns if it is for 60 Hz only and 6 turns if it is for 50Hz )

If the core is a strip wound toroidal, then the numbers is reduce to 4 and 5
turns.

So, for 1 turn per volt, the cross section needs to be around 5 square
inches.


.... Phil

 

Daniel, John, Phil and Tim, thanks to each one of you for your reply.

Let me assure you, this isn't a homework question that I am trying to
push on to the newsgroup. (If it was homework, I wouldn't do it. )

I stumbled on these doubts while trying to develop a negative DC voltage
source. Amongst the other features in the requirement spec., one
requirement is that we would like to make this negative voltage source
as far as possible without using a center tapped transformer. As it so
"turns" out - its difficult.

What was irritating is that I could not find (printed or online) any
standard text which discusses these points:

a. A transformer with its primary terminals connected to Vp and
ground (zero volts) will have its secondary terminals as +Vs/2 and
-Vs/2. (For a small bit of time, I continued to think that the
secondary terminals would be: +Vs and 0V, mimicking the single ended
primary.)

b. If the flux crossing all the coils is the same (as in same
density), what causes the gradient in the voltage across the coil? You
see, the gradient on the coil is a change in the voltage magnitude and
polarity from one end to the other! I have concluded that this happens
because at one end, the flux is *entering* the coil and *leaving* at the
other. This *entering* and *leaving* causes the opposite polarity (some
"thumb rule"), but what about the drop in magnitude? Then again I
"imagine" that the voltage generated all across the coil is such that
the positive voltage is completely worn out by the time it reaches the
other end and vice versa for the negative voltage. The positive and
negative voltages meet at the mid-point, cancel each other and give us a
beautiful zero! But why?

I will continue to be worried that my understanding is a figment of
imagination until I see some text/formal/mathematical treatment which
confirms/denies all this.

I will be embarrassed to see any URLs which discuss these points
straight away. But then they are welcome anyway.

Thanks once again,
Anand

 

To "use a couple of bridge rectifiers.....", you are assuming that the
bridge rectifiers are galvanically isolated from each other. That is,
the source for positive voltage regulator and the source for the
negative voltage regulator should not come from the secondary of the
same transformer.

If you do "tie one of the bridges negative terminal...." when the source
of the positive and negative regulators are the same there is a problem.
For a full wave bridge rectifier, the capacitor charges to 2Vpeak (no
load condition) as in: +Vpeak on its positive terminal and -Vpeak on its
negative terminal. So connecting the positive of one bridge to the
negative of one bridge is not a solution.

You could simulate this or build this. Just measure the voltage between
the two points you want to tie as common before connecting them. You
will see a 2Vpeak voltage between them. (Obviously, you don't go ahead
and connect them.)

If using two separate windings or separate transformers is your
solution, then I agree. It is easy.

Thanks,
Anand

 

I don't see where I have assumed or stated that the secondary is
galvanically related to the primary. All that I have said is valid with
or because of the galvanic isolation.

Sorry I wasn't clear about this, the idea is to make a positive *and*
negative supply from the same source (as in from the same secondary
winding) without using a center tap. (In which case "calling the "+"
terminal ground doesn't help. And flipping the diode's orientation just
moves which side of the bridge the positive and negative voltage appears.)

Thanks,
Anand

 

I am sorry, I haven't been clear. We need to generate a positive *and*
negative voltage using the same transformer secondary without a center tap.

(I kind of started writing where I was irritated thinking about the
negative voltage.....)

 

But you have to examine the whole cost, and it may be simpler to go with
the centre tapped transformer.

You could use a second transformer. This works especially well if the
second voltage is relativley low current, you can use a much smaller
transformer.

Got to a switching supply, more complicated but with multiple voltages it
may work out cheaper.

Especially if the second voltag is low current, use a dc to dc converter,
ie connect to the positive supply an oscillator and rectify the output for
the lower current negative voltage.

Use a higher voltage supply and split the DC voltage in half. This works
well if the two voltages are the same but of different polarity, and works
best if the current is relatively low. You in effect are setting an
artificial ground at the output of the rectifier.

Michael

 

OK can I add more components to the none ceter tapped secondary?

Then make yourself (say) a 30V supply. Add a power opamp rail
splitter*, and tie the center point to ground. Viola... +/-15 Volt
supply.

George H.

*in it's simplist opamp is a buffer with non inverting input driven
from two resistors one tied to each rail of the power supply.

 

Don't forget interwinding capacitance. A 10 meg DMM will almost always
see something.

 

"Anand P. Paralkar"

** It is ridiculously simple to make equal "+" and "-" voltages from a
single winding.

The circuit is called a "full wave voltage doubler". Two diodes and two
filter caps, each pair wired as a half wave rectifier gives two DC rails of
opposite polarity from one winding.

http://electrapk.com/wp-content/uploads/2011/08/full-wave-voltage-doubler-1.jpg

The centre of the two caps is the zero volts point.


..... Phil

 

"Phil Allison" wrote in message

Agreed. Here is a simulation of a circuit that I actually built:
http://www.enginuitysystems.com/pix/120Sine-320DC_Doubler.png

Here's the ASCII file if you want to play with it:
http://www.enginuitysystems.com/pix/12V-300V_CT_Doubler.asc

Paul
www.muttleydog.com

 

sure that's easy..

use one leg of the secondary as the common and the other leg will have
2 diodes branching from it. Each diode will be such that you'll get a
+ and a - source...

THere is one big problem with this how ever, you need to use a cap
double or more than what you'd normally would to filter it on the DC
side. This is because you'll only get 30 hz per diode. Basically, you'll
have a gap on each side when the opposite side is conducting.

Of course, if your requirement for the - source is a low current type
that could be done with a added circuit that operates from a single
power source.

Jamie

 

Another way, that I saw in an Elektor project, is to use a full wave
rectifier for the main supply and use two half wave doublers hooked
to each end of the secondary. (Or you could call it a capacitivly
coupled full wave rectifier).


------|(-----+----|<---------+----- V-?
| |
+-->|--gnd +--|(--gnd
|
+-->|--gnd |
| |
------|(-----+----|<---------+


Mark Zenier
Googleproofaddress(account:mzenier provider:eskimo domain:com)

 

Hi Everybody,

Firstly, thanks a lot for your detailed replies. Although this
started-off as a thread on my doubts on the transformer, this thread got
(unintentionally) drawn to another topic. One that I was planning to
post here anyway.

The reason I said that getting both - the positive and negative source
from a transformer without a center tap is difficult is that I tried
what a lot of people have recommended here. I built a circuit that had
the two diodes connected to the transformer secondary. One diode had
its anode while the other had its cathode connected to the transformer
secondary. The other end of these diodes were connected to a capacitor
each. These capacitors had a common point which we could call the ground.

I was surprised to find that this ground actually drifted! When I
measured the voltage across the ends of the capacitors (the end
connected to the diodes), the voltage measured was constant. However,
when I measured the voltage across the ground and the other end of the
capacitors, I saw that this voltage changed. So the V+ source and V-
source with respect to the ground was not constant!

I don't know the exact reason what causes this drift. But as a remedy,
I put a resistor in parallel to each of the capacitors. (P. E. Schoen
has posted this).

This stopped the ground from drifting but I don't think one could use
this solution in a practical circuit. A resistor in parallel at the
output of a voltage source will not hold up in case of a heavy load (low
load resistance).

That's why it seemed difficult.

There are ofcourse many other circuits that have been suggested here.
Thanks a ton gentlemen.

Thanks,
Anand

 

For some reason that circuit just does not look appealing..

Not knowing the application it is kind of hard to come up with the
proper solution however....


Transformer Bridge
+-+-++---+--------+
-. ,+--------+ A A + | |
)|( +--++ | --- | |
)|( +--+(-+ --- | |
-' '+--------+ A A + | |
+-+-++ + +------------------+
| .-. | | |
| | | | + |
R1, R2 100K | | | | |\| Ilimit |
| '-' +-+|-\ ___ |
| | | >--|___|-+---+ Commom
| +-----+|+/ GND
| .-. |/+
| | | |
| | | |
| '+' |
+---+--------+

Jamie

 

Is that how "Fly Backs" got their name ?

Jamie

 

unbelievable.

Jamie

 

I think you got that backwards, bud!

Jamie

 

Please:

E = -L di/dt

Opposing the change that produces it.

This is .basics, so we need a little pedantry.