| |
G8MNY > TECH 25.08.10 23:05l 359 Lines 16966 Bytes #999 (0) @ WW
BID : 60608_GB7CIP
Read: GUEST
Subj: Understanding Transformers
Path: IZ3LSV<IW8PGT<ON0AR<GB7CIP
Sent: 100825/2153Z @:GB7CIP.#32.GBR.EU #:60608 [Caterham] $:60608_GB7CIP
From: G8MNY@GB7CIP.#32.GBR.EU
To : TECH@WW
By G8MNY (Update Dec 07)
(8 Bit ASCII graphics use code page 437 or 850, Terminal Font)
Although the principles have been known since Faraday, here are some practical
pointers. Transformers have the ability to change AC voltages (& hence
impedance match), isolate several different windings, saturate the core to
protect against large current spikes.
THE IDEAL TRANSFORMER
All the copper must be wrapped around all the iron, & all the iron wrapped
around all the copper, as this is not possible the most efficient shape for a
transformer would be a two piece chain, where one end of each link is thinned
down to passes through the hole in the other link.
__ _ ZDDDDDDDDDDD?
/' ~~=-~~ '\ 3 F ____ 3
3 .-~ 3 Practical 3 E 3 ZDADDAD?
] ( ^ Option used 3 R 3 / \
3 ~-_ 3 in Line & EHT 3 R 33 WINDINGS 3
\._..-=-.._./ Transformers 3 I 3 \ /
Copper Iron 3 T 3 @DBDDBDY
MOST EFFICIENT 3 E ~~~~ 3
@DDDDDDDDDDDY
Maximum efficiency happens when the copper losses equals the iron losses.
Typical power efficiency is about 95% for a modest transformer, with smaller
power ones (plug top) down to about 70% & a large substation sized 3 phase ones
up to 98%.
SYMBOL
DD? ZDD DD? ZDD
o )||( o )::(
)||( the DOT o )::(
)||( indicates )::(
DDY @DD same phase DDY @D
IRON CORE IRON DUST/FERRITE
The windings can The core can be tuned
have tapings with a movable part
EQUIVALENT CIRCUIT
As transformers are in the real world they have short comings, this equivalent
circuit demonstrates most of them. (for primary only!)
Copper Interwinding C
Resistance Stray C ZDDDD4CDDD?
DDDBDDDDRDDDDD(((()DDDDDDBDDDDDDDDDBDDDDBDDDADD? ZDADD
3 ===== )|| === 3 o )||( o Ideal
=== Leakage )|| ZDD4 Riron )||( Transformer
3C Inductance )|| _3_=== 3 )||(
DDDADDDDDDDDDDDDDDDDDDDDDADDDDDDDDDADDDDADDDBDDY @DBDD
winding Primary @DDDD4CDDDY
capacitance Inductance
The iron losses are often indicated with "Riron" like this, but it represents
both eddy current losses & BH curve losses, where the value is not linear with
voltage but a curved relationship with the magnetising current.
Flux __.
IRON 3 _.-~~/~ A MODERN 3 ..-DD
LAMINATE 3./' ./' FERRITE 3 //
./3 ./' 3 //
DDDDDD/DDEDD/DDDDDD +H DDDDDDDD/E/DDDDDDD +H
./' 3/' Amp Turns // 3
./' _./'3 // 3
_.:..-" 3-B --D'' 3-B
D~~
With ferrite the eddy current & BH losses are very small but have a sudden
limit, & the BH curve magnetic cycle area which determine the power loss per
cycle is also much smaller than for iron. (difficult to draw here).
The high frequency possibilities of ferrite means much smaller lighter
transformers. For most ferrite cores the power handling is proportional to
frequency up to about 250kHz. So a core handling 50W @ 25kHz can handle 500W @
250kHz, but can the copper & insulation!
CONVENTIONAL MAINS TYPES
These are made of layers on iron sheet with a central bobbin for the windings.
to reduce eddy currents..
Solid iron Laminated Welded Laminated Iron Dust
__
/|\ 3 3 | ::: RRR 000
| 3 3 | ::: ::: 000
| 3__3 \|/ ::: ::: 000
<--
High circulating Little eddy Still little Very little
eddy currents current in eddy currents eddy currents
flow around sides each sheet as as only welded as only tiny
of large cross little area & on one side. area per
section like a Insulated (oxide) particle.
shorted turn. Surface to next
Laminate.
ZDDD?
_3 3_ The weight of the transformer determines the energy
/ 3 I 3 \ through put for 1 cycle, so a 60Hz transformer is 20%
3 3 R 3 3 smaller & lighter than a 50Hz one of the same rating.
3 3 O 3 3 Ratings also depend on maximum safe enamelled wire
\_3 N 3_/Bobbin temperature & the dissipation. E.g. a fan cooled one
3 3 in a projector would catch fire without the fan!
@DDDY
The laminates are either E I or U T shaped to form the core, with a copper wire
wound bobbin placed on the middle leg. Note the size of the centre limb is
about 2x the outer part as it has to cope with 2x the magnetic flux. These
transformers are normally run close to the core saturation point, so increases
& voltages or a drop on supply frequency soon cause problems.
E.g. a 2x 110V primary 60Hz designed transformer run at 242V across the winding
on series & at 50Hz, would have the magnetic flux increased by 10% from the
voltage, & a further 20% from the frequency change, This could take the iron
beyond super-saturation & cause the inductance to drop may times, producing a
heavy magnetising load current & associated heavy heating of the core & copper
windings!
A heavily saturated cores cause high 3rd harmonic line currents & increased no
load powers in the transformer. But properly utilised & choke fed, a saturated
tuned transformer makes the heart of a good voltage regulator/line conditioner.
ZDDDDDDDD?ZD? ZDDDDDDDDDDDDDDD?
3 ZDDDDDDY3 3 @DDDDD? ZDDDDDY
The EI format is often 3 3 3 3 ZD? 3 3 ZD?
made with 2 x I shapes 3 @DDDDDD?3 3 3 3 3 3 3 3
cut out of 2 Es, so it 3 33 3 3 3 3 3 3 3
is efficient in 3 ZDDDDDDY3 3 3 3 @DDDY 3 3
material usage. 3 3 3 3 3 @DDDDDDDDDDDY 3
3 @DDDDDD?3 3 @DDDDDDDDDDDDDDDY
@DDDDDDDDY@DY Not Popular
Each lamination is oxidised on one side to produce a poor conductor to the next
lamination to reduce eddy currents. Also to minimise magnetic losses further
the laminations are normally interlaced so that the joining gaps are covered in
the adjacent layers.
To hold the transformer together you can get away with welds across the
laminates provided no electric loop is formed.
___________ ______________
3o o3 3 ______ 3
More conventional methods 3 ZDDDDD? 3 3 3Small 3 3
are to use bolt holes in the 3 3Large3 3 @D? 3Trans-3 ZDY
corners or centre of the 3 3 3 3 ZDY 3former3 @D?
laminates sides, but the 3Transformer3 3 3______3 3
hole area must not degrade 3 @DDDDDY 3 3______________3
the magnetic circuit. 3o o3
~~~~~~~~~~~
The bolts hold down 2 clamping plates to squeeze the laminations together, as
when magnetised the laminations will oppose each other & try to push the
laminations apart. The end plates can be fully enclosed to provide full
electrical & some magnetic shielding.
_____
_//~~~\\_ Outer Further magnetic shielding can be
^|~~~~~~~|]Copper proved by a copper strip around
^|_______|]Shorted the whole bobbin former, but
~\\___//~ Turn outside the laminations.
~~~~~
Transformer acoustic hum is made from the steel & copper movement due to the
magnetic forces, both at 2x the frequency (6x for 3 phase), but also from
magneto-striction where the steel parts actually shrink & expand as their
magnetic domains in each crystal of iron rotate, so clamping does not reduce
this noise!
Quite often the whole lot can be varnished dipped/impregnated to provide the
best low noise transformers & corrosion & dampness resistance. But repair is
normally impossible afterwards!
End plates or crimped folded steel tabs casing, provide mounting holes either
around frames or just 2 holes on Tabs.
_n_n_n_
____/_______\___
Side >]0 0 0 0^ _ /~~~~~~~\ _ ZDDDDDDD?
frame 3~~~~~~313~~~~~~3 Tab 3 3~~~~~~~~~~~3 3 ZDADDDDDDDAD?
holes 3 313 3 holes303Folded Case303 3 P C B 3
as 3______313______3 3_3___________3_3 @DBDDDDDDDBDY
well >]0 0 0 0^ \_______/ @RDRDRDRY
~~~\~~~~~~~/~~~
~u~u~u~
PCB mounts can be by soldering the connection pins moulded on to the plastic
bobbin former.
The bobbin can be either over wound with an insulating layer separating primary
& secondary, or a single bobbin with 2 wind areas, so the insulation integrity
is maintained even in a burnt out state (meets double insulation safety
standard).
____________
/ Secondary \ 3~~~~~~~~~~~~~~~~~~~~~~~~~3
/ __________ \ 3 _____ _____ 3
3 / \ 3 3 3 QMMMOMMMMMOMMMQ 3 3
3 3 ________ 3 3 3 3 3 Primary 3 3 3
3 3 3 3 3 3 3 3 3 3 3 3
3 3 3 CORE 3 3 3 3 3 XMMMMMMMMMMMMMX 3 3
3 3 3________3 3 3 3 3 3 Secondary 3 3 3
3 3 Primary 3 3 3 3 3 3 3 3
3 \__________/ 3 3 3 OMMMQMMMMMQMMMO 3 3
\ Insulation^ / 3 ~~~~~ CORE ~~~~~~ 3
\____________/ 3_________________________3
LAYERED BOBIN SPLIT BOBBIN
Another safety feature often included nowadays is a thermal fuse fitted inside
the transformer primary layer. These protect from fires & should not be
bypassed, replacements are available.
The enamelled copper wire windings must be wound fairly tight as like the
laminations there is forces between wires & any movement will eventually rub
through the enamelling & cause a shorted turn & an instant failure.
Connections
These can wired remotely or tags mounted on the bobbin etc.
TURNS PER VOLT (from Tom GM4PRO)
A blanket approximation of 7 turns per volt, even for the fairly small
transformer.
E = 4.44*f*B*A Volts
Where "E" is the e.m.f. per turn (NOT turns per volt),
"f" is the supply frequency (i.e. 50 Hz.),
"B" is the peak flux density of the core material (typically 1.5 Tesla
for cold rolled grain orientated silicon steel, as used in most
modern transformers),
"A" is the cross sectional area of the core in square metres.
For a typical small mains transformer, this can be reduced to:
Turns per volt = 4.65/A
Where "A" is the core sectional area in square inches.
TURNS RAIIO & MATCHING
For AF transformers in audio stages, an output transformers tapings sets the
load line saw by the output valves. Impedance ratio is just the winding ratio
squared. So a 20:1 transformer turns or voltage ratio gives a Z ratio 400:1
e.g. used on a 15j LS, the Anode sees a 6k load.
Or to match a 17j ribbon mic to 600j AF preamp input, that is a Z ratio of 36:1
or 6:1 turns ratio, so an AF transformer with 1mV input that gives 6mV output
may do.
UNCONVENTIONAL TYPES
In some applications other designs have reduced fields or offer low profile
etc.
L Shaped Core
These are magnetically inferior depending whether the primary & secondary
overlap on common bobbins or are on different legs.
ZDDDDDDDDD?
3 ___ 3
ZADDA? ZADDA? The core can be made from flat laminations or
3 3 3 3 from U curved sections where each lamination
3 3 3 3 is a different shape! In that case the magnetic
3 3 3 3 flux leakage is much reduced.
@BDDBY @BDDBY
3 ~~~ 3
@DDDDDDDDDY
Toroidal
These are very popular nowadays & offer a much lighter transformer as less
steel is used & they have quite magnetic low flux radiation.
The main problem is in winding them, as this has to be done once the core is
put together. The core is just very long a strip of laminated steel with an
oxide layer on one face to reduce eddy currents, wound on itself in spiral
fashion with an insulating layer on the outside.
.-~~~~-. The primary is wound first, as it is normally the thinner wire
/ .--. \ & more able to bend around the square cross section of the
( ( ) ) core. Very thick wire for a high current secondary can be
3\ '--' /3 difficult & either several thinner wires used, or strip used,
\'-,__.-'/ as the whole secondary has to be out on a winding machine's
'-.__.-' bobbin shuttle & has to pass through the centre of the toroid.
The cross section looks the same as a conventional layerd bobbin transformer.
Primary to secondary insulation failure is always a possibility with a
toroidal. So using an earth on kit using them is recommended!
Due to the reduced amount of steel & efficient use of copper they also suffer
"randomly" from very high fluxing up turn on currents, as a full half cycle at
turn on looks like "DC" & will keep the core saturated over several cycles,
greatly reducing the inductance. E.g. a 230W transformer, will take about 1A on
full load, including a very low 50mA magnetising current, but up to 13A surge
on turn on. So fusing MUST be a slow blow type.
Mounting them is a problem, as too high a clamping force can damage the
windings. & too little pressure will allow for transit movement & winding
damage "CATCH 22"! So a bolt clamp & plate with rubber mats are normally used.
Circular Shaped metal plate
rubber mat ======? 1 Z======= _ W A R N I N G
/~~~~\ \_[~]_/ /~~~~\ |\ Low AC Voltage @
3toroid3 1 3toroid3 VERY HIGH CURRENT
Circular 3 3 1 3 3 available over 1 turn
rubber mat \____/ 1 \____/ |/_ DO NOT SHORT
-----------==========1===========-----------
Chassis ___Bolt
OTHER TRANSFORMER USES
In audio work they are still used where... good isolation, balancing, phantom
powering, impedance matching are needed, well as provide good RFI protection.
Phantom supply
3 1:10 10mV 3\
Mic DDDDDDDDDDDDDDDDDDDDD)DD? ZDDDDDDDDDDDDDD4 >DDD
with 1V of RF & hum 3 )||( Z=10K > 3/
Preamp pickup on a 3__)||(
100j long screened )||( Magnetically AF MIXER
nominal balanced cable )||( Screened With no RFI
1mV DDDDDDDDDDDDDDDDDDDDDDDDY3 3 Case & no hum
-------------------------CDADDDDDDDDDDDDDD
Dirty Earth return _3_ RF Ground
There are costs of course, as it is difficult to proved a wide frequency range
needed for really super audio, but 40Hz - 15kHz @-1dB is quite achievable, with
>60dB common mode hum & RFI rejection.
RF TRANSFORMERS
These iron dust cores or shaped ferrite cores. Shapes can be EE or UU, pot
core, toroidal, or even just a rod.
ZDDDDBDDDD? _________ .--~~~~--.
3 ZDDADD? 3 3 3 / \ _.--._
3 3_____3 3 3 ZDDD? 3 ZDDDDDDDDDDDD? / \ _______
3 3 3 3 3 3 3 33~~~3 3~~~33 3 /~~\ 3 () )
3 __3__ 3 CDD4 CDD4 C4 CDD4 C4 ] 3 3 ^ ~~~~~~~
3 3 3 3 3 3 3 3 33___3 3___33 3 \__/ 3 rod type
3 @DDBDDY 3 3 @DDDY 3 @DDDDDDDDDDDDY \._ _./ (radiating!)
@DDDDADDDDY 3_________3 section through ~~
EE UU a pot core Toroidal
For switch mode applications at 10-100kHz the power capability can be enormous
compared to conventional mains frequency type of the same size.
Several stacked ferrite rings can be used to provide useful broadband HF power
output transformer.
.=====. O/P 1__ ____
O/P 1_____ //_ _\\ ____O/P 2 )::(
( | | ) ( | | ) +12V___)::( Aerial
( | | ) ( | | ) -6x rings )::( Secondary
( | | ) ( | | ) O/P 2__)::(____
+12V~~~~~~ \\~~~~~// ~~ 2x 1/2 turn 5 turns
Primary is '=====\\ Secondary
brass tubes inside \\ turns inside
rings connected the tubes
to end pates.
For higher frequencies & low power, bead types (miniature toroid), or double
hole versions give extended frequencies up to 1GHz. The 2 hole one offers
complex RF hybrid possibilities used in mixers etc.
__ _ __
((o) ( (oo)
~~ ~ ~~
Why don't U send an interesting bul?
73 De John, G8MNY @ GB7CIP
Read previous mail | Read next mail
| |
