Star Tech Journal

Issue: 1987-September - Vol 9 Issue 7

STAR*TECH Journal
Sapte ■ ber
1987 (10)
Inside Switching Regulators
(Part 5)
By Janes Beck
Greater southern Distributing Co.
Atlanta, Georgia
.FINM.fllDa
Figure 1 shows the control IC and its
connections to our power supply. Even
though it looks more canplicated than the
simple error anp/nooulator, that I
presented in the last installnent, it is
still the sane basic arrangenent.
If you will refer to the pin-out of the
TL949C that was printed in Part 4 (SirTJ,
July 87), you will see that pin 14 is the
reference voltage. This reference voltage
is canpared to the +5 volt sanple that is
applied to pin 1. (Please note that I did
not take advantage of the outp.it control
function on pin 13. Instead I have pulled
it high by connecting it to pin 14.)
'!he rest of the circuit is very straight
forward. C9 and Rll are used to set the
frequency of the sawtooth wave generator.
The voltage dividers forned from RS, R6
and R7, R8 are used to apply the reference
voltage to the error anp inp.its.
'!he unregulated power supply that T4, D5,
06, and Cl0 form is used to power the IC
and to provide a current source for the
drive transforners. T4 is also performing
another very important function; It is
maintaining the isolation between the AC
input and the regulated output. (Wouldn't
you know it; power supplies within power
supplies!) Fortunately the IC and drive
transformers use little current. That way
T4 and Cl0 don't need to be large.
IIDlLAR OIERVIEif
you have seen, the basic building
blocks of a pulse width nooulated suwly
are generally straight forward. These are
power handling, rectification and control.
Even though there are areas of overlap, I
have found using this nooular view during
troubleshooting helps to make things
easier.
As
If you will pull the copy of STAR""Iml
JOORNAL (May 87) that has the other half
of our schematic, we can recap sone
troub 1.e shooting that was discussed in Part •
2 (SirTJ, April 87) and also cover sone new
ground .
PCllm &-m-tt:M
you might guess, the most comnon
problem area is the power handling system.
It is the nuscle of our supply, and takes
a lot of stress. If you have a "dead"
supply, and the fuse is blown, you can
just about be assured that there is sone
problem with this section.
As
I, at one tine, \\Uuld first replace the
fuse•to see if maybe it was just one of
those fuses that was destined to blow.
After wasting a couple of boxes of~
fuses, I came to a conclusion; If the fuse
has blown in a switching regulator, there
is probably a good reason for it.
CBlDt 'DIE CAPS
In many cases, the problem is not even a
semiconductor failure. I have found many
bad supplies with shorted capacitors in A
the noise filter. Oleck these first. It's •
an inexpensive quick fix. If those check
OK, then move on to other causes.
SWI'lUIERS
switching transistQr failure is a comoon
problem, but don't just replace the failed
part and power the supply up. There might
have been a reason for it to go. Oleck its
associated despiking diode. It may have
been the original problem that led to the
destruction of the transistor.
SJPH.Y
Of course you can't rule out the bridge
rectifier or the filter cap. Although I
have never had a filter cap short, and
have only seen two bridges short out, and
one of them had a good excuse; It seems a
switching transistor had shorted, and
someone had replaced the fuse with one of
those 200 anp no-blow types!
A problem with the rectifier section, or
an excessive load, can cause all kinds of
havoc. A shorted diode or capacitor here a
may cause the switching transistors to 9
fail. At best, the supply will appear to
be conpletely dead. Renenber, check toose
diodes and capacitors in the outp.it
STAR*TECH Journal
September 1987 (11)

section! If the supply has multiple output
voltages, check those sections also. It is
better to check too many things than not
enough. Once u:p:m a time, a shorted -5
volt regulator cost me a couple more
switching transistors than it should have.
(OK, maybe more than a couple.)
A cornnon problem with the rectifier system
is leaky or dried out capacitors. Even
switch mode power supplies get hot under
heavy loads, and this can cause the filter
caps to dry out. This results in a noisy
output that cah cause problems that seem
to be logic related. The only effective
way to check this is with an O'scope.
Although I have found some supplies will
emit an audible "squeal" during operation
\'.hen the filter caps are going bad.
cnma.
Fortunately, there are very few problems
with the control logic. If you have
checked out "everything" and the supply
refuses to work, you might try looking at
the power supply for the control IC. If it
is OK, look for a resistor that is way out
of tolerance. I have yet to see an IC
failure, not that it doesn't happen. It
just hasn't happened to me, yet.
IN CDD:DSICE
I do believe that should wrap the whole
thing up. Just remerrber -think of cause
and effect, break the supply up into
digestable chunks, and have at it. Sure
you're going to run into the one that just
flat out refuses to return from the dead,
but don't get discouraged. It is the forty
dollar supply that had a ten cent part go
out that makes it all worthwhile .

- .
-----IC1
--------~
,L'-\C\L\
----1-8

Download Page 10: PDF File | Image

Download Page 11 PDF File | Image

Future scanning projects are planned by the International Arcade Museum Library (IAML).

Pro Tip: You can flip pages on the issue easily by using the left and right arrow keys on your keyboard.