# Simple power solved I hope.

Discussion in 'Electronic Basics' started by Scott Wiper, May 25, 2004.

1. ### Scott WiperGuest

This curcuit apparently does not need inductors for C2, C7 and C8.
Apparently ZI went to Carleton Univeristy and asked a graduate student who
is specialising in switching type power supplys. He said to me that a series
inductor for the above caps is not required because the final transitor I
will be useing a power darlington pair known as a 2N6284. This apllies to Q2
and Q3 and well. The 2N6284 has built in base/emitter shunt resistors shown
on Q5. This should absorb spaikes on C2 and C7/8 in This schmatic. The
2N6284 has a gain of typical 2000.

Here is the datasheet:
http://www.onsemi.com/pub/Collateral/2N6284-D.PDF

I have dicaded to use this transistor because it's high current gain. I am
building the and thanks to the Univerity I can test with a scope to see if
their is a spike on C2 # 120/50%Duty cycle. I am hoping this Transistor can
switch and the frequency I have set between 120 and 144 HZ. I have resplaced
the two pulse drive Transitors with TIP41C meduim power transistor that are
NPN so the 2N6284 will switch at 180 degrees out of phase with the Q outputs
on U2A and U2B.

Should I place a varible inductor in series with the voltage stablising caps
if I exeed 1.5A at 13.8VDC and 1.5A at 5.3VDC to remove spaikes or ripple
during the switching? I have inductors that can be adjusted from 60mH 360mH.

Note I will only 1.5A at 12VDC (Panel lights and fans) and 2.5A at 5.3VDC
(LED indicators and all logic IC's including total hours under power.)

Here is the schmatic: (This image is non-interlased it should be clearer to
view.)

http://www-travel-net.com/~swiper/switch.gif

Express pcb sheeet file

http://www-travel-net.com/~swiper/switch.sch

2. ### Fritz SchlunderGuest

Greetings Scott.

Keep that schematic (which BTW the posted URL is slightly off in that it
should be http://www.travel-net.com/~swiper/switch.gif instead of with that
hyphen) in a safe place for years to come. If you stick with electronics
and continue to learn about it, someday you'll hopefully find yourself
digging through some old stuff and find that schematic. You'll take a quick
look at it and instantly be filled with nostalgia as a grin comes to your
face. I know I've had this happen to me when I look at some of my first

Meanwhile we don't live in the future or the past, but rather the present.
Therefore it might be more useful to you for me to provide comments on the
schematic itself.

If you hook the circuit up to power as drawn, you will be disappointed to
find that it does absolutely nothing useful or spectacular. If you place a
meter across C1 you will likely see that indeed it gets charged to
approximately the right voltage, but you will be disappointed to find that
nothing else in the circuit seems to do anything or have any power.

You appear to have a fundamental misunderstanding of how PNP transistors in
general work. The combinations of components Q1, D9 (one example set) and
Q6, D10 (another example set) and Q4, D5, R8, R9 (another example set) are
arranged in such ways that they serve no useful purpose. I can see what
your intended purposes are, but they do not function as you think they
should. A PNP transistor only turns on when current is sunk from the base
of the PNP (the emitter supplies all the current for both the base and the
collector). D9 and D10 prevent any current from ever flowing out of the
base of the PNP transistors, so they will always remain off.

LED D6 will never turn on as wired, but if you reconnect the anode of D5 to
go to ground instead of +12V you will find that it does perform some
function although not necessarily quite your desired one.

I'm not sure what your overall objective is exactly with this schematic, but
it appears you are trying to build a high efficiency switch mode powersupply
which charges B1 as well as provides nice clean 12V and 5.5V outputs. What
you have concocted unfortunately won't provide high efficiency by any means
even if you fix the above mentioned problems with the PNP transistors (and
any other problems I didn't immediately notice that may be in there
somewhere). That is, Q3 will most certainly get very hot regardless of R4
when under load. What you have concocted has all the disadvantages of a
switch mode PSU while maintaining all of the disadvantages of a linear
regulator (namely low efficiency).

To understand why that is you need to read some books on switch mode
powersupplies. Focus your research on the buck mode topology since that is
the most relevant to you right now. Don't bother asking people at
universities anything. Although you could get lucky and find diamonds in
the rough, the probability is slim. For the most part neither the students
(including graduate) nor the professors know anything at all that is
directly practical. Don't get me wrong, some of the stuff they teach in an
electrical engineering program can come in handy for a practicing electrical
engineer, but none of it is directly practical (which is what you need for
manufacturer's application notes.

Despite what the graduate student may have said, all switch mode
powersupplies (with the exception of capacitor based charge pumps,
piezoelectric transformer based converters, or other similarly exotic
circuits) that expect to exceed the efficiency of a linear regulator must
include at least one inductive device.

That said, good luck Scott. If at first you don't succeed don't be
discouraged just keep at it. When a component fails don't replace it
without closely scrutinizing the design and trying to find the possible
cause first.

End Transmission (but no need to beam me up Scotty).

3. ### ~^Johnny^~Guest

It might help to type the link like this:

http://www.travel-net.com/~swiper/switch.gif
--
-john
wide-open at throttle dot info

~~~~~~~~
"The first step in intelligent tinkering is to
save all the parts." - Aldo Leopold
~~~~~~~~