[sci.electronics] TV to receiver: a different angle

phil@osiris.UUCP (Philip Kos) (05/05/88)

Okay, I know much of this issue was thrashed out Very Thoroughly not too
long ago, but I've got a slightly different situation from the (last)
original poster.

I just picked up a stereo TV yesterday (I didn't intend to buy a stereo
model but that's another story) which has RCA (phono) jacks on the back
for hooking up a pair of external speakers.  The manual warns you right
out to connect these jacks *only* to 8 ohm speakers, so I don't figure
that hooking them up directly to the AUX inputs on my receiver would be a
good way to go.  :-)

Anyway, I do have an idea for a relatively simple and straightforward
interconnection mechanism, and I'd like to see what y'all think about it.

Basically, the TV output drivers want to put a (relatively) high voltage
across a (relatively) low impedance.  I want to connect these drivers to
a (relatively) low voltage, (relatively) high impedance device.  The
obvious passive solution is to use an L-pad to attenuate the signal, then
an audio matching transformer to "raise" the impedance.

I can't think of anything wrong with this scheme (although I'm probably
missing something obvious...), and since it has the added benefit of
isolating the receiver from the TV with no extra trouble, I'd love to use
it if I can.

Assuming I'm wrong about something, what is it?  Tell me why this idea
won't work and win a thank you.  If there's nothing fundamentally wrong
with the idea, is it really as simple as it seems, and if so, what are
some ballpark figures for the amount of attenuation needed and the ratio
of the transformer?  (Assume 5W power output per channel into 8 ohms
nominal.)  I can tweak an L-pad to find the "right" amount of attenuation
and then build a fixed network so just being close would be enough; I'd
want to have a pretty good estimate for the impedance ratio, though...

Thanks to anyone who can help me out with this.  I'd like to ask that any
discussion be kept in both of the groups I've posted to, or at least
sci.electronics, because I don't get any rec groups anymore.  Also, I'll
keep (and maybe even summarize, if there's enough) private mail, so if
you don't want to go public that's fine with me.


                                                                 Phil Kos
...!decvax!decuac!\                                   Information Systems
  ...!uunet!mimsy!aplcen!osiris!phil           The Johns Hopkins Hospital
...!allegra!/                                               Baltimore, MD

bks@unisoft.UUCP (Brian K. Shiratsuki) (05/06/88)

In article <1590@osiris.UUCP> phil@osiris.UUCP (Philip Kos) writes:
>I just picked up a stereo TV yesterday (I didn't intend to buy a stereo
>model but that's another story) which has RCA (phono) jacks on the back
>for hooking up a pair of external speakers.  The manual warns you right
>out to connect these jacks *only* to 8 ohm speakers, so I don't figure
>that hooking them up directly to the AUX inputs on my receiver would be a
>good way to go.  :-)...
>[discussion of using L-pad and matching xformer to connect to aux in on
  preamp]
>...(Assume 5W power output per channel into 8 ohms
>nominal.)...

since the television has rca jacks on the back which are relatively
accessible, it would seem reasonable that neither side is connected
directly to the ac line, ala common ``hot chassis'' televisions.  so
isolation probably isn't an issue.  five watts into eight ohms comes to
2 sqrt(10), around 6 volts.  this seems like a reasonable signal to
feed directly into your preamp.  since the aux input probably precedes
the volume control, you can use that for your attenuation.

if you're worried about properly loading the amplifiers in your tv,
then you could connect 5W, 8 ohm resistors in parallel with the preamp's
inputs.

before you actually do this, you might measure the dc across the tv
outputs, to see if there's a significant amount of offset.  if there is,
then you should at least take the care to turn on the tv before turning
on the amplifier, or capacitor rather than direct couple.
-- 

				brian