Resistors at the inputs and outputs of ICs ???
Sep 13, 2003 at 4:57 PM Thread Starter Post #1 of 7

00940

Headphoneus Supremus
Joined
Nov 6, 2002
Posts
4,493
Likes
47
I was searching for a buffer to drive some IC and felt on this :

buffer.gif


What seems mostly strange to me is R10 and R12. They have to be put like this :

bufcab.gif


Why is it done so ?

Website : http://www.pha.inecnet.cz/macura/buffer_en.html
 
Sep 13, 2003 at 5:31 PM Post #2 of 7
It's a voltage divider. Two things it might be good for is:

By putting out double the voltage from the buffer and splitting it at the source any noise picked up in the cable is halved.

Input impedance of most amps is very high, by doing this you can omit the effect of variations in that impedance since the amp is parallelled with 50Ohms.

In combination with a common mode choke this might be very favorable in noisy conditions.
 
Sep 13, 2003 at 7:14 PM Post #4 of 7
Since the designer has chosen values of R10 and R12 that corresponds to the impedance of the cable it sound like a good idea. Resistors are cheap so I'd suggest you try both. Impedance is dependant on the frequency so bigger resistors will lessen the effect of the cables frequency dependent attenuation. Some one else can probably supply you with the formula but the effect is probably pretty small, atleast in theory.
 
Sep 14, 2003 at 11:11 AM Post #5 of 7
Yeah, it's a divider, but more importantly they're matching the output impedance of the buffer to the input impedance of whatever is being driven. This gives you optimum power transfer over all else.

Why one would want to do this for audio is a mystery, though...it's mainly an RF trick, to keep noise low and avoid reflections. That doesn't matter in audio. The only time I can think that this sort of thing matters is for the "120 ohm output impedance" crowd, and in that case you don't have a resistor across the far side of the cable, you have a pair of headphone drivers.

I advise you to try it without adding output impedances and see how that works. Adding a little bit of impedance inside the loop may be necessary in some cases, but that's an entirely different deal.
 
Sep 14, 2003 at 12:27 PM Post #6 of 7
The main purpose of this driver is to drive long IC, as sometimes needed in a biamp system. Those IC could get some noise.

With some gain (the seller of the pcb advices up to a gain of 10 without loss) and a pot it can also be used as a preamp and headamp. I'm considering building it inside a cdp (those cabinets are awfully empty) with a pot and a variable gain switch. So I wouldn't need a preamp and would also have a decent headphones out for the lower impedance cans my old RKV cannot drive.
 
Sep 14, 2003 at 7:45 PM Post #7 of 7
I think this cable debate is more on philosofical level than technical. Do we talk longer cables than 10 meters/30 feet here? If no, this reflexion thing is a non-problem really.

It's cool to have impedance matched cables but hardly necessary.

I don't believe for a minute that the voltage divider is any good. Better to have good 75 ohms and the needed gain. If you have noise problems there are other solutions.
 

Users who are viewing this thread

Back
Top