Headphone/Mic splitter, dual input

All,

I have a project I've been working on on and off for the last year or so, and I can't seem to figure out what's up. I'll admit right now, I'm a bit of a hack when it comes to this stuff, but I find it extremely entertaining/fun.

I play computer games with a friend, and we talk on the phone at the same time (rather than using Skype, or the like). I want to be able to wear my nice headphones, and use a mic as well, and be able to hear the phone conversation, and the computer through my headphones.

I've designed a schematic using the info on this page: http://rane.com/note109.html by combining his design for a combiner, and a stereo to mono combiner linked together.

My design is attached.

So, there are two issues. The big one at this point is that the person on the other end of the phone convo gets a bit of echo of their voice back. From talking to someone else, sounds like some caps (22µF or so) in the circuit would help with this. Not sure where they would go. Maybe on the computer input? Any thoughts?

The second is that, the computer is much louder than the phone even when the computer is turned way down, and the phone way up. Guessing this is just because the computer can drive far more than the phone. Any thoughts on how to adjust this? I was thinking something like a stereo potentiometer (like this: http://www.parts-express.com/images/products/standard/023-646_s.jpg) that adjusts the balance between the two sources, but not sure how well that would work.

Anyway, any help would be greatly appreciated. I'd really like to get this built. My initial breadboarding has been very promising, I just want to work out these couple bugs.

Thanks!

The right channel output of your IPhone is connected directly to the right channel output of the computer's sound card. Probably not the way to go. If you're following the concept in the rane.com circuit, you should attach 470 ohm resistors to both IPhone channel outputs and both computer channel outputs (4 resistors); then you'd attach a 10k resistor to ground from each headphone channel output. In the schematic, rane is showing balanced input jacks which you're not using so the whole top rail is actually connected to ground if you were to plug in a standard cable.

Also, is your mic a stereo mic? You could consider just plugging that, or one channel of it, directly into your IPhone's mic jack.

On the pot, you'll have to experiment and see how it actually sounds with your devices, but you could replace each pair of 470 ohm resistors with a single pot channel... like this: Left(White) --> POT edge --> POT wiper --> POT edge --> Left(green), and hook the POT wiper terminal to the node where the two 470 ohm resistors currently connect. A 1k pot would give you two 500 ohm resistors when in its center position. You could also consider adding a small fixed resistor in series with each computer output if the computer is always much louder than the phone. Just some ideas to try!

okay, that makes sense. I was going by the unbalanced combiner schematic, rather than the balanced. But that also answers some questions about why the right side is louder than the left in the headphones. =P The 10k's to ground might also help with the echo, maybe. I was wondering if the echo was basically coming from the common ground... =/ I know one design I found for something similar had some caps to, I guess, filter that echo, but I can't see where the echo is actually occurring. =/

I don't understand the Mic thing. It is a stereo mic, which is why I was using the stereo to mono sum-er design, since the iphone doesn't like stereo mic's.

With the computer being louder, I'd guess I'd just have to experiment with resistor values until it was more reasonable? Guess that's the advantage of breadboarding!

Thanks for the input. I'm going to pick up some more 470's tonight and re-layout this and see what I can get. Wish they made breadboard compatible stereo 1/8" jacks. =)

staze said:

Wish they made breadboard compatible stereo 1/8" jacks. =)

Click to expand...

They do! Buy some of the 22 gauge hookup wire and solder three pieces to the jack and strip the other ends.

I'm not sure about the echo problem. It implies the audio signal from your phone output is somehow getting back to your phone input. Usually a capacitor is used to route signals from some node to ground but if the signals are already on the ground itself...? Make sure all the ground connections to the devices are solid though.

yup, that's what I do! I just wish it was a simple plugin. =P

And yeah, that's what I thought... I can't figure out where it's looping back. I thought caps were for smoothing signals, and removing AC, which confuses the hell out of me in this case.

The one "similar" system I've found is this: http://awaitinginspiration.com/blog/2009/11/iphone-call-recorder-v2 for recording calls, and he has caps on the recording end because that's where he was seeing the echo created, but from talking to him, it seems like this still didn't fix it completely.

If there's errant signal on the ground, would installing something like a 10k-20k resistor on iphone ground connection do anything? Is it generally bad to have any resistance on ground?

Let me update my schematic with your suggestions and post it to confirm it's what you're saying.

Standby...

Actually, if you google it, a lot of people complain about echo on an iPhone when simply using headphones or the "apple headset." It may not even be related to your circuit.

And no, you don't want to add resistors between any of your grounds and the rest of your circuit. All your devices' grounds should be connected together with as low an impedance as possible.

Maybe experiment with just headphones and see if you can repro the issue and also see if it's related to the volume level of the device.

Okay, updated. Though, seems like if I just had the "Left (Green)" input from the iphone connect to the lines from the computer to the headphones before the 470Ω, I could remove a resistor requirement, but for now, I've got a separate one in there. Or should I move the "Right (Red)" from the iphone to the right, and add another 470Ω to it?

Anyway, any thoughts would be awesome.

As for the echo, really makes me wonder how Apple avoids this.

I hope you're not on a tight budget where the extra 5 cents will make or break you! I don't entirely follow your description of the circuit mod, but the left channel is correct and the right channel is incorrect. You still have a direct connection between your iphone output and computer output on the right channel. The resistors should be connected directly to each of the four incoming channels and then should connect together in pairs on their other end, just like you did with the left channel. Also, you can remove the 10k resistor that goes across your output channels. Then you should have a perfect match to the rane circuit!

Okay, how about this!

staze said:

Okay, how about this!

Click to expand...

Perfection!

Cool. Just re-breadboarded, and couldn't get the mic to work for the life of me. Tried putting a 4.7kΩ in the line, and nothing.

On a whim, I disconnected the right side of the stereo mic, since I know the iPhone only sees mono anyway, and viola, it worked. No idea why... thought that's what the summing part was for.

Anyway, computer is still quite a bit loader than the phone, so I'll try ramping up the resistors on the computer input and see what I get.

On the plus side, no echo that we can hear, which is awesome. =)

Anyway, updated schematic attached. Let me know what you think. I guess for the loudness, I'd just adjust the computer input 470Ω resistors? Just for a test, I just "doubled" them to 1kΩ and that seems to have helped. I guess a simple Pot might be better, but I'll mess with that later.

Thanks for all the help!!!

so, just for S's and G's, I just replaced the 1k's with 4.7k's, and it's even better. Damn human ears.

So, my question, I guess, is there any threat of burning something out on the computer by putting the impedance too high? So right now, measuring with my meter, it's about 4.62kΩ (meter actually says 4.6211k-4.6225k)* from the computer input, to the headphone jack, on both left and right channels.

Thanks!

*Yes, I just picked up a frickin awesome Fluke 189 on ebay a couple weeks ago. =) It's the big reason I'm working on this project again.

Sounds like you're close to interactive video gaming bliss then!

As far as _input_ impedance (the impedance your circuit provides at its input to other devices' outputs), the higher it is, the less current the source has to provide and therefore the less the source has to work. So there's no danger whatsoever in going too high, you just lose more signal amplitude.

If you're just going to use one mic channel, you may just want to do a direct connection and remove the resistors. You're just hooking up the mic to the jack at that point.

hmm... I just wonder if removing the resistors would make the iphone not see the mic again. But I could try.

Would I still want the 20kΩ resistor (or rather, 2x10kΩ resistors) between the mic and ground?

So, okay, I'm still confused on the iPhone not seeing the Mic when it was in stereo through the suming circuit since the impedance seems to be high enough. But hell, if it works, it works.

Logically it makes sense about the input impedance since that's all a rheostat is, is a variable resistor. I'll make sure to post back once I have this re-soldered up to let you know if it works, but at this point, I'm pretty happy.

Version 3 will just add a stereo POT I think to adjust the balance between inputs (if I can manage to make that work). I'd think just each input on the ends, then the output on the wiper. I would THINK that would work. But I'll breadboard that tonight if I get the other soldering done.

Thanks very much for all your help. Out of curiosity, did you just learn from doing this stuff (which is pretty much the answer I give when people ask me about computers), or did you read any good books that you might suggest?

Thanks!

staze said:

Version 3 will just add a stereo POT I think to adjust the balance between inputs (if I can manage to make that work). I'd think just each input on the ends, then the output on the wiper. I would THINK that would work.

Click to expand...

Yeah, that will work for sure. If you want to make it a bit lopsided there are ways to do that, so that your computer is always somewhat 'turned down' compared to the phone but where you can still adjust the balance. If you think of the pot as two entirely separate resistors with an arbitrary connection point in between I bet you'll figure it out.

Just make sure you get the basics down, too, Ohm's Law, all the primary components, resistors, transistors, diodes, capacitors, and what they do. It makes things much easier. I got an older edition of this gem from abebooks.com for like $3, "A Practical Introduction to Electronic Circuits" by Martin Jones and, for audio, Douglas Self's "Small Signal Audio Design" answers a world of questions. If you want to get into active rather than pure passive circuitry for these same kinds of circuits, you should check out "Op Amps for Everyone," whose author also frequents this board -- it's a must-read.

Yeah, figure if I wanted to make it a direct 50/50 in the middle, I'd just add some resistors to the computer side. I guess my question would be, there should be some way to actually deduce this. Like, measure the current output from the computer sound output, and the iphone sound output, and use resistors on each side to actually get close to matching the inputs. Just put the meter inline between the input and the rest of the circuit and measure AC mA (isn't analog audio all over AC, or is there a DC bias)?

Ah yeah, know ohm's law (though applying it in some cases isn't always obvious), and pretty much what the basic components do. You forgot inductors. =P

Awesome on the book list. =)

Sure, that sounds right. Or try to get an AC voltage reading right across the terminals.

Still haven't soldered up the resulting circuit... was a bit dismayed when I found my perfboard won't fit in the project box I had. =/ But working with the breadboarded version seems to work great!

Also, just purchased a copy of "Practical Introduction to Electronic Circuits" off ebay for a darn good price considering it's the current edition (~$30).

Thanks for the recommendation. Should be a good read. =)

Oh you'll love it! My copy is outdated by at least two editions but still a favorite of mine. The author really has the gift of clear explanations of the subject matter.

When all else fails, there's always the Walmart tupperware aisle