Audio TRRS to TRRS crossover cable

[Gavin]

Hi. I have two devices, both with standard TRRS (stereo audio output + microphone input) sockets. They are both expecting a standard headset with microphone to be attached but I would like one to talk to the other via a cable.

Can I just take the two output channels from the output device and connect them (together) to the microphone channel on the receiving device? Or should I only use one channel? Are the input and outputs even at the right ‘level’?

Gavin

[Richard Ibbotson]

My thoughts are:

Do not assume the TRRS are standard. Some devices use different pin allocation. Best to check

Do use both stereo outputs obviously result will be mono though. Not so important for music but important for some sound effects.

You will need to attenuate the headphone outputs so as not to overdrive the microphone input. A resistor from each output channel to the microphone input and a resistor from the microphone input to ground should be enough. The resistor values can be adjusted to give the best microphone input level. Maybe use a potentiometer in place of resistor to find best values.

Richard

Sent via Richard's Phone

On 31 Dec 2021, at 11:35, Gavin <gavi...@gmail.com> wrote:

Hi. I have two devices, both with standard TRRS (stereo audio output + microphone input) sockets. They are both expecting a standard headset with microphone to be attached but I would like one to talk to the other via a cable.

Can I just take the two output channels from the output device and connect them (together) to the microphone channel on the receiving device? Or should I only use one channel? Are the input and outputs even at the right ‘level’?

Gavin

--
You received this message because you are subscribed to the Google Groups "rLab / Reading's Hackspace" group.
To unsubscribe from this group and stop receiving emails from it, send an email to reading-hacksp...@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/reading-hackspace/CANjezMdF1TrStQw0YoCLnQwamqULnJgdGoOT3mQDHeakzxdJ_g%40mail.gmail.com.

[Richard Ibbotson]

You should also check if the microphone has a DC voltage across it to power a condenser microphone. If so then a DC blocking capacitor in series with the microphone pin should be added. 100nf probably OK if you keep your resistors in the kilohm range.

Cheers,

Richard

Sent via Richard's Phone

On 31 Dec 2021, at 12:11, Richard Ibbotson <richard....@btinternet.com> wrote:

My thoughts are:

[Gavin]

Many thanks Richard. So if I have got this right, the resistors are effectively acting as dividers to bring the voltage level of the line out to microphone levels. I have tried to draw as a schematic below (though obviously I would use higher quality resistors than those shown). Is this right? Or should the cap (if required) be on the path between I and R3?

Any guidance on ballpark resistor values very welcome

Gavin

To view this discussion on the web, visit https://groups.google.com/d/msgid/reading-hackspace/E5E39F81-7F35-41A9-BFEA-90B1BF8B1571%40btinternet.com.

[Richard Ibbotson]

Hi Gavin,

Yes, I would move the cap to the right to be on the on the path between I and R3. I would make R3 1K ohm to start and  R1 and R2 to be around 22K ohm to start. Better if you have an oscilloscope to measure the voltages or just adjust value based on the level meters in the devices or distortion.

Cheers,

Richard

To view this discussion on the web, visit https://groups.google.com/d/msgid/reading-hackspace/EDDB49A6-6379-42F5-897B-68B2F207BC25%40gmail.com.

[Gavin]

So I went out and bought a cheap headset for the receiving device, which I have disassembled and established that:

- the connector appears to be a TRRS following the CTIA specification- L, R G, M - from tip to cord

- when the headset is plugged in, there is a voltage of 1.28V between the mic and the ground lines, confirming your suspicion Richard that it uses a condenser mic. So the cap will be required.

My plan of action (I’d welcome views) is to:

- solder some additional wires on to the mic and GND contacts in the headset to take out to an oscilloscope

- hope that the headset still works with the receiving device when I plug it in

- use a frequency generator app on my tablet to produce a continuous sound with a steady waveform

- my expectation is that it will produce a waveform in the voltage on the scope(?) around the 1.28V bias voltage

- if the signal is clear and smooth, I should be able to turn up the volume and assess how large a waveform a 'voice level' noise makes

- this will be my target range (but around 0v) when I then divide the line level signal from the sending device down with the resistors

Does this make sense or have I fundamentally misunderstood the nature of audio signals / how all this works? 

Gavin

[Gavin]

Principally out of thanks to Richard, but also for anyone else who might stumble across this post, I have made up the circuit in a breadboard and tested on both an oscilloscope and also with the actual target devices. The circuit as recommended by Richard works and I am going to have a board made up according to the following schematic (outputting device plugs into TRRS1 and the receiving device into TRRS2):

The receiving device recognised the circuit as a headset with a microphone, and when viewed on the scope the potentiometers did a good job of attenuating the signal. Resistance of ~2k per channel brought the signal down to a level similar to what I measured off of the headset that I bought, so I have designed in 2x2k resistors, 1 per channel, to provide a base level of attenuation. There will then be a dual potentiometer to further attenuate the two channels if required.

As it happens, when testing with the target devices, I found that the receiving device auto levels the input signal so that however the potentiometers were set (I didn’t test below 2K per channel) the sound was about the same level. The difference was in the sound quality with (minimal) signs of clipping at low attenuation and noise at high levels of attenuation. It is tempting to just design the potentiometer out and use fixed values of resistance but I though I would leave it in in case I want to use it for other devices that do not auto-level in this way.

Gavin

On 3 Jan 2022, at 01:32, Gavin <gavi...@gmail.com> wrote:

Testing on an oscilloscope with the mic close to the tablet speaker, and the tablet making quite a loud noise, only gives a waveform with a range of 60mV either side of the bias voltage. Anyone know if this is about the right range?

[Richard Ibbotson]

Thanks Gavin,

Good to hear it is sounding out well for you.

 

I think you might also be gratified to hear that following your presentation to RADARC Reading Amateur Radio Club back a few years and plagues, that they are now using Kicad for several individual and club projects. Some of them are impedance-controlled RF designs on 4 layer boards mainly SMT.

 

Cheers,

Richard

 

From: reading-...@googlegroups.com <reading-...@googlegroups.com> On Behalf Of Gavin
Sent: Sunday, January 16, 2022 4:58 PM
To: reading-...@googlegroups.com
Subject: Re: [RDG-Hack] Audio TRRS to TRRS crossover cable

 

Principally out of thanks to Richard, but also for anyone else who might stumble across this post, I have made up the circuit in a breadboard and tested on both an oscilloscope and also with the actual target devices. The circuit as recommended by Richard works and I am going to have a board made up according to the following schematic (outputting device plugs into TRRS1 and the receiving device into TRRS2):

--

You received this message because you are subscribed to the Google Groups "rLab / Reading's Hackspace" group.
To unsubscribe from this group and stop receiving emails from it, send an email to reading-hacksp...@googlegroups.com.

To view this discussion on the web, visit https://groups.google.com/d/msgid/reading-hackspace/F6FBEF9F-4136-44D8-A10D-D356CB86C51C%40gmail.com.