New project! I have a Honda Goldwing motorcycle and it has a factory audio system that includes a headset jack and an AM/FM radio with an auxiliary input and a CB radio and intercom system. Way cool. Works like a charm.

Honda, however, does not believe that modern motorcycles should have a cell phone interface, for obvious reasons, but many riders are forced to the side of the road to answer the phone, which often means removing riding gear, helmets, etc. It's a major pain.

So...the bike's system jack is pretty straightforward, it likes a 600 Ohm dynamic microphone, and produces a maximum output voltage into two 75-Ohm speakers of about 3.5V AC.

The cell phone, on the other hand, at least the few I've tried, will easily use a similar dynamic microphone (tested using a voice recorder app on my Galaxy and HTC Evo, etc) as the bike. I've even used Thom Engdahl's 'converter circuit' to make a dynamic look like a condenser, and that works too.

It's really easy to safely split the helmet microphone signal to provide a signal (around 10mV typically) to both the bike system (direct input) and the cell phone (isolated input via transformer) using a good quality audio transformer as a 1:1 tap or even a 1:2 tap to the cell phone. Mouser sent me a variety of very nice ones and this part works great. I can speak in the helmet and hear myself since the intercom has sidetone, so you know you're speaking as well to the phone. Testing confirms I can do intercom, CB radio and cell phone voice recording, and hence outgoing phone calls this way. Cool there.

Here's the hitch: the phone has stereo media audio output, not sure the voltage/current level, but assume it's in the 18-20mW range for typical phone headsets. I'm thinking the incoming phone call audio part is also 'stereo' via two mono streams since my bluetooth and wired headsets play calls in both ears.

Problems come in when I combine the stereo audio output from the bike (which has a convenient mute function button on the handlebar) with the stereo audio output from the phone.

Since the phone uses a single ground for mics and speakers (using a TRRS 4-circuit plug/jack), but the bike has a separate ground line for the speakers (common gnd) and a separate ground line for the mic. The passive resistive network I used was the typical 10K passive resistor approach. That left a feedback loop that is more than unpleasant when I patched in the phone audio output!

My question is how to combine (actually mix) the two speaker-level audio streams which have differing grounds properly.

I considered transformers on the cell and bike outputs to lift grounds and then combine them with a resistive network, but the resistive loading seems wrong intuitively, and there's really no isolation once you wire them together.

I could possibly use a 4-input mixer to combine the two stereo streams into a single one, but I am not sure if op-amp mixers like you'd typically use on line level signals would do well with higher voltage and current levels that drive speakers, so really the question is whether anyone can recommend a 'speaker level' mixer approach. As long as the levels are inside half of 12 Volts, I could use opamps to combine the speaker outputs even if I have to drop the levels with negative gain, but not sure they can source the current needed, so an output stage is probably called for.

Thanks for any feedback, no pun intended...

Ed

 

The phone uses a single ground for mics and speakers, but the bike has a separate ground line for speakers (common gnd) and a separate ground line for the mic.

That seems odd, that there would truly be two. Have you put a meter on the two "grounds"? Maybe the mic needs power?

To answer your op-amp question - it would not be a problem to use speaker level inputs. You might need different resistors but maybe not even that. I'd try capacitive coupling to maintain isolation. You might lose some bass if your coupling caps are too small.

 

The factory system provides two independent grounds for the mic input and the speaker output common ground. In any case, if you connect the two grounds, it always generates a feedback loop, so I'll definitely need to 'float' the grounds so the cell phone's common ground won't make that same connection ending in a feedback loop. I'm sure those grounds on the motorcycle are eventually "battery(-)", but the internal circuit is not known. I'll put a scope on the two grounds during operation today and see how much difference there is between them.

The bike system has an internal pre-amp and is designed for connection of a plain old dynamic mic, so it's not a condenser or electret, tho I have used them via putting a AAA battery inline with a condenser mic and it works nicely, but not as good as the dynamic anyway.

Based on your input and my own thoughts, it might be good to just use some op amp unity gain/buffers for the 4 audio streams that need to be combined into a single stereo feed. Might even add some level control, but since the op amps will work in single supply mode, I have 6VDC swing, that might be enough. Worst case, I can get some 12V single supply opamps.

 

Makes sense. Just come on back and post your plan before you move on to building the project. The pros here will likely have good inputs for you.

 

The speaker output amps may be bridge type, thus the speakers cannot be grounded.

But it's a problem to combine the audio output of the bikes system with the audio output of the phone directly, since they are both power signals. Doesn't your bike have an external input for the audio?

If not you may need to use two op amps connected as differential amps to pick off the speaker signals and then combine that with the phone signals using two more op amps connected as two-input summers (you can buy quad op amps so only requires one package). This would then go to a stereo audio power amp to drive the headphones. Most op amps don't have enough output power to drive headphones.

 

Here's a tip. It is ok to build a AM-FM Radio/CB/MP3/phone/intercom combo. Just don't mention the words auto, car, bike or motorcycle.

 

Thought this is what the "blue tooth" cell phone interface was made to do?

Most states are making it illegal to use cell while in motion. Guess while inside a helmet no one will know.

 

I've been on the bench trying different approaches. I had some good luck with a mixer that has an attenuation stage (resistive divider) to drop the speaker level signals down to about 1V, then the mixer combines it all, using the typical op-amp mixer scheme. Problems are mostly the op-amp instability, probably due to the capacitive loading of the speakers. I tried it direct and capacitive loaded, it's unstable on both counts.

True, it's not legal to use a phone in many states, but it does not mean you can't carry one with you and use it at the gas station sitting still, which most riders do. Bluetooth is of course an option, but it has its own issues. I don't have the facility to develop a bluetooth module that will work with the bike, it would face the same problems as my wired solution does in interfacing. There is a stereo AUX input on the bike, but switching over to it eliminates the use of radio and weather features. So I just want to piggyback the phone on top of the headset stream, multiplex style if you will.

Thanks for the input, and specifically any tips on stabilizing an op amp. The oscillations are terrible using circuit topoligies from TI, Fairchild, and many other sources. Something I'm NOT doing is making it unstable.

 

I smell ToS violations here. You should never talk in a cellphone while driving, in fact in sweden where i live, such action is illegal and leads to heavy fines if caught.

 

I've been on the bench trying different approaches. I had some good luck with a mixer that has an attenuation stage (resistive divider) to drop the speaker level signals down to about 1V, then the mixer combines it all, using the typical op-amp mixer scheme. Problems are mostly the op-amp instability, probably due to the capacitive loading of the speakers. I tried it direct and capacitive loaded, it's unstable on both counts.
......................

What is the load that you are driving with the op amp?
Post a circuit if you can.

 

Of course u can talk...just wear a head set. U should have ur hands free, even though ur mind is occupied somewhere else.

It's what goes on here.