Jump to content


Photo

Audio comms nightmare


  • Please log in to reply
6 replies to this topic

#1 Kilrah

Kilrah

    RC-Cam Mentor

  • Trusted Member
  • PipPipPipPipPip
  • 2,079 posts
  • Location:Switzerland
  • Interests:Modelling, electronics, computers

Posted 27 September 2010 - 10:39 AM

OK, not really R/C related, but I'm having a hard time with what looked like an easy electronics job.

The goal is to take this: http://www.savox.com...data=HC-1&sub=1
and connect it to this: http://www.mypilotst...tstore/sep/3583

Doesn't sound so complicated, but...

Firstly, testing the mic. It's a "bone" mic, that's supposed to be put at the top inside a helmet, so that it touches the top of the head. This makes it a decent pickup device, while being insensitive to wind, which is the goal. Playing with it, hitting it etc shows that it's capable of a very big dynamic, still giving good and very ample signals (1 Vpp is easy) when hit. OK, looks good. Except that once put in condition, on top of the head, the vibrations and thus signal are actually tiny. 1.5 mVrms is what you'll get by talking LOUD without shouting.

Now the radio. Despite quite a bit of searching, I haven't been able to find what kind of mics, amplitudes etc are expected by aviation radios. All I've seen (in the radio's specs) is that the input impedance is 150 Ohms. To determine amplitude, I've simply sent some music into the radio, then some sine waves, while listening and then watching on the scope the signal that comes out of a receiving radio. It seems 50-100 mVrms gives a decent signal without too much distorsion. At this point, to avoid flooding the aviaiton freqs with my tests I have a 20dB attenuator and dummy load on the transmitting radio instead of the antenna. Perfect for bench testing, I'm not disturbing anything past 200m or so.

So, we see there's a serious level mismatch. Hitting the mic will clearly show on the other side, but talking as intended doesn't give anything. An amplifier is needed. So I make one, firstly a simple single-transistor mic preamp. Perfect, my level is all good, and I can hear something. It's pretty distorted, turns out it's because the radio decides to put a bias of 5V or so on the mic input when it's transmitting. OK, will take care of that later. Time to go test outside.

Remove dummy load, install antenna - That's where heck breaks loose. The 5W of 120MHz RF wreak havoc with the poor little 1mV signal. All that comes out of the receiving radio is some ~1kHz peeping. Put dummy load on again - no problem. Back to the bench.

I have since then tried replacing my transistor am with some slightly better controlled opamp based one, shielding it in a hermetic metal box, with SMA input/output plugs with coax/shielded cable everywhere possible, with no single change. Tried to filter a bit, probably badly (I'm a digital guy, analog isn't my thing at all)...
Simply installing all the mic part (mic, connected to amp with a <5cm cable, and power supply) on the bench, with scope on amp input and output, NO connection to the radio, then going 2 meters away with the radio and keying it will throw 2V transients in the audio path...

I've already spent a few days on that thing with no luck, and pretty much about to give up and find some other way to use a relay, i.e mic goes into a small PMR radio, other PMR radio is connected to aviation radio and keys it when it receives audio, and vice versa, that relay being put somewhere far away within the useable radius... but before that, would anyone have an idea on what I could further do to get decent behavior of the "normal" configuration?

#2 Mr.RC-Cam

Mr.RC-Cam

    RC-Cam Mentor

  • Admin
  • PipPipPipPipPip
  • 4,695 posts
  • Location:USA
  • Interests:R/C video and photography.

Posted 27 September 2010 - 11:06 AM

Did you try applying the typical bias voltage to the Mic's audio out signal? Usually a 2.2K - 10K ohm resistor, fed by 3-5 VDC, is used to bias mics with integrated FETs. You will need to AC couple (~100uF) to the ham rig , and may still need a preamp to correct the impedance match, but the bias voltage trickery might be the place to start.

Also, see page 54 in the instruction book. :)
- Thomas

#3 Kilrah

Kilrah

    RC-Cam Mentor

  • Trusted Member
  • PipPipPipPipPip
  • 2,079 posts
  • Location:Switzerland
  • Interests:Modelling, electronics, computers

Posted 28 September 2010 - 01:42 AM

Did you try applying the typical bias voltage to the Mic's audio out signal? Usually a 2.2K - 10K ohm resistor, fed by 3-5 VDC, is used to bias mics with integrated FETs. You will need to AC couple (~100uF) to the ham rig , and may still need a preamp to correct the impedance match, but the bias voltage trickery might be the place to start.

Yes, tried with and without. Response is better without, suggesting the mic capsule might be dynamic.

Also, see page 54 in the instruction book. :)

Internal mic is set as disabled.

#4 Kilrah

Kilrah

    RC-Cam Mentor

  • Trusted Member
  • PipPipPipPipPip
  • 2,079 posts
  • Location:Switzerland
  • Interests:Modelling, electronics, computers

Posted 28 September 2010 - 05:52 AM

OK, I've bought a $50 set of 446MHz PMR radios, even if not needed as a relay like my backup "solution" suggests, they can at least help me troubleshoot.

The mic/amp work perfectly with them. Mic direct works but slightly low, with the amp it gives a perfectly audible signal.

With the mic / amp connected to the PMR radio, sticking the PTT in TX mode and walking around with the 2nd PMR and the aviation radio, I can now defintiely confirm it's the amp not liking the aviation radio's RF. The amp will saturate and "peep" whenever the aviation radio is keyed within a 10m radius (with no electrical contact). The PMR radio transmitting and with the antenna right on the amp causes no problem at all though, and nothing else is disturbed by the aviation radio.

Now, how could I shield the amp (all previous tries didn't help), or make it immune to that intense 120MHz RF?

In any case the relay seems to be a perfectly viable solution, but I still do want to learn something from that issue...

Edited by Kilrah, 28 September 2010 - 06:06 AM.


#5 Doofer

Doofer

    RC-Cam Mentor

  • Trusted Member
  • PipPipPipPipPip
  • 277 posts
  • Gender:Male
  • Location:Nottingham, UK
  • Interests:Gliders (especially slope), electric, heli, submarine, rockets, quadrocopters

Posted 28 September 2010 - 05:53 AM

You didn't say what sort of helmet you're using.

It seems the most 'suspicious' feature you report is the lack of audio signal in normal use - suggests to me that it's not working as intended.

'Ear bone' mics fit tightly in the ear and pick up from the skull, 'jaw bone' mics pic up from close contact with the jaw... but a 'top of head' bone (skull) mic is only going to work well if you're bald, or you move your hair out of the way of the transducer. Just sitting on top of your hair, getting good enough contact to get bone vibrations might be a big problem.

Looking at that device, and the services they provide for (emergency services) I suspect those are all fairly enclosed-design helmets. I wonder whether it is more an air acoustic device working within the confines of an enclosed space (the helmet), rather than bone contact.

I would contact Savox and ask if there are any special requirements about the helmet type, and perhaps in the meantime try some other helmet types...

#6 Mr.RC-Cam

Mr.RC-Cam

    RC-Cam Mentor

  • Admin
  • PipPipPipPipPip
  • 4,695 posts
  • Location:USA
  • Interests:R/C video and photography.

Posted 28 September 2010 - 07:18 AM

Now, how could I shield the amp (all previous tries didn't help), or make it immune to that intense 120MHz RF?

In harsh RFI installations you would place the circuitry in a shielded metal box. All I/O signals and power must enter via feed-thru chassis capacitors. Internally there would be a common mode choke (Toroid) on them. The amp circuit would have high freq suppression too (e.g., cap across OPAmp's feedback resistor). Building the amp on copper proto board type PCB (grounded layer of copper on one side) is helpful too.
- Thomas

#7 Kilrah

Kilrah

    RC-Cam Mentor

  • Trusted Member
  • PipPipPipPipPip
  • 2,079 posts
  • Location:Switzerland
  • Interests:Modelling, electronics, computers

Posted 28 September 2010 - 10:57 AM

You didn't say what sort of helmet you're using.

A skydiving helmet (i.e very tight).

but a 'top of head' bone (skull) mic is only going to work well if you're bald, or you move your hair out of the way of the transducer. Just sitting on top of your hair, getting good enough contact to get bone vibrations might be a big problem.

That sounded weird to me too, but we met and discussed with the Savox representative and tested the unit in the helmet with their radios and all the "standard" equipment, and it was perfect. They said we'd be surprised how it works, and we were. It picks up through hair perfectly if it's pressed against it. Anyway the person who will wear it has no hair, so it can only be better than my tests ;)

It seems the most 'suspicious' feature you report is the lack of audio signal in normal use - suggests to me that it's not working as intended.

It's more complicated than this. The standard usage scenario of this thing uses the headset, and a radio-specific remote speaker-microphone or push-to-talk unit that connects to the radio. So they have those things that are adapted to standard police/emergency radio models, with the appropriate signal conditioning and connectors. In our case, as aviation radios of course weren't one of the options they had, and their units were too bulky, we went for the headset only, planning to adapt what was needed ourselves. So I expected having some work to do, the amp does the job well for the audio matter, it's only the RF that's causing me big trouble.

At this point I think I'll try the relay solution for now.

Edited by Kilrah, 28 September 2010 - 11:00 AM.