Palm TX VOIP headset adapter

actually this is my first msg on this forum,
I am not sure of the method to be proper..

but just want to ask:

assume mobile voip works well with palm + headset on tx.
Has anyone tried this with a bluetooth headset?

I checked softick's Audio Gateway on the web, sounds interesting..
but yet no mike support in that.

possible with T|X?

thanks to anyone with an idea

I think Softick's been working on trying to get your typical bluetooth headset to work with their software for some time, but currently it does not. Motorola makes a nice pair of bluetooth headphones that also have a plug, just in case you want to use with something that's not bluetooth, plus they support the AVRCP profile (as does the softick software), so if you're using pocket tunes, you can control it right from the buttons on the headphones. Also has a built-in bluetooth headset so you can use it with your BT cell phone too (both paired at the same time). You can pick them up at Radio Shack for around 130. Kinda pricey, but well worth it if you listen to a lot of music with your TX. Definitely not a mainstream solution, but I've been doin' it for several months now and I couldn't go back. As far as bluetooth headsets go with the T|X, your best bet is this mod coupled with the Jabra A210, assuming that'll even work. I wouldn't go for it. The A210 has worked rather lousily (is that a word?) for me and my cell phone. Hopefully Softick will get Audio Gateway working with the Handsfree and/or Headset profiles, but until then no good options.

Where did you get the jack? I've looked all over for 2.5mm jacks, but I want to make sure the one I buy will work for a headset. It looks like a lot of them are only for output.

1 for $12 here.

(Sorry I never responded. I haven't looked at this thread in a long time, but someone PM'ed me with a question about it, so I had another look.)

The 2.5mm jack is available where I am in many forms. Regardless of which shape it is, the only truly important thing is that it is a three connector jack: tip, middle, and ground. (Basically it is no different than a standard 3.5mm stereo headphone jack you find in all of your MP3 players; it is just smaller.) But just make sure the jack matches your headset plug.

If memory serves me right, the tip is the mic and the middle section is the mono earphone on a Nokia headset. But you need to test your headset before you assemble the whole thing to make sure you have it right.

You don't even need the jack to make one of these work. If you just cut the plug off of a cheap headset, you can connect the wires directly to the Athena connector pins. I just wanted the flexiblity of connecting different headsets.

Wow I'm so glad I discovered this thread. I just happened to have all the same parts as twrock, and was about to do the same thing based purely off the pin chart in the Palm Developer Guide.

That resister is out of nowhere I would not have RTFM'd enough to see that.

Now that I'm looking at chapter 14.3.1, I'm wondering why you chose to ground TX through the 10K ohm resister, rather than RX.

I'm also wondering why you bothered with the right channel. Isn't the convention for mono audio to use strictly the left channel

Well..., first let me say that I am no electrical engineer and it was quite some time ago that I worked it all out, but I do hope I got it right in the end. The resistor came as a surprise to me too. I thought the whole thing would be a simple issue of connecting the SPKR_R, SPKR_L, MIC_IN, and AGND wires and it would work fine. Makes sense though that there has to be some way to cut out the internal speaker other than inserting a headphone plug into the 3.5mm jack.

Your second question first. I wired both channels together only because if for some goofy reason someone is listening to a stereo recording (music, whatever) they get both channels through their mono earphone. No other reason whatsoever.

I downloaded the current Developer's Guide and noticed that it is dated November 2006. So this is much newer than the one I was using when I did my initial work. I can't find my older guide, but I can see that there are at least some changes. However, I don't think that there is a change to which (TXD or RXD) should be grounded through the resistor, but of course I could be wrong. As I look at 14.3.1, I seem to recall that I interpreted the RXD to ground to be about inserting a set of earphones into the 3.5mm jack. However, I was connecting an audio device at the hotsync connector. So I interpreted that to mean that I should use TXD to ground. You can also make reference to 14.5 to see if that reinforces my interpretation or says something different. In any case, it's what I did and it works. YMMV.

Regarding the 10K or 6.8K issue I mentioned above, I find it quite interesting that Palm put out a revision way back in June 2005 that recommended using 6.8K instead of 10K, but in this manual that is obviously newer, they are back to saying 10K. Go figure. I have used both, and both work. I know of someone who is using 4.7K and having that work just fine. I really don't know what the high and low cutoffs are, but I think you'd be safe just to go with the 10K mentioned in this latest manual.

Good luck with your project.

It doesn't make sense to me (yet), because I can plug into the built-in 3.5 stereo jack and it cuts off the internal speaker. Unless that jack has some kind of internal switch, shorting the left or right channel to ground is enough to switch the speaker off.

Perhaps Palm wanted to leave open the option of using the internal and external sound simultaneously. If you neglect to ground TX, does it play through both the internal speaker and your connector? If so, then it would all make sense to me.

That was my first thought. However the manual mentions a "minimum series resistence to ground on pins 14 and 15" as being 8 Ohms. I'm a software guy that has forgotten EE 101; so I'm uncertain what shorting the left and right channels does to the impedence requirement. I would prefer to mix the right channel in only if it means I can get the correct impedence. I still need to find out the impedence of the Jabra A210.. it could be 4 Ohms for all I know.
Thanks for your help. I'll probably just connect the TX to ground via resistor as you did, since it works.

The key is the impedance. The Treo's headphone jack is \"smart\" it checks the resistance between the L & R channels and if it's too low the Treo assumes that you have plugged a hands free mic into the jack. The Treo then combines L&R and makes the remaining channel mic. If the resistance is high enough then the Treo assumes that you have plugged in a stereo headset or all in one device into the jack and reacts accordingly. I have tried a number of different stereo adapters, headsets, all-in-ones, etc and there is no way that I can see to tell which will work until you try it. I have had success with a Belkin car cassette adapter, and a couple of headsets. Good luck.

\"If you neglect to ground TX, does it play through both the internal speaker and your connector? If so, then it would all make sense to me.\"

In my testing and use, if I did not use the resistor, the internal speaker stayed on and there was no sound out to my external speakers/headphones/headset connected to the Athena connector. YMMV.

I decided to just connect the left and forget about the right. The motivating factor was to save power (less wattage is consumed on a single channel).

The other thing I did differently is leave off the power connector.. I couldn't find a purpose for it without wiring it to a PSU. If I had more motivation, I would have installed a tiny dip switch in the spot where the power connector goes, and used it to give me the option of combining the right channel on the fly.

I had a skilled jeweler do the soldering with a propane torch.. the area was just too small for my soldering tip. (BTW- if you attempt this, you'll first want to clip off all the unused pins to get them out of the way).

I tested the outputs and it worked. I have yet to install Articulation to test the mic. Here are my pics if it helps anyone:

I only left the power connector in to increase the number of contacts between the plug and jack. It just makes it a little less likely to come unclipped, but it is certainly unnecessary.

Yeah, the soldering is fairly delicate. My iron has a small enough tip, but my hand isn't always the steadiest. Fortunately the worst that can happen is to melt something down and have to start over again; no real harm done. I was more scared when I recently installed a mic internally in my TX! There isn't much room for error on that side of the connector, and the consequences of "cooking" something is a lot higher. I'm glad the deed is done and the Palm still works.

Enjoy!

BTW, if you want a free voice recording program, you can try SoundRec. Seems to work well for most people.
http://mytreo.net/downloads/details-112.html?SoundRec

Note quit true, since both the right and left drivers would be driving the same load. This would be logically equivalent to putting two batteries in parallel driving the same light bulb. Since P=V^2/R, the voltage is unchanged and the independence is unchanged, the power is also unchanged.

Note: The above logic is assuming ideal sources. However, no source is ideal. I will have to think about all the math involved for non-ideal sources. My intuition is that the above analysis will still be close.


Aaaaaa!!!! Electronics work with a propane torch! I would not let a propane torch anywhere near any electronic work I was doing! However, in this case we are only dealing with connectors and resistors, which can all take this kind of heat, save the possible melting of the plastics.

It's beyond me; I briefly studied that over a decade ago. However, this post claims otherwise.

You're probably picturing a large torch that a plumber would use. The size of the flame was a fraction of a millimeter with the oxygen nearly completely cut off. It enabled extreme precision, without forcing the solder joints to be heated. A conventional soldering iron requires the wire and/or connector to be heated enough to melt the solder (otherwise the solder will creep toward and stick to the iron as opposed to the actual target surface). The solder has no chance of sticking to a flame, so very little heat needed to be applied to the surface of the components / wire.

I studied that over two decades ago, but then never used it. So my opinions may be completely flawed.


The equations given are for computing impedance when putting two speakers in series or parallel. What we are trying to do is compute the power when two sources are put in parallel driving one speaker.

Actually, the more I have though about it, what we want is the power consumed by the system (for that will effect battery life) and the power delivered to the speakers (for that will effect volume and efficiency).

The more I have thought about this, the more I realize that the battery-light-bulb model given above is overly simplistic. Typical sources (amplifiers) are designed to have the same internal impedance as the speaker they are meant to drive. I will have to think what happens to the effective impedance when two sources are put in parallel. Intuitively I would guess the effective impedance would drop. Which would mean the efficiency of the system would drop (less of the power that is consumed by the system is delivered to the speaker), unless one also drops the impedance of the speaker.
_________________

Edits: Changed 'than' to 'that' in last sentence.