Ok, turns out i lied in the previous post, as this part 2 won't be about the transmitter - so sue me...
Either way, into the meat of it: upon some listening tests with music, it turned out the expanders were misbehaving just awfully. They made the volume jump up and down virtually all the time, in a very disconcerting manner - basically briefly increasing the volume (noticeably) upon every louder bass note (and only for the duration), and dropping back down in between.
I took apart one of the receivers, and lifted the schematic of the expander chip and the associated passives around it. Pretty much exactly what's in the datasheet of the TA31101, with the exception of having omitted the coupling capacitor going to the "Comp In" (pin 11), thereby turning the compressor side of the chip into another 1:2-ratio expander, just like the other half.
My solution to this problem, in the end, was to remove the coupling capacitors going to the two inputs of the internal rectification circuits (to pins 2 and 15, respectively), which provide the control signal for the gain cell. At least the signal's nice and consistent now.
That being said, i kinda wouldn't mind if it could go a bit louder though. Towards that goal, i ended up deactivating the "channel balance" trimpots from the circuit (by removing the 4.7k ressitors in series with them), and reducing the 4.7k in series with the expander outputs down to (by "piggybacking" a couple of 4.02k resistors on them). That didn't turn out to do much, but there's also a group of resistors, capacitors and a couple clamping diodes between the tuner chip and the expander, and i've got a hunch there's some attenuation going on there, as well. I oughtta see what the scope says about the signal levels in that area... If push comes to shove, i'm tempted to tack on some dead-simple P-channel MOSFET switch to power the headphone amp chip straight from the lithium cell (ie. before the series diode i had tacked on its output, to bring the voltage more in line with the series pair of AA's it would normally use).
Upon further investigation, there's a good chance the power-mod won't be necessary. Did a quick little test with a wire soldered on to the expander chip's supply line, and the other end manually temporarily connectable to one of the rectification circuits' "storage" capacitor (in this case, the one connected to pin 1). Result - right side went REALLY damn loud. Perhaps a little bit too much so, actually...
Normally there seem to be about 0.53V on both those pins / capacitors, regardless of the signal level applied to the transmitter. Since the 3.3V or so that were on the Vcc line proved to make things a bit TOO loud, i'm thinking i'll tack on some resistive dividers there, to provide maybe 1V or so on those pins. I'll have to first mess around with a trimpot and find something close to what could be called a "sweet spot".
Some 'scoping later, and after tacking on a 10k trimpot to adjust the input voltage to pin 1 of the compander chip (which handles the right output channel), it turned out that, for a seemingly maximum-before-the-transmitter-overcompresses level, about 0.95V are needed on pin 1 to just-about drive the output into clipping, at full volume. Given a 3.2V or so Vcc supply for the chip, that would mean the voltage needed would be aboout 30% of the supply voltage. I have some 4.02k resistors, and some 10k, and those would seem to be juuuust the right ratio (29.7% of Vcc calculated required voltage; 28.6% with the 10k/4.02k divider). I'm also thinking i should be able to get away with only using one divider for both halves of the chip, since they're supposed to do the same job (no expanding, just straight-up gain).
<~10mins later>
Hunch confirmed, mod-wiring done, results satisfactory - we now have the maximum volume in the "pretty damn loud, gonna have my ears ringing tomorrow" territory. The first volume step (out of a total of 16) above "off" is perhaps a bit less than what you'd need as a "normal" level for headphone listening, in a quiet environment. Damn i love DIY... <insert demonic laugh here>
Right, now on to applying the mod to the other two...
Hi I'm intrested to mod our 6 Reseivers and 3 Transmitters the way you did. Is deactivating the trim pots still neccessary? On the pic it's still done...
ReplyDeleteThat depends on what you're trying to achieve.
DeleteAt first the pumping of the receiver needs to be eleminated..I guess it's the rectifier mod (C at pin 2 and 15)and of course the more power of the Transmitter. If the deactivating of the trim pots doesn't do so much why should I do so? Thanks fro your reply.
DeleteThis comment has been removed by the author.
ReplyDeleteWhat can be achieved with the power mod? You wrote there is a good chance that the power mod is not necessary and then found the bias mod. So do I get enough volume just with the bias mod?
ReplyDeleteAnd if I look on the PCB as shown in #17 there are traces fron pin1 and pin16 going to other PCB layer with vias. Did you check where they come from? Maybe you can already change the value of any other component on the other PCB side to get the same effect?
DeleteI Already did the Transmitter mod. But I'm still not sure about the receiver mods. If the caps on pin2 and 15 are removed don't you disengage the expander and hear the compressed signal comming from the transmitter? Also does the volume mod summarize the stereo signal? Why have both path a signle line? There is no balance opotion?
ReplyDeleteYes, the idea on the receivers was indeed to deactivate the expander, because especially in conjunction with the compression going on in the transmitter, it just sounded attrocious, with the levels jumping all over the place.
DeleteBetween re-reading the text here, and a hazy recollection of the mod, the volume "setting" thing only affects the gain applied to the signal (which you'd want to be the same for both sides, right?) - it does NOT mean it mixes left and right together.
Feel free to read through the datasheet of the compander chip, to try to maybe get a better grasp of its workings.
Hey, thanks man!
ReplyDeleteYour right after following the traces I ended at the µPC. But I wonder why they use two independent DO's of th µPC...
Hi, thanks you so much for this post, could you please describe how did you connect resistors on the last photo ? and which nominals they have, I didn't get how that spot of resistors connected to the chip pin at bottom
ReplyDeleteThank
Hello
ReplyDeleteThanks for this.
I’m about to try this mod.
I’m new to this so need a help.
When removing coupling capacitors do you mold a wire instead there or just leave the place empty?
Cheers
If i wrote "remove", then i meant "remove", and NOT "bypass" or "jumper".
DeleteCool. Thanks 😊
Delete