Push-To-Talk-Box Schematic *Please have a look!*

[xeawr]

Dear fellows,

this is an epic moment in my life: This is the first time that I dare to actually post something online, asking a smart guy or two, to have a look at this schematic. Unfortunately where I live, I don't know anyone thats good with audio electronics, so I'm clueless when it comes to the evaluation of my circuit diagrams.
I spent the last years reading a lot about electronics and spent hours digging through unbelievably comprehensive knowledge bases available online. I had absolutely NO clue about electronics just two years back, but by now I think I know a liiiiiittle bit about this and that - well just enough to get some very (!) basic low-voltage audio stuff done.

I've drawn a schematic of a Push-To-Talk-Box, which I would use for doing FOH work. As you can see, it is made up of very cheap and readily available parts, yet has some "comfort features" (switchable +48V, LEDs, etc).
Although I conceived this box myself, it is basically a compilation of a number of different schematic that I added into this PTT (Push-To-Talk) schematic (I'll try to post them down there).

As you can tell I struggled with a good looking layout, not knowing if there are any rules where to place what and so on. Please don't mind the mess. Please ask, if there is anything unclear.

May I ask humbly if somebody is willing to quickly have a glance at this and tell me where I have mistakes (before I run off and build it). I've done some breadboarding on parts of it (such as the negative voltage generator), but I don't know if the whole thing makes sense.
I don't plan to amplify the signal a lot, maybe somewhere between 6-12 dB, just enough, so the mic preamp in the next stage needs only a comfortable amount of gain. I know I'm missing half a volt on the negative supply rail of the opamps: I'll either drop some voltage with a diode, or I'll add two simple zener diode (+ resistor) regulators.
Do I have to worry about this 0.5V offset?

Thanks A LOT in advance for your help!!!

 

[xeawr]

Here you can see most of the schematics I used:

 

[Hideki]

May I ask humbly if somebody is willing to quickly have a glaze at this and tell me where I have mistakes (before I run off and build it). I've done some breadboarding on parts of it (such as the negative voltage generator), but I don't know if the whole thing makes sense.

I don't understand the reason for all this circuitry when you don't really want any gain and the only real function is to short circuit the input (or not). Couldn't everything be replaced by a single button?

The input impedance and the entire opamp structure seems more suitable for a line input and not a mic preamp.

TL-072 for a mic preamp? No, please... don't do that. Not only is it rather noisy, and you have no need for fet inputs, it is also rather terrible at driving low loads, since it can't output much current before it starts distorting. It will absolutely fail when trying to drive your 300 ohm load and output transformer.

LM317 can't generate 48V from 12V. It's simply a linear regulator that will output less than what you feed in. To step up the voltage you need a different solution, like a switching boost regulator (or an AC supply and a voltage doubler/tripler).

Don't worry too much about the 0.5V (or more) loss, but worry about the noise that a 555 timer will make. You may be better off using a more dedicated voltage inverter instead of a 555.

 

[ruffrecords]

I think you need to think carefully about exactly what you want this thing to do. Remember that 20dB of gain for a mic is not much. This means the output noise of the 20dB amp will be predominant which in turn means if you then  plug it into another mic pre, all its hiss will be amplified too. Basically, a 20dB mic pre feeding another mic pre is a recipe for very poor signal to noise ratio.

Cheers

Ian

 

[PRR]

OTOH, a talkback mike may tolerate a hiss-level which would be annoying in many other uses.

I do not see how you get 48V from 12V with the voltage reducer regulator. Also do not see -12V.

90% of phantom mikes will work with 12V supply. You may not be able to run "any" phantom mike, but you sure can find one voice-worthy mike to eat 12V. The opamps can run with single 12V. Less headroom, but you don't need much, and light clipping does not hurt (actually improves) talkback.

Me, maybe I am old-fashioned, but I would get the SM-58-like mikes (dynamic, no phantom) which have on-off switch, go right to available mike input. A close-talked dynamic has plenty of output to talk to the stage. The switches are fairly reliable. These switch-mikes often lay around pawn-shops (maybe the bottom of your mike-bag?) now that "better"(?) mikes are sold for low prices. The lack of LED is a minor thing IMHO. Yes, it would be good to know when a muttered curse might roar across the stage. But when you mutter curses, you don't always check the LED first. Discipline, switching ON only when talking, works.

Last such job I did, I used an old CB radio PTT mike. Operation was obvious even for guest techs. This ran to a hacked $13 cassette player to a stupidly efficient 900Hz-2KHz horn dedicated to talk-back. That however was for a light booth where connection to a PA system was not assured. A PA tech talkback could be far simpler.

 

[xeawr]

I don't understand the reason for all this circuitry when you don't really want any gain and the only real function is to short circuit the input (or not). Couldn't everything be replaced by a single button?

The input impedance and the entire opamp structure seems more suitable for a line input and not a mic preamp.

TL-072 for a mic preamp? No, please... don't do that. Not only is it rather noisy, and you have no need for fet inputs, it is also rather terrible at driving low loads, since it can't output much current before it starts distorting. It will absolutely fail when trying to drive your 300 ohm load and output transformer.

LM317 can't generate 48V from 12V. It's simply a linear regulator that will output less than what you feed in. To step up the voltage you need a different solution, like a switching boost regulator (or an AC supply and a voltage doubler/tripler).

Don't worry too much about the 0.5V (or more) loss, but worry about the noise that a 555 timer will make. You may be better off using a more dedicated voltage inverter instead of a 555.

Hello again, thank you SO MUCH for your insightful comments. That thing with the LM317 is really embarrassing - I did a poor  research job... just forget about that whole section. I really want the full 48V phantom, so I'll come up with something else.

For a second I really questioned my objective and wondered why I had all this circuitry instead of the "simple button solution" . Thanks PRR for your insights (I usually don't air my displeasure, but quite amusing). You and Hideki hinted, that it could be so much simpler. True indeed.
I think the way it is now, it's not really smart, as u guys have pointed out - especially with one (pseudo)preamp "feeding" another preamp. As I reflect on it now, I think it would be good to build a microphone preamp (which it almost is) with PTT functionality. That way I can directly feed the output of the PTT box into the Line Inputs of the mixer, avoiding an additional unnecessary stage.
Apart from the practical use that this little box would have, I also enjoy building something useful, instead of just spending money on electronic components, without real results. Do you kinda know what I mean?

What's a good input and output impedance for a mic pre? I know JFETs have very high input impedances, but I failed to find out what a typical maximum input impedance for a BJT is, can anyone enlighten me?

The problem is also (apart from knowing very little ;D ) that I would like to keep the parts count down. That's why I reached for opamps in the first place (also I have little experience with transistors). To me opamps are so easy to understand and to use, especially when it comes to the powering aspect. Also opamps of the same models are more or less all the same, however transistors vary from one to the next, sometimes quite dramatically. I have the impression that therefore they require more "fine tuning" and this gives me a hard time understanding how to set them up and run them with great results (also I don't get much gain out of them). Just a side note.

If I remember correctly, a dB gain of 60, which is a common value for a preamp, equals a voltage gain of 1000. So I would need an opamp/transistor or two with this capability. Is this even possible for one opamp? Would the NE5532/4 be a better choice, especially regarding that TL072 is a bit unfavorable because of its noise?

I have a general question: Is it even possible to build a transistor-based preamp with decent SNR, with a few jellybean transistors? I do have a couple of standard BJTs, JFETs and MOSFETS by the way. I've spent hours searching for a handful of 2SK170s - which to me seemed like to holy grail of preamp-transistor - but couldn't find any without having to break the bank.

Regarding the noise of the NE555, is it not possible to filter the noise and just by adding a simple RC network? I chose the NE555 simply because it was available and cheap (1/4 of a TL7660)

PRR, btw, the -12 is supposed to be the -11,5 coming from the NE555 set up as negative voltage generator, so I would have 24 Volts from + to - , probably not necessary, but it seemed right to me.

That's all I can think of right now. THANK YOU AGAIN for your contributions and pointing out the flaws in my schematic. I really really appreciate that.

 

[ruffrecords]

It might be worth your while searching for the SSL talkback circuit which includes a compressor. Does just about everything you want.

Cheers

Ian

 

[PRR]

> What's a good input and output impedance for a mic pre?

I suspect there's books and even web-pages.

Beware: there's all different opinions, more than is explained by the fact that it isn't very critical.

> what a typical maximum input impedance for a BJT is

Under 1 Ohm to over 1Meg. Any mature designer can push it any way you want.

Jellybean transistors are NOT that variable, except when you do not understand it is the *circuit* not the transistor.

In real work, hiss may be more the issue than impedance. Here we learn OSI, not loading impedance. In talkback, bah!

> dB gain of 60, which is a common value for a preamp

I've used that much gain once or twice over the decades. Never for close-mike talkback. Don't you have a console? Does it have dB numbers on the gain trim? Set it up and see.

I would cave-man thus. I would figure SM-57 several feet from a talker to get 74dB SPL and output 0.2mV. Close to lips on a serious Talker, 2mV. Line Level may be 1,230mV or more, but any board's Line In must accept a 1983 cassette deck outputting 200mV. 2mV in to 200mV out is gain of 100, 40dB. Things that eat Phantom "generally" have 10dB-20dB higher output in the same sound field, so 30dB even 20dB may be just-enough, and 40dB fixed would lead to gross clipping.

BTW: if this talk-box is on/near the console (adjacent rack), an unbalanced output is perfectly clean and tons cheaper and simpler. (Same for short microphone lines, except Phantom strongly likes balanced, which IMHO is why to grab an old Dynamic or a $1 Electret instead of even a $69 Neumann-imitation.)

> apart from knowing very little

Everybody is born that way. We all have to learn. But do you let children (any beginner) design a river bridge? Car air-bag? Cellphone battery? Clothes washer? Operating system? OK OK, it does seem like there's a bunch of beginners in the design rooms of the work. Your talkback amp won't drop folks in the river or shatter the laundry walls. I would even encourage some try-and-see (staying out of the 120V/230V wall-power side for now). But you can throw 10 parts together thousands of ways and only a couple of them work. Your design team gets less frustrated if they have some understanding of basics.

 

[thismanysounds]

Just a suggestion, but you might consider doing the muting after the amplification stage.

This avoids 60db gain from any switch noise, and any adverse affects of shorting the phantom supply.

 

[Andy Peters]

I would suggest as a first-order thing that you don't need phantom power for a talkback mic.

Most of us just use an SM58S (the version with the switch), patched into an unused channel on the console.

Also, pretty much every live console I've seen in the last couple of decades has had a built-in talkback circuit, with mic preamp, push-to-talk button, and output assignments. And nobody ever uses it, preferring to patch an SM58 into an unused channel.

If you insist on building something, consider using a THAT1510 or INA163, powered by a couple of 9V batteries. Add a pot for gain/level, XLR female for input, XLR male or ¼" TRS for output, and your push-to-talk button which just acts as a mute by shorting the mic pins 2 and 3 together.

 

[xeawr]

Guys thank you SO MUCH for your all your very worthy input. I'll go and redesign the whole thing and I have again learned a lot about circuits. I already had  a quick look at the SSL Talkback circuit, but I definitely need to spend some more time trying to understand the whole thing...

By the way, my FOH console is not really a console, its a Yamaha 01V96i with an additional 8ch input AD-Card (MY8 AD24). It's really a small project studio mixer, that's why they omitted several functions that would come in handy for FOH work (I contacted Yamaha about this).
By boosting the Mic-Level up to Line Level (with adjustable gain) I could use the AD-card-inputs that don't have a preamp (or gain control other than the channel fader). That way I would leave the valuable 16 preamp channels available for the really important microphones ... .

Please let me know briefly - yes or no:

- Can't I filter out any noise on a power supply rail (keywords: Noise coming from the NE555)?

- Will a NE5534 (cheap, low noise opamp) work well in 1-stage talkback preamp configuration with moderate gain (in a well designed circuit - without having hiss all over the signal)?

Thanks again so much for your help, you fellows are great!!!

 

[Andy Peters]

Please let me know briefly - yes or no:

- Can't I filter out any noise on a power supply rail (keywords: Noise coming from the NE555)?

It's possible to filter the noise from the charge pump, but how about doing something a lot easier? Use a 16 VAC (not DC) wall wart, a couple of diodes, a couple of 1000 uF caps, and build up a voltage-doubling half-wave rectifier:

Hang LM317 and LM337 regulators off of the rectifier outputs, and now you have a nice regulated bipolar supply.

- Will a NE5534 (cheap, low noise opamp) work well in 1-stage talkback preamp configuration with moderate gain (in a well designed circuit - without having hiss all over the signal)?

INA163[/quote] works better and is easier to configure as a mic preamp.

 

[Whoops]

By the way, my FOH console is not really a console, its a Yamaha 01V96i with an additional 8ch input AD-Card (MY8 AD24). It's really a small project studio mixer, that's why they omitted several functions that would come in handy for FOH work (I contacted Yamaha about this).

Hi, welcome to the forum.

You should have said that in the beginning, explaining with detail what you need and what you were trying to achieve will help people helping out.
Until now we though you wanted to do a talkback system for live sound.
I'm a Studio and FOH engineer and I always used the SM58 with a switch, it always works great so I was not seeing the point also as other members.

So you have a studio and want to have a talkback mic for the control Room with a mute switch and mic preamp included so the output is balanced line level, is that right?

something like this?

If so, for the Mic I recommend you to use Panasonic Omni electret capsules, easy to DIY and a lot of info on the web on how to implement them. Sound great and are pretty cheap.

For the Mic preamp I second what Andy recommended  use THAT1510 or INA163 IC's
For the balanced output you don't need a transformer , use THAT 1646

For the Switching I would only use Momentary type of switching,  Latching switching for studio talkbacks are usually forgot in the On position and bring out serious problems in a recording session.
The smaller problem is that musicians will play with the talkback open so it will affect their monitoring during play time.
The Big Problem is that people in a control room will make comments thinking the musician recording on the other side of the glass is not able to hear them, if a bad comment is made when the talkback switch is by mistake left in the On position the session or the band can be over in a Pinch.
Normal comments bands and producers make could be "this drummer can't play on time", "She can't sing in tune", but "I had sex with his girlfriend" can also appear at some point.

Momentary switches seem boring because you need to have your finger there, but will for sure make your sessions smother and avoid unnecessary conflicts.
I learned this many years ago, I questioned it in the beginning, "why do I have to keep pressing the SSL talkback?", but after so many years recording I really wouldnt use it any other way.
Maybe a momentary Footswitch, that would be nice.

 

[xeawr]

"Whoops" thanks you so much for your time and thoughts. I'm really sorry If I didn't make this clear enough:

I'm using a Yamaha 01V96i for live/FOH work, not for studio work. I just mentioned that this mixer is designed for the environment in a small project studio (I was told by a Yamaha representative).

Anyway, this little box you show on the picture looks really nice. Looks a lot like how I  imagine my PTT-box will be (with a preamp inside)

Thanks for pointing me in the direction of THAT1510, INA163 etc. The problem is, they're not really on the cheap side. I just checked the INA163 and found out it's about 8 bucks (Euros) a pop. It might sound ridiculous to you, but I want to get there by using low-cost readily available parts, which is why I had put a TL072 there in the first place  :-[  (I of course discarded this idea).

Is this possible after all? I hope so... what do you guys think of the circuit I attached? Also check out: http://www.circuitdiagramworld.com/amplifier_circuit_diagram/Balanced_Microphone_Preamplifier_NE5534__3193.html

By the way, I'm now (whenever I have some time) breadboarding circuits as much as possible. Thanks Andy, for recommending the 16VAC wallwart idea, I think I'll go that way... as it is very convenient and practical

 

[Whoops]

I'm sorry Xeawr, I got confused when you said it was a small studio mixer. I though it was the mixer for you small studio.

For the supply you can even have a really small 2x secondaries toroid inside the Box  and then you don't need an external wallwart.

I see the circuits you've posted, you ditched the 48V?

why didn't you go with the circuit posted in the website link you posted?
http://www.circuitdiagramworld.com/amplifier_circuit_diagram/Balanced_Microphone_Preamplifier_NE5534__3193.html

 

[xeawr]

Hey Whoops! I didn't ditch the 48V phantom, I'll plan to build it around the schematic I attached recently (of course breadboarding it first). Still having trouble finding a simple and elegant way to raise the DC voltage with discrete parts. I know they do it in phantom powered mics all the time, but alas that doesn't help at the moment . Once I did some extensive testing I'll redo the schematic and upload it so you can see 'the whole thing'.

Thanks for your consideration tough!

 

[Whoops]

I didn't ditch the 48V phantom, I'll plan to build it around the schematic I attached recently (of course breadboarding it first). Still having trouble finding a simple and elegant way to raise the DC voltage with discrete parts.

Don't do it with discrete parts, save your self space.

You can use this schematic for your build,  read Note 3 on the schematic.
With the tripler reference you can have 48V from a 15V secondary transformer (single or dual)

http://www.jlmaudio.com/ACDCv4%20Schematic.pdf

 

[xeawr]

Hey Whoops, I can't see the schematic - is it me?

 

[Whoops]

sorry,

http://www.jlmaudio.com/ACDCv4%20Schematic.pdf

 

[Andy Peters]

I'm using a Yamaha 01V96i for live/FOH work, not for studio work. I just mentioned that this mixer is designed for the environment in a small project studio (I was told by a Yamaha representative).

I know a lot of live guys whose first digital console was the 01V96.

Thanks for pointing me in the direction of THAT1510, INA163 etc. The problem is, they're not really on the cheap side. I just checked the INA163 and found out it's about 8 bucks (Euros) a pop. It might sound ridiculous to you, but I want to get there by using low-cost readily available parts, which is why I had put a TL072 there in the first place  :-[  (I of course discarded this idea).

For a one-off project, the shipping cost to get the parts to you will be a large part of your cost! Also, the cost of the enclosure might surprise you, especially if you opt to get one drilled and painted out of house.

Also, why do you need phantom power? For talkabout, I stand by my assertion that the SM58, or indeed any dynamic mic, with a switch is more than good enough, and then you don't need the complication of phantom power.