Article Index   Home

Article #2a: Myriad / IRN News mixing & buffering circuit

copyright ©1999 Paul Blitz

 

Introduction

As described in Article 2, many Hospital radios are starting to use PC automated playout systems. one of these, Myriad, is very popular.

Myriad uses 3 sound cards, and there for has 3 stereo audio outputs which must be combined. In addition, Myriad can control an IRN news feed, and this 4 audio source needs to not just be mixed with the 3 Myriad audio channels, but must be switched on / off by the PC.

The circuit shown here simply takes the 4 audios, combining them to give a single (eg station output) signal: if you also need to buffer the audio to feed a mixer, then a slightly more complicated circuit is needed... see below for details.

Block Diagram

Block Diagram Description

The 4 stereo input signals come in on the left. The first 3 are from the soundcards, the 4th is from an IRN tuner. The IRN signal is switched (using a reed relay) before being combined with the 3 Myriad audio signals.

It is very likely that the IRN signal level is different from that output from the PC: there is therefore a preset potentiometer on each IRN input to allow the levels to be balanced.

The combined audio is buffered, and output.

Let me download the circuit details!

Circuit in GIF format (19k)

Detailed Circuit Description

You need to look at one of the above pictures to understand this....

Starting at the left of the circuit, the stereo input signals are capacitively coupled (to eliminate DC offset problems on the inputs), then mixed to mono, and buffered. It is possible to change the gain on the input buffer (eg if mixing 3 soundcards plus news, and the news is not at the same audio level) by adjusting the feedback resistor value, or even replacing it with a preset pot.

The gain of the buffer is Rf/Rin (Rf is the feedback resistor, R3, and Rin is the input resistor, R1 and R2). Currently, the circuit has a gain of 0.5 for each input... but as there are two input signals, they add together. The end result is that if you have a 0dBu signal on EACH input, then the MONO signal will remain at 0dBu. If you are only feeding in a mono signal, then you can either (a) use a single Rin, and set the gain to 1; or (b) feed the mono input to BOTH resistors (don't worry, this DOES work!)

After the input buffers, the signals go 2 ways: first of all, they are all mixed together to provide a single "station output" signal. The other way each signal feeds a further output buffer amp. The outputs are NOT capacitively coupled - if you use the specified op-amps, any DC offsets will be so small that they can be ignored. Besides, most equipment you connect the buffer unit to will have their own input capacitors.

Each buffer inverts the audio signal phase: because all audio paths have TWO buffers, the phase of all outputs is the same. "Why is this important" you may ask: well, by keeping the phase the same, it means that if you feed the signals via a desk, and then switch between the desk output and the "station output" of this buffer circuit, there should be no clicks (assuming that your mixer doesn't invert the phase!!)

Bear in mind that the summed output will be simply the sum of all the inputs: if you decide to simultaneously play out two (or more) signals, then, unless you drop both levels a bit, your output will be a bit too high (this is no different from fading up two channels on a normal mixer: if they are both "loud", the output will then be "very loud").

Power Supply

A power supply is needed: this should be a + / - 15 volt regulated supply, capable of providing around 100ma to each power rail. It is quite ok to use 2 PSU's, one for the +ve, the other for the -ve. Suitable pre-built PSU's are available from RS or Maplin. Another option is to but a transformer & a coule of "regulator cards", also available from RS / Maplin.

Do NOT use a switch-mode supply! SMPSU's often create audio noise... that's the last thing we want on this circuit! At the powers we are talking about, a linear PSU is not a big deal.

The circuit will work just as well with any supply from +/-12 volts to +/- 18 volts: the higher the supply, the more audio headroom you will have. However, make sure your op-amp IC's can take the voltage: some versions of the TL071/2/4 range only go to 15 volts)

It may be that you have a suitable PSU in another bit of equipment: if so, then you can possibly use that. The 100ma quoted above is a very conservative estimate.

If you feel up to building youir own PSU, then take a look at Article #3, which describes how to build a simple linear PSU.

Alternative bits

The op-amps specified are the TL071/2/4 range. The 4 input and 4 output buffers are shown using the TL074 quad op-amp. Feel free to use 2 dual (TL072) or 4 single (TL071) devices instead... it will work just the same. Similarly, if you choose to build more or fewer channels, just select a suitable device. If you end up with one or more op-amps unused, then simply connect the -ve signal input to the output, and the +ve signal input to 0v (this makes what is called a "follower" circuit).... that will stop the op-amp going noisy, latching up, or consuming excess power.

The TL06x range of op-amps are low-power versions (with a matching lower performance) of the TL07x, whilst the TL08x range are a lower cost version, with a worse noise figure. Wilst they may seem to be ok, I'm not going to vouch for how they cope with longer cable runs (that comment applies especially to the TL06x range) that are often needed in radio studios.

Many other modern op-amps are also suitable: however, it is generally NOT worth paying huge amounts for "state of the art" high-spec devices: at the end of the day, your station output is feeding either (a) a BT landline, then a 100 v line amplifier and finally piezo earpieces at the patient end; or (b) a MW transmitter which is listened to on a low-grade radio ("Walkmans" may have a good FM performance, but their MW performance is usually very close to crap!). In all cases, the audio quality of the TL07x range of op-amps far exceeds the quality of all the other downstream equipment!!!! (I have won that arguament MANY times!!!)

And Finally....

Finally, if you need any help with the circuit in any way, then feel free to email me with your questions.  

 

(plb, rev 1, Sept 1999)  

 

Article Index

Home