Aggregator

Over the years I’ve become a student of mic preamp design, building and modifying several along the way and learning a little more each time. Usually, I worked from a kit or published set of plans. Recently, I’ve tried some designs from “scratch,” researching various components, studying earlier designs, and incorporating them into raw schematics, followed by circuit layout, design tweaks and final fabrication.

Since my last two builds were vacuum tube devices, I wanted to do a simple, solid-state design this time. I came across some old preamp ICs in a parts box and almost used them but discovered they had been obsolete for years.

Was there a viable updated replacement? Enter THAT Corp., a relatively small IC manufacturer that specializes in chips for audio applications. THAT makes a few chips that are direct replacements of some popular preamp ICs like the Analog Devices SSM2019 or Texas Instruments INA163. If you’ve ever cracked open a broadcast console, you may have seen one. THAT’s website is a treasure trove of design notes and white papers on mic preamp design, with plenty of ideas to get a project going.

This project uses two ICs from THAT: the 1512 Low-Noise Audio Preamp, and the 1646 Balanced Line Driver. Using design notes from THAT and other sources, including advice from several more experienced DIYers, I was able to come up with a relatively low-cost design that has plenty of gain and good performance numbers for most applications.

The mic preamp can make or break a recording. Aside from the microphone, it’s the first stage in the signal chain before the recorder, and in some cases the only stage. It has to be clean and have ample headroom (unless noise and distortion are your thing), yet have sufficient gain to handle a wide variety of microphones.

Professional microphones have a balanced output, so the preamp will have a balanced input. Normally this is accomplished either with transformer balancing, which is expensive, or by using a standard op-amp as a differential amplifier, usually involving two op-amp stages with their attendant gain feedback loops, etc. The THAT 1512 takes care of this within the chip, providing its own balanced input. All that’s needed is a pretty standard input stage that can provide phantom power. The phantom power is sent to Pins 2 and 3 of the input XLR jack through a matched pair of 6.81K resistors, R1 and R2. These limit the current of the phantom supply.

The phantom power section

In order to preserve common mode noise rejection, any components that are mirrored between positive and negative signal paths must be matched in value as closely as possible. SW1 [switch] allows for turning off phantom power when it is not needed, and LED1 illuminates to show the actual presence of phantom voltage. R9 limits current through the LED to keep it from going “poof!” Capacitor C13 is there to smooth out any ripples from the 48 V supply. Between Pins 2 and 3 of the input jack and ground, ceramic capacitors C1 and C2 shunt any RF noise that might hitch a ride on the mic cable. Bad mic cables make good radio antennas!

Keeping stray static at bay is the job of the diodes.

Obviously, we need to keep 48 VDC out of our audio circuit. In a transformer-based design, the transformer would handle this, as transformers only pass AC. Likewise with capacitors, which are much cheaper and take up less space. This is why inexpensive designs use them. The problem is that inexpensive designs tend to skimp on these coupling capacitors. Years ago, I hot-rodded a mic preamp that originally had 4.7µF tantalum capacitors in the coupling stage. I replaced them with nonpolar electrolytics of a much higher value, and performance was improved.

Here, for C3 and C4, I use the same ones. At 100µF it’s overkill, I’ll freely admit, but the higher value reduces low-frequency phase shift (the LF response here is in the single-digit Hz range). Anything around 22µF or greater will work. Besides, it’s very difficult to match capacitors to such tight tolerances.

The high-pass filter is engaged by a switch — SW2.

Here’s where R5, R6, and R7 come in. They form what THAT calls a “T-bias” circuit, which boosts low-frequency common mode impedance. C14 is another ceramic capacitor across the inputs to clean up any remaining RF noise. By the way, R3 and R4 are there to limit any fault currents that might sneak by the capacitors. Their low value prevents input impedance issues.

Additional protection from stray static charges and other voltage transients is provided by diodes D1 through D4. This is a simplified version of a number of protection circuits I’ve seen. Anything ugly gets shunted to ground.

Now, it’s on to the preamp IC, which does the heavy lifting in terms of gain: up to 60 dB of gain, in fact. While a lot of designs will set the chip at a fixed gain level and introduce level controls somewhere between subsequent stages, ours is a simple mic preamp. It would be a simple matter of just inserting a potentiometer (VR1) across the gain setting pins of the chip, right? Not that easy!

Rapid changes in that resistance can introduce DC offset in the chip, which translates to thumping and popping on the output. This is where C5 comes in; a very large capacitor to kill DC offset. Why so large? Because VR1, R8, and C5 comprise a high-pass filter, so the capacitance has to be large enough to bring the low-frequency response down. In this case, it puts it around 5 Hz at maximum gain, keeping any rolloff well below 20 Hz. VR1 is a reverse-log pot, which provides the correct gain vs. position curve.

Capacitors C7 through C10 filter RF gunk out of the power rails to each chip.

Speaking of high-pass filters, I included one here to roll-off any mic or room rumble. C6 and SW2 provide a HPF, but this one has a twist. (Special thanks to the folks at www.groupdiy.com for this idea.) Because the changing resistance of VR1 naturally changes the characteristics of the HPF, this filter’s rolloff actually increases somewhat at higher gain settings. At first, this may seem undesirable, but think about it — low frequency artifacts are more likely to be a problem at higher gains than at lower gains. At any rate, C6 is small enough to rolloff the low end, but not to the point of sounding thin.

Now on to the output stage, handled by the THAT 1646. It’s one of the simplest I’ve ever seen. One IC and a couple of nonpolar capacitors. Caution must be used if inserting any other stages or components before the 1646, as it is very sensitive with regard to impedance. C11 and C12 are there to address any common-mode DC offset on the outputs. From there, it’s on to the output XLR jack, passing through a simple polarity switch, SW3, to reverse phase if needed.

Finally, capacitors C7 through C10 filter RF gunk out of the power rails to each chip, a very important consideration in any design. Clean audio has to have clean power.

Since this whole thing is built around THAT ICs, I decided to simply call it “THAT Thing.” Tune in next time, and we’ll talk about the power supply, breadboarding the prototype, and putting it all together.

More information about the THAT 1512 and 1646 ICs, as well as design notes and other information can be found at:

• www.thatcorp.com/Design_Notes.shtml
• www.thatcorp.com/datashts/THAT_1510-1512_Datasheet.pdf
• www.thatcorp.com/datashts/THAT_1606-1646_Datasheet.pdf

Curt Yengst, CSRE, is a contributor to Radio World and an assistant engineer with WAWZ(FM) in Zarephath, N.J.

Email us with your own DIY ideas at radioworld@futurenet.com.

The post “THAT Thing” — A Solid-State Mic Preamp Project appeared first on Radio World.

First Meeting and Chairs of Advisory Committee on Diversity and Digital Empowerment

.

.

.

Petition for Declaratory Ruling and connected assignment applications granted

.

.

.

.

The Media Bureau Announces a Freeze, effective immediately, on the following Filings (collectively, Filings)

.

First I’ll tell what you already know. Back in the day, AM broadcasting was king and FM was commercial-free. Things changed in the 1970s as FM grew in popularity. Here we are 40+ years later with many AMs struggling. Some have gone away because they were no longer financially viable. To make matters worse, AM directional stations are more time-intensive and costly to maintain, especially when compared to FM stations.

On the positive side, I know a number of smaller AM/FM combination and stand-alone AM stations in Minnesota that are doing well. One town has a 1 kW AM with a 100 kW FM. The AM brings in 40% of the sales revenue because it has always been locally programmed with live announcers until 1 p.m., then is live again during afternoon drive.

AM radio isn’t supposed to sound bad. It can be a clean and pleasurable listening experience, even when there is only 3 kHz of audio bandwidth. On the other hand, AM can be ugly to the ear when there are maladjustments.

SCIENCE

Modulation is the process of adding audio to a transmitted signal. Amplitude modulation is aptly named. A station’s carrier (transmitter power) is varied by the station’s audio. Carrier power is depressed to zero watts to achieve 100% negative modulation. It increases to 1.5 times carrier power when 100% positive modulation is reached. That is why a thermocouple antenna ammeter reading rises with modulation. You read it during a programming pause to get an accurate measurement.

METERING

AM modulation monitors have –100% and +125% lights indicating overmodulation. You really don’t want those lights to come on. More is not better.

First, be sure to set the monitor’s RF carrier level control so the carrier meter needle is in the right spot, as per manufacturer’s instructions. A carrier meter misadjustment will result in inaccurate modulation monitor readings.

Fig. 1 shows an AM modulation monitor. The –100% and +125% lights are on and yet the analog modulation meter reads only 94%. It is normal for an analog meter to read lower than actual modulation. In fact, 85 to 90% is a more realistic meter display, because it cannot track peaks as lights do.

Fig. 1: AM modulation monitor showing overmodulation.

A monitor’s audio output will sound excessively bright or harsh if a de-emphasis audio circuit is not included. Monitors traditionally do not have this, but often a simple capacitor and resistor modification will do the trick. The idea is to undo the high-frequency boost that is a part of the audio processing, per the National Radio Systems Committee (NRSC) standard. As you probably know, the transmitted audio has increased high-frequency response to overcome high-frequency rolloff in most receivers. The goal is to restore flat frequency response to the listener. Some audio processor manufacturers are using non-standard pre-emphasis curves to suit their taste. That complicates getting a realistic feel for frequency response. At least they are trying to make the best of receiver frequency response roll-off.

ON A SCOPE

An article I wrote regarding the operation of oscilloscopes, “Your Scope Is a Tool for all Seasons,” appeared in the Jan. 13, 2013, edition of Radio World.

To refresh your memory, a scope has a display where a dot that travels from left to right is deflected up and down with voltage. In this case, we will look at a transmitter’s RF output.

Fig. 2: An AM RF carrier wave on an oscilloscope.

I’ll begin with Fig. 2. It shows an oscilloscope with a view of the transmitter’s carrier with the scope sweeping at high speed (0.2 microsends per horizontal screen division) to see the actual carrier wave of an AM radio station. By carrier, I mean the transmitter’s power output. What you see is an almost perfect sine wave at the station’s operating frequency.

 

Fig. 3: A carrier with no modulation.

Let’s zoom in to the scope’s screen. Fig. 3 shows the carrier when the oscilloscope is slowed down to view audio (0.2 milliseconds per division). No modulation was present at that instant. Fig. 4 shows a 1 kHz sine wave modulating the carrier 100% positive and negative. The positive parts are the top and bottom peaks. They are mirror images of each other. The negative modulation part is where the carrier is just pinched-off at zero power in the center of the screen. This sine wave is relatively clean/undistorted, with less than 0.5% audio harmonic distortion.

Fig. 4: A carrier modulated 100% with a 1 kHz sine wave.

Many receivers do not reproduce it that way. The last 5 or 10% of negative modulation, between 90 and 100%, is where receiver detectors have trouble faithfully reproducing what the transmitter is sending. The result is audio distortion. We all know that unwanted audio artifacts are a listener turnoff.

Fig. 5: 100% modulation with receiver detector output.

In Fig. 5, I’ve switched the oscilloscope to dual trace mode. It shows the transmitter at 100% modulation on the top trace. The bottom trace was sampled at the receiver’s detector. I made the measurement there so it rules out additional audio harmonic distortion, which might be added in the output stage. By definition, harmonic distortion is where this 1 kHz audio tone will have unwanted audio products at 2 kHz, 3 kHz, 4 kHz etc. because of non-linear system performance. In this case, distortion from transmitter through the receiver detector measured 5.1%. It was only 3.1% at 90% modulation.

Fig. 6: 125% positive modulation, 100% negative modulation with receiver detector.

Fig. 6: Traditional analog audio processing used diodes to clip the negative side of audio before it went to the transmitter so it would not attempt to overmodulate the negative modulation while allowing positive modulation to go to 125%. The downside is that it added as much as 6.5% harmonic distortion in the process. Add the receiver’s problems to the mix and you have a whopping 10.2% distortion. Ouch! You’d never allow that on FM.

Newer digital processors reduce but may not eliminate the problem. Yes, the station can be a bit (about 0.9 dB) louder on the dial, but it is irritating to many listeners. They don’t know how to describe it, but oops, there goes another tune-out! Again, some people hear it and some don’t. Best not to penalize the station with high modulation.

Fig. 7: The transmitter is being badly over-driven at 100% negative modulation.

Fig. 7 shows the transmitter being modulated at over 100% negative modulation. I’ve moved the scope’s trace up a bit so you can see detail. Negative peaks go flat to the center, which is no carrier at that instant. Modulation like this will not pass the required NRSC occupied bandwidth nor will it pass my ear test for listenability. It is tiring to hear.

Fig. 8 is where you want to be. No more than 95% negative modulation, the sweet spot between loudness and listenability.

Fig. 8: 95% program modulation of the carrier.

It is a shame to lose listeners for that last 5% (about 0.5 dB) of modulation. Few if any will hear the loudness difference. Likely most will hear grit in the audio of transmitters modulated to the max. You can make up much of the modulation percentage difference with careful adjustments of the audio processing, before it goes to the transmitter. Software-defined receivers eventually will solve much of this problem, but we need to deal with today’s radios.

When I was installing AM stereo years ago, negative modulation was usually set at 95% and positive modulation at 95% for stations to sound clean. It was positive +125% if the client preferred it. That extra positive modulation comes as “forced asymmetry” where the negative audio peaks are soft clipped so the positive peaks can go higher. Ouch!

Surprisingly, bad-sounding audio with less than 100% modulation will usually fit into the NRSC occupied bandwidth mask, in the FCC required annual measurement. That is because of the required 9.5 kHz low-pass filter in audio processing.

AM stations competed in loudness wars to beat the other guy years ago. Now it is time to give listeners a pleasant experience with natural-sounding audio. Don’t drive them away.

I grew up in a broadcasting family that owned two AM stations and no FM. Success was dependent on keeping listeners. Loudness was not the answer.

Comment on this or any article. Write to radioworld@futurenet.com.

Mark Persons, WØMH, is an SBE Certified Professional Broadcast Engineer. He recently retired after more than 40 years in business. His website is www.mwpersons.com.

The post Find Your Modulation Sweet Spot appeared first on Radio World.

The BBC has announced that its two flagship radio programs for Afghan audiences will now be carried live by Radio Television Afghanistan, the country’s national broadcaster.

BBC News Dari

“Majale Shamgahi,” which is broadcast in Dari, and “BBC Naray Da Wakht,” broadcast in Pashto, will have the first half of its hour-long evening news programs every day on RTA’s FM networks in all 34 provinces of Afghanistan, as well as on medium wave.

The BBC programs examine key local and international issues with daily reports, interviews and analysis.

“A partnership with the BBC further reinforces RTA’s mission of informing the Afghan nation,” said RTA Director General Ismail Miakhail.

Pictured from left to right are Diva Patang, RTA presenter based in London; Ismail Miakhail, RTA director general; Jamie Angus, BBC World Service director; and Ismael Saadat, planning and commissioning editor, BBC News Afghan.

“Adding BBC programming to our output will contribute to the provision of trusted and impartial news about Afghanistan and the wider world.”

[Read: Radio TechCon Opens Registration]

BBC News Pashto

“We are delighted that the new partnership with RTA will allow our content to reach more people in Afghanistan, on channels they already know,” added Jamie Angus, BBC World Service director.

Miakhail also said that the RTA Academy would use the BBC as an example as it looks to train its country’s journalists on ethical journalism.

“Majale Shamgahi” will air from 6:30–9:30 p.m. Kabul Time, and “BBC Naray Da Wakht” will air from 3–4 p.m. Kabul Time.

 

 

The post Radio Television Afghanistan Rebroadcasting BBC Radio Programs appeared first on Radio World.

Announces auction of FM broadcast construction permits and seeks comment on auction procedures

.

.

.

.

.