When my company began purchasing solid-state AM transmitters back in the late 1980s, we immediately began to see a considerable reduction in operating costs. Solid-state transmitters, even the early ones, were considerably more efficient than their tube-type predecessors. Not only were the power amplifiers more efficient, they didn’t require kilowatts of power to heat filaments, and they didn’t have big cooling blowers that likewise consume a lot of power.

In the summer months, the savings were even greater because the heat load was reduced and it didn’t take as much air conditioning to maintain a safe operating temperature (although it might be argued that tube-type transmitters didn’t require as cool an environment as solid-state designs).

Succeeding generations of solid-state transmitters have become more efficient, using fewer devices and a smaller number of combiner stages to get the final power output. Later transmitters also employed higher voltages and lower currents, which often resulted in lower I²R losses in the DC wiring and connections.

The NX50 offers several MDCL algorithms to choose from.

Perhaps one of the biggest improvements we saw was the use of Litz wire in RF circuits, particularly in filter components, which not only reduced losses but also allowed for a much smaller footprint, since large copper tube coils were no longer required.

Current generation AM solid-state transmitters boast a range of 83 to 90 percent AC-to-RF efficiencies; impressive and very “green!”

A BETTER MOUSETRAP

With the cooperation of the FCC, some transmitter manufacturers have gotten aboard with Modulation-Dependent Carrier Level (MDCL) technology, which has been a staple in European MW broadcasting for many years. In just the last few months, stations have begun to employ this technology and evaluate its power-saving benefits and effects on the signal.

As soon as the public notice came out, I filed the required paperwork with the FCC for one of our West Coast 50 kW stations, KCBC, which already had an MDCL transmitter in place. The Media Bureau issued a modified license within a few weeks and we were good to go … sort of.

The issue for our digital AM station was that the firmware in the Nautel NX50 we were using would not permit simultaneous MDCL and HD Radio operation. We could have simply turned the digital carriers off for a while to test, but the good folks at Nautel got us a firmware update that would permit both. Our staff at the station got the update installed and the MDCL turned on in short order.

AMC in action. Note that the instantaneous output power was 29.0 kW while the nominal TPO was 52.65 kW, about 2.6 dB of compression

After considering all the available algorithms, we opted for the Amplitude Modulation Companding (AMC) algorithm with 3 dB of carrier compression. Our station serves the Bay Area with a talk format from a site near Modesto. As a result, while most of the service area is within the 5 mV contour, it is also at a considerable distance from the site. My concern was that receiver AGCs in the distant coverage area would pump up the gain (and thus the noise) while the carrier was reduced, thus lowering the signal-to-noise ratio of the demodulated audio. That would be unacceptable.

As it was explained to me, the British AMC algorithm takes care of that by restoring full carrier power during periods of low or no modulation, thus holding receiver AGC levels where they should be for the full-power signal. Further, with the sideband power unchanged, the loudness and reach of the station should not be changed.

That is exactly what we observed. In short, we have heard no discernible degradation of the signal, audio or coverage at all, analog or digital.

SIGNIFICANT SAVINGS

What we have observed is a big reduction in our power bill.

We activated AMC operation at KCBC the first of November. When the November billing cycle closed at the end of that month, I was able to look at the electrical usage for that month and compare it to the same period for 2010. This showed a 21 percent decrease in electrical usage and an $800 decrease in the amount billed. That’s significant.

I looked at the December usage in early January and found similar savings -- right at 21 percent -- for that month as well (compared to December of 2010). And as I prepared to write this, I checked the January usage: It showed 23,988 kWh compared to 30,306 kWh for January of last year, again, right at a 21 percent savings in power (and $882 less in energy costs).

In late December of last year, we replaced the 16-year-old Nautel ND50 at KLTT in Denver with a new NX50, the same model transmitter that we use at KCBC. Within a few days of installing it, we got the AMC working and I was personally able to observe the effects of this on the sound and coverage of the station.

As with KCBC, I could not detect any observable effects, even in the fringe area on the west side of the continental divide. What I could detect was a 21 percent decrease in energy usage at the site for the month of January compared to January of 2011. That is right in line with the KCBC savings.

KLTT Chief Engineer Amanda Alexander with her new MDCL-equipped NX50 and trusty ND50 aux (also MDCL-equipped).

Since those unqualified successes, we have added two more stations to the MDCL list. Our 50 kW station in Birmingham uses a 2000-vintage Nautel XL60, which by itself does not have MDCL capability. The good news is that Nautel offered us a firmware upgrade to the AM-IBOC HD Radio exciter that drives the transmitter with all the MDCL algorithms in it. So we jumped on that.

We applied for and got FCC authority, upgraded the exciter and turned on the AMC with 3 dB of carrier compression. As with the California and Colorado stations, we have observed no degradation in signal, coverage or audio in either the analog or digital domains. I don’t yet have power savings numbers for that station, but my guess is that we will see something in line with what we have observed elsewhere.

We also activated MDCL on a 5 kW AM in Birmingham, Ala. Presumably the savings will scale. We’ll be watching in the months to come.

DEMAND WARNING

One anomaly that I noticed on the KLTT electric bill was that while total kWh was down 21 percent, the total cost was down only 11 percent.

It didn’t take me long to figure out why. I noticed that the demand, which had been running right at 99 kW, dropped to only 90 kW for January. The reason was that during the early days of January, we had to operate both transmitters for testing. During this testing, I had limited the ND50 to 10 kW, but we had the new transmitter running into the dummy load at various power levels for awhile. It only takes 15 minutes to set the peak demand for the month, and that’s just what I inadvertently did.

That brought to light an issue. If we have to use the auxiliary transmitter for more than a very short period of time, we’re going to face a big jump in the electric bill if we run it at full power. At KLTT, we had a solution — that aux transmitter, like the XL60 in Birmingham, is driven by an AM-IBOC HD Radio exciter. We performed the same firmware upgrade and now the KLTT auxiliary is MDCL/AMC equipped. We are good to run it at full power if needed, without any peak demand penalty.

KCBC’s aux, also an ND50, does not have an AM-IBOC exciter that we can upgrade, so the solution there is to simply limit its maximum output power to 25 kW. If our chief engineer needs to test it at full power for any reason, he can do it on the generator.

We will undoubtedly continue to learn more about MDCL. We may experiment with some of the other available algorithms to see what their advantages and liabilities are. But at this stage, I am prepared to call this experiment a success. I can’t see ever going back.

W.C. “Cris” Alexander is director of engineering at Crawford Broadcasting and a past recipient of SBE’s Broadcast Engineer of the Year award.