Many modern ham operators are not familiar with the Ballast Tube. You will find them in high end receivers like the Collins R-390 and the Hallicrafters SX-88. You also find one in the NC-303. There are highly qualified engineers who swear that the performance of these (pricey and short lived) ballast tubes is superior to the (even more expensive) special solid state replacment units sold. A ballast tube at full ratings is specified at 1000 hours. If you reduce it to about 80%, that increases to 5000 hours. If you cycle the power on and off frequently, that also reduces life span. If you drop it or jar it, it will fail immediately like any light bulb. I am not going to debate the efficacy of the use of the original ballast tube or the specialized replacement here. You will have to do your own research. A very scholarly thesis comparing the ballast tube and its solid state replacement can be found in Electric Radio Magazine ER# 321, February 2016, which is part 2 of the series; contact Ray at Electric Radio to order your copy and ask for the whole series: https://www.ermag.com/

I tend to think things out in a scientific fashion, as I was schooled to do. The NC-303 has a B+ regulator, to prevent line fluctuations from causing oscillator frequency changes. I would believe that regulating the filament voltage would also be beneficial to a degree. National itself states the expected drift with certain line changes is on the order of 20 Hz on 10 meters. If you replace the 4H4C Ballast tube with a 6V6GT vacuum tube, the drift increases to 250 Hz or so on 10 meters. I would consider that quite acceptable on AM mode. Also, I primarily use this receiver on 40 meters and below. Given that the AC line voltage these days in America is much better than the 50s, I also assert that this is a non problem. However, I would also assert that using DC on the filaments by way of an appropriately rated flavor of the LM317 regulator chip for the NC-303 would work even better than the 4H4C Ballast tube. It only has to operate one tube filament, the HF oscillator. With the R-390 or SX-88, you will have to figure out the load current and select the proper regulator, possibly one in a TO-3 package, with an appropriate heat sink. I will stipulate to being an agnostic on the belief that the 4H4C Ballast tube is essential to the proper operation of the NC-303. OR ANY OTHER QUALITY RECEIVER. My bovine fecal matter detector has gone off on this issue. You make up your own mind after you do your research.

Here is one more idea for the NC-303 that solves two problems. On the higher bands, the NC-303 and especially the HQ-170 can possibly have a hum FM modulation that appears on SSB and CW signals that is caused by using AC on the filament of the oscillator tube. While you can maybe ameliorate the problem by careful selection of oscillator tubes, the best way to go is DC filaments. In the NC-303, there is another problem, the availability of the ballast tube 4H4C. This is conveniently on a socket. In fact, the service bulletin says if the 4H4C fails, just plug in a 6V6GT tube. That is exactly what I did, to save the unobtainium ballast tube from wear. I put it in a box and placed it near the unused octal accessory plug at the right rear of the chassis, in case I ever needed it. There is a real fetish about this old buzzard technology. It was used to address the bad AC line stability back in the day. We no longer have this problem. The 6V6GT works just fine. Consider this idea: You build your own regulator on an octal plug that includes a rectifier, filter, and 6 volt or so regulator made from an LM317 or LM340. This eliminates the hum and you never have to worry about where you are going to get a ballast tube again. BOTTOM LINE: Quit worrying about a replacement ballast tube (save your money for useful toys) and plug in a dead flat "pull" 6V6GT that you put in your box of tubes for target practice.

A more elegant solution that does not require designing a regulator module is to rewire the oscillator tube filament to the regular 6.3 VAC supply; then you do not need the 4H4C OR the 6V6GT replacement. If you are worried about the increased current drain on the filament supply, wire the dial lamps to the source for the oscillator filament and 4H4C, and change the bulbs to the right voltage for that winding; an automotive version of the bulb should be perfect. This will rebalance the filament and regulated filament supply loads to duplicate the original drain on both, by simply swapping the loads.

On the R-390, you will find discussions related to this problem which involve changing the tubes to a higher filament voltage rating and eliminating the Ballast tube. Remember that the military used the R-390 in very severe environments, and required such extreme levels of performance that the R-390 still ranks very high on the Sherwood receiver test web site.

Since this is a matter of faith, I have encouraged you to research the matter and decide what is appropriate to your value system on vintage gear and preservation of it; please do not threaten to burn me at the stake for heresy. Whether it is pristine and stock, or reasonable modern adaptation to increase the time these treasures can be enjoyed, it is still appreciation that we all can share for the revered ancient ones who created them.