The Fairchild 255A Amplifier Project

The Fairchild 255A is a well designed, 30 watt workhorse amplifier utilizing a 6AB4, 12AU7 and EL34 Pair.

The amplifier pair arrived in pretty bad shape, having been water damaged:

The schematic diagram, from the original instruction manual, is shown below:

Step I - Damage Assesment

The first step in resurrecting the amplifiers was to assess the damage.  There was a small amount of rust on the steel chasis and tube sockets, most of the aluminum was oxidized, and the main electrolytic filter capacitor appeared to be leaking. Not bad, certainly not a disaster!  The paint on the transformers was chipped.

I decided to plug the unit into the  Variac and see if it worked.  I connected the output to an 8 ohm 60 watt resistive load, and the input to the oscillator section of my HP 339a Distortion Analyzer.  I was pleasantly surprised that the distortion was less than 1% across most of the audio spectrum, but with some annoying hills and valleys. Overall,  however, the distortion is  more than claimed by Fairchild in their instruction manual.

Step II - Cosmetics

I dissasembled the amplifier completely, removing the tube sockets by drilling out the rivets with a drill press.  The fuse holder was to be re-employed in the rebuilt version, but new tube sockets, input and output jacks, potentiometers and line cord were the order of the day.  After the sockets and transformers were removed, I removed the metallic decal attached to the front of the chasis by heating with a heat gun on a low setting.  This melted the adhesive and allowed me to gently peel the decal away for reapplication.

The chasis was sanded with 220 grit paper, primed and painted with three applications of Krylon enamel.  Between painting steps I used 0000 steel wool to buff the chasis and then removed the dust with a tack cloth.  The transformers were similarly attacked.  Surprisingly, Krylon "Khaki" was a near exact replacement for the color of the transformers although the gloss grey enamel is much lighter.

I employed a new line cord and strain relief, replacing the OEM zip cord and light switch that were in place.  Where Fairchild had originally placed a two prong 120 VAC outlet on the amplifier I installed a small rocker switch.  

Step III - Modifications?

It seemed to me that we would be better off with silicon rectifiers and a beefier capacitative reservoir.  The electrolytic cap which was in place had to go anyway.  Using it in its present state seemed to invite disaster. (I have had caps blow up, sounds like a shotgun shell going off, travels at about the same speed, and leaves behind all kinds of debris)  I had some Mallory low profile 180 uF/450V electrolytics on hand which would fit on a PC board underneath the chasis. Conveniently, the PC board is attached to the studs of the power transformer via threaded 8-32 standoffs.

I decided to regulate the preamplifier and phase splitter stages with an LM317 circuit modified with the inclusion of a high voltage NPN transistor and zener to prevent the LM317 from seeing more than its allowable "input/output" voltage.  The adjust leg of the regulator is lifted 325 volts above ground with a zener diode string and the appropriate filter and bypass capacitors are all in place.  The regulator delivers a ripple free 340 volts to the node which feeds the phase splitter.  There is just enough load on the regulator to insure optimal performance.  The schematic for the new power circuit is shown below:

The B+ for the pre-amplification and phase splitter stages are further filtered by a 33uF/450VDC capacitor at each end adjacent to the tubes.  Star grounding is used at one lug of the terminal strip mounted near the preamplifier stage.

The input potentiometer was replaced with a low noise variety.  The 250N input capacitor was replaced with a high quality polycarbonate.  All of the paper and ceramic caps and the old carbon composition resistors were jettisoned and replaced with silvered mica's, Orange Drops and Xicon metal film.

The final modification was to allow biasing each EL34/6CA7 separately.  I did this by tapping the bias supply (new 220uF filter caps and rectifier) and added two 10K PC mount trimmer pots.  Each pot is connected to the bias supply via its own 22k resistor.  Were the pot to be damaged, the bias will go to a full -116VDC, cutting off the output tubes and preventing damage.  The bias is controllable over a range of -26 to -44 VDC. 

In order to measure the cathode current of the output tubes I placed a 1 ohm resistor in the cathode path to ground by means of a 1/8" normally closed phone jack.  The ground path is uninterrupted unless a plug is inserted. The voltage drop across the resistor can be measured.  The balance adjustment was also replaced with a PC mount trimmer. It is important to carefully check the resistance of the cathode resistor by subtracting the resistance of the test leads in situ .  Each of the 1 ohm resistors I used actually measured 1.18 ohms.

Step IV - Checkout