Here is the schematic from the Eico 667 tube tester. It is very similar to the 666, except now there are a pair of extra levers, and a few test sockets are missing. But the meat and potatoes are identical. If you are planning to restore your 667 to peak operating condition, here are a few pointers.

• Replace the selenium diodes CR1 and CR2. I rivet in a small terminal strip, using the mounting hole from the 'old' diodes. It is now easy enough to install 1N4007 diodes across the terminal strip.
• While you are at it, replace filter capacitors C1 and C2. Since modern-day capacitors are much smaller than their ancestors, I replace C1 with a 22uF/150VDC or possibly a 47uF/150VDC capacitor, while C2 is now a 22uF/25VDC capacitor. Typically I use whatever I happen to have on hand.
• Carefully check the calibration potentiometers R7 and R18. If you are having any difficulties calibrating your tester, change these potentiometers! Replace them with any quality 100K linear-taper potentiometer. I make sure I use a control with a nylon shaft, and avoid those with split and/or splined shafts. If necessary, I also cut a notch in the top of the shaft, to facilitate using a small screwdriver for any adjustments. Eico used very poor quality controls here, and they are prone to failure.
• When doing the calibrating for the Line Adjust, I will set it so 'High Noon' gives me a reading of 6.3VAC with a 6L6 inserted in the socket! This may seem obvious, but I sometimes carefully calibrate the Line Adjust exactly as the manual describes, and I end up testing a 6L6 with barely 6VAC on the filaments. I choose a 6L6 as opposed to a 6V6 because the filament draw is higher, and I want to be sure this procedure is done under a decent sized load! You can measure the filament voltage while testing say a 12AT7 or a 6SL7 after, and be sure your filament voltages are consistent, and correct! Tweak the Line Adjust calibration as necessary.