[Guest]

First, align the IF section. Set your oscillator to the radio's IF frequency, probably 455 kc, and tune the trimmer caps in the IF cans to peak output.
You can use your ears, but all proper manuals will tell you to put an audio signal on the oscillator, and measure the voltage out of the audio section.
We could find 10 different opinions on which IF can to start with, but you'll have to go back and forth one after another, since the cicuits within each transformer interact with one another. With this radio, you MUST set the variable bandwidth to the minimum (counterclockwise?) position. Otherwise, you will not be accurate. The wide bandwidth position broadens the response curve to a flat-top, not a peak. Ideally, we would all have RF wobbulators, which SWEEP the output signal through a 20kc range from 445 to 465 kc, and watch the output on a scope, tuning for for a peaked curve (best selectivity) or a flat-topped hump (best wideband audio). Lacking this, imagine you are looking through the keyhole on a door, down a hallway, into th keyhole on another door, more hallway, another keyhole, etc. You are trying to get them ALL in a line so that a laser beam could pass through the series of holes and shine on a wall at the end. Reduce the output on the oscillator as you get closer and closer to alignment, so the AVC does not make the tuning seem broad.
Once completed with the IF section, set the radio to the AM BC band at about 600 kc, set your oscillator to 600kc if you have a counter to verify with, or if the radio actually picks up stations, set the dial to that stations frequency and on the multi-gang tuning condenser turn the trimmer for the local oscillator section until your oscillator signal (or station selected) is peaked. With most radios the oscillator section has smaller cut plates, but I think the radio in question has all the same size and has trimmers and padders which complicates the issue. You may have to refer to docementation to verify which is which. After you have the signal of interest at the correct setting of the dial, you must then peak the trimmers on the other 2 (RF) sections for maximum output.
Now pick a frequency at the upper end of the dial and find which trimmers
adjust the upper end. They are different from those for the low end.
Repeat the procedure at the upper end of the dial. Now go back and recheck and/or adjust the lower end of the dial. This is a general simplification, as this radio has MANY trimmers and padders, and they all interact with one another, from one band to other bands, and you can really throw things way off if you just start twisting randomly. You really should try to follow Philco's recommended order. If someone hasn't already "tightened those loose screws" you shouldn't have to move them much. Accquire or make a non-metal hex driver to adjust with, or prepare to go nutty fighting the detuning effects of a metal tool.
Most radio manufacturers recommend starting on the lowest frequency band, align that, the next higher, and so forth. Finally, check the midpoint of each band with an oscillator and frequency counter to see if the dial scale error is acceptable. There is normally some error at the 25% and 75% tuning mesh positions, one above, one below. But if you've done the job right, the dialscale error should be nil at 920 kc or so on the BC band.
Having not done this before, expect some frustration. If you can, practice on a much simpler radio first, to get feel for how it all works.
If the radio is out of the ballpark, STOP and get the factory documementation and READ it 5 times before you get deeper into a hole.

Best of luck..