I guess we should start with a little history of the Gibson PAF pickup. By the mid-1950s,

Gibson wanted to counter the latest electric guitars introduced by Fender. Leo Fender had built a company that was a sizable competitor in the solid-body guitar market place.

Gibson believed they could beat Fender with their high quality Les Paul, and by developing a low-noise pickup.       

The problem with Gibson’s P-90 and Fender’s single-coil pickups was inherent in their designs, allowing 60-cycle hum (noise) to interfer with the sound.

Seth Lover was the Gibson engineer assigned to solve the problem. Seth connected two single coil pickups in series (opposed to parallel) and connected the coils out-of-phase electrically and magnetically. Thus the signal noise of each separate coil canceled out the noise of the other coil. That is how the pickup came to be known as a "humbucker".

Seth/Gibson filed their patent for the pickup design on June 22, 1955. Gibson added the new pickups to steel guitars in 1956, and in 1957 on electric solid-body and arch-top guitars including the Les Paul Model. During late 1957, a small black decal with gold lettering was added to the underside of the pickup that read, "PATENT APPLIED FOR" (hence the PAF abbreviation).

Seth Lover received his pickup patent #2,896,491 on July 28, 1959. By mid to late 1962, Gibson changed the pickup decal to read, "PATENT NO 2,737,842". Interestingly the patent number listed on the decal was not for Seth’s pickup design but was for Les Paul’s trapeze tailpiece! Perhaps this was a research roadblock for the competition, or maybe just a mistake?

Billy Gibbons with his 1959 Gibson Les Paul (Pearly Gates):

Photo by Kim LaFleur ~ Vintage Checkout ~ 

From 1956 until 1961 Gibson used different Alnico magnets in their PAF pickups. Alnico magnets (alloys ALuminum, NIckel, and CObalt) come in a different grades based on their magnetic strength. Gibson generally used the same magnets (size/grade) which was available for their P-90 pickups. But Gibson randomly used Alnico 2,3,4,5 grade magnets in PAFs until 1961 (remember the higher the magnet’s number, the higher the magnetic strength). This can often account for how two PAF pickups can sound quite different.

In July 1961 Gibson began consistently using a smaller Alnico 5 magnet (smaller as in the flat top side of the magnets were smaller length-wise). Since inconsistency was king at Gibson during this time, Alnico 2 short magnets are sometimes seen too. By 1965 though Alnico 5 was the standard for all Gibson humbuckers.

The original PAF magnet length was 2.5″ long, which was decreased by 1/8″ to 1/4″ to around 2.25″ in July 1961. But the “short magnet” PAF can be seen as early as 1959 and is still original. Gold plated guitars (ES-345, LP Custom, etc) seem to use the short magnet PAFs before nickel plated guitars (like the ES-335, LP Standard, etc). Just from a consistency point of view, July 1961 is the date considered by most as when short magnets were the norm for PAFs. Generally speaking decreasing the length decreases the power of the pickups, but this was somewhat counteracted by the Alnico 5’s added strength.

When new, the shorter A5 magnet is more powerful than the longer A2 magnet. So do short magnet PAFs sound worst than 1957-1960 long magnet PAFs? NO. In fact, they may sound better in many cases. But there are lots of things that effect sound, with the magnet only being one piece of the equation. Dimensions of PAF magnets follow (measured using a micrometer, and obviously this will vary a bit from magnet to magnet): 2.509″ long (”long magnet” version), .506″ wide, .131″ thick. The “short magnet” PAF length was the a bit different: 2.371″ long, .491″ wide, and .121″ thick. Another interesting point are the magnets in 1950s P-90 pickups (remember P-90 pickups are single coil predecessors to PAFs). There are *two* magnets in the P90 pickups, and yes they are identical to the 1950s PAF magnets (rough sand casted). Because of this, there has been a fair bit of “magnet hijackings” where players take p90 pickup magnets and put them into newer pickups, hoping to get that original PAF sound. Albert King with his 1959 Flying V equipped with original PAF Pickups:

The pickup were wound with #42 plain enamel wire. On original PAFs the bobbin wire appears purple, versus later PAF and patent# pickups that appear reddish. Gibson eventually switched to polyurethane coated wire around 1963. When wire coatings change, the sound of the pickup does change, contributing to the PAF following. The amount of wire (and coating) wound on each bobbin determines the pickup’s resistance. When the bobbins are wound with more than a nominal amount of wire (either on purpose or by accident), they are more powerful with fatter midrange but less treble. Due to the human factor and the wide tolerance of the manually-run pickup winding machines used by Gibson from 1956-1961, PAF pickups usually measure between 7.5 and 9.0 thousand ohms (K ohms). By 1962 (the end of the PAF era), Gibson was making pickups very consistently with 7.5k ohms of wire (give or take .25k ohms). The separate bobbins of a PAF can measure very differently due to Gibson’s manufacturing techniques. For example one bobbin could measure 3.5k, and the other 4.5k ohms (for a total of 8k ohms). This mis-matched ohms is actually a good thing, as certain frequencies will stand out if both bobbins have different resistance. This contributes to why two PAF pickups can sound quite different. The coil winder was a Leesona 102, and did have auto stop counters to keep pickups windings consistent. But these winders ran using a fiber gear and were prone breakage. The work around to fixing the counters is to time the winding process. That is one reason for the randomness of PAF pickup resistance. Around 1965 to 1968 (exact date unknown), Gibson changed from a manually-run pickup winding system to a fully automated system. Because of this their humbucking pickups all became a consistent 7.5k ohms from 1965 and later. The manual-run system had a machine operator that decided when a pickup bobbin reach about 5000 turns of wire. So there was plenty of room for under and over-winding. When the fully automated system came into place, the pickups were very consistent in their windings (and hence total ohms). A fantastic clip by the Allman Brothers Band featuring Duane Allman and Dickey Betts: