[AmFlyer]

We need to begin with the motor design. Initially in 1946 all Gilbert engines had universal, open frame, series wound motors. These run on either AC or DC, hence "universal" in the name of the motor. In 1947 Gilbert introduced DC motors. these were just the old universal motor with a permanent magnet for the field in place of a field winding. They run on DC only, AC power will cause armature failure in a time period measured in seconds. Unfortunately Gilbert chose to not add DC after the engine number of the engines that were built with the DC motor, who knows why???!!! Therefore engines numbered 332 or 342 made beginning in 1947 through 1950 can have either a universal motor or a DC motor. There were some engines with DC after the number, these were the 332DC, 334DC and the 342DC. These always have a DC motor. In 1949 Gilbert began numbering some engines with AC after the number, for example 322AC. All this meant was it had a universal motor. But all 322's had a universal motor.

The only engines ever built with a DC motor were Northerns and 0-8-0 Switchers.

The use of suffixes came to and end with the introduction of knuckle couplers. A 302 or 302AC became just a 303 when built with a KC. The 332AC became a 336. In 1957 Gilbert adopted a 5 digit numbering system but that is a story for a different post.

 

[AmFlyer]

The DC motor armatures are different than the Universal motor armatures. The only real difference is is the universal armature has about 1.5 Ohm resistance and the DC armature has about 2.5 Ohm resistance, to compensate for the lack of the Field resistance in the formerly series wound motor.
Way back in 1954 my dad bought a Gilbert AlNiCo DC field from a 332 DC motor and swapped it into my 322AC with no other changes. That engine ran great for a number of years. It is possible to make that swap and use the original armature. There are two different DC armatures, so if the armature is also swapped the correct one must be used. The early DC armature has no oil slinger and is 1/16" shorter than the later armature with the oil slinger.
As far as power, the easiest DC supply is an original transformer with a cheap 10A DC bridge rectifier and a DPDT switch for direction control. By cheap I mean under $5.
The only real issue is the open frame universal motors used in Gilbert (and Lionel) engines are three pole designs. These are not inherently smooth running. Modern can motors with 7 poles are way better, especially for smooth, low speed running. The best way to get optimum low speed performance from a three pole universal motor is to use a power supply that outputs a chopped AC waveform. I use the MRC AH101 for this reason. The downside is the motor may be noisier and engines with built in horns/whistles cannot be used because the higher frequency harmonics make them sound continuously.
I believe Gilbert introduced DC postwar for several reasons. One, Gilbert had to develop a DC motor anyhow for their HO line so it was easy to scale that up for S gauge. Two, the AlNiCo permanent magnet material became cheaper and more widely available after the war. Three, it offered positive direction control over the engine and eliminated the need for the troublesome and expensive reverse unit. The DC line was eliminated in S for several reasons. First, with the Korean war AlNiCo again became controlled and expensive. Second, AC was just too pervasive in the non-scale toy train world so the DC engines were not big sellers. Third, there were no cheap, high current, DC supplies. Look at the expensive vacuum rectifier tubes Gilbert had to use in their #14 and #16 rectiformers. Anyone with AC trains had to buy one of those or a bulky and expensive #15 selenium disc rectifier just to run their DC engine. AC track power was easier and cheaper for toy trains.