My take:
Since there are some posts that seem to be based on article 430....
Under article 430: For a motor circuit, the breaker is only there to clear short circuits and extreme overloads ( stalled motor). It's not designed to thermally protect a motor.
Why does article 430 exist? Its to prevent nuisance trips when starting a motor. Since the load from a motor is almost purely inductive, its starting current is very high ( up to 6X run current). Article 430 allows you to increase the breaker size to account for this momentary loading without having to increase the wire gauge as much as a similar circuit with a less inductive load.
While 430 lets you up the breaker to prevent startup tripping, it also specifies that you also need to provide thermal protection for overloading (a separate system from the breaker in the box). If your motor has a built in overheat sensor, you should be good. If you don't, you need to supply additional circuit protection.
If you were to redesign your current breaker and wiring system to meet article 430, you would also be allowed to increase the size of the breaker, that would increase the time it took to trip (assuming no additional thermal protection is added), and how hot the motor gets before its power is cut.
Since you don't have a problem starting your motor on a 20 amp circuit, the inrush current is not that high. Basically, stay with a 20 amp breaker.
Also, my read of article 430 ratings is a 25 amp motor circuit requires 12 AWG wire ( table 310.16, 60C column). Also see 240.4(D)(3), that limits you to a 20 amp breaker on 12 awg wire ( fine in this application).
A Few Points:
1: Your wire gauge meets the requirements of article 430.
2: You breaker is adequate for your application since its not tripping on startup ( no need to upsize it).
3: In your case, a 20 amp break is better, since it is currently working as an overload protection device.
Is the motor under load when it slows down (aka are you making a heavy cut?) I'm guessing no.
Why is your motor tripping its breaker?
1: Its running past its rated output ( to large a load).
2: Its damaged (probably a bad run cap or a shorted winding).
Since this is a new problem, look for a damaged motor or cap. As to figuring cause and effect, thats tough. It sounds like your system lacks thermal protection and at some point you could have overheated your motor. Or you have a run cap that went bad.
I would start with the run cap.
See the link for a step by step on a 1.5 HP motor.
http://www.ecmag.com/section/codes-standards/branch-circuits-lampholders-motor-controllers-and-more
To determine the minimum branch-circuit-conductor ampacity, use Sections 430-6 and 430-22. Section 430-6 requires the use of Table 430-148 to obtain the full load current of a 1.5 horsepower, 120-volt single-phase motor. Although there is no 120-volt column, the commentary that is part of the Table allows the 115-volt column to be used for motor nameplate voltages of 110 to 120 volts. Therefore, the full load current of the 1.5 horsepower motor is 20 amperes.
Section 430-22(a) dictates the minimum-branch-circuit conductor ampacity. This section requires a minimum ampacity of 125 percent of table full load current. Therefore, the branch circuit conductor ampacity cannot be less than 25 (20 x 1.25). According to Table 310-16, the ampacity of 60 degrees C or 75 degrees C No. 12 copper wire is 25, which satisfies Section 430-22 (a).
The next step is to size the overload relays in the motor starter. Part C of Article 430 has the title: “Motor and Branch-Circuit Overload Protection,” and part of Section 430-31 has this sentence: “Part C specifies overload devices intended to protect motors, motor-control apparatus, and motor branch-circuit conductors against excessive heating due to motor overloads and failure to start.” Notice that the overload relays are intended to protect the branch circuit conductors as well as the motor and controller.
As far as run caps failing, over voltages and heat are typical causes.
http://www.temcocontent.com/capacitorfaq.html#fail
Why did my run capacitor fail?
The answer may be simple, but depending on how close the run capacitor is to its design life, it may also be difficult to nail the reason down to a single factor.
Time - All capacitors have a design life. Several factors may be interchanged or combined to increase or reduce the life of a run capacitor, but once the design life is exceeded, the internals may begin to more rapidly decay and drop in performance. Simply put, a failure may be attributed to it being "just old."
Heat - Exceeding the design limit of operating temperature can have a big effect on run capacitor life expectancy. In general, motors that are operated in hot environments or with little ventilation will experience a dramatically reduced lifespan on their run capacitor. The same can be caused by radiated heat from a generally hot running motor causing the capacitor to run hot. In general, if you can keep your run capacitor cool, it will last a lot longer.
Current - Motor failure causes the capacitor to be overloaded. This scenario is less commonly noticed, as it would usually be accompanied by a partial or complete failure of the motor. The motor is overloaded or has a failure in the windings, causing the current to climb. This can have an effect on the capacitor.
Voltage - This single factor can have an exponential effect in shortening design life. A run capacitor will have a marked voltage rating that should not be exceeded. Let's use 440 volts as an example. At 450 volts, the life may be reduced by 20%. At 460 volts, the life may be reduced by 50%. At 470 volts, there is a 75% life reduction, and so on. The same can be applied in reverse to help increase life by using a capacitor with a voltage rating significantly higher then needed, although to a lesser dramatic degree.
