your scope input capacitance is most likely about 20pf. if you ever look at probe cable (do destructive testing like this on a known BAD cable) on the inside it is a coaxial cable with a SINGLE STRAND of wire in the center, which gives it a capacitance of about 1pf/foot. with the 17 pf/foot cable you mentioned, you will be degrading your signal after the first 1.2 feet. go to www.linear.com and do a search for Application Note 47. there is a whole chapter in that document about probes, cables and compensation.
as to your cap problem, you need to measure the DC voltage across it and check to see if there's any AC going across it. you want to do these tests with the load disconnected (i.e. disconnect the scope from the power supply). the first test will be with the cap connected normally. turn the power supply on and measure the DC voltage across the cap. it should be at least 10% below the rated voltage of the cap. if so, shut the power supply off, and lift the + side of the cap and put a 10 ohm resistor in series with the cap (one side of the resistor to the + lead of the cap, the other end to where the cap was connected). power the supply back up, and measure the DC voltage across the resistor. for every volt across the resistor you are reading 100ma current (or for every millivolt it's 100 microamps). this is the DC leakage current through the cap. next measure the AC voltage across the resistor. with the same scaling, calculate the AC current through the cap. also with the 10 ohm resistor in circuit, measure the AC voltage where the cap normally connects (referenced to ground this time). (btw, with the 10 ohm resistor in circuit, you are simulating a cap with 10 ohms of ESR, which while not exactly a normal condition, it will work normally enough to make your measurements).
your DC leakage should be less than 1ma (the lower the better). your AC current through the cap should be less than 100ma. if it's much higher than that (like 500ma or higher) this might be a sign something is seriously wrong. your AC voltage at the point where the cap normally connects should be less than 20% of the rail voltage. if it's much higher and the AC current through the cap is high, i would suspect the rectifiers. if the DC voltage is too high, is there a regulator between the rectifier and the cap? the regulator might be outputting enough voltage to exceed the voltage rating of the cap. if the DC leakage current is too high, the cap has too much leakage and should be replaced.
if it were me, i would have replaced the rectifiers when i replaced the cap.
i would try hooking some wires to the + and - sides of the cap and run them out to a multimeter and monitor the cap voltages (AC and DC) until the cap begins to overheat again.
as to your cap problem, you need to measure the DC voltage across it and check to see if there's any AC going across it. you want to do these tests with the load disconnected (i.e. disconnect the scope from the power supply). the first test will be with the cap connected normally. turn the power supply on and measure the DC voltage across the cap. it should be at least 10% below the rated voltage of the cap. if so, shut the power supply off, and lift the + side of the cap and put a 10 ohm resistor in series with the cap (one side of the resistor to the + lead of the cap, the other end to where the cap was connected). power the supply back up, and measure the DC voltage across the resistor. for every volt across the resistor you are reading 100ma current (or for every millivolt it's 100 microamps). this is the DC leakage current through the cap. next measure the AC voltage across the resistor. with the same scaling, calculate the AC current through the cap. also with the 10 ohm resistor in circuit, measure the AC voltage where the cap normally connects (referenced to ground this time). (btw, with the 10 ohm resistor in circuit, you are simulating a cap with 10 ohms of ESR, which while not exactly a normal condition, it will work normally enough to make your measurements).
your DC leakage should be less than 1ma (the lower the better). your AC current through the cap should be less than 100ma. if it's much higher than that (like 500ma or higher) this might be a sign something is seriously wrong. your AC voltage at the point where the cap normally connects should be less than 20% of the rail voltage. if it's much higher and the AC current through the cap is high, i would suspect the rectifiers. if the DC voltage is too high, is there a regulator between the rectifier and the cap? the regulator might be outputting enough voltage to exceed the voltage rating of the cap. if the DC leakage current is too high, the cap has too much leakage and should be replaced.
if it were me, i would have replaced the rectifiers when i replaced the cap.
i would try hooking some wires to the + and - sides of the cap and run them out to a multimeter and monitor the cap voltages (AC and DC) until the cap begins to overheat again.