Ok, i understand your meter then probably measures impedance instead of ESR ( like most do, but its easy to test, a real esr meter has no prolem at all measuring the ESR of a 100 nF cap). They do so because it is the most easy way and Xl is to neglect at 100 KHz so you can measure impedance instead, many datasheets even state impedance at 100 KHz, only the smaller the capacitance, the less you can neglect Xl. And if it does measure Z and also not uses 4 wire technic i would not trust it other then bad/good! like a ESR of 10 Ohm will be bad in 99% of the cases but under 1 of 2 Ohm you have no choise then take the gamble, especially if there is no datasheet or you have no lcr bridge to measure D.
Never disconnect the earth of your scope. The mains potential is then on the chassis and you touch that. Use an isolation transformer to feed the patient ( let the scope allway earthed) or differential probe. ( suited for this voltage/frequency)
I have scopes that can meaure this even with a 1:1 probe but most of the Chinese toys will not survive that. Look in the manual for max voltage and also for the max value of the probe and the right CAT ratings.
My Agilent multimeter has no problem measuring the AC ripple on 350V, but if yours beeps ect do not try again, if protection is not very well inside it can be over in a flash, i used to have a Voltcraft meter that " flashed" that was the last B-brand meter I ever will buy (and it was a top of the line voltcrap scopemeter costing >150 euro)
DC leakage is just DC leakage. An ideal cap has an infinite DC resistance ( Rp) in real life this is very high. If you connect a cap to a DC powersupply, use a current limmiting resistor and connect a sensitive currentmeter ( or measure the voltage dop over the resistor) in line you will see at first a higher current, hem max that the resistor allows and following the RC time the cap gets loaded. You see the voltage over the cap rising and the current decreasing. After some time ( like a minute or more ) the current must be within the datasheet specs but there are some "rule of tumbs.
Like max 0.001 x V x uF = uA but if a cap leakes over 50-100 uA i replace it.
My standard procedure is measure ripple current ( not if I must know if caps are right before start up)
then I lift one leg of the cap, measure DC leakage ( and if neccesary reform the cap if the instrument is long not used) then measure capacitance and DF using a HP or GR bridge. From the DF and some calculating i decide for an unknown cap what to do. And then for fun and to see if they are right I take one of my ESR meters and measure ESR.
I have seen caps with perfect capacitance and ESR, not leaking DC at half the working voltage but at the working voltage leaking over 15 mA. If DC leakage is real bad, ESR will be super good.
Caps that have perfect ESR, do not leak DC but lost most of theire capacitance ( most time electrode ripture, a chemical problem.
If it is lost capacitance because total loss of oxide layer then there often is DC leakage.
Upto now I have found lots of caps leaking DC, most of the rest lost capacitance and i think about 10 with a high ESR and 4 of those were in an old HP bridge, the others in consumer electronics, but 99% of my repaires are instruments. I know bad ESR is a bigger problem in consumer electronics, but not so bad as many people think, our ESR meter just let us think it is a big problem. And because this way we replace many more as nessecary we most times also replace the DC leaking or lack of capacitance caps so the result is a working monitor and we are happy ;-)
i know, i'm wrong because you all have good ESR meters and I talk bullshit, but for everyone who does not agree, i do not care,
only if you measure DC leakage, capacitance with a real good meter ( and ESR not in situ and at he right frequency, mine is adjustable from 10-100KHz, so I can measure at the frequency the datasheet gives and under 10KHz I use a bridge) and then if you still disagree tell me %-B