If it's nominally 0.15 Farads and you don't have a tester, but you a have a 45 minutes to spare, you can:
1) measure the out-of-circuit cap DC Voltage with a cheapo DMM to verify its state of charge. I'm rashly assuming it's less than 15 Volts.
2) find a 1/2 Watt or larger power, 1k resistor and record its actual resistance to 3 significant figures.
3) discharge the cap through that same resistor and wait 15 minutes for it to fully discharge.
4) meanwhile, using the DMM's 2 Volt DC range, measure and record the unloaded Voltage of a fresh C or D size, 1.5V Alkaline battery. It should be 1.55 to 1.65 Volts, Record the actual Voltage.
5) calculate and record 63% of that measured Voltage
6) using the 2 Volt DC range, connect the DMM across the cap and
7) observing proper polarity, connect that series RC network across the battery
8) noting the start time, watch the cap charge and record how long it takes it to reach the Voltage recorded in step 5). You've just measured one RC product or tau, one time constant.
9) let's say it's 120 seconds. Let's say the recorded R=1.02k or 1,020 Ohms. That means C = (120 seconds)/(1.02k)_or .118 Farads. Round the result off to 2 significant figures or .12 Farads
C, in Farads =
(charge time to 63% of applied Voltage, in seconds)/(recorded resistance in Ohms)
This simple measurement can be thrown off if the capacitance is non-linear with applied DC Voltage, the capacitor is leaky or has a high Equivalent Series Resistance (ESR), Most of the problems can be resolved with technique. You can measure nominal leakage with the DMM on a resistance range. It may take an extra hour. The leakage resistance could also be nonlinear with applied voltage, requiring additional techniques.
For nonlinearity, you can re-measure and recalculate the capacitance at 1.6 V, 6V and 12V and see if the calculated capacitance changes significantly.
ESR may be a bit difficult with a DC technique, unless you charge and/or discharge the cap through a low resistance, high power resistor. Be prepared for some sparks.
There are faster AC methods using a computer's sound board output as a signal source driving the series RC, then feeding the resistor or capacitor voltage back to the mic input and measuring the Voltage and/or phase with an oscilloscope program. Works just fine if you pick the right resistor value and frequency, so the sound board's low frequency roll-off doesn't mess up the measurement. You'll still have the polarize the cap with that same 1.5V or higher Voltage battery.
Just be aware that if you're charging a high C capacitor much above 1.5 V, you're in arc and spark territory. Be careful. A .15 F cap charged to 12 VDC stores ~ 11 Joules of Energy.
Energy = (1/2)(C)(V squared)
If it's discharged into, say, a 10 milliOhm short, that's a nasty 1,200 Amp spark or a peak power pulse of 14 kiloWatts.
Ron
Clearly structured and simply explained method, based on basic laws of physics. This is a wonderful answer. Thank you !
By the way, the four 0.15F capacitors used in this amplifier power supply (two per rail) are charged to 33,5V. This is likely enough to instantly vaporize anything getting in contact between a rail and ground, including a horse.
How do I interpret the measurement results that I get when I connect a 50 mΩ resistor like those under
https://www.isabellenhuette.de/fileadmin/Daten/Praezisionswiderstaende/Datenblaetter/RUG-Z.PDF
in series with the electrolytic capacitors and in series with the subsequent consumer (e. g. power amplifier) when there is a high current requirement) and measure the voltage drop with a scope at this 50 mΩ resistor during dynamic operation (between idle and current bursts of different strengths) ?
Despite the overall increased ESR, meaningful comparative measurements should be possible here if one is able to carry out the correct interpretation with them.
https://www.isabellenhuette.de/fileadmin/Daten/Praezisionswiderstaende/Datenblaetter/RUG-Z.PDF
in series with the electrolytic capacitors and in series with the subsequent consumer (e. g. power amplifier) when there is a high current requirement) and measure the voltage drop with a scope at this 50 mΩ resistor during dynamic operation (between idle and current bursts of different strengths) ?
Despite the overall increased ESR, meaningful comparative measurements should be possible here if one is able to carry out the correct interpretation with them.
Attachments
Last edited:
My choice of a 1k, 1/2 Watt resistor was a compromise between smoking the resistor and the time required to discharge/recharge the cap. Regardless, the cap's discharge time is much greater than the thermal time constant of a 1/2 watt resistor, so if the cap(s) are charged to 33.5 Volts, the discharge 1k dissipation will start out at 1.1 watts and the resistor will dissipate beyond its ratings. Use a 1k, 2 Watt or larger power resistor.
I've done a number of experiments with short periods of resistor and semiconductor overload, using convection, conduction and liquid cooling. In the dark ages, semiconductor transient dissipation data, like Page 9 of the Attached TI App Note 1028, weren't published. We had to roll our own.
In the case of semiconductors, typically forward-biased diodes and Vbe junctions, measure and record Vfwd or Vbe vs temperature at a current low enough to avoid a meaningful IR contribution by the bulk resistance.
The trick is to put the device in operation for a time >> the device's thermal time constant, then quickly drop the current to the previous Vfwd or Vbe test current level, measure Vfwd or Vbe and let the junction Voltage, itself, tell you how hot it is. Then restore full dissipation. It was actually done by taking a few hundred microsecond reading every few seconds, and correcting the power dissipation data for the short loss of duty cycle.
Typical 130kV Linear HV supplies for X-ray tubes had 100 Watt, 1k carbon arc-limiting resistors in cooling/insulating mineral oil. Sometimes the supply would arc and subject a resistor to ~ 150 Amps, i.e. 22 Megawatts, during the arc. That's > 20,000 times the resistor's steady-state power rating in air and certainly > 2,000 times the resistor's power rating in oil.
Carbon resistors manufactured at room temperature can have a tiny amount of internal moisture, i.e. water. In those big resistors, that super-heated water, i.e. steam, would rise to the top of the oil tank as a bubble. You could hear the arc. Little bits of black carbon would blow out of the resistor body and settle in the bottom of the oil tank.
Resistors could survive dozens or even hundreds of arcs until the final one blew the resistor open. That's why they were mounted in metal end clamps.
A 2 watt carbon comp resistor in motor or cooking oil can easily handle a kiloWatt for a hundred microseconds.
Another thing, unmentioned, is dielectric absorption. The Voltage on a fully discharged cap can rise from the dead - sometimes to a dangerous level - or at least high enough to mess up a low Voltage measurement. For safety reasons, HV capacitors are usually shipped and stored with shorting wires.
Last edited: