So a while back a 250W mercury highbay i dragged home from outside blew up, the capacitor was bad... if memory serves me correctly it was a 13.5 MFD condenser, could i use a 13MFD Condenser as a temporary substitute? or will the lamp dislike it?

It will work so long as the voltage rating is higher or the same, but motor starting caps don't last anywhere near as long as a high quality lamp cap..

Do you have any other size light caps because you can combine them in series to reduce the value or in parallel to increase the value.. here is a simple calculator tool http://www.electronics2000.co.uk/calc/series-parallel-capacitor-calculator.php

And if your good you can add a relay to a multi capacitor setup and add dimming to the light, or if your real good you can get multi-step dimming (with the correct ballast.)

Dont use starting caps, but run caps would be fine. Use higher voltage caps, not only for the higher voltage, but also for the (implied) current rating (for the rated voltage at 60Hz)

the capacitor I'm using came from a 250W HPS lowbay fixture i found at the junkyard, also thanks....

it seems to run very dim, it actually runs better and WAY brighter on 175W gear which leads me to believe that i need to raise the MFD's a tad... maybe i was mistaken, perhaps it was a 16.5MFD Condenser rather than a 13.5MFD? love te yellowish white light this C lamp gives off...

Measure it, some capacitors just degrade over time to useless capacitance values

Don't have capacitance measuring equipment, can it be done with a multimeter??

It can but very crudely. i dont know how far off the result would be...



Way 1 :

Measure the input impedance of the multimeter. This is done by measuring voltage of a known source (say stabilized 5V) through a resistor, and picking a resistor value such that the measurement shows 1/2 of the real voltage. For example, if you measure 5V through a 10meg resistor and the multimeter shows 2.5V, then the multimeter's impedance is 10meg as well. Most multimeters are between 1meg for the cheapo ones to 10meg for the good ones

Charge the capacitor to known voltage

Connect the multimeter to the capacitor and see how fast it discharges. Time the moment at which the voltage drops to some intermediate value. (this can be done by fiming it on video and analysing later on the computer in a video editor)



Ct [ sec ] = C [ F ] * Rmultimeter [ Ohm ]

Vcap (at time t) [ V ] = Vcap (initial) [ V ] / exp ( t [ sec ] / Ct [ sec ])

exp ( t [ sec ] / Ct [ sec ]) = Vcap (initial) / Vcap (at time t)

t / Ct = ln ( Vcap (initial) / Vcap (at time t) )

Ct = t / ln ( Vcap (initial) / Vcap (at time t) )

C = Ct / Rmultimeter



Way 2 :

Take another capacitor of known value of about the same capacity

The good one charged to known voltage, the one being under test discharged

Momentarily short them together, so some charge transfers to the bad one

Measure the remaining voltage on the good one

Initially :
Q [ Coulomb ] = Cgood [ F ] * Vinitial [ V ]

In the moment of shorting the capacitors :
Q [ Coulomb ] = ( Cgood + Cbad ) [ F ] * Vremaining [ V ]

Cgood * Vinitial = ( Cgood + Cbad ) * Vremaining

Cgood + Cbad = Cgood * Vinitial / Vremaining

Cbad = Cgood * ( Vinitial / Vremaining - 1 )



Bad capacitors also often have leakage, if there is leakage in your capacitor, that would at least screw the test results very badly

A very simple test from the Advance pocket HID Ballast troubleshooting guide.. 

CAPACITOR CHECK
1. Disconnect capacitor from circuit.
2. Discharge capacitor by shorting between terminals.
3. Check capacitor with ohmmeter set at highest resistance scale:
If meter indicates a very low resistance which then gradually increases, the capacitor does not
require replacement.
If meter indicates a very high resistance which does not diminish, it is open and should be
replaced.
If meter indicates a very low resistance which does not increase, the capacitor is shorted and
should be replaced.

None of the tests indicate reduced capacity, which is the main way how dry film caps EOL

Ash, I didn't write it, I gave Advance the credit they wrote it in there pocket guide for HID ballast troubleshooting, and it's really just a simple test, but it does give you a good idea if the cap has failed...

But to simplify this process, short the cap leads then you charge the cap with 5 volts from say a USB charger, and then apply a 2.2K resistor across the leads to discharge the cap in 60 seconds the 16.5 mfd will be at just under 1 volt (0.96) across the terminals and the 13.5 will be about 0.66 volts across the leads.. if it's a large difference just let me know the voltage and I'll tell you the cap value, and check the cap to see if it has a self discharge (bleed) resistor already attached to it because this will affect the calculation.. 

No good idea at all. Most PFC caps that are a few years old will pass Advance's test fine, but they are at 1/2 their rated capacity or less..

The Advance test is really a hard-failure go-nogo test. It is true, quite before the "individually fused segments" selfhealing method, the capacitors failed very often that way. The "metalized paper plus oil" capacitors reduced their capacitance really only very tiny bit with each recovery, so before significant capacitance loss it way sooner ended up either as a short circuit, or a runaway reaction caused the capacitor to really go open circuit.
The problem comes with the modern plastic ones, using separate sections (in 100's pF till nF range), each having its own fuse in the metalization structure. There the repeated "recovery" just gradually reduces the capacitance, without tendencies of going hard short and/or runaways. So the result is, the capacitors are kept really till their capacitance drops significantly, without any hard failure.
And if a hard failure occurs, it uses to be such, the damage is plain visible without any electrical test at all...

What I would do is first do the "Advance" test, to screen out hard failures.
Then connect the capacitor in series with an incandescent lamp to 120V mains and measure voltages across the cap and the passing current.
From the current and the capacitor voltage you calculate the capacitor reactance (Xcapa = Vcapa/Icircuit)
so then the capacitance itself (you know your mains frequency; C = 1/(2*Pi*Freq*Xcapa) )

The bulb in series is necessary for a case, when the capacitor tends to go short circuit only above some voltage level (so such fault is not visible on the low Ohm-meter voltage). The power rating should correspond to the expected capacitor value.