Short Arc Electronics Capacitor question/Help

It looks like I will get a couple of new lamps in the near future for my two ORC X-500-14 Gunfire Simulators. In the mean time, I am proactively replacing all the power diodes, encased resistors and a bunch of ceramic disk capacitors - items that can fail just sitting over time. My question is, in the pic, there is a 400 Volt, 2 uF Aerovox axial/tubular capacitor sitting above the big orange square power supply. You can make out the "VOX" printing on it. Siliconed to it above are 3 ceramic disk and one axial orange colored capacitors. All of these are hooked in parallel and connect back to insulated terminal pins. There is also what I think is a "drain" resistor wired across the caps. It's kind of a wiring mess. I am wondering why I could not just aggregate the Capacitance of all caps and buy one axial cap of the summed capacitance and achieve the same result? The Aerovox cap is of the "metalized paper" type as I found in some very old PDF data sheets.



Added: After removing some of the black silicone, I can see that the ceramic disk caps all have different Voltage ratings, some are NP0 dielectric, some are Z5U dielectric so it's probably best to replace as originally configured. One cap I can't find:

1D
NP0
1KV

I know lines 2 and 3 but can find nothing on 1D which must designate capacitance. Any thoughts?

Added: After reading maybe 40 sites of Cap code deciphering, I'm finding only 1 site that gives a 2-digit code definition. It would appear to be a 1 Picofarad cap with a "D" tolerance - .5pF. Happy to hear others thoughts.

There are a couple reasons why capacitors in parallel can be better than a single one of the same total capacitance. I don't know what's at play here, but here are some thoughts.

Real capacitors aren't like ideal capacitors. They have series inductance, so at some frequency they start to look like inductors instead of capacitors. Smaller physical size and smaller capacitance both tend to have smaller inductance, thus pushing the crossover frequency higher. Paralleling several different caps can give you effective performance at low, medium, and high frequencies.

It's possible to tune a resonant circuit by paralleling several caps to get a precise value that's either not available, or not within tolerances, or perhaps not known until the circuit is assembled. There's a limit to how effective this can be, as the temperature coefficients and aging of the larger values can easily swamp the 'precision' achieved by the smaller ones. I'm not that familiar with metallized paper caps, but it seems to me they might be very stable over time and temperature.

It's likely that the second largest value is probably a film cap, the top one in the pic. Film and ceramic caps age very well, and I wouldn't replace them just because they are old.

I doubt very much that you have a 1 pF cap. Looking at the pic, I'd guess the wiring alone contributes more than a few pF, and the lead lengths have enough inductance to swamp any low-inductance characteristics of that size cap. Unfortunately, I can't help you with what it actually is.

I'm not sure what you mean by 'encased resistors', but some types of resistors have poor life (when measured in decades) and some have excellent lives. Maybe you could be more specific?

i would leave those alone.
most likely they are custom selected for a critical value.
i doubt those are supply bypass caps but it is common for those to have multiple types and values.
like a big lytic to suppress supply ripple and several sizes of ceramics and or tantalums to get rid of noise/rf from a stage.

Sometimes you see ugly "crutches" in a design, but if you remove them everything collapses

Timing is everything. Just this afternoon before reading your responses, I had received all the caps I ordered to match what is installed. I vacillated for a week on whether I would change them out or not. Had I read your responses, I probably would not have begun the replacement process for one of the lights. To get it disassembled so that I could check capacity of each cap, I had to cut all the leads very short such that I most likely can't re-use them. The new caps readings are fairly close in each case. Unless ORC hand-tested and selected each cap for the purposes of fine tuning these lights, then as long as I replace each cap with a new cap of the same Farad capacity including Tolerance spec, Voltage rating (higher Voltage rating being OK) and dielectric rating (Z5U & NP0's used), then I should be no worse off having replaced them.

Here's another general question. When I tried to find the 2uF Aerovox cap, I had no luck so I ordered a new, 1% tolerance higher Voltage capable cap. The OEM Aerovox is 400VDC rated and the new precision cap is 600 VDC rated. I wouldn't think this would be an issue but I'm not an electronics expert. Any thoughts on that?

You're right about temperature-stable caps. In all of my research to find these replacements, I found out that the NP0 designation (on a couple of the ceramics) is the top of the line with regard to being stable at a very wide temperature range. I used the wrong terminology "encased resistor". It's a chassis mount resistor - those found in the gold metal cases.

There is an AC to DC conversion process being done via a 25 Amp/200 VDC spec bridge rectifier. I was going to replace the Rectifier also. I found the exact part number. However, one light's BR is 25 Amp rated, the other is 30 Amp rated. All other specs are the same. I found the same series BR in a 25 and 35 Amp version and have both available to use. All specs are the same except the current limit. Any reason not to use the 35 Amp unit?

The 400vdc rating is it's breakdown voltage (the voltage at which point it could arc over and self destruct). It generally doesn't hurt to replace it with a higher voltage rating as long as it's the same type of capacitor (ceramic, paper, electrolytic) with the same farad rating. There are a few exceptions in tuned circuits where the change in breakdown voltage could mean a change in it's overall impedance but I doubt that this is the case here.