If you use a smaller cap the crossover frequency will be higher. Larger, the frequency will be lower. Film caps are better if you have space .
Stay as close as you can to the intended value. You can go over the Voltage.
You can replace electrolytics with film caps but you must pay attention to measuring the ESR correctly. Especially with caps that go towards ground and are not in series. A Dayton Audio DATS will measure this for you.
I've heard good things about replacing electrolytics with film but I wouldn't go too crazy. Mundorf MKP or Clarity SA is about as expensive as I'd go.
I recap many Klipsch crossovers and use mainly Audyn red and Dayton 1% poly caps from Parts Express. I also use them in the woofer circuit rather then those electrolytics (normally high ESR) so many companies seem to use. They can be pretty large so sometimes I have to fasten them down where I can and run insulated leads to them. You have an odd value you can’t find then use two that add up. 30uf + 38uf soldered in parallel for instance will give you 68uf and the bonus is two hooked up in parallel will give you lower ESR then one 68uf cap will. I have found this to be true with any combination of values in that two caps always are lower in ESR then one.
Use the values on the board and be safe.
Now just for the heck of it with a two way speaker I designed recently I wanted passive crossovers. It seemed with the components I selected that the crossover calculators never had those nice round even numbers that so often appear with commercial passives. So figuring the math meant something I used a combination of caps that would get me right on calculated values and not round them off. It might take three to get my 13.74uf value for instance. Then since the difference in db efficiency between woofers and horn drivers is normally pretty large I did a fixed resistor L-Pad doing the same thing to come as close as possible to calculated values. The difference in fidelity is quite apparent and on those L-Pads a .5db attenuation to much or to little can make a big difference.
Anywhere you put a decent polypropylene cap, it will be better than a good electrolytic, period. In old speakers the argument of esr is normally not as valid except maybe in some correction circuits. The old drivers and crossover parts have all drifted a bit. ESR is the resistance of the part itself and it is possible to effect the total resistance of the circuit. Normally, you are not looking at more than a 1/4 ohm or so.
If ESR is a concern, when replacing electrolytic with a MKP (Polypropylene) you can round down a bit, say you are replacing a 10 mfd electrolytiic, if you use a 9.7 or 9.8 mfd mkp, you should be fine. The larger the cap value the more that ESR could change. So in larger sizes, you may drop .5 mfd. For those that may not understand "why" a driver, may be called 8 ohm, but in reality it varies and may very well average more like 6 to 7 ohms, every part that you add to the crossover adds resistance, between caps, coils and resistors, you could very well end up with 8 ohms or more.... Good designers figure this in when designing crossovers. If you are really worried, i would recommend that anywhere there was an electrolytic to use a New, good quality Electrolytic and use a value a bit smaller, say 10 to 20 percent lower than the original and then add a good film & foil or mkp to bring the total micro farad rating up to the original part.... So for a 50 mfd electrolytic, use 40 to 45 mfd, electrolytic and add 5 to 10 mfd to match the original rating. This is still preferable to no film at all. This should make a difference, good luck, Tim
I disagree, based on hands on experience with @timlub a little.
It's important to note exaclty when ESR matters a lot and when it doesn't so much, and also how much we are talking about. It's usually under an ohm for all types of signal caps, but higher for electrolytics generally, especially older one's.
In series with a driver half an ohm difference isn't a big deal. The problem shows up with caps going to ground. 2nd and 4th order caps in low pass filter and zobels. A change of 0.4 Ohms there can really change the impedance curve. This is honestly hard to imagine without the right simulation. I am afraid I lack the files and memory right now to find them. In a circuit you'll often see these caps have a very small value resistor in series. That's where you want to be especially careful.
In the middle of the road here is caps in series with a woofer, whose ESR may act as a bit of a baffle step compensation. See if you can find the writing by Troels Gravesen where he discusses using this effect.
@jim94025 I just wanted to reiterate again what I posted. Its hard to know who to listen to or not. I've been building..... building speakers for 45 years and also remember, I said "except in correction circuits"
The benefits of replacing electrolytic capacitors is noticeable, Make the upgrade and don't look back. Again, as I posted earlier, if you are really worried, use new electrolytic and use a polypropylene to get your value, you'll still hear it. Good luck, Tim