Adding bypass capacitors correctly to a crossover?

[Dr Morbius]

Not sure if this is the right place to ask but in the picture below ( I will shorten the leads permanently ) I have added a yellow Auricap (0.015uf) to the tweeter and midrange circuits, and wondered if this is the correct way? They are connected in parallel to the two Jantzen capacitors (1uf each for a total of 2uf in each circuit. Btw the crossovers are Klipsch type AA. I’ve also went with black lead to black and red lead to red. Thanks for any help!………….Steve

 

[DVDdoug]

It looks parallel to me! "Red" or "black" doesn't matter for this type of capacitor.

Only electrolytic capacitors are polarized, and only special non-polarized electrolytics are used in speaker crossovers.

You shouldn't need the additional capacitors. What you're adding is insignificant and you probably don't want to change the crossover characteristics.

 

[Dr Morbius]

It seems these bypass caps have helped in that the attack has gotten quicker - not by a lot but noticeable.

 

[Oddball]

It seems these bypass caps have helped in that the attack has gotten quicker - not by a lot but noticeable.

Now that is a curious thing. I can't comment on engineering as have no clue there, but "attack" is interesting? Do you mean dynamics, or some other word that is understandable more easily?

 

[JeffS7444]

Many years ago, inspired by International Audio Review (Wonder Caps!) and the original Paul & Stan PS Audio, I spent hours fussing with bypass capacitors, mostly in the amplification. At the time, I wasn't measuring anything, just evaluating by ear. At the time, I felt I could hear a difference, but was never entirely satisfied, and haven't fussed with bypass capacitors since.

 

[mhardy6647]

The value added, if any, by bypassing larger value (crossover) capacitors with small value caps is even subject to heated debate between subjectivists.
It is another one of those kinda sorta sounds like it makes some objective sense if you squint your eyes just right notions!

 

[G|force]

It seems these bypass caps have helped in that the attack has gotten quicker - not by a lot but noticeable.

Confirmation bias?

 

[Bruce Morgen]

Some professionally designed passive crossovers have small value caps in parallel with the larger NP caps that figure into the LCR arithmetic -- e.g. "C3" and "C4" in this JBL schematic from the 1990s:

 

[Redacted]

Bypass? In the diagram above I see caps in parallel, in which case their values add. Is this not just to get close to a specific desired value when using standard cap values?

 

[Dr Morbius]

Confirmation bias?

Adding the bypass caps to one speaker, switching to mono and letting the wife (who I must admit has good hearing) - she picked the modified speaker with 4 different song correctly each time, and when I said attack, I actually mean dynamics, which is closer to what I meant.

 

[Bruce Morgen]

Bypass? In the diagram above I see caps in parallel, in which case their values add. Is this not just to get close to a specific desired value when using standard cap values?

It is not.

 

[Dr Morbius]

The value added, if any, by bypassing larger value (crossover) capacitors with small value caps is even subject to heated debate between subjectivists.
It is another one of those kinda sorta sounds like it makes some objective sense if you squint your eyes just right notions!

Well, in this instance it seems to work for me. And as an aside, I make my own 14 gauge copper cables with Nakamichi twist lock rca banana plugs instead of spending $1,000 for a 1 meter cable……….. try to save on expenses where I can.

 

[mhardy6647]

Bypass? In the diagram above I see caps in parallel, in which case their values add. Is this not just to get close to a specific desired value when using standard cap values?

No, the notion is that adding the low value bypass (yes, parallel) capacitors has some presumably positive effect on the frequency dependent impedance of the crossover compared to the "main" XO capacitor. Again, it's one of those "on paper, yeah maybe" effects. It certainly shouldn't do any substantial harm.

 

[NTK]

In RF, using multiple different types of capacitors in parallel is used to reduce impedance when frequency is above the range (f_r, the self resonance frequency) where the capacitors no longer behave ideally (the yellow colored straight line in the figure [Fig. 2, a log-log plot] represents the ideal capacitor).

In the example below, the aluminum electrolytic capacitor (330 μF) starts deviating from being ideal at ~10 kHz. Adding a smaller ceramic capacitor (22 μF) can help reducing these non-ideal effects of the larger electrolytic capacitor. However, I doubt these effects are highly significant at audio frequencies.

Source: https://fscdn.rohm.com/en/products/..._capacitor_impedance_characteristics_an-e.pdf

 

[OldHvyMec]

Adding a bypass or (frequency collection cap) has been around for a long time.

Be aware that there is a correct direction to install a non-directional cap. Especially if there is a lot of RF/EF in the area.
They have a natural noise rejection in one direction and will induce noise in the other direction. Anode foil (+), Cathode foil (-)
You will note the cap you use has a red and black lead for a reason. Some are unmarked, and some have just arrows.

FYI, one of the best caps to use is 1 or 2% box WiMa. They may not look like some of the fancy caps, but they sure are consistent and have excellent QC.

For optimal noise rejection using a Non-Polarized (NP) film capacitor (such as in audio crossovers or high-gain circuits), the standard recommendation is to install the capacitor so that its outer foil (or outside foil) lead is connected to the lower impedance path to ground, or closer to the receiver (like a tweeter) rather than the source.

NEVER assume there is no reason for installing an NP cap in a certain direction. Especially in an XO or as coupling and decoupling caps in valve amps.
Valve amps and SS alike can become a nightmare of noise, especially with large boutique caps installed backwards or Anode + side OUT.

I've installed a lot of frequency collection caps, but because of the added wire leads' length, I keep the wire tails as short as possible and mechanically join the mini cap legs as you should by cleaning the main cap leads thoroughly, tightly wrapping the larger cap leads with the FCC or Bypass cap leads, and then heat-sinking each side of the larger and smaller leads before you add solder. The mechanical joining alone should work fine, but the dab of solder is to hold your joint in place and keep the mechanical joint tight.

If you want to do a presteen job, don't forget the shrink tube. I've never noticed a huge change by adding FCC or Bypass, BUT I've surely added a few for some of my over-the-top buddies. I'm currently doing a Cary SLP-05 for shits and giggles, I'm sure the next owner will love the added hupla. I'll stick with my older Mac C20s and C11s (wicked tone controls) they have FCC or Bypass, and so do the older Marantz 7 and 9 series preamps

Just so you know, the term Bypass Cap is normally associated with filter caps, not in the sound path.

Enjoy

 

[G|force]

Thanks @NTK for the thorough explanation. I knew there was good reason to use bypass caps in some applications, but I could not have explained it half as well as you did. And with visuals no less! Good stuff!

 

[A Surfer]

Well, in this instance it seems to work for me. And as an aside, I make my own 14 gauge copper cables with Nakamichi twist lock rca banana plugs instead of spending $1,000 for a 1 meter cable……….. try to save on expenses where I can.

If you were trying to save on expenses would you not just purchase ready made cabling that is available in Walmart or other big box stores?

 

[Dr Morbius]

If you were trying to save on expenses would you not just purchase ready made cabling that is available in Walmart or other big box stores?

Mine has lower resistance. I prefer it that way.

 

[A Surfer]

Mine has lower resistance. I prefer it that way.

I can't imagine there would be the remotest chance of their being any audible difference.

 

[Redacted]

"filter caps, not in the sound path."

So a filter that is not in the "sound path" does what? Sounds like a cost-saving right there, if we are not talking about crossover filters....