[gsvarney]

Hi all,
I was thinking about bypass caps on a pair of nice speakers I have from the early 90s. The original crossover is still in place, but I did add a couple of .1uF Miflex KPCU-01 caps to the tweeter filter. All original caps are non-electrolytic. These has a dual voice coil woofer and one of them has a filter with a 12uF capacitor in it. It's a Solen and the speakers still sound very nice. However, I am considering replacing that one, as well. So, I decided to ask Gemini and ChatGPT about this. They came up with suggestions for caps based on my listening preferences and suggested these for lower frequency use:

Mundorf MCap EVO Oil
ClarityCap ESA/CSA
Jantzen Audio Superior Z-Cap
Audyn Cap Plus
Solen Fast Caps (Polypropylene)

The Jantzen looks good, and are about $70 each. So, if I wanted to reduce cost, could I bypass something else with a smaller Jantzen. The response suggested 10uF Solen with a 2uF Jantzen. I questioned the ratio and wondered if for lower frequencies this makes sense. This started a discussion about how this works, which had not really registered  in my brain before. Chat said that the bypass cap operated in a frequency range based on its value, like caps do. So a larger bypass for bass would allow the effects of the bypass to start lower and then go through the upper bass/midrange to the crossover point. So, for my HF bypass, the .1uF Miflex actually appears to start its work at or above the upper audible frequencies or beyond, but have some effect lower. This is what adds the air and details. It also suggested up to .47uF (on 8 Ohm driver) would not affect the crossover frequency much at all, but move the sound quality of the bypass lower which may improve the sound, if that's what I like. Let's say it started acting at maybe 15KHz. I do have, at 60 YO, some high-frequency hearing loss. Would this be a reasonable way to tune the speaker a little more to my needs?

I just never thought about what was actually happening with a bypass cap as far as how the values could tailor the sound in the upper frequencies. What do you guys think?

Thanks,
Geoff

[Early B.]

What do you guys think?

We think you should stop getting audio advice from ChatGPT.

Take a few steps back and tell us what speakers you have from the early 90s. Also, post a pic of the crossover. This will determine if they're worth considering for an upgrade. You may need a complete crossover redesign. 

[gsvarney]

Early - Yeah, I don't just take Chat at face value, trust me. It is a good way to get some info fast. I've often corrected Chat when it gives bad info. I use it to get me started on things as it is fast. That's why I came here. I posted over on the ACI forum and here at GR about my speakers. See here: https://www.audiocircle.com/index.php?topic=187643.0. These are worth working on, they really are good speakers. I'm looking to make them better, if I can. If you've been around audio DIY (speakers) as long as I have you may have heard of Mike Dzurko. Was at the early end of speaker kits. Quality stuff. Quality parts in the crossover. I'll post some pics. Not very well known outside of DIY, especially from the 80s and 90s. I am going to replace the resistors also.
Audio Concepts Sapphire (original)
Do you think Chat was correct in its explanation of how the bypass caps behave?

Thanks

[Early B.]

See here: https://www.audiocircle.com/index.php?topic=187643.0.

You were already given good advice in that thread. Either enjoy what you have with no changes or do a complete crossover overhaul by shipping one of the speakers to Danny. Otherwise, you'll be wasting time and money on band-aid approaches such as by-pass caps and resistors.

[Tyson]

IME, upgrading all parts to premium parts (wiring too) will improve the sound.  For mid/tweeter circuits I find Miflex copper the best.  For larger caps on the bass, the Alumen-Z caps are great.  Duelund resistors are best resistors.   Jantzen wax coils are the best inductors.  And since we're in the GR Research circle, their copper hookup wire is great, too.  I also recommend changing from your standard binding posts to the GR Research pure copper tube connectors. 

The more 'crap' you can remove from the signal path, the better. 

[gsvarney]

Thanks for your input, Tyson. I will look into Alumen-Z as well. I do have good copper wire for this and will rewire everything as part of this project.

[Tyson]

Nice.  Also bypass caps - I've used a bunch of different values and you really don't want to go above .5uF ever.  I usually use anything from .1 to .5 for my speaker bypass caps, depending upon what's on sale  

And don't use anything but copper caps in the bypass position.  IME, the more copper you have in the signal path, the better.

[Hobbsmeerkat]

Nice.  Also bypass caps - I've used a bunch of different values and you really don't want to go above .5uF ever.  I usually use anything from .1 to .5 for my speaker bypass caps, depending upon what's on sale  

And don't use anything but copper caps in the bypass position.  IME, the more copper you have in the signal path, the better.

This.

The main reason we don't recommend larger bypass caps is due to the phase shift it causes within the audible range, as the smaller cap discharges faster than the larger cap.

For the mid and bass components, you can get away with a ratio of 1:100, so if your mid cap is say 33uF, I wouldn't recommend above 0.33uF as you're keeping the phase shift out of the range of the driver(s). some woofer circuits use 100uF values, so you could potentially get away with a 1uF as that network is likely crossing at 200-400Hz range.

It's the same reason we don't recommend pairing 10uF + 5uF caps if you need a 15uF value, as that will definitely cause phase issues within the audible band.
(the better option would be 3x 5uF caps)

Danny always recommends 0.1uF, with 0.22uF being the largest he's okay with for a tweeter cap.

[gsvarney]

Thanks for the clarifying information, guys. So was ChatGPT correct, or close, on the small bypass value impacting very high frequencies, sometimes beyond 20KHz, to add air, etc.? I don't care if it's right or not, but I would like to have experts weigh in. Chat can give food for thought, and I don't consider it to be the expert. I have a tweeter circuit in these speakers with identical models of caps at 3uF and .68uF. In this case, it appears this was done to reach the appropriate total value. However, due to the second cap being .68uF, this does not have a bypass effect, it just behaves as a 3.68uF cap, or close to that. Is this correct.

Thanks.

This.

The main reason we don't recommend larger bypass caps is due to the phase shift it causes within the audible range, as the smaller cap discharges faster than the larger cap.

For the mid and bass components, you can get away with a ratio of 1:100, so if your mid cap is say 33uF, I wouldn't recommend above 0.33uF as you're keeping the phase shift out of the range of the driver(s). some woofer circuits use 100uF values, so you could potentially get away with a 1uF as that network is likely crossing at 200-400Hz range.

It's the same reason we don't recommend pairing 10uF + 5uF caps if you need a 15uF value, as that will definitely cause phase issues within the audible band.
(the better option would be 3x 5uF caps)

Danny always recommends 0.1uF, with 0.22uF being the largest he's okay with for a tweeter cap.

[Hobbsmeerkat]

To some extent, but it's mainly using generalized information that was fed/taken from across the internet, so it takes both good info and bad info and uses the context around it to come up with an answer, the more consistent that information is online, the more accurate it is likely to be, so I would take any answers it gives with a generous grain of salt.

Bypass caps can be used to either bypass caps or bypass resistors so it could be confusing the two.
So say you have a speaker that is a little soft above 8-10KHz, you can add a 1uF or 2uF cap across a resistor in the tweeter circuit to lift the top end up higher, but I wouldn't bypass the tweeter cap with anything larger than a 0.22uF

A 3uF with a 0.68uF can be used to get a ~3.7uF value, but you're still going to get a phase shift near the top of the tweeters range as the 0.68 cap still discharges faster than the 3uF cap will. The better option would be a 3.6uF cap with a 0.1uF as a bypass, you still get the same approximate value, but without the phase shift being in the audible range.

[oldhvymec]

If your on a budget and you want to try something that HELPS, just add Bypass caps to the mids and highs.

There is seldom a need for any adding to the sub/bass section other than making sure you have high enough
voltage just in case.

I was taught WAY back the proper term was a "frequency collection cap" and "Bypass" was in the power supply
for cleaning up the noise and voltage spikes, to maintain a stable voltage.

I suggest you study the crossover and just add .1 to .01 "Frequency collection caps" to the mids and highs using
good old 2% WiMa polies and put a smile on your face. FEW caps in that position sound better than that cap. Believe
it or not. All the fancy caps are wonderful but very few are better for doing exactly what you're trying to do than a
1-5.00 usd WiMa box cap.

The expensive way and one that I currently use on my Cary SLP-05 (https://partsconnexion.com/duelund-0-01uf-600vdc-jdm-series-silver-foil-wax-oil-precision-capacitor/) was a nice addition to there already upgraded parts list
that cost close to 1800.00 dollars if they do the work and of course shipping. I spent under 300.00 and used better parts. I went with tinned copper and a silver frequency collect cap that I posted the link to. BTW the term is an old Marantz term used on the 7 and 9 series pre and power amps. 7C etc.

I'll tell you another difference if you like to add a razor-sharp finish is a "Teflon". CJ amps and power amps used them
on coupling and decoupling positions for the valves. They take a while to form but they are extremely transparent, but they take patience . 2-400 hours sometimes.
I've seen them used in XO too. EX; "TRT" in VMPS speakers

Have fun!

Regards