Tweakers' Asylum

Tweaks for systems, rooms and Do It Yourself (DIY) help. FAQ.

Return to Tweakers' Asylum

Message Sort: Post Order or Asylum Reverse Threaded

Input coupling caps as high-pass filter

78.29.224.52
Posted on February 2, 2017 at 11:34:26
KanedaK
Audiophile

Posts: 2519
Location: Brussels
Joined: April 27, 2010
 I have a trends ta10.1 that drives the midrange and treble on my 3 way horn speakers. Trends goes full range and the low pass is filtered out by the speakers passive filter.
The passive x-over between bass and mid is 400Hz.
Bass horns are powered by a different amp trough an active crossover.
I tried full active but this configuration sounds so much better.
I would like to remove some bass frequencies going to the Trends to release it from unnecessary stress and distortion.
I was thinking changing the input capacitors so that it creates a 300Hz high-pass filter.
Would this work and which caps would you recommand for maximum sound quality?
My trends is completely modded and uses high quality film caps but I forgot the brand. They are yellow with an AN logo - but no they're not audio note caps, if such a thing exists.

 

Hide full thread outline!
    ...
RE: Input coupling caps as high-pass filter, posted on February 2, 2017 at 18:54:52
Tre'
Industry Professional

Posts: 17293
Location: So. Cal.
Joined: February 9, 2002
 That's what I do.

The input cap value is computed against the input impedance (resistance) of the mid/tweeter amplifier (300b SET in my case) and creates a 6db per octave high pass filter at 200Hz (my crossover point for my JBL 2231 woofers).

The woofer amp is feed from a active (opamp based) 200Hz low pass filter.

Equation for simple 6db per octave high pass filter,

-3db crossover point in frequency = 159155 divided by (cap value in uf times resistance)

In my case the input impedance (grid resistor of the first stage) is 220k ohm. A .003535uf Russian SSG silver Mica cap in series with the input signal forms a 6db high pass filter with a -3db point of 204Hz (close enough). The cap is installed in the amplifier between the RCA input jack and the 220k grid resistor.

This keeps a lot of bass frequency out of the 300b amp.

Most series feed single ended (gapped) output transformers don't have enough inductance to play bass well and even though the speakers won't respond to the lower frequencies, if those frequencies are allowed to get to the output tube (and the output transformers) all the frequencies will suffer in terms of distortion because the lack of inductance causes the load line to become elliptical.

In the presents of low frequencies all the frequencies will follow that poor elliptical load line.

Blocking those bass frequencies from the amp made a large improvement in the sound of the mids and highs without resorting to adding opamps to the mid/tweeter signal chain.

BTW The Russian SSG cap are the most transparent caps I've ever heard.

Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"

 

RE: Input coupling caps as high-pass filter, posted on February 3, 2017 at 09:46:01
Hornlover
Manufacturer

Posts: 2529
Joined: March 8, 2002
 This certainly can work. It is best to use a scope and SG. If no scope, a multimeter will work to, since you are just looking for the -3db point. I did this on the maggies I used to have, making them -3db at 70, and using a sub below. By removing some of the low end from the amp, it will also clean up the mids and highs somewhat as well.

 

You would be crazy not to, posted on February 3, 2017 at 12:20:50
rickmcinnis@dogwoodfabrics.com
Audiophile

Posts: 3040
Location: Atlanta
Joined: December 15, 2003
 This is the most elegant approach.

I am using Room Equalization Wizard to see what my horns are actually doing in my room.

Textbook values will be close but it is good to be sure.

I am using the JUPITER copper foil caps at the moment. I have had to parallel another small cap to get the right value.

Sometimes the nominal input impedance is not as constant as you would think so it is a good idea to measure what is happening in the room and at the crossover point.

My amps require input and output caps so I have sized them both for my high pass crossover.

 

RE: Input coupling caps as high-pass filter, posted on February 3, 2017 at 18:50:19
KanedaK
Audiophile

Posts: 2519
Location: Brussels
Joined: April 27, 2010
 Thanks for the encouraging answers!
Online calculator gives me a 7,96pF for a 400Hz -3dB point at 50k impedance.
I'm unsure about the calculator. Does this sound like the right value?
I guess if it is, I need to hunt for 8pF caps.
Would non-audiophile silver-mica 5% tolerance caps be worthwhile? Or do I need to aim at exotic vintage Russian rarities?

 

300hz is too high, posted on February 3, 2017 at 18:56:56
madisonears
Audiophile

Posts: 1587
Location: midwest
Joined: September 6, 2006
 Of course this makes sense, and is what I use, along with the acoustic roll off, in my own system as the entire high pass filter to my mid/treble amps and speakers. Be aware that any electrical filter will act in conjunction with existing speaker system response. In other words, if there is already a filter as part of the speaker system, it has a transfer function that will be affected by whatever other filters you add. You don't state what the 400hz filter's slope is, but it will be affected by the 300hz filter at the amp input.

Just as a goal of speaker xovers is to set any filters outside a two octave window of flat response, so should this amp filter be designed accordingly. I wouldn't try anything higher than 100hz. You still will be removing a considerable portion of low frequency energy demand from your amp. For a 100Hz corner freq, you would need a .03uF cap into 50k ohms. Standard cap value is .033uF, which would give 96Hz corner.

And yes, the Jupiter copper foil/wax paper caps are superb for what you're doing. There might be better caps, but I have not heard them.

Peace,
Tom E
berate is 8 and benign is 9

 

RE: Input coupling caps as high-pass filter, posted on February 4, 2017 at 15:20:26
madisonears
Audiophile

Posts: 1587
Location: midwest
Joined: September 6, 2006
 Not sure what calculator you used, by I get a -3db of almost 40k hz with an 8pf cap into 50k ohms!

Tom E
berate is 8 and benign is 9

 

7,960pF not 7.96pF (nt), posted on February 4, 2017 at 15:58:56
Tre'
Industry Professional

Posts: 17293
Location: So. Cal.
Joined: February 9, 2002
 .
Have Fun and Enjoy the Music
"Still Working the Problem"

 

RE: 300hz is too high, posted on February 5, 2017 at 06:58:06
KanedaK
Audiophile

Posts: 2519
Location: Brussels
Joined: April 27, 2010
 Damn.

I followed your advice and bought a bunch of NOS EROMAK polycarbonate film/foil on the big ayction site.
but I misread and made a mistake: they are 0,33uF, not 0,033uF :( so stupid.
So i guess tht would give me a useless corner frequency of 9,6Hz, right?

Maybe I can cancel my purchase. Except if I can use them for something else, but the value is too small for a CD player output coupling cap, right?

By the way the passive high-pass filter at speaker level is 6B/octave.

 

RE: 300hz is too high, posted on February 5, 2017 at 09:09:09
madisonears
Audiophile

Posts: 1587
Location: midwest
Joined: September 6, 2006
 That's happened to everyone at least once! Don't sweat it. We all waste more money than we should on this obsession. How much are they worth, a few bucks? You could keep them around for trying as bypass of other caps.

Eventually, if this configuration works properly and sounds good, you will need to spend serious money on really good copper foil caps because the entire mid/high signal will pass through them.

What is the configuration of the current passive filter? Please supply complete, specific information. Why do you want to add another filter if you're already shelving low freq's by 6db/octave below 400hz?

Peace,
Tom E


berate is 8 and benign is 9

 

RE: 300hz is too high, posted on February 5, 2017 at 18:09:02
KanedaK
Audiophile

Posts: 2519
Location: Brussels
Joined: April 27, 2010
 I want to add another filter because the 400hz high-pass is after the amplifier.
The amp is fed full range and I wanna remove some low frequencies to put less stress on it.

All is well, I've just ordered some Russian PIO caps, apparently the "best" ones.
I'm curious to try them - besides filtering the lows, they will most likely sound better than the Arcotronic caps I have now.

 

RE: 300hz is too high, posted on February 6, 2017 at 10:45:12
madisonears
Audiophile

Posts: 1587
Location: midwest
Joined: September 6, 2006
 You didn't mention the value of the PIO caps you ordered. If you use the .033uF (that's 33 nano) as I recommended, the input to your amp will be down 1/4db @ 400hz, 1/2db @ 300, 3db @ 100, 9db @ 50, 15db @ 25hz, so that is a decent reduction of low freq energy without significantly impacting existing filter slopes, although the response of the speakers will be affected somewhat in that their lowest frequencies will be reduced as well. This will probably have a greater impact on perceived clarity than reducing input to the amp. All small speakers distort when tasked with reproducing low frequencies, whereas most decent amplifiers do not.

Depending on what it is now, you might need to tweak the response of the bass xover/amp by raising the xover point (not terribly difficult), adjusting the slope (pretty complicated), or merely increasing the volume a bit (easy!) to compensate for the increased roll off of low freq's from the mids.

I doubt that Russian PIO is the best because there is no such thing (I assume that's why you used quotes), although they may be the best value. Improving on the sound of Arcotronics caps is no tall order, so you should be pleased. If you care to eventually invest more, I guarantee you would be even more pleased with copper foil.

Peace,
Tom E

berate is 8 and benign is 9

 

RE: 300hz is too high, posted on February 6, 2017 at 13:55:25
KanedaK
Audiophile

Posts: 2519
Location: Brussels
Joined: April 27, 2010
 When my speakers were fully passive, they relied more on natural roloff of the various horns to create the slopes, than on the first order crossovers - mainly there to stop the robust Atlas midrange driver from frying. So I don't think whatever happens lower than 400Hz is making really much difference ( wére not talking extreme slopes here).
The reason I went active on the low pass is because I changed bass horns, and the new ones don't roll off so quick above 400Hz - I had to have a steeper slope than 6dB/oct because it sounded like crap.
But my 8 watts amp is now driving the high pass only and there's 50watts on the bass horns, and I like the added power for sure!

 

RE: 300hz is too high, posted on February 6, 2017 at 15:31:08
KanedaK
Audiophile

Posts: 2519
Location: Brussels
Joined: April 27, 2010
 But yes I ordered 0,033uF as you said :)

 

Page processed in 0.038 seconds.

[ Top Top of Page ] [ Contact Us ] [ Support/Wish List ] [ Copyright © 2024, Audio Asylum, all rights reserved ]
[ General ] [ Speakers ] [ Tubes ] [ Vinyl ] [ Digital ] [ Hi-Rez ] [ Video Asylum ] [ Cables ] [ Tweaks/DIY ] [ Music ] [ Films ]