Suggestions for a passive line-level high-pass crossover?


Looking for suggestions on a passive line-level RCA high-pass crossover that I can insert between my preamp’s outputs and the amplifier that is driving the mids and highs of my bi-amped speakers. I don’t need anything fancy or overly complex, but I am looking for audiophile-grade circuitry that will have the least sonic signature.

 

Specifically, I need it to be a 100 Hz high-pass crossover point, with a slope at of at least 12 db per octave. Doesn’t need to have a level control, adjustable settings, etc. Simple, but high-performance parts and sound.

 

Any suggestions?

 

Thanks gang!

 


128x128mhwalker

A passive high-pass filter is commonly accomplished via the insertion of a single capacitor on the input jacks of the main speaker’s power amp, the value of the capacitor determined by: 1- The desired x/o frequency; 2- The input impedance of the power amp. However, this method creates only a 1st order, 6dB/octave filter. For a 12dB/octave (2nd order) x/o, filtering is best accomplished via an active x/o.

A reasonably priced but also high-quality active x/o is the First Watt B4 (all discrete circuitry---no IC’s, no Opamps), which provides all the filters one should need: 1st, 2nd, 3rd, 4th order (6, 12, 18, 24 dB/octave) slopes from 20Hz to 3200Hz, in 25Hz increments. The B4 retails for $1500, but Reno Hi-Fi (a main retailer of First Watt products) sells it at a discounted price. It is about to be discontinued, so get one while you can!

The following manual(for the Dahlquist DQ-LP1) contains a formula for, and a chart of, the values(capacitors and if needed, resistors) that will give you your desired crossover point(BUT- 1st order), based on the input impedance of your mid/high amp. Keep in mind that greater than a 1st order(6dB) slope will cause greater than 90 degrees of phase shift, in your mid/high drivers. For instance, with a 2nd order(12 dB/oct) filter, you may find the best sound with either the mid/high or bass driver connected out of phase, unless they were already designed for a higher order slope. I had always used polypropylene film and Vishay metal film resistors(when required), to achieve the absolute minimum signal distortion/degradation. I have installed such filters inside amps, at the RCA terminals, to eliminate the need for additional connectors and cables. See, "technical" on page 4: (https://sites.google.com/site/mpbarney/home/dahlquist-dq-lp1)  Don't know if you can use any of that, but- what the hell.......  =8^)

Bi-amping vertically (rather than horizontally) does not negate the need for a x/o to accomplish. Neither of the two amps in a bi-amped system knows whether it is being used vertically or horizontally (other than in the sense of how the amp and speaker interact with each other), and that consideration has nothing to do with the x/o itself.

Using one amp for both the low-pass and high-pass on the same speaker (vertical bi-amp) has it’s advantages, and using one amp instead for either the low-pass or high-pass on both speakers (horizontal) has it’s. Using a solid state amp for the low-pass (bass) signal on both speakers, and a tube amp for the high-pass (mids and highs) on both, works great with many speakers.

" Bi-amping vertically (rather than horizontally) does not negate the need for a x/o to accomplish. "

Unless you have some really odd equipment, that's not true. There's no xovers on my vertical biamp system, or any other one I know of.

" Neither of the two amps in a bi-amped system knows whether it is being used vertically or horizontally (other than in the sense of how the amp and speaker interact with each other), and that consideration has nothing to do with the x/o itself. "

The amp may not know, but you need to.

" Using a solid state amp for the low-pass (bass) signal on both speakers, and a tube amp for the high-pass (mids and highs) on both, works great with many speakers."

You'll get sound, but its usually an unbalanced train wreck. When you put an active xover in the system, you can get the levels right, but now you have to listen to one amp on the mids and highs, and another on the lows. Who would want that? Do it right the first time and be done with it. Also, don't forget that the active xo will have a big impact on SQ. It will negate a good portion of what your source, cables and preamp are doing.

The lucky owners of Vandersteen 7’s, for one. That speaker comes with a powered (via a solid state amp) subwoofer, and many 7 owners use a tube power amp for the upper drivers. People have done that successfully for decades, with speakers like the Infinity RS-1b, for one.

sfall, if you don’t have an external x/o splitting the signal into two or three sub-signals, you are sending a full-range signal to all the drivers---not a good idea! Are you sure you’re not just using a pair of stereo amps to in effect bi-wire your speakers? Bi-amping absolutely requires an external x/o, whether passive, active, or a combination of both. If an external x/o (external in the sense of not inside the speaker. A passive filter can be inside the power amp, on it’s input jacks) is not present, bi-amping is not, by definition, being accomplished. The outboard x/o replaces the speaker's internal one, splitting the full-range signal from the pre-amp at line-level and feeding the divided signal to one amp for the low-pass signal (bass), and one for the high-pass (mids and highs, typically), rather than at speaker-level (after the power amp---inside the speaker). Each amp then is fed only the frequencies intended for the driver(s) it powers. That’s the whole purpose of bi-amping!

Bi-amping with an active x/o (or passive) can be done vertically with a pair of identical amps if one so desires.

if you are doing it to integrate a subwoofer don't bother.  just find out what the acoustic crossover of your speakers is (as measured in your room) and set the subwoofer accordingly 
" The lucky owners of Vandersteen 7’s, for one. That speaker comes with a powered (via a solid state amp) subwoofer, and many 7 owners use a tube power amp for the upper drivers. People have done that successfully for decades, with speakers like the Infinity RS-1b, for one. "

I thought it would be self evident that I was referring high's, mid's and low's, and not sub frequencies.  I'm one of the people you are referring to that's been doing it for decades. I did it with Vandersteen, Genesis, Martin Logan, B&W, Unity, Audio Physic and probably a few others that I'm forgetting. 

" sfall, if you don’t have an external x/o splitting the signal into two or three sub-signals, you are sending a full-range signal to all the drivers---not a good idea! "

That's just not true. The only way that would happen is to bypass the internal xover.

" Are you sure you’re not just using a pair of stereo amps to in effect bi-wire your speakers?"

Yes. The only way you can biwire a speaker is to use speaker cables. That's why they call it biwiring.

" Bi-amping absolutely requires an external x/o, whether passive, active, or a combination of both. If an external x/o (external in the sense of not inside the speaker. A passive filter can be inside the power amp, on it’s input jacks) is not present, bi-amping is not, by definition, being accomplished. "

That's completely false as well. For starters, you can look at my system. No expernal xover to be found. I don't know where you are getting these definitions from, but you may want to find a new source.

" The term “vertical biamping” was coined to describe a biamp configuration where one channel of a stereo amplifier drives the bass section of a single loudspeaker and the other channel drives the treble section of the same speaker, as opposed to the commercial biamp configuration where one channel of a stereo amplifier would drive the bass section of the left speaker and the other channel would drive the bass section of the right speaker, and another amplifier would drive the treble sections of these same speakers.

Passive biamplification means that an electronic crossover is not used and the signal is divided into frequency bands by the passive crossover within the speaker. This allows the passive crossover network to perform its job of equalizing the response of individual drive elements and correcting for phase anomalies.

One channel of the left stereo amp drives the bass section of the left speaker and the other channel of the same amplifier drives the treble section of the left speaker. One channel of the right stereo amp drives the bass section of the right speaker and the other channel drives the treble section of that speaker.No external crossover is required and the internal crossover network in each speaker is still utilized. The advantages are many.The heavy current demands of the bass sections of each speaker can now be shared by two amplifiers and two power supplies. Each midrange/tweeter section will now have a dedicated amplifier channel. This will reduce cross talk for better imaging and provide a substantial increase in dynamic range while minimizing the potential for speaker damage caused biamplifier clipping. If the bass amplifier clips it won’t deliver high energy, high frequency distortion to the tweeters—the primary cause of speaker failure." 

APJ 4 P.16

It's pretty clear, but if you have info that says otherwise, I'll certainly look at it.



sfall, I complete understand what you’re saying (except for one matter, that of you using a two channel power amp on each of your speakers, without a line-level xover), but you aren’t getting what I’m saying. First of all, the op stated he was looking for suggestions for a passive LINE-LEVEL high-pass crossover for use at 100Hz. I used upper-case for line-level to draw attention to the fact that the op is talking specifically about that type of xover. A line-level xover is completely different from the speaker-level x/o that you are talking about. Not only that, the suggestion that the op "Do a vertical biamp instead. You won’t need the xover." makes no sense, and for a couple of reasons.

First, his speaker contains no internal, speaker-level x/o with a filter at 100Hz! The op states he wants a line-level x/o to put between his pre-amp and power amp, to filter out the sub-100Hz frequencies from reaching his power amp and speakers. 100Hz being a very common frequency at which to add a sub, we can safely assume that is the reason for the need for a line-level xover.

Let’s define our terms: A speaker-level xover is the common filter installed on the inside of a loudspeaker. It receives the full-range (all frequencies) amplified signal from the power amp, separates it into either two or three parts (for a 2-way or 3-way speaker respectively), and sends the filtered signals to the speaker’s drivers. In this scenario, the power amp is "seeing" and amplifying all frequencies. Many speakers can be used in ONLY this fashion---the speaker’s one set of binding posts receiving the full-range signal from the power amp, the speakers internal xover then dividing the signal into frequency bands which are then sent to the appropriate drivers. The filtering is done with passive parts---capacitors, resistors, etc. These parts not only divide the already fully-amplified signal received from the power amp, they are also used to create the balance between all the drivers, the output level of each driver. The parts are also used to create the "order" of each filter---1st-2nd-3rd-4th, which is 6-12-18-24dB/octave.

A line-level xover (whether passive or active) is a separate component, external to the speaker. It receives the full-range (again, all frequencies) line-level signal from the preamp, separates that signal into two or three parts, and sends each signal to a separate power amp, one amp for each frequency group. Each power amp then amplifies only the frequencies it receives, and powers only the driver(s) it is hooked up to. Line-level xovers can be active, or passive. The xover frequencies and filter slopes can be set---no choices provided, or adjustable.

Here’s an example of the difference between a speaker-level xover and a line-level one: My first bi-ampable speaker was the original Magneplanar Tympani T-1. It came with an internal speaker-level symmetrical 1st-order (6dB/octave) xover at 3200Hz. There were two sets of input jacks on the speaker, with a jumper cable connecting the two. If not bi-amped, the jumper was left in place, and a single power amp connected to one of the set of jacks. Audio Research, the Magneplanar distributor at the time, recommending bi-amping the T-1, and offered a passive line-level xover, the PC-1, with which to do so. The PC-1 created a simple, symmetrical 1st-order filter at 1000Hz, the recommended bi-amp xover frequency. The front panel contained four level knobs, high and low frequencies for two channels. When bi-amping, the jumper connecting the bass and m/t panels was removed (which defeated the speaker-level xover), each speaker jack receiving it’s own amplifier channel output cable.

The ultimate ARC/Tympani T-1 system at that time was with the speaker bi-amped with a D-51 on the m/t panels, and a D-75 on the bass panels, in horizontal bi-amp fashion. I couldn’t quite swing that, so had a pair of D-51’s, which I used in the vertical bi-amp fashion you recommend. Yes, I completely understand your rational for recommending the vertical arrangement, and it has it’s advantages. My dealer (Walter Davies, now known for his Last record care products) suggested it as it allows the left and right channels of the amp to not be in competition with each other on power-demanding bass transients, as they are if used horizontally. That leaves more of the amps power reserves available for the bass channel of each vertical amp. On the other hand, having one channel of an amp seeing low frequencies and it’s other channel seeing the mids and highs, as it does in the vertical arrangement, negates one of the advantages of bi-amping---the freedom from low frequencies inter-modulating with mids and highs. Those are trade-offs, the user free to choose between the two forms of bi-amping.

The op stated he wants a passive, 12dB/octave, line-level, high-pass xover. I have installed a passive (a single capacitor on the input jacks of a power amp) filter to create a 1st order-6dB/octave xover at 100Hz, but I believe it requires an active circuit to achieve 12dB/octave. I could be mistaken, so perhaps someone else knows better.

So sfall, I assume your speaker has two sets of binding posts? And you connect a 2-channel power amp to the speaker, another identical amp to the right, each amp channel feeding one set of speaker binding posts? And both channels of each power amp receive the same full-frequency signal from your pre? Is it not then obvious that your speaker’s internal xover is still doing the filtering---at speaker-level, not a line-level xover doing the filtering? If you want to call that bi-amping go ahead, but that’s not how the term has traditionally been used. You’re right about one thing, though: This type of "bi-amping" absolutely requires the use of identical stereo power amps---the input sensitivity of all power amp channels, and the amp’s gain characteristics, must be identical for the speaker’s internal xover to filter as intended!