[GarMan]

Hi Jim,

Wondering if you've ever considered adding active crossovers (tube and/or opamps) to your product line.  Won't be too much of a stretch from your phono amps and preamps.

Gar.

[Brinkman]

When I purchased my Bugle Pro and Clarinet I told Jim this was something I too was interested in seeing, especially given the reputation of Hagerman and the hole in the DIY market when it comes to this. Jim obviously has his hands full at the moment, and I'm not quite "having to have" an active crossover right now, so I can hold my breath for a while. Funny that you mention this, I just mentioned wanting to do a crossover in my prior post...

[hagtech]

So how flexible should an active crossover be?  Two band or three band?  Do you need your choice of rolloff steepness?  This isn't too hard with opamps, but using tubes might make this difficult.  All depends on the requirements, how far you want to take it.

jh

[Scott F.]

Jim,

I bi-amp using an active XO on my main system. I've thought about DIY'ing one of these for a while now and I know what features I'd be looking for.

Personally, I'd integrate your preamp and XO together. You could use the Clarinet as the gain for the upper frequencies and design an opamp based gain stage for the bass frequencies. The bass section would 'follow' the Clarinet that way you wouldn't need huge amounts of gain for bass, just 6 to 9db or so. Just enough to level out some of the worst (bass shy) recordings. The other features I would look for is....

- Adjustable crossover points from around 40Hz up to around 320Hz (stopping every octave), that also gives the Ribbon guys a chance to use it too
- Those adjustable points would be manually set in both the bass and upper frequencies (in other words, you'd have to crossover frequency switches, one for the bass and one for the upper frequencies)(Oh, I'd give the option on the upper frequencies to run full range just for mini-monitors and the like)
- Adjustable slopes, 6, 12 and 18 (or 24)db. This is fairly easy to do with a selector switch

This one is completely wishful thinking but a full function remote with bass gain controls would be super-cool but that could be asking too much.

This is definitely a niche market product but I think there are a good many of us that would love to integrate our pre and XO to remove some of the clutter.


just my $.02 

[Jon L]

So how flexible should an active crossover be?  Two band or three band?  Do you need your choice of rolloff steepness?  This isn't too hard with opamps, but using tubes might make this difficult.  All depends on the requirements, how far you want to take it.

jh

What the world needs is something that will sound as good or better than the Bryston crossover at lower cost.  I experimented with many active crossovers, including ones way more $$ than Bryston, but none of them clearly was better and most much worse, especially when it comes to dynamics and bass foundation. 

I've inevitably come back to Bryston several different times now.  I believe their success comes from the use of discrete components for X-over and stout PS.  The weakness of Bryston, however, is the volume matching controls. Any attenuation or added gain on these knobs significantly degrade the transparency compared to 0dB, which is why I'm using my EVS attenuators instead.  I also don't like the fact Bryston's volume control changes the upper frequencies, not the bass volume.

So if the Hagtech HangOver crossover (you can use that, Jim ) can accomplish all that, then heck, that's my next crossover!
 

[Brinkman]

Jim,

The active crossover DIY I've been eyeballing is here: www.snippets.org/filters/crossover.htm

I am in a situation where I love using my ST70 for the mid-hi and will bi-amp for my low-end expectations (solid state amp here). I'm really open-minded about mixing tubes and SS, which is why I went with the Bugle/Clarinet combo. In fact, my original intention was to house a MC step-up tranny/Bugle/Clarinet/DIY active crossover as a all-in-one chassis, but am so ignorant about circuitry this is a pipedream and not possible at the moment.
That said, I think an op-amp based active crossover is perfectly reasonable for those on a budget. Tube rolling for a Clarinet, crossover AND tube amp is a bit overwhelming a situation for me. Moreover, if one's using tubes at all, one probably (or should) have reasonably sensitive speakers. I don't know if there's a correlation about speaker sensitivity and slope filters, but my Klipsch Fortes have a woofer at 6db/octave slope and a squawker and tweeter at 12db/octave. What I'm trying to say is perhaps the active crossovers should be aimed towards the most likely speaker designs of folks using your pre-existing line of products.
This obviously doesn't simplify things any, but I can only speak for myself anyhow.

I think the Clarinet build-in idea posted by Scott is very novel and close to what I was initially had in mind.

Thanks for asking!

Benjamin

[analog97]

So, a X-O is a device that goes after the pre-amp, then separates highs from lows, then feeds 2 amps?  Then, you can remove the speaker jumpers to bypass the speaker X-over?  Is this the concept?  How does this improve the performance? 

[JoshK]

Indeed, opamp based xo would be easier and cheeper, but tube xo would be da bomb.  Typically when I've run across tube based xo, they were done with a CF followed by passive xo (SK probably) parts then by another CF on the output.  Shelving, BSC and other fancier EQ's would require cascading multiple sections together (starts adding up with tubes, but would be uber cool). 

I think the issue with actives is always the same, someone ultimately wants something different than someone else.  It would be ideal to have modular boards that one could piece together and add PSU to, to get where ever they were heading.  This could allow the user to place into existing preamp or your clarinet provided the B+ and heaters had sufficient current output. 

[miklorsmith]

I'd like to have a simple hi-pass filter, it could have a fixed frequency of, say, 65 hz with a steep slope.  I can cross the subs with their parametric EQ.  How "not there" could such a device be?

[Brinkman]

So, a X-O is a device that goes after the pre-amp, then separates highs from lows, then feeds 2 amps?  Then, you can remove the speaker jumpers to bypass the speaker X-over?  Is this the concept?  How does this improve the performance? 

Analog97,

A crossover can go anywhere in the chain between your source and your speaker drivers. Crossovers can add gain to the signal (active, powered) or simply divide the signal (passive, like the crossover circuit inside your speakers). An excellent article to read would be this: http://sound.westhost.com/bi-amp.htm
In short, the "crossover" is the frequency in the crossover circuit (such as for a three-way speaker) which decides whether the frequency is low-pass (woofer), band-pass (squawker), or hi-pass (tweeter). In the case of my Klipsch Fortes, everything up to 700hz goes to the woofer, 700hz to 6Khz goes to my squawker and above 6Khz to my tweeter.
How it "improves" performance is completely subjective, in that for an infinite amount of money, one could acquire the "world's best" speakers and "world's best" stereo amplifier and laugh at the poor folks milking the last ounce of performance from their measley mid-fi set-ups. However, if one were to pick up a tube stereo from a flea market and restore it to good-as-new, one might find that it's mid-range performance is exceptional, but it's high-frequency and low-frequency extension just aren't as good as a more recent average-quality solid state amp. Using a crossover allows one to say, 'hey solid state, alls you gotta do is take care of the bass, nothing too complicated' and to your tube amp 'hey you old geezer, you're not as strong as you used to be, times have changed, but I know you got some tricks up your sleeve, so do some freakin' mid-range magic before you blow up, okay?'

And voila, your speakers are singing to you. It's as close to something-for-nothing as home stereos get.

Best,
Benjamin

[miklorsmith]

Good explanation.  I'm running 2 watts of 45 SET magic from 65 hz up and 500 pro watts below that.  Sounds nuts but works like a dream.

[analog97]

Dear Benjamin,

Thanks for a wonderfully lucid and damn funny explanation.  I missed just a point: when you remove the jumper on speakers, does that disengage the X-over?  Then, when you bi-amp you gotta have another x-over to replace the speaker x-over.  Hopefully I've got it.  Thanks, larry

[SHIF]

There are loads of links on the internet for "Tubed crossovers".  You can find schematics, kits, assembled units, etc; 2-way, 3-way, 4-way, symmetric slopes, asymetric slopes, whatever you want..

I like the units offered by Marchand Electronics:
http://www.marchandelec.com/xovers.html

If I ever decide to biamp my Maggies, I will probably buy his XM26 with asymetric slopes to mimic my existing passive circuit (low-pass: 3rd order@250, high pass: 1st order@200.)

Of course I'd rather build a Hagtech half-kit, especially if it offered a simple method to alter the cross-over points and slopes "on the fly", but only if it utilized tube amplification 

-S

[Brinkman]

Dear Benjamin,

Thanks for a wonderfully lucid and damn funny explanation.  I missed just a point: when you remove the jumper on speakers, does that disengage the X-over?  Then, when you bi-amp you gotta have another x-over to replace the speaker x-over.  Hopefully I've got it.  Thanks, larry

Larry,

If one were to have an amplifier supplying a signal to the woofer only, the speaker wire could feed into the speaker input terminals, and then directly to the driver itself; the division of frequency to the woofer has taken place "further upstream", so there is no need to utilize that portion of the passive crossover circuit, though one certainly can. In this same scenario, a tube amp could be sending a signal to both the squawker and the tweeter. You would then need to utilize the portion of the passive crossover circuit inside the speaker that divides the signal between the two drivers, unlike the woofer portion of the circuit which can be scrapped.
You would need 2 pair of inputs on each speaker; a pair for each amplifier. On a typical three-way speaker, one could also "tri-amp". This would require 3 pair of inputs, and as you probably figured out, means one amplifier per driver. The logical conclusion is a complete bypass of the passive crossover inside the speaker, with each driver being directly powered by it's own dedicated driver.

IMAGINATION TIME: It's 1975. You have a quadraphonic cartridge on you record player. You have a quadraphonic amplifier... One could use the pre-outs on the amp, splitting the four channels for each of the speakers to be utilized. Using active crossovers, each of these channels is being split into three frequency chunks (for each driver in the speaker). One could have 3 pair of monoblocks for each pair of speakers. That's a dedicated amplifier for every driver on all four of your speakers. Twelve amplifiers, plus the the quadraphonic receiver, plus two active three-way crossovers.
I wonder how Dark Side of the Moon sounded like this.

The technical specs regarding how the individual amplifiers performance improves in your typical bi-amping scenario is beyond me. I would imagine that the decreased frequency dynamic of the input would result in a more linear amplification, but I could be wrong. There's also the advantage of letting each amp deal with a phenomenon called "reverse EMF" (the tendency of drivers to push a signal "back" upons due any number of factors). I'll let the technical folk explain this one, or you could read the algebraic madness in the article I linked to in my prior post.

Best,
Benjamin

[analog97]

Benjamin, again my thanks.  I got it now.  I see now how an external x-over, more "upstream" than the speaker x-over obviates the need for a speaker X-over.  I am using these really big speakers, Polk SDA 1.2Tl's.  They have 13 drivers per box and a x-over that is embedded in the speaker that is in a box about 20" wide and 6" high....I've never pulled it out, but it must be pretty complex.  I don't fully understand the benefits of bypassing this massive speaker x-over, but I understand your teaching.  The final proof of the pudding must be determined by experience.  Thanks again for a thoughtful response.  Larry

[Brinkman]

BUMP

So how flexible should an active crossover be?  Two band or three band?  Do you need your choice of rolloff steepness?  This isn't too hard with opamps, but using tubes might make this difficult.  All depends on the requirements, how far you want to take it.

jh

A three band opamp-based active crossover sounds perfect to me. The suggestions made for adjustable filter slopes and adjustable rolloff steepness (can this be made adjustable?) seem like great ideas as well. I suppose the hi- and low-pass frequencies are made semi-permanent by the installation of certain resistors. At least, that's what I've gathered from some DIY crossover designs I've looked at. Of course, if one wanted to be able to adjust low-pass and hi-pass via a knob, it might be difficult to provide enough "pass frequency stops" to please everyone. I could be wrong.
On my three-way Fortes (according to the Klipsch page), the low-pass is 700hz and the high-pass is 6Khz. I wonder what the most typical crossover freqeuncies are for three-way...