[abernardi]

Hi guys,
  I spent the weekend ripping out the passive crossovers, rewiring the speakers, making an additional speaker cable run, adding a Marchand XM9 active crossover, an additional amp and am now letting it all burn in a bit.  I'm not much of a DIYer, but this hobby makes me do things  .  I'm posting this for anyone who might be interested and maybe some advice, since I don't really know what I'm doing 

  When I initially fired it all back up late this AM, I was immediately depressed.  It sounded worse than before.  I'm hoping burning in the new additions will improve things, so I'm in a wait and see mode.

  I decided to do this because, besides being a hopeless addict, the man who co-designed these speakers, Jerry, has said from the beginning that they respond very well to biamping.



  These are Avlar speakers, a two way with an 8" SEAS woofer and a folded ribbon AMT tweeter inspired by the Heil AMT, but redesigned.  Very high sensitivity: 100dB/1W



   When I opened it up and took out the crossover I saw that they had used JPS Labs Super Blue 2 cable for the internal wiring.  I looked it up and it had some decent reviews, but mostly from home theater installers.  Nothing against home theater, but I think I had some better cable. 




   HT cOz made my speaker cables and when I told him I was biamping, he helped me make another run on my own (thanks Robert!).  It's basically a Quad Litz braid of 16 gauge PCOCC copper wire.  Pretty labor intensive since I have 23' runs! 
   Anyway, I thought I'd use the same wire for the speaker innards.  It's not a particularly easy wire to work with, it has several different conductor widths, supposedly to accomodate different frequencies better.  But it doesn't twist cleanly.  I could tin the ends for crimping on the quick disconnects for the woofer, but the binding posts and the AMT both had screw terminal connections.  I did some surfing and read somewhere that it's better NOT to tin the wire for the screw type connectors for a better contact.  But these wires are flakey.  I suppose I could have soldered them anyway, but I wanted to try it this way first and so far all seems good.





   I picked up a used Marchand XM9, (would have preferred the tube version XM126, but that's big bucks).  The crossover is a standard 24dB/Octave Linkwitz-Riley network, these speakers crossover at 500Hz. 







For amps I'm using a Dodd battery push/pull EL84 amp at 15w/channel for the highs and a Virtue Sensation w/Dodd buffer and battery at about 30w/channel for the lows.  Preamp is the latest Dodd battery pre. 

   So it kind of freaked me out initially because I couldn't get the sound I was getting before.  The biggest question I have is about the bass.  It's really weak and the low lows are less full AND less defined than before.  The Marchand lets you set levels for the highs and lows.  It gives you a fair amount of gain, it looks like +8dB over the center position.  I need to play with it more, I haven't adjusted the damper control at all yet, but to get it to sound close to right I have the high gain at -3dB and the lows all the way up to +8!  I didn't expect that.  I thought with the more powerful amp I'd have an abundance of bass.  I tried turning down the gain on the highs more, but things started sounding muffled.  If I didn't know better I'd say the crossover frequency is wrong, but Jerry insists it's 500Hz.  Clearly I don't understand some basics.  I've heard that sometimes a passive crossover is preferred because it can be more customized than just a simple crossover at a specific frequency, but this is the same parameters that Jerry had on his speakers at the shop.  So I've put it all aside for a now while I let it all burn in.



   I still haven't put on the banana plugs, but I had to hook them up to see if it would work. 

   NOW, besides the bass issue, I am noticing a few very impressive improvements in the sound:  The image seems larger, I mean, it's like the singers suddenly doubled in size  , it was startling.  Also listening at low levels is much more detailed and listenable.  I can have it at practically a whisper and I'm still hearing a full range.  Before, everything got smaller and harder to hear.  Even though the balance is thinner because of the lack of bass like before, the music seems to have more weight.  I don't know if it's more detailed, but it's more powerful and punchy.  One drawback is that I'm hearing more low level system noise.  It's not a ground hum or anything, it's just got more system noise.  It must be the Marchand.  I'll have to play a little with the cabling and make sure the power transformers are well out of the way. 
  So that's it.  If anyone has any input I'd be grateful to hear it.

[JohnR]

Hi, a crossover is more than just splitting the frequency range between two drivers - they also equalize the drivers' response and correct for baffle step - the latter may account for the lack of bass. Also, check the phasing of the drivers. You may want to look into getting some acoustic measurement gear.

By the way, what does the middle control on each channel do?

[Davey]

You've completely "redesigned" your speaker system.    It's not unexpected it would sound quite a bit different.  You can let things burn in all you want, or make changes to speaker wire....it won't make any difference. 

As John mentioned, the stock crossovers provide more than just frequency dividing.  You could now have EQ changes, polarity issues, relative level differences, etc, etc.

The best way to approach "conversions" of this sort is to measure the electrical response of the existing network before starting.  This allows to save a "baseline" response you can use as a reference when evaluating relative changes in future configurations.

Unfortunately, the Marchand XM-9 is extremely limited in capability since it only allows symmetrical LR24 slopes and no native equalization.  It won't even allow to change HP and LP filter frequencies independently because of the state-variable architecture.

John,

The middle control is a "damping" control on the state-variable topology that allows shape changing of the curves near the crossover frequency, but not in the pass bands.

Cheers,

Dave.

[Ethan Winer]

You can let things burn in all you want, or make changes to speaker wire....it won't make any difference. 

The best way to approach "conversions" of this sort is to measure the electrical response of the existing network before starting.

I agree with this. However, I think the best place to measure is in the room at the listening position with a microphone using suitable software such as the freeware Room EQ Wizard. This takes into account amplifier volume settings and everything else, rather than just the crossover's response and output levels. This might help further:

Room Measuring Primer

--Ethan

[orientalexpress]

can you Bi amp with passive xover that came with?or does have to be with eletronic xover?


lapsan

[sunnydaze]

can you Bi amp with passive xover that came with?or does have to be with eletronic xover?


lapsan

Yes you can, it's called "passive bi-amping".  You simply run one amp into the speaker's lower posts, and the other amp into the upper binding posts.  The speaker's internal x-over is still in use.  If the two amps are different, the louder one must have a volume control on it to adjust for spl diffs, speaker top vs bottom.  I have tried this many times with many different amp combos, but I always keep going back to single amp full range.  Sometimes it impresses initially in terms of dynamics, size and wallop, but in the long run always seems a bit "off" to me.  Like a lack of coherency and integration.

Most of the reading I have done says that passive is really a waste of time.  To get the full benefits of bi-amping, one must go "active"  = disconnect the speaker's internal crossover, and replace it with an external one.  Perhaps those more technical and experienced than me can chime in.

[JerryLove]

Most of the reading I have done says that passive is really a waste of time.  To get the full benefits of bi-amping, one must go "active"  = disconnect the speaker's internal crossover, and replace it with an external one.  Perhaps those more technical and experienced than me can chime in.

In a best case scenario: Passive bi-amping gains you 3db over just using the more powerful of the two amps.
One could also make a case that, in the event that you are clipping, you might isolate the clipping to the (likely more challenged) LF amp and so avoid the sound degradation to the HF driver.

Both could be accomplished better with a sufficiently powerful amp. There's no real-world case where I've seen passive bi-amping as a good idea.

Yes. To multi-amp (I've got a speaker set that runs on 4 channels each) you need to crossover before amping.

But as pointed out: the crossover in your speaker does a lot more than the crossover in your AVR. It controls beaming and comb filtering, it deals with phase, it compensates for baffle reflections, it adjusts for uneven sensitivity both between drivers and across the driver's range.

And yes, 500Hz sounds like a very low crossover for an AMT.

[orientalexpress]

Thank you,i was thinking Bi amping with KT88 mono block for the high and DAC cherry plus for the bass.Glad i got excellence advice before i dive into it.Thanks again 


lapsan

[randytsuch]

I would say that a passive xover COULD do a lot of compensation, depending on the drivers used and what is required.

If the picture in the first post is the entire xover, it can't be doing much other that acting as a low pass/high pass filter.
I only see two caps and two inductors, so unless I'm missing something this is a pretty simple xover and it is not doing that much.

I think it's a first order, so it's only 6db/octave slope if that's right, don't know if that is making a difference in what you hear.

Randy

[Davey]

You're correct, that it doesn't seem to be a complicated crossover....just an LC on both drivers.  But, we don't know. 

Replacing said crossover with an XM-9 which has LR24 electrical slopes is a complete seat-of-the-pants approach.... from a relative point of view.

That's why I mentioned taking an electrical measurement of the existing crossover to establish some sort of "reference."  Not that you have to recreate that reference with an active/alternative approach.....but at least it's something to start from.

But heck, we don't even know how well engineered the stock, passive crossover is.  It might be a total cluster. 

So many issues. 

Dave.

[abernardi]

Great responses!  According to Jerry, the passive crossover is just a simple low pass/high pass, same as the XM9.  But Randy, you bring up an interesting point, I need to check because now I'm not sure if the passive crossover is a first order network or a fourth order.  I thought Jerry said it was a Linkwitz-Riley fourth order, but I also remember him saying it was a gradual slope crossover, not a steep slope.    The XM9 is a fourth order crossover and isn't adjustable.   

The low crossover point of 500Hz is one of the aspects of this particular AMT that they're most proud of.  It does have excellent imaging.

[mboxler]

Hi

If it's 1st order, the two inductors would need to be in series with the woofer, and the two caps (paralleled) in series with the tweeter.  If we knew the values of these components, and the impendance of the drivers, we could calculate the -3 dB points of the low and high pass filters.  BTW, the XM9 can be modified for different slopes...see page 11

http://www.marchandelec.com/ftp/xm9man.pdf

Mike

[Ethan Winer]

can you Bi amp with passive xover that came with?or does have to be with eletronic xover?

I agree with the others that suggest staying away from "passive" bi-amping. Good quality active crossovers do not have to be expensive, and they have slopes of 24 dB per octave which (AFAIK) few if any passive crossovers use.

--Ethan

[abernardi]

  Update:  I left everything alone for a few days so I could come back with a clear head and I had a chat with Jerry who confirmed that it should indeed be a fourth order simple network and that the lack of bass was to be expected because the ATM has about a 100dB sensitivity and the woofer is in the low 90's.  And the two amps might have different sensitivity, etc.  So I cranked up the low frequencies and dramatically decreased the highs.  I also had the speakers toed in a little, and I now have them facing straight forward.  It's still not quite right, BUT, it's close and there is plenty of bass. 

  The reality was that I was freaking out a little.  They sounded so different and I knew I was in over my head.  So when I initially turned down the highs I thought it was degrading the quality as well.  This is an excellent example for me of how powerful the psyche is!  I was not hearing accurately because I was upset.

  It's clear I need to tune the room.  The drapes are being made and the rug is going down, that will tame the room to some extent, about all I can do in the living room. 

  The good news is that the glare and harshness is gone.  Before with the passive crossover, a single female voice with sparse instrumentation sounded incredibly life-like, but an orchestra sounded screechy and artificial.  Rock or big band horns were glaringly bright.  Now the orchestra is much fuller and three dimensional and natural.  Much more detail, lyrics now understandable that weren't before.  The mids and highs are not etched, in fact it seems smoother, yet more detailed.  The bass is very full and warm and the shape is smoother and more refined, but I'm not sure the Sensation has as much control of it as the Dodd did.  Again, I may not be describing it just right. 
  The image is still big, I'm not sure I can describe it, but it's like the image is enlarged and going beyond the edges of the screen or window, so that I feel like I'm not seeing everything, does that make sense?  Straightening out the speakers helped a lot and widened the soundstage, like opening the french doors wider to look out on a garden.  But it's still a little crowded, as if the room is too small for the speaker.  The room is 16 x 26, I would think that's big enough.
  Also, I have to take a little off the high end of the left speaker to center the image.  I'm not sure where that's coming from.  I think it has to be the Marchand.  The low end seems right.  I'll be putting on the plugs this weekend.  Thanks for reading!