[EthanH]

My system is not that bright, but on many recordings the sound is just a bit more forward/lean than I would prefer.  Problem area seems to be the upper midrange/low treble, but I don't know for sure as I have no measuring equipment.  Stuff like Dire Straits can be just a bit forward at times, while Radiohead sounds absolutely perfect, oddly.  I'm just really sensitive to brightness and love bass impact.

System is:  Onix XCD88 - Cayin TA30 - DIY speakers.
Cables are: Iron Lung Jellyfish and Volex PC; DIY yardmaster SC; cheap molded IC that came with the Onix. 

The speakers are DIY monitors using seas L18RNX/P woofers and 27TBFC/G tweeters.  Overall they sound very good, so I'd like to just smooth out and fatten up the sound a tad if possible.  From what I gather my options are:

a) fiddle with the crossover.  sort of a hassle and could get expensive.
b) buy some ICs.  don't know what to expect here.
c) mod the Onix with new opamps.  not sure if this will do anything.
d) buy new cd player or dac - expensive, so I'd rather not do this.
e) buy sub. again, expensive so would rather not go this route.

Any ideas or suggestions?  I'd like to keep this under $200, the cheaper the better.  I feel like there should be some way to get what I want, but am not sure of the best route to go.  Thanks.

Link to Audiogon System:  http://forum.audiogon.com/cgi-bin/fr.pl?vopin&1161351433

[JohnR]

Have you trying toeing the speakers so they cross in front of you?

[TRADERXFAN]

I vote for this one...
a) fiddle with the crossover.  sort of a hassle and could get expensive.

The speaker is 90% of the equation...

How was your speaker designed?

Edit** ok, I googled the designers name and found your speaker... is this it?
http://www.zaphaudio.com/audio-speaker17.html

found this little tidbit... maybe try this first?
"Response shaping options


Top end response shaping - If you wish to tame the top two octaves a bit, rather than mess with the L-pad, I recommend adding a shelving circuit. This will be an .15 mH coil, paralleled with a 1.5 ohm resistor. This circuit should be placed in series right before C6 on the schematic. This will reduce the tweeter output gradually in the frequency range above 5kHz, resulting in a laid-back tonality that some may prefer. Decreasing the coil value moves the corner frequency up, and increasing the value will move the corner frequency down. "

-Tony

[Wind Chaser]

Not everyone's cup of tea but... EQ?

[JackD201]

What's the resonant frequency of the 27TBFC/G tweeters?  A notch might be another option to consider.

[EthanH]

Thanks for the feedback so far.

JohnR:  Crossing the speakers in front of me doesn't improve things.  thanks though.

TRADERXFAN:  Those are indeed my speakers.  I figured the crossover might be one of the best routes to go.  I'm not sure if the response shaping option would address the frequency range I'm talking about though, as the area I'm sensitive to seems to be the lower treble (although I'm really not sure about this, as I don't own any measuring equipment).  Would it be better to gradually reduce output above 5khz, or reduce output between 2.5khz and 4khz with the tweeter lower rolloff option?

Wind Chaser:  I'd be open to an eq but I don't have one and a good one sounds like it would be expensive.

Jack:  the tweeter's resonant frequency is up around 27khz so I don't think that's the issue.

Oh yeah, about room treatments:  I have hardwood floors, a glass door, and three windows with one directly behind the listening area.  Room treatments would probably fix the problem, but aside from putting a curtain on the window behind the listening area there's nothing else I can do due to WAF.  The other windows on the side wall already have curtains, and the glass door is located next to the speakers on the rear wall.

[Wind Chaser]

Just a few more ideas...  THESE may or may not address your specific concern, but if they do... they are a "cheap" fix and will IMO make a substantial positive improvement in other areas.  There is a LONG THREAD HERE on Audio Circle about them.

Perhaps this is a system synergy thing?  Not familiar with your amp, but some amps are more characteristically forward than others especially with certain HF drivers.  I generally warmer sounding amp and presentation which can sometimes times seem to relax top end aggressiveness.

Did the Radiohead recordings that sound perfect now previously sound restrained or lacking?

[Curly Woods]

What capacitors have you used for the treble Xover?  This would be an easy way to tailor the treble other than damping the tweeter a little bit. 

[Wind Chaser]

http://www.silcom.com/~aludwig/images/diffdem.gif

The above animation illustrates whats going on with diffraction.  Some people say addressing the problem affects / attenuates the FR by as much as 1.5 db. 

[EthanH]

Wind Chaser:  thanks for the idea about the felt.  I'll read a few threads to get a better idea.  I'm not sure if this is a system synergy thing; the amp is an EL34 integrated, and the cd player is not great but is supposedly pretty good for the price point.  My old speakers, Totem Arros, were more bright than the ones  I have now.  The Radiohead recordings that sounds so good now previously sounded congested and too forward.

Curly Woods:  I use Janszen Superior-Z caps throughout the crossover, plus mundorf and mills resistors, and Janszen air-cores with the exception of an Erse Super-Q steel laminate in series with the woofer.  Internal wiring is Supra.

[BobM]

I second trying the felt around the tweeter idea. It will definitely hep reduce and baffle diffraction, which can sometimes sound like forwardness and a brittle type of sound. It will not lower the level of the tweeter but should help to blend it with the mids/woofers better. You should be aware that the proximity to the actual tweeter cone/dome itself can dramatically affect the effect, so play with closer vs farther away to see what works best for you, as well as the amount of felt. Sometimes just a few small dots in close proximity works better than a lot farther away. However it can alter the character of the tweeter as well, YMMV.

If that doesn't quite cut it for you then I would suggest increasing the value of the series resistor in the tweeter crossover. Making it larger by perhaps .5 or 1 ohm will reduce the level of the tweeter without changing its inherent character.

Good luck,
Bob

[richidoo]

You have good drivers and crossover parts, good amp and source. Acoustics are not the problem. Crossover is the culprit.  It is difficult to make drivers sum flat, in phase, with good dispersion and lobing. A pile of quality parts is pretty much required to tune it by ear. Simulation only gets you so far. Steep low frequency filters make perfect simulated graphs, but they are very difficult to implement with real parts. 2nd order at higher frequency is much easier to integrate and often sounds more open if the drivers are comfortable doing it. There is always a trade off. Steep filters can solve a lot of problems.   

Some testing can be done without expensive equipment. Read Vance Dickason Loudspeaker Recipes to see his method of working through the design and tweaking of passive crossovers on 4 different speakers.  http://openlibrary.org/b/OL1127984M/Loudspeaker_Recipes  Joe D'Appolito's book "testing loudspeakers" will also give you some insight.

You have a computer, it came with a mic, you seem very smart. A SPL meter and apiece of paper can get you a long way in the freq domain at least. Audacity can generate specific frequency tones. Also this http://www.download32.com/sound-generator-i3032.html Free software like Room EQ Wizard can help you record and graph freq sweeps, and time domain impulses too. It's more accurate if done outside.

You could always hire someone to help. Rick Craig (Selah Audio) does custom crossover design and is very good at it and reasonable. I have heard half dozen of his passive crossover designs, they are consistently excellent. You get the tweaked crossover, built, tuned ready to play for less money than you would spend on the collection of tweaking parts. But you have to ship the speakers. And  it is fun to DIY. Enough rambling.  Good luck

[Curly Woods]

Curly Woods:  I use Janszen Superior-Z caps throughout the crossover, plus mundorf and mills resistors, and Janszen air-cores with the exception of an Erse Super-Q steel laminate in series with the woofer.  Internal wiring is Supra.

Well those are certainly good capacitors, so it does look like that some massaging of the xover might be in order.  Have you contact Zaph to discuss your issues with him?  Maybe he could assist you in reworking the xover to address what you are hearing?  The suggestion about looking into Rick Selah is also a very good idea, but I really doubt that he can do it for less than $200.00 for finished xovers

[S Clark]

Try taking the R9 value up by .5 to 1 ohm or the R10 value down by the same amount.  It should do what you are looking for.  http://www.zaphaudio.com/audio-speaker17.html

[EthanH]

Looks like a lot of people like the felt around the tweeter.  Maybe I'll order a pair of these if I need to get something from madisound in the near future:  http://www.madisound.com/catalog/product_info.php?cPath=404_116&products_id=770

I agree with many of you that the crossover would likely be the most effective means of altering the sonics.  I'd like to hear what Rick Craig could do with regard to a crossover redesign (it would be interesting to compare), but like Curly Woods said I suspect his services would stretch my budget a bit.  I'd like to take in-room measurements but I don't have  a mic for my laptop and don't really feel like buying any software since I'm not sure when my next diy project will even be.  As I said before, the leanness is not that offensive, so I'm first going to try a cheap fix before taking more costly approaches. 

Changing the R10 value would be relatively painless and would be cheaper than replacing C6 so I may try that.  In the meantime I constructed an impedance flattening circuit from some parts I had on hand that should smooth out a broad impedance bump centered around 1khz.  I have a tube amp, so I figured it's worth a shot to see if there's some audible bump occurring as a result of the amp/speaker interaction.  There's a very subtle difference, but I'll have to listen to it for awhile before determining whether I prefer the system with the impedance flattening circuit in or not. 

Thanks for all the great ideas so far.

[jimdgoulding]

http://www.silcom.com/~aludwig/images/diffdem.gif

The above animation illustrates whats going on with diffraction.  Some people say addressing the problem affects / attenuates the FR by as much as 1.5 db.

The passive component I make would do that for you in the area you describe, in the lower bandpass of your tweeter just above the crossover where waveforms are most severely diffracted by ends of cabinets with right or near right angles*.  There is no change to the true output of tweeters just the removal of what is arriving late out of time and phase and summing in with what there is. 

* That's the conventional wisdom.  I treat the entire surface so waveforms have no interaction with any part of the exposed reflective surface.

[JackD201]

In the meantime I constructed an impedance flattening circuit from some parts I had on hand that should smooth out a broad impedance bump centered around 1khz. 

Oooooooooh great idea! Do let us know how that goes.

Yo Jim, does the fact that the tweeters have hex grills and are set in a mirror imaged configuration make a big difference when it comes to using diffractionbegones?

[jimdgoulding]

In the meantime I constructed an impedance flattening circuit from some parts I had on hand that should smooth out a broad impedance bump centered around 1khz. 

Oooooooooh great idea! Do let us know how that goes.

Yo Jim, does the fact that the tweeters have hex grills and are set in a mirror imaged configuration make a big difference when it comes to using diffractionbegones?

I need some help with your question, Jack.  I would hope the any grill cover is microscopically fine (I make the cutout 2.5" in diameter so as not to grab anything vital which should, hopefully, allow for tweeter grills) and I don't understand what is meant by mirror imaged is a sense that would matter.  Can you clarify?  Thanks.

[JackD201]

Take a look at the OP's pics Jim, he's got a link on the first post. His tweeters come with built in grills. They are also set mirror imaged aka not centered on the front baffle. Think vintage ARs or some ProAc models. I'm curious if the grills which are meant for diffraction themselves as well as protection from kiddy fingers make a difference and also if in the case of mirror imaged tweeters your pads will surround the tweets equally left and right or if the pads will cover the entire wider length to either side and if the latter is the case, why.

Thanks for the patience and apologies if I inadvertantly hijacked this thread 

[jimdgoulding]

Oh, I dig, offset.  Yeah, that helps.  Don't know how the grills deal with diffraction but off setting the tweets will make edge diffraction arrive at different time intervals and less noticable is my bet in the end.  I THINK that's the reason some designers (a lot of Brits, for example) do it.  I make pads that extend out to the cabinet ends regardless.  We good?