Jump to content

Active Bi-Amping/Tri-Amping FAQ


Chris A

Recommended Posts

For the $160/each I paid for them (literally brand new used from ebay), the Ashly's are very nice.  But yes, I obsess over graphs too much and want more accurate control with time delay, crossover slopes, etc.

 

On 1/24/2017 at 10:41 AM, Khornukopia said:

Have you considered getting digital active crossovers sooner, rather than next time?

 

Link to comment
Share on other sites

@Chris A

 

It's done?   Please feel free to critique.  REW file is found here: https://drive.google.com/open?id=0B1J0a4OV_WGLeTZGTGZyTEdKaGM

 

I'm now using the 4x8 Xilica XD4080, tri-amping P-39f's (passive xo's removed), and also integrating two JL Audio F212 subs using the Xilica.  I did this so I can use other pre-amps (such as the Oppo HA-1 via AES from my server in the next room) and not be "stuck" with the Emotiva XMC-1 when I want to use my subs.

 

All drivers are time aligned to the best possible/nearest Xilica time delay setting.  I used band limited measurements to fine tune the phase at each crossover region after first relying on a band limited impulse response measurement.

 

Impulse, waterfall and phase is the best response I've ever achieved without using Dirac, and I'm only using a total of 6 PEQ filters for all front channels (2 subs and 2 for each of the front left and right LF drivers).  After/if I calibrate using Dirac, I will post a new file.


The 200 Hz issue is not a crossover, but a room issue and I can't move my main speakers to fix it.  LF to MF is FIR XO at 500Hz (per Klipsch spec sheet).  MF to HF is FIR XO at 3.8kHz (to avoid some >1% THD spikes in the HF driver).  THD is very very low too at 90dB, measured with a Steinberg UR22 mk ii USB audio device and an Earthworks M30 mic at the main listening position, pointed upwards at the ceiling.

 

However, I did not use the all-pass filters to correct for the phase shift for the Sub and LF 48dB/octave butterworth crossover at 80 Hz, and would like to explore this if it will give even better results.  If there is a beginner tutorial on generating all-pass filter settings, I'm open to trying it.  Rephase only does auto FIR filters which I can't use (these are only for crossovers and only 1500 taps are available on the XD4080).

 

Also, there is still a slight amount of hiss from the XD4080, but no more than the Ashly XR1001's, probably less.  If you have any ideas, please advise.  Unfortunately the only amps I have to tri-amp with are the Emotiva XPR-1's (two for LF), and XPR-2's (two for MF+HF).  The preamps I've tried are the Oppo HA-1 and Emotiva XMC-1.  If I take the XD4080 out the slight hiss goes away for the most part.  The MF horns are 106dB efficient, but I already added series resistance (so the amp sees ~10 ohms for HF and for MF drivers); this eliminated the amp's hiss for the most part.

 

Thanks for all your help on this project!  Pictures of the XD4080 settings are attached.

Sub settings.png

LF settings.png

HF settings.png

MF settings.png

fir settings.png

gain settings.png

Room Front.jpg

  • Like 2
Link to comment
Share on other sites

Bottom line: how does the Xilica sound relative to the passives?  I've found that documenting the subjective effects of these type of changes is valuable in the future, when you forget the differences. 

 

I know that my Xilica XP-8080 and Yamaha SP2060 pretty much added back any micro-detail that I might have missed using the Dx38s.  It was like the step up using the First Watt F3 over the D-75As, except that the difference was about 2X greater than the First Watt (really).  The Xilica was considerably quieter than the Yamaha, as I've mentioned, and has more bi-quads (available filters) per channel so I could really clean up the bass bin performance of the Jubs and the K-402-MEH.  I cranked down the FR to ±1.5 dB from 100 Hz to 10 kHz ((flat response, ignoring room modes), and cleaned up the response considerably below 100 Hz (subject to room modes). The difference in listening caused me to go through my music and movie collection once again, and experiencing probably the first time that I feel that I've exceeded anything else that I've heard before: my setup is now my reference for judging music and movie soundtracks critically.  That's a big leap ahead--and one that I wasn't sure that I was going to achieve. I feel confident in what I'm hearing now is the recordings, not the setup...

 

I did a quick look at your REW file and really haven't generated any big questions.  I'll keep gazing at the plots.  The only thing that stands out besides that 248 Hz room mode is the excess group delay growth on the woofers below 200 Hz (peaking at 14.5 ms at 70 Hz).  Other than that, everything looks super.

 

Chris

  • Like 1
Link to comment
Share on other sites

@Chris A

 

Here is the Dirac REW file:  https://drive.google.com/open?id=0B1J0a4OV_WGLY2hMSV9YNDNwRmc

 

Some really pretty plots in there for you to examine!  Dirac did a fantastic job for me this time and did fix the 80Hz ringing.  I also bypassed all PEQ on the XD4080 since I know Dirac can EQ things so well and also improve phase response. 

 

I am not worried about the slight hiss (may not hear it in a normally treated room, but I have a lot of absorption in mine).  Well worth the upgrade in clarity versus passive crossovers, although the Palladium crossovers sound pretty decent stock, I can still hear details I had missed before on almost every track; especially strings and symbol decays.

 

I definitely don't regret purchasing the XD4080 and would buy another if it was reasonable.  The FIR filters are super nice!

 

4. waterfall.jpg

3. etc.jpg

2. THD.jpg

1. FR.jpg

  • Like 1
Link to comment
Share on other sites

Thanks for taking such a detailed look!  It is definitely true EGD is not flat, and this raises a question I had:

 

Would you sacrifice a flat response for a better matching/flatter excess group delay (EGD)?  Or do you think the way I matched phase is ok?  I unwrap phase about the crossover and zoom in (using a band limited sweep around the crossover), then I'll change delay on the impulse plot incrementally until the angle matches (keeping track of how much delay I apply).  I know I can get EGD to be ~-2ms which is a lot closer to the rest of the frequencies, but this comes at a price of phase alignment (if sub polarity is not swapped).

 

However, I think previously I swapped polarity on the subs and obtained an even better EGD match, but then delay was very unrealistic (like 28ms+).  I remember this option not affecting phase as much though.  Maybe I should go with that one?

 

EDIT: this Dirac calibration combined with accurate phase alignment sounds incredible.  Still open to trying to improve EGD depending on your advice, but bass strings, drums, etc sound perfect and 100% life-like.

 

One thing I did different on the Dirac calibration is I left the mic at the center of the mlp and didn't move it for the 9 measurements.  I think this works since my room is well treated, and gives to a ruler flat response.  Please see prior post above for new Dirac REW file.

 

4 hours ago, Chris A said:

I did a quick look at your REW file and really haven't generated any big questions.  I'll keep gazing at the plots.  The only thing that stands out besides that 248 Hz room mode is the excess group delay growth on the woofers below 200 Hz (peaking at 14.5 ms at 70 Hz).  Other than that, everything looks super.

 

Chris

 

  • Like 1
Link to comment
Share on other sites

6 hours ago, etc6849 said:

EDIT: this Dirac calibration combined with accurate phase alignment sounds incredible.  Still open to trying to improve EGD depending on your advice, but bass strings, drums, etc sound perfect and 100% life-like.

 

I believe that you're there.  I'd just spend time listening and enjoying the fruits of your work. 

 

I bet it sounds amazing.  Great job!

 

Chris

  • Like 1
Link to comment
Share on other sites

Unfortunately, I think only a few people will ever hear what a calibrated and tri-amped system in a well treated room can do.  Thanks so much for your help Chris!  I had sunk a lot of free time and money into my system since I was at your place to capture the clarity of your system in my small room.

 

I really want to tri-amp my rear speakers and my center later, but I think the money I've spent recently on the tri-amping project is enough for a few years!  Even the shelves were several hundred in raw materials to build, another $80-100 for XLR cables and some DIY speaker cables, extra amps, Xilica processor, etc... 

 

Not a cheap solution, but well worth it in performance gains.    I will setup alerts for one or two more Xilica XDxxxx's.  Hopefully if I wait a few years I will find a surplus deal.

 

For better or worse, this may be the end of tweaking things as I have reached the point where I don't believe there is a way to improve things audibly for stereo content in my particular room.  It definitely sounds world class and life-like on every track I've played.

 

6 hours ago, Chris A said:

 

I believe that you're there.  I'd just spend time listening and enjoying the fruits of your work. 

 

I bet it sounds amazing.  Great job!

 

Chris

 

  • Like 2
Link to comment
Share on other sites

Very nice audio system. Your pictures and charts look like a high end professional installation. I am still learning to utilize all the different tools in REW, so it is good to see what you have done.


P.S. for other readers; you may not be able to download and open the linked REW files without having the free REW program on your computer.

Link to comment
Share on other sites

i have a question you must forgive me I'm a novice I was wondering why there is six binding Post on my forte i?

I called klipsch and they said you should only be two I am I'm guessing somebody done modifications to the speaker. Be easy on me I'm new to the speakers game

Thanks for any info in advance

IMG_0147.JPG

Link to comment
Share on other sites

I believe you're looking at separated HF and LF binding posts that have been added.  This would indicate that someone modified your Fortes to bi-amp or tri-amp them. 

 

Here's a shot of a Forte I (rectangular access port) with a broken binding post:

 

bindingpost.jpg

  • Like 1
Link to comment
Share on other sites

You need something that will split the high frequencies from the low (an active crossover), and at least two stereo amplifiers.

  • Like 1
Link to comment
Share on other sites

Do you have a plot of the frequency response for each driver overlayed with the full speaker response? It would look a lot like the screenshots of your filter settings. This will give a good indication for how much summing you're getting at the xover frequency.

 

One thing about phase is that it will change dramatically if you move the microphone around.

 

Are you driving all three woofers in parallel? Didn't they do a tapered array in the passive xover? Are you sure you're not seeing polar lobing at 250Hz? Or is the mic within 14 inches of a reflective boundary? That could also be true of the speakers too. Is there a 28" path-length difference between the bottom and top woofer?

Link to comment
Share on other sites

On 2/6/2017 at 9:27 PM, DrWho said:

Do you have a plot of the frequency response for each driver overlayed with the full speaker response? It would look a lot like the screenshots of your filter settings. This will give a good indication for how much summing you're getting at the xover frequency.

 

One thing about phase is that it will change dramatically if you move the microphone around.

 

Are you driving all three woofers in parallel? Didn't they do a tapered array in the passive xover? Are you sure you're not seeing polar lobing at 250Hz? Or is the mic within 14 inches of a reflective boundary? That could also be true of the speakers too. Is there a 28" path-length difference between the bottom and top woofer?

 

HI thanks for your help.  Sorry for not seeing this sooner.  

 

New driver plots are coming next week once I move the speakers to their final resting place (TBD).  Klipsch did do a tapered top woofer, but the parts list from Canada I found only seemed to list one part number for the P-39f woofers, so I figured it was ok to tri-amp?

 

Under this folder there are several REW files (.mdat) if you want to see the driver measurements used in the previous post (view details and pick the one prior to that date).  There are also plots (jpeg images) that show the 240Hz dip has been there before I went active (thnk these were dated March 2016).

https://drive.google.com/drive/folders/0B1J0a4OV_WGLS2lYYl85N1BRNkE

 

I found last night that if I switch my speakers and subs around (plus move the speakers out more) it becomes a 200 Hz dip though.  You speak of a path difference of 28"; could this be related floor and/or ceiling bounce?  The direct path difference for the top and bottom woofer to the mic was 3-4" at most.  I will attempt to debunk the floor bounce theory as I made some 8" x 2' x 4' bass traps I can lay on the floor and see.

 

The mic was as pictured in my last post, so the only thing within 14" of it was the chair back which is well padded.

 

PS: Also got rid of any hiss whatsoever in the right MR and TW by purchasing a member's used Benchmark AHB2 amp as even adding some H-pads I made previously (~ -12dB) wasn't eliminating it.  Have it hooked to the right for now just for testing purposes, but found another one that is coming next week for the other speaker.


EDIT: I think this is caused by room shape.  Laying 8" OC 703 on the floor in front of the speaker helped a tiny bit, but dip was still there.  You can hear the drop out when playing a 201 Hz sine wave and moving around.  Judging by ear and test tone, if I move my chair forward 2 feet, it would fix it, but this would compromise imaging.  I'm thinking of trying the speakers corner loaded up front as close to the front wall as I can get them?

Link to comment
Share on other sites

Using the REW functions as you and Chris A have described, I ran some tests and discovered that I had a 70 Hz null at my main listening position, but not other areas of the room, when operating just my RF Klipschorn. I can't very easily move that speaker's position, so I moved other objects in the room while playing a 70 Hz sine wave test tone, until that acoustic null disappeared.

Link to comment
Share on other sites

@Khornukopia

 

There are some directions for bi-amping and using REW that I found in one of user jtalden's posts at: http://www.hometheatershack.com/forums/rew-forum/72727-new-function-required-implementation-low-pass-high-pass-filters-perfectly-simulate-response-curve-eqs-7.html#post776346

 

Definitely worth checking out the REW technical thread at home theater shack.  Also, all of ChrisA's and DrWho's posts here are obviously very very helpful!

 

Jtalden was able to re-time things things to get my phase response to look even better!  Here is a comparison for the left speaker (red is Dirac, green is without).  You can see how the slopes match up very well.  His process changes a little based on the situation, but it is well documented if you read his posts on home theater shack.  He recommends measuring the drivers at the MLP, which worked out well for me (but my room is well treated).

 

If you have dual subs, I would crossover higher than 70 Hz (over a 100Hz?) to make it easier to match sub and front speaker phase tracking.  You could also place the dual subs ideally and hopefully avoid the 70 Hz room issue altogether if possible.  There is clearly an art to getting a nice phase response (knowing when to invert polarity, when to know when it is optimized for your room, etc....), so hopefully someone with more experience than me helps you!

 

 

 

 

Dirac phase.jpg

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...