Jump to content

B&C DCX464 as the mid & hi drivers for a K-Horn


Wirrunna

Recommended Posts

This is a re-write of going to active crossovers on a B&C DCX464 to an Eliptrac horn in the tophat in my K-Horn
The B&C DCX464 is a 1.4" driver requiring an adapter to mount it on the Elliptrac - https://www.parts-express.com/eminence-ha14-2-1-1-4-aluminum-horn-adapter--290-5626
Here is a recent REW plot of the Left K-Horn. This is the current working config. The rest of these posts attempt to tell how I got to this response.

1288379698_Oct11FullSysInputPEQsTL82.8dbIRDelay0.thumb.jpg.0893eccc7c6ff7895dff6d4c5ba92e46.jpg

 

These measurements are taken at 1m from the centre of the Elliptrac using miniDSP UMIK-1 and 1/6 smoothing.

 

 

 

 

Equipment.

Sources are FM Tuner and a Squeezebox server with Raspberry Pi based streamers feeding a NAD C725BEE Receiver. REW is run on an Asus laptop feeding the AUX input on the NAD.
To do the crossovers I have 2 miniDSP 2X4HD and a pair of https://www.audiophonics.fr/en/power-amplifier/audiophonics-mpa-s125nc-rca-stereo-class-d-power-amplifier-ncore-2x125w-4-ohm-p-12756.html for mid and hi amp channels, the bass is done by the NAD.  
Due to noise on the line outs from the miniDSP the DCX464 is attenuated by an l-pad of 22 ohm in series with each driver and 8.2 ohms in parallel. I first tried the l-pad resistors the other way (8.2 in series) but while the noise was reduced it was still audible at 1m. When SPL and phase are sorted I will experiment with different gains in the equipment chain.


The K-Horns are in a room 6.5m wide, 10.8m long with 2.6m ceilings (height).


The process I used for each driver :
1. Measure SPL and phase
2. EQ to flatten the SPL.
3. Measure the flattened (PEQed ?) SPL and phase
4. Load the PEQed measurement into rePhase
5. Use rePhase to set the crossover and flatten the phase
6. Measure the rePhased SPL and phase.

Then when happy with measurements of individual drivers, run the system with all drivers :
7. Run a 20-20,000 sweep for all drivers for SPL and phase. Determine gains needed.
8. Display GD to see needed delays and also Spectrogram to confirm.
9. Apply gains and delays in DSP and re-run sweep for SPL and Phase.
10. Examine phase response and go back to step 5 and fix recalcitrant drivers. Also fine tune SPL. Experiment with driver invert.
11. When happy with combined SPL run  EQ to fix SPL anonamilies and apply to the input PEQ.
12. Confirm SPL and phase and play some music.


Notes : All measurements were taken with a miniDSP UMIK-1 calibrated microphone positioned 1 m from the Eliptrac centre at a height of 1.18 m. All graphs are smoothed to 1/6.

 

Raw measurements.
Jul 14 M051R1 Bass 20-700 90db Raw.jpg

image.thumb.png.4a9460c302587b4be11f70c26bea6b9c.png

 

B&C DCX464 Mid in Eliptrac Horn - Jul 14 M063R1 Mid 200-7000 90db RAW.mdat

image.thumb.png.c02f1dd20c07dc1b81fcb19b7fc2a67e.png

 

B&C DCX464 Hi - Jul 14 M071R1 Hi 3000-20000 Raw.mdat

image.thumb.png.0318983ec1883ef79a21849cc7d76c67.png

 

Now to EQ
K-Horn - EQ Jul 14 M051R Bass 20-700 90db TL89db 20-500 1db

image.thumb.png.0e0a92037710d553952cacfab8717448.png

Equaliser: miniDSP 2x4 HD
Jul 14
Filter  1: ON  PK       Fc   36.05 Hz  Gain   5.20 dB  Q  5.803
Filter  2: ON  PK       Fc   42.00 Hz  Gain  -6.10 dB  Q  3.846
Filter  3: ON  PK       Fc   69.80 Hz  Gain   2.80 dB  Q  2.002
Filter  4: ON  PK       Fc   91.00 Hz  Gain  -4.30 dB  Q  3.266
Filter  5: ON  PK       Fc   132.0 Hz  Gain  -5.00 dB  Q  4.573
Filter  6: ON  PK       Fc   159.5 Hz  Gain  -5.30 dB  Q  5.218
Filter  7: ON  PK       Fc   210.0 Hz  Gain  14.40 dB  Q  5.712
Filter  8: ON  PK       Fc   239.0 Hz  Gain  -7.70 dB  Q  4.130
Filter  9: ON  PK       Fc   282.0 Hz  Gain   2.70 dB  Q  6.892
Filter 10: ON  PK       Fc   355.0 Hz  Gain  -1.80 dB  Q  8.950

 

B&C DCX464 Mid - EQ Jul 14 M063R1 Mid PEQ TL86db 330-4000 1db 1on6S.txt

image.thumb.png.deb68f83fe72a43c3b622417a21dd57c.png

Equaliser: miniDSP 2x4 HD
Jul 14
Filter  1: ON  PK       Fc   422.0 Hz  Gain  -2.00 dB  Q  8.662
Filter  2: ON  PK       Fc   466.0 Hz  Gain   2.00 dB  Q 10.000
Filter  3: ON  PK       Fc   548.0 Hz  Gain  -3.40 dB  Q  8.148
Filter  4: ON  PK       Fc   771.0 Hz  Gain  -4.90 dB  Q  5.059
Filter  5: ON  PK       Fc    1065 Hz  Gain  -4.70 dB  Q  6.909
Filter  6: ON  PK       Fc    1318 Hz  Gain  -2.50 dB  Q  5.577
Filter  7: ON  PK       Fc    1712 Hz  Gain  -4.70 dB  Q  3.909
Filter  8: ON  PK       Fc    2047 Hz  Gain  -3.80 dB  Q  2.607
Filter  9: ON  PK       Fc    2800 Hz  Gain  -2.80 dB  Q  4.412
Filter 10: ON  PK       Fc    3810 Hz  Gain   3.00 dB  Q 10.000

 

B&C DCX464 Hi - EQ Jul 14 M071R1 Hi TL 84.5 3600-20000 1db

image.thumb.png.6b67aae793767fb9956dd6d08c1a06f3.png

Equaliser: miniDSP 2x4 HD
Jul 14
Filter  1: ON  PK       Fc    3600 Hz  Gain   2.00 dB  Q 30.000
Filter  2: ON  PK       Fc    4180 Hz  Gain  -1.70 dB  Q 10.000
Filter  3: ON  PK       Fc    5030 Hz  Gain  -1.20 dB  Q 10.000
Filter  4: ON  PK       Fc    5650 Hz  Gain   5.10 dB  Q 15.000
Filter  5: ON  PK       Fc    7000 Hz  Gain  -1.00 dB  Q 10.000
Filter  6: ON  PK       Fc    8800 Hz  Gain   0.50 dB  Q  8.000
Filter  7: ON  PK       Fc   10000 Hz  Gain  -2.00 dB  Q  1.000
Filter  8: ON  PK       Fc   12200 Hz  Gain   2.00 dB  Q 15.000
Filter  9: ON  PK       Fc   14130 Hz  Gain  -1.50 dB  Q 10.000
Filter 10: ON  PK       Fc   17380 Hz  Gain   3.00 dB  Q 20.000

 

Edited by Wirrunna
Re-write and update full system REW plot to current
  • Like 2
  • Thanks 1
Link to comment
Share on other sites

rePhase contd.

 

image.thumb.png.ee67c28bf6038f8653d70f4eafc53d3f.png

 

Now for all drivers.
SPL and phase - M040L Full system - the phase will clean up after time alignment

image.thumb.png.f35582fc405158ccf2ba4cbf8a62204c.png

 

GD or Group Delay - showing Hi at 0ms (Y axis), Mid at 7ms and Bass at around 14ms. M040L Full system - GD

image.thumb.png.de9ba7c9879d1214336cfc652ce1df95.png

 

 

 

 

 

 

 

 

 

 

 

  • Like 3
Link to comment
Share on other sites

Spectrogram - scale -3 to 20ms to show delays - M040L Full System - Spectrogram

 

image.thumb.png.9b50fe456354aed247687c82d5d71461.png

 

Full system after input PEQs applied and delays and gain adjustments.

Jul 16 M083R2 Full System Input PEQs Bass -3.5db 0ms Mid +5.5db 6.9ms Hi -1db 13.9ms.jpg - IR delay of zero

image.thumb.png.b5ac2dca01ff92ef41659c524411e82c.png

  • Like 2
Link to comment
Share on other sites

It seems like there is some serious R&D going on in a few companies right now and B&C is one of them. Get a load of these! Down to 300hz for a crossover point is seriously interesting. L-Pads for sure on most of these horn drivers as they get way ahead of the woofers in a heart beat. They are promising crossovers in December for this driver  and I assume tuned to this horn.

https://www.bcspeakers.com/en/products/horn/1-4/0/ME464

https://www.bcspeakers.com/en/products/coaxials-hf/1-4/8/DCX464-8

 

  Don't like to hear about that hiss in the Mini DSP HD and if you manage to tinker and tame it I want to know what you did.

 

Link to comment
Share on other sites

Hiss in the miniDSP.

Dave, I slammed these components together using RCA leads from the box of spares, some have "gold" plating, a couple have tin plating etc. They are a temporary setup stuffed in an old Ikea VCR media stand with the miniDSPs sitting on top of the n-core amps. Cable routing is not optimum!

I suspect some of the hiss will be induced by having the digital gear too close. When I first heard it I thought the noise was the fan in my Asus Zen Book laptop but then realised it was coming from the speakers. The noise disappeared when I put the B&Cs on back on the L-Pads I was using with the passive xover. I will follow it up later as I was having too much fun playing with the DSP. The L-Pads knock about 8db off.

Early testing was done using an old pair of AR 18BV speakers just to make sure that there were no earth loops, loud clicks when the amps turned off etc and the hiss was not audible, but the ARs are probable mid 80s db/W while the B&C is 112 db/w in a horn. A speaker with that sensitivity is going to expose any inherent noise in the amplifier chain.

Link to comment
Share on other sites

jorjen, my understanding is that the DCX462 is a DCX464 with an adapter. I got the DCX464 second hand, so no choice.

 

As a self reward for getting this far I purchased a miniDSP UMIK-1 as my Mitey Mike is probably 25 years old. The two mics were reasonably close with their readings , but the USB connected mic is far more convenient than a mic plus a preamp and then 3.5mm plug.


The hiss from the miniDSP has been eliminated by an L-Pad of 18 ohms in series and 8.2 ohms in parrellel with the drivers. Those values were the first 10watt resistors I could find in my parts box that I had 4 off each (2 L-Pads each channel).

 

In the mean time, I have re-read Chris A's guides, explored the miniDSP FIR filters and rePHASE and worked out how to build a FIR filter crossover with EQ for the miniDSP 2X4HD but the limit of 1024 taps per output appears to limit the effictiveness.
So I set about to re-measure everything with the UMIK and explore the many crossover slopes available
I found that BW24db seemed to work pretty well, eliminating a couple of bumps and holes. Rather than graphs I have attached the REW measurement as a zipped file. The measurement is with the UMIK-1 at 1m on axis with and aiming down the throat of the Elliptrac horn. The crossovers are Butterworth 24db at 400Hz and 3500Hz,  delay is 6ms for the mid and 6.2ms for the hi. Levels adjusted for flat response.


Next rainy day I will try the FIR filters.

 

Nov 4 Mnt26 Bass 0db 0ms BW24db 400Hz Mid 6db 6ms BW24db 3500Hz Hi 0db 6.2ms BW24db.zip

Link to comment
Share on other sites

FIR Filters.

miniDSP suggest using rePhase freeware to build linear phase crossovers. There is not a lot documentation available but there are a few hints around.
After giving it a go I have settled on LR48 xover filters at 350Hz and 3500Hz. The DCX 424 is rated to 300Hz and is only 2db down at 350Hz in the Eliptrac horn before falling steeply.

 

 

Nov 10 Left Khorn.jpg

Nov 10 Right KHorn.jpg

Nov 10 Right Khorn Spectrogram.jpg

Link to comment
Share on other sites

  • 1 month later...

Further tweaking.

Equalisation was an iterative process using REW measurements and the EQ function in REW to calculate the settings then import the REW EQ settings into the miniDSP, run another REW measurement, tweak the EQ settings in REW EQ, import .......... Equalisation is by PEQs in the miniDSP.
I picked crossovers at 350Hz and 3,500Hz , the crossovers were set using FIR filters in the miniDSP by rePhase Linear Phase Filters and are Linkwitz-Riley 24db per octave at 350Hz and 3,500Hz. rePhase was also used to flatten phase response. As only 3 of the 4 available outputs of the miniDSP are being used, FIR taps are allocated only to the active outputs.


How does it sound ? Well I'm not a speaker reviewer, but there is a noticeable increase in clarity and sound stage and the system can be played quieter than with the ES passive crossovers. I have aimed for a flat response and add a "room curve" with setting bass boost to 2 o'clock on the NAD receiver.


I will try some gentler slope crossovers after the holiday season - it is summer down here in Australia and I have family staying for a few weeks.

 

Jan 7 Right speaker spectrogram.jpg

Jan 7 Right speaker.jpg

  • Like 2
Link to comment
Share on other sites

Looking at the frequency plot above I realised it was before applying input PEQs after a full range EQ, so here is the further EQ'd full system, right speaker then left.

 

 Chris A and the Dude, I'm interested in your experience with the plate amplifier DSP software. I found that the miniDSP integrates easily with REW and rePhase, however rePhase takes a bit to get your head around and has the ability to make spectacularly bad filters very easily.

These N-Core amps are the closest thing to a straight piece of wire with gain that I have encountered.

Jan 8 Right Full after add input PEQs for full EQ.jpg

Jan 8 Spectrogram Right Full after add input PEQs for full EQ.jpg

Jan 8 Left Full after add input PEQs for full EQ.jpg

Jan 8 Spectrogram Left Full after add input PEQs for full EQ.jpg

Link to comment
Share on other sites

My experience with dsp is as as follows.

 

Minidsp for mainly sub integration, there are a lot of tutorials on this, which makes it a bit easier. I've tried it with 2 way active systems in the past with lacking success (but I'm not an expert, so don't take that with a grain of salt). I didn't feel that the q's had the ability for fine details (if I recall).

 

I have been running the Ashley Protea 3.24 3in6out on my main system with more success than the minidsp. However, I've only used filters recommended for Jubscala, never took the time to tweak.

 

I've recently picked up a Fushion Fa123 for my center channel. This software has been a learning curve. I am preparing myself to go down a rabbit hole this afternoon to see if I can follow some of Chris's guides and set this up more accurately as I have recently received a pair of BMS dc4590 coaxial drivers, so stay tuned for that.

 

One question I have for you and anyone else.

How do you know when, where, and why you use FIR filters versus regular filters? This is one of the first things asked when starting a new program with the Hypex Fushions.

 

Sent from my SM-G986U using Tapatalk

 

 

 

 

Link to comment
Share on other sites

5 hours ago, The Dude said:

How do you know when, where, and why you use FIR filters versus regular filters?

One of the guys over on the diyAudio forum summarized the use of rePhase this way:


1. use parametric filters/gains (PEQs, i.e., "minimum phase") EQ to first flatten SPL response in the passband
2. then use minimum-phase filters 'compensate', and 'time offset' in measurement to further flatten phase, and to continue flattening SPL.
3. Apply Linear-Phase Filters (FIR filters) to flatten phase and SPL more.

 

Chris

  • Like 2
Link to comment
Share on other sites

Dude,

I had read either on the miniDSP site or somewhere on diyForums that you use PEQs to flatten the passband, FIR for the xover and phase flattening.

 

Dave A,

That you tube link is what I did, but the miniDSP 2X4HD has a different interface.

 

  • Like 1
Link to comment
Share on other sites

  • 3 weeks later...

This post has been edited with a new .pdf file.

 

Rainy day down under today after a few days of 35+C temps, so added a bit to my own instructions on using REW and RePhase.

 

 

Using REW and rePHASE to create a flat phase crossover for miniDSP 2X4 HD.

 

 

1 Use REW to measure and EQ.pdf

Edited by Wirrunna
Revise pdf
  • Thanks 3
Link to comment
Share on other sites

Thanks John.  I think that many miniDSP crossover owners probably don't know much about the FIR filter capabilities of their units.  This is a good place to show just how that can be explained.

 

I've talked about the use of DSP crossovers to not only flatten the SPL response of the loudspeakers/rooms, and upstream and downstream electronics such as the effects of non-flat response preamps and amplifiers, including mostly single-ended tubes/valves having no circuit feedback but also some solid state designs such as using headphone amplifiers with very high output impedance to drive compression drivers, etc., but also the phase response.

 

I find that those using loudspeakers not having controlled directivity down to ~200-300 Hz (horn-loaded bass bins with directivity control using large mouth bass horns like Jubilees, Khorns, La Scalas, Belles, Peavey FH-1s, etc.) usually cannot hear the effects of phase flattening.  This is because in-room reflections from nearby room objects and boundaries interfere from direct-radiating woofers destructively cancel out directed acoustic energy from these types of bass bins.

 

ray-tracing-early-reflections-order-1-3.jpg.8b684e7c64fecf79aeb3930f688614ca.jpg

 

Horn-loaded bass bins having directivity control to below the room's transition frequency (the Schroeder frequency):

 

PeaveyFH-1 next to JubeKhorn.jpg

Klipschorn and FH-1 bass bins

 

maxresdefault.jpg

Belle Klipsch

 

Klipsch%20La%20Scala%20Cherry-750x750.jp

Klipsch La Scala

 

Multiple_entry_horn.png

Sound Physics Lab Unity/ Danley Sound Labs Synergy Horns

 

Klipsch - Jubilees w zebrawood and guard dog - DSC_2907-resize3.jpg

Klipsch Jubilees

 

These type of loudspeakers have an advantage that other loudspeakers don't have: the ability to control early in-room reflections down into the midrange and below, and when coupled with DSP crossovers having FIR filter capability, produce an almost unbelievable soundstage image fidelity that is seldom experienced by listeners.  I've talked about this effect before here. Here's a summary of the effects:

 

1. significantly increased perception of bass--so much so that I had to remaster my stereo recordings

2. overall subjective depth and seamless soundstage improvements,

3. an indescribable naturalness of sound that's easy hear with live acoustic instrumentation recordings, and

3. elimination of harshness, particularly of acoustic instrumentation like acoustic guitars, violins/violas/cellos/double basses, and wind instruments (brass & woodwinds).

 

Chris

  • Like 3
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...