• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Multi-Channel, Multi-Amplifier Audio System Using Software Crossover and Multichannel-DAC

Have you compared Benchmark AHB2 to your Accuphase in full range passive mode? If you so can you described what you've heard?
 
Have you compared Benchmark AHB2 to your Accuphase in full range passive mode? If you so can you described what you've heard?

Yes I have done it, as I described in my post #253 entitled;
Benchmark AHB2 Tests and Trials: Part-1
Comparison with ACCUPHASE E-460 in Single-Amplifier + LC-Network Reference Sound System


BTW, tomorrow or the day after, you will find here two or three of my new post sharing tests and evaluations on ROTEL RB-1582 MkII together with Benchmark AHB2 and ACCUPHASE E-460 in my single amplifier system and also in multichannel multi-amplifier system.
 
Last edited:
ROTEL RB-1582 MkII Tests and Evaluations: Part-1
Comparison with ACCUPHASE E-460 and Benchmark AHB2 in Single-Amplifier + LC-Network Reference Sound System

Note: I am an end user audio enthusiast and I have no conflict of interest at all with any of the manufacturers, import companies, distributors and audio shops relevant to this post. This post is not intending to intensively share objective and/or subjective evaluations of the candidate amplifiers, but I would like to share about how I would test and try candidate amplifiers in this project.

Hello friends,

One of my audio enthu friends kindly brought his ROTEL RB-1582 MkII conventional class-AB power amplifier for tests and evaluations in my multichannel multi-amplifier project. Very fortunately, RB-1582 MkII stayed for two days while one Benchmark AHB2 was still working at my home.

RB-1582 MkII is a rather big and heavy power amplifier of 431mm W x 150mm H x 407mm D, 18 kg, 200W + 200W (20-20,000Hz, 0.09%THD, 8Ohm), and the detailed specs and inside photos can be found here;
http://www.rotel.com/product/rb-1582-mkii
WS000942b.JPG


You may find a nice review article on RB-1582 MkII by Erich Wetzel dated 15 February 2014 at;
https://www.soundstageaccess.com/index.php/equipment-reviews/533
to which ROTEL pasted the access link in their above RB-1582 MkII product page.

As usual, in this Part-1 post, I first tested RB-1582 MkII in my Single-Amplifier + LC-Network Reference Sound System comparing it with ACCUPHASE E-460 and Benchmark AHB2.

Now we have Benchmark AHB2, YAMAHA A-S301(B) (not in use in this Part-1 post) and ROTEL RB-1582 MkII tentatively all stacked;
WS000944.JPG


And, ACCUPHASE E-460, my reference amplifier;
WS000945.JPG


You can see all the DACs and amplifiers;
WS000946.JPG


And, here I show you again the current speaker alignments looked from my listening position;
WS000947.JPG


As for the unique position of super tweeters FOSTEX T925A (STs), please refer to my post #27. Also please refer to my post #84 for the differences between YAMAHA NS-1000 and NS-1000M.

The backside of the speakers is like this;
WS000948.JPG


Using my audio sampler tracks, we, my friend who kindly carried RB-1582 MkII and myself, first had an intensive listening session with my ACCUPHASE E-460 reference system;
WS000949.JPG


Please refer to my post #248 for the 22 Ohm resistors in Be-SQ, Be-TW and ST circuits.
The entire system scheme is;
WS000950b.JPG


The software crossover EKIO's configurations remained unchanged as those for always in my reference sound system;
WS000951.JPG


For the listening sessions with AHB2 or RB-1582 MkII, the input gains, gains for Be-TWs and STs were slightly changed, but all the other parameters remain unchanged;
WS000952.JPG


Before to go into listening sessions with RB-1582 MkII, we listened to AHB2 just in the same setup I did in my post #253 to re-confirm the excellent sound quality of AHB2 in the single-amplifier system;
WS000953.JPG


The entire system in this setup is;
WS000954b.JPG


Then, finally we carefully replaced AHB2 with RB-1582 MkII for intensive listening sessions;
WS000955.JPG


The entire single-amplifier system with RB-1582 MkII is;
WS000956a.JPG


What was the total sound quality with ROTEL RB-1582 MkII in this single amplifier reference system in comparison with Benchmark AHB2 and ACCUHPASE E-460?

We fully agreed with Erich Wetzel's review that RB-1582 MkII is a powerful and very stable class-AB amplifier giving rather warm and soft sound in LOW to MID range which is easily distinguishable from the sound of AHB2 and E-460. Since we listened to RB-1582 MkII just after the sessions with E-460 and AHB2, we felt that the RB-1582 MkII's sound would be a little bit too warm and soft, slightly blurred, and hence not so suitable for our highly sensitive, high-speed highly-efficient SP drivers especially for Be-SQ. I can say that if we did not experience the preceding listening sessions with AHB2 and E-460, we would be quite satisfied with the sound given by RB-1582 MkII though...

As for the HI to Super-HI sound range, i.e. above ca. 7,000 Hz, we recognized RB-1582 MkII gives very nice silky smooth, transparent, clear, high resolution and high-speed pleasant sound probably due to its much driving power and very nice "phase stability and continuation"(?) even though we had little time to do intensive objective measurement with microphone ECM8000 and REW. This excellent sound quality in HI Fq range given by RB-1582 MkII highly encouraged us to try this rather heavy duty powerful amplifier to drive only Be-TWs and STs at the final session in the multichannel multi-amplifier configurations which will be shared in my next post.
 
Last edited:
ROTEL RB-1582 MkII Tests and Evaluations: Part-2
Incorporation in Multichannel Multi-Amplifier System and Listening Sessions


Note:
I am an end user audio enthusiast and I have no conflict of interest at all with any of the manufacturers, import companies, distributors and audio shops relevant to this post. This post is not intending to intensively share objective and/or subjective evaluations of the candidate amplifiers, but I would like to share about how I would test and try candidate amplifiers in this project.


In this Part-2 post, I share with you my tests and evaluation of ROTEL RB-1582 MkII in my present multi-channel multi-amplifier system just like my tests with Benchmark AHB2 shared in my post #258.

The software crossover EKIO's configurations remained unchanged;
WS000957.JPG


and;
WS000958.JPG


We first tried RB-1582 MkII driving directly and dedicatedly the Be-SQs;
WS000959.JPG


Please refer to my post #248 for the 22 Ohm resistors in Be-SQ, Be-TW and ST circuits.

Just like we recognized in single amplifier tests described in my previous post #263, in comparison with Benchmark AHB2, RB-1582 MkII gave a little bit too warm and soft, slightly blurred MID range sound which seems not suitable for the highly sensitive, high-speed, highly-efficient Be-SQ.

Then we also tested by mutual exchange of RB-1582 MkII and ACCUPHSE E-460 in this scheme so that RB-1582 MkII drives WOs and E-460 drives the Be-SQs. Now the MID range sound quality was considerably better given by E-460, but the LO range was again a little bit too warm and soft given by RB-1582 MkII.

Finally, since we were much impressed the day before by the HI range sound of RB-1582 MkII as described in my previous post #263 in single amplifier system, and because all of RB-1582 MkII, AHB2 and E-460 are capable of receiving XLR balance input from DAC8PRO, we, including my wife who joined the listening session, unanimously agreed to have powerful RB-1582 MkII to directly and dedicatedly drive the highly efficient Be-TWs and STs as shown in this scheme;
WS000960.JPG


Here the RCA input into L and R sub-woofers YST-SW1000 (SWs) was achieved by the DAC8PRO's AES/EBU (CH1+CH2) sync digital out to ONKYO DAC-1000(S), and, therefore, now all the channels are completely in sync.

I slightly modified the gains of EKIO in Be-TW and ST range into this;
WS000961.JPG


Please note that the gains for Be-TWs and STs to be driven by RB-1582 MkII were slightly increased to -5 dB and -10 dB respectively for better matching with Be-SQs driven by AHB2.

Now, the entire system diagram is;
WS000962b.JPG


All of us were much surprised and deeply impressed with the total sound quality given by this setup. It was just most impressive, amazing and fantastic audio listening experiences at home we ever had.

Now I well recognized that even for highly efficient Be-TWs and STs, we should carefully select and install rather powerful (or enough power) excellent amplifier, in this case RB-1582 MkII, to achieve superior total sound quality together with E-460 and AHB2.

In this setting, we also very much carefully compared the ON or OFF of the 22-Ohm resisters in Be-SQ, Be-TW and ST lines, and we found that the use of 22-Ohm resistors gives slightly better sound, I can say slightly better smooth and clear "departure" of the sound from the drivers and better sound continuation between the drivers. As discussed in my post #248, it would be very difficult to scientifically or objectively justify the use of these 22-Ohm resisters in SP lines, but I assume that a little bit of extra power load to AHB2 and RB-1582 MkII given by these 22-Ohm resistors would allow the amplifiers to work very smoothly and efficiently in slightly above the zero power level for Be-SQs, Be-TWs and STs.

I well know that AHB2 has no zero-crossing distortions, and also RB-1582 MkII should have no or negligible audible zero-crossing distortions; it would be difficult to explain the favorable effect of 22-Ohm resistors in terms of "zero-crossing distortions", I assume. Even though, the entire system with Be-SQs, Be-TWs and the STs, are now very much responsive and sensitive for this kind of really fine tuning.

We also carefully compared the ON or OFF of the protection capacitors (the blue Jantzen Audio Standard Z-CAPs 400 VDC), 68 microF for Be-SQs, 10 microF for Be-TWs and STs, and we found again that they have no inferior effect on the total sound quality.

Consequently I decided to keep these 22-Ohm resistors and protection capacitors in the SP line, at least until I would fully decide the amplifiers in present multichannel multi-amplifier system.

Edit: I had very interesting and invaluable discussion on validation and justification of this issue in #99(remote thread), #100(remote thread), #101(remote thread) on the remote thread entitled "ASR dummy load configuration".


Using this final setup in this post, we really enjoyed listening to various "music" (not always the physical "sound" itself) in our intensive, but stress-free and tireless, listening sessions for half-a-day with great pleasure and satisfactions.

In any way, based on my tests and evaluations with Benchmark AHB2, ROTEL RB-1582 MkII and ACCUPHASE E-460 in present multichannel multi-amplifier project, and also having the confirmation of full utilization of 8 channels of DAC8PRO's output and AES/EBU digital (CH1+CH2) in-sync operation of sub-woofers L and R YAMAHA YST-SW1000 with RCA line level input via ONKYO DAC-1000(S), I have now almost decided to configure the total stereo 5-way 10-channel system in this scheme with Hi-Fi Amplifier-X, -Y and -Z;
WS000963b.JPG


Now I can say that the total system should be almost ideal and perfect for me if I would use three of Benchmark AHB2 as Amplifier-X, -Y and -Z in this scheme; Japan domestic sales of AHB2 through sole import company and distributors, however, has just started in July this year, and the price is still considerably (about 1.3 to 1.6 times) higher than that in the USA. For the time being as second option, therefore, I may use just one AHB2 as Amplifier-Y and use my existing ACCUPHAE E-460 as Amplifier-X.

I may agree with your possible thoughts that the dedicate use of the rather powerful, large and heavy RB-1582 MkII as Amplifier-Z in above scheme for highly efficient Be-TWs and STs would be just "too much" in this multichannel system. I myself somewhat feel the same... I would like, therefore, to further test and evaluate rather nice XLR-input-capable "small" amplifiers, like SOULNOTE A-0, TEAC AX-505 and TEAC AP-505, to drive Be-SQ and STs in this multichannel multi-amplifier system, even though such arrangements and schedule have not yet been fully fixed.
 
Last edited:
Note: I am an end user audio enthusiast and I have no conflict of interest at all with any of the manufacturers, import companies, distributors and audio shops relevant to this post. This post is not intending to intensively share objective and/or subjective evaluations of the candidate amplifiers, but I would like to share about how I would test and try candidate amplifiers in this project.

Hello friends,

In my above post #264, I described that I have almost decided to configure the total stereo 5-way 10-channel system in this scheme with Hi-Fi Amplifier-X, -Y and -Z hopefully using one, two or three Benchmark AHB2;
WS000976b.JPG


Please refer to my post #248 for the 22 Ohm resistors in Be-SQ, Be-TW and ST circuits.

In parallel with "Class-A(H)B Path" including AHB2 for Amplifier-X, -Y and -Z in this scheme, I am also seriously considering another "Pure Class-D Path" of amplifier exploration before to make final decision on amplifiers in my multichannel project.

I have already tested Hypex NC400 based heavy duty amplifier(s) as shared in my post #228, and I was impressed by the total sound quality given by this rather expensive NC-400 based class-D amplifier. I also would like to test and evaluate, therefore, Purifi 1ET400A based class-D amplifier(s) in the same multichannel multi-amplifier system. The interesting thread entitled "Purifi 1ET400A - decisions, decisions." is also encouraging me to try 1ET400A amplifiers in my project.

I understand ca. 200W@8Ohm power rating would be much more than enough in my multichannel system for dedicate drive of WO, Be-SQ, Be-TW+ST using possibly three of "Purifi 1ET400A stereo class-D amplifier". And, just like as I purchased DAC8PRO directly from OKTO Research in Prague, I can arrange international personal import purchase of 1ET400A class-D amplifiers from manufacturers in Europe, in the USA or in Austraria, I believe.

I also found that a few small domestic audio manufacturers in Japan are designing 1ET400A based class-D amplifiers; I just started contacting with them for possible beta-test at my home in my multichannel multi-amplifier project.

On the other hand, when we, my wife and myself, were enjoying the wonderful sound with ACCUPHASE E-460, Benchmark AHB2 and ROTEL RB-1582 MkII as Amplifier-X, -Y and -Z respectively in listening to our favorite piano solo (Iddo Bar-Shai) and period instruments (Purcell Quartet + organ) albums,
WS000981.JPG

I came to another idea that I should not fully exclude rather conventional and conservative "Pure Class-A Path" in amplifier exploration e.g. using one, two or three units of ACCUPHASE A-46 and/or A-36, and other pure class-A amplifiers available from several other manufacturers.

As Balle Clorin wrote here, these nice ACCUPHASE class-A amplifiers are rather expensive outside of Japan, but fortunately I am living in Japan and the domestic prices are comparable to the still very expensive imported Benchmark AHB2.

In any way, I am really happy that I could almost decide the multichannel multi-amplifier configuration as shown in above scheme with Amplifier-X, -Y and -Z, and I have several options for actual amplifiers in the three paths; "Class-A(H)B Path", "Pure Class-D Path", "Pure Class-A Path", and mixture of these paths would be also possible.

Since I am still not in a hurry to fully decide multichannel amplifiers in my project, my exploration on amplifiers will be steadily and carefully continued for the coming several months or longer. I would highly appreciate, therefore, having your suggestions and advice for my amplifier exploration.
 
thanks for sharing your experience. Would you mind rating amplifiers you have auditioned in full range setup? Would you trade Accuphase for Benchmark? Where does Hypex solution falls in on this chart?
I am considering multichannel setup for my DIY based project (using yamaha tweeters and midrange). Though some of the commentary I have read on Hypex amps is that they lack the weight/bloom typically present in Class A/B amps. I also read similar comment on Benchmark amp. I am referring to lower midrange/upper bass region. For instance when you stay next to musician playing violin you can feel it with your body. When I listen to most recordings in conventional (modern) system violin sounds like it is missing the resonance chamber as if someone applied high pass filter. I can hear each individual string but there is lack of acoustic "density". It sounds clinical and lacks any engagement. BSO sounds like bunch of kids warming up before the concert. I can follow each instrument independently but missing the overall picture. In my opinion chasing the ultimate resolution, soundstage and other "audiophile" aspects we have lost the sound presentation as a whole. What's going on here? Is it out equipment? Recording? For me the ultimate quality of the system is not 3 minute first impression, but rather how I perceive it long term. Do I want to hold my breath and enjoy listening to every track from beginning to the end? With hyper detailed setups I find myself skipping tracks. I want a setup that plays anything and showing the best part of the recording vs deficiencies. There are too many setups that can play well, but they play well only 4 reference recordings and everything else triggers the "next track".
 
Note: I am an end user audio enthusiast and I have no conflict of interest at all with any of the manufacturers, import companies, distributors and audio shops relevant to this post. This post is not intending to intensively share objective and/or subjective evaluations of the candidate amplifiers, but I would like to share about how I would test and try candidate amplifiers in this project.

Hello friends,

In the English Brochure for ACCUPHASE A-36, a pure class-A power amplifier, they described;
"Current feedback principle assures excellent phase characteristics in high range. As shown in the illustration, the A-36 uses the output signal current rather than voltage for feedback. Since the impedance at the current feedback point is very low, there is almost no phase shift. A minimal amount of NFB therefore results in maximum improvement of circuit parameters."
WS000985.JPG


In my comparison between YAMAHA A-S301(B) and ROTEL RB-1582 MkII in dedicated drive of high range with Be-TWs and Horn STs, I found and heard better high range sound with RB-1582 MkII probably and mainly due to "less phase shift".

As I am not an expert on "high range phase shift depending on the NFB feedback", I would appreciate having your thoughts and technical input on this issue, especially whether I can expect further better audible high range phase features/stabilities with using A-36 in my multichannel multi-amplifier system, or not.
 
Last edited:
thanks for sharing your experience. Would you mind rating amplifiers you have auditioned in full range setup? Would you trade Accuphase for Benchmark? Where does Hypex solution falls in on this chart?
I am considering multichannel setup for my DIY based project (using yamaha tweeters and midrange). Though some of the commentary I have read on Hypex amps is that they lack the weight/bloom typically present in Class A/B amps. I also read similar comment on Benchmark amp. I am referring to lower midrange/upper bass region. For instance when you stay next to musician playing violin you can feel it with your body. When I listen to most recordings in conventional (modern) system violin sounds like it is missing the resonance chamber as if someone applied high pass filter. I can hear each individual string but there is lack of acoustic "density". It sounds clinical and lacks any engagement. BSO sounds like bunch of kids warming up before the concert. I can follow each instrument independently but missing the overall picture. In my opinion chasing the ultimate resolution, soundstage and other "audiophile" aspects we have lost the sound presentation as a whole. What's going on here? Is it out equipment? Recording? For me the ultimate quality of the system is not 3 minute first impression, but rather how I perceive it long term. Do I want to hold my breath and enjoy listening to every track from beginning to the end? With hyper detailed setups I find myself skipping tracks. I want a setup that plays anything and showing the best part of the recording vs deficiencies. There are too many setups that can play well, but they play well only 4 reference recordings and everything else triggers the "next track".

Hello tpaxadpom,

I really thank you having your above important and nice post in this thread.
You wrote; In my opinion chasing the ultimate resolution, soundstage and other "audiophile" aspects we have lost the sound presentation as a whole.

I now feel exactly the same after my series of comparative listening sessions in the past several months with ACCUPHASE E-460, YAMAHA MX-A5200, DENTEC DP-NC400-4-EXP, Benchmark AHB2, YAMAHA A-S301(B) (only for tentative use) and ROTEL RB-1582 MkII.

In my single-amplifier + passive LC-network system (with full elimination of attenuators) full range system, I still love E-460 best for our music listening experiences, as you may guess and agree!

As I wrote in my post #265, recently I started very nice communication with several top people of small excellent audio manufacturers in Japan who are designing Purifi 1ET400A based class-D amplifiers. After careful insights on my current project, one of them kindly advised me to stick on to Class-AB path or Class-A path in my amplifier exploration. He said very honestly, "class-D technology still not yet has achieved catchup with excellent class-AB and class-A amplifiers, and there are obvious limitations in class-D approach; I think it would be quite difficult or even impossible to push class-D amps to the stage of 'excellent' class-AB or class-A amps."

He continued, "Benchmark AHB2 is basically class-AB amp with new pulse power supply technology, and the measured specs of AHB2 are really wonderful. If you are really satisfied with AHB2 in your music listening, not the sound listening, you may go with AHB2. I really strongly suggest you, however, to try excellent class-A amplifiers as amplifier-X, -Y and -Z in your multi-amplifier system before to decide on AHB2."

One of the people also pointed about possible less durability and stability (depending on selection of parts such as resistors, capacitors and coils), lack of enough warranty, possible problems in future maintenance and repair, and inferior consistency of parts availability, associating with many of the new class-D amplifiers.

These are also the reasons for my careful reconsideration on exploration not to exclude excellent class-A amplifiers in my multichannel multi-amplifier system.

Even though I know well that this is "ASR Forum" and many people here like "objective measurement based" discussion and rather hate "subjective arguments on total sound quality", I really would like to discuss further on your points and perspectives after I would hear more from people visiting my thread. I will hopefully further respond to your nice post, therefore, during the coming weeks.
 
Last edited:
Before you go for any low Class A amplifier like my A-36 from Accuphase you must be sure that your speakers have high efficiency. I used to have 95db/2.8V speakers and power was not an issue, now I have Revel F36 with 90-91db/.2.8V which is OK but on the borderline for listening at a bit over my normal listening level . That is keeping peaks under 0db on VU meters (30 watt class A, goes to 60W+3dB in AB before clipping occurs . I never actually heard any clipping)
If you have speakers with efficiency below 90db/2.8V you probably need more power than A36 can give , so maybe Benchmark would be a more safe option regarding power reserves.

I am curious where do you set the crossover frequencies for the different frequency ranges?
 
Last edited:
Before you go for any low Class A amplifier like my A-36 from Accuphase you must be sure that your speakers have high efficiency. I used to have 95db/2.8V speakers and power was not an issue, now I have Revel F36 with 90-91db/.2.8V which is OK but on the borderline for listening at a bit over my normal listening level . That is keeping peaks under 0db on VU meters (30 watt class A, goes to 60W+3dB in AB before clipping occurs . I never actually heard any clipping)
If you have speakers with efficiency below 90db/2.8V you probably need more power than A36 can give , so maybe Benchmark would be a more safe option regarding power reserves.

I am curious where do you set the crossover frequencies for the different frequency ranges?

Hello Balle,

I thank you indeed for your encouraging post!

As shown in my recent several posts, I am considering the possible use of two or three units of ACCUPHASE A-36; one for dedicated and direct drive of highly efficient Beryllium dome mid range squawker (Be-SQs), another for highly efficient Beryllium dome tweeter (Be-TWs) plus very high efficient horn super tweeter (STs). Furthermore, in my multichannel multi-amplifier setup, I already eliminated the passive LC-network, i.e. eliminated all the coils and capacitors, and the crossover is done upstream by software crossover EKIO.
WS000976b.JPG


The crossover configurations by software crossover EKIO are;

15--50 Hz for sub-woofers (SWs): -12dB/Oct LR filters: driven by active SWs YAMAHA YST-SW1000
45--500 Hz for woofers (WOs): -12dB/Oct LR filters: to be driven by ACCUPHASE E-460 or Benchmark AHB2 or A-36 (or A-48??)
500--6,000 Hz for Be-SQs: -12dB/Oct LR filters: to be driven by A-36
6,000-- Hz for Be-TWs and STs : -12db/Oct LR filter (Lo-cut at 6,000 Hz)+1.5 microF -6dB/Oct Lo-cut CAP for STs: to be driven by A-36

(note: I already have E-460 as my reference amplifier)

I believe there should be no power problem with A-36; one for BE-SQs, one for Be-TWs+STs, even though I do not have the actual efficiency values for these high efficient drivers.

I am now searching for the efficiency value for the WOs of YAMAHA NS-1000 which is YAMAHA JA3058(A) cone woofer to assess whether A-36 would be powerful enough to drive it or not. I do hope Coach_Kaarlo who is driving YAMAHA NS-2000 with two of Benchmark AHB2 would post his thoughts on this issue. Please note that I will dedicatedly and directly drive the WOs by one A-36 with no LC-network, i.e. no coils nor capacitors for WOs.

Of course, if I would have one A-36 (purchase or demo unit), I could easily check all of the above.

From "save the forests" and "bill for electricity" points of view, I think it would be better for me to use E-460 or AHB2 for WOs though.....
 
Last edited:
Hello Balle,

I thank you indeed for your encouraging post!

As shown in my recent several posts, I am considering the possible use of two or three units of ACCUPHASE A-36; one for dedicated and direct drive of highly efficient Beryllium dome mid range squawker (Be-SQs), another for highly efficient Beryllium dome tweeter (Be-TWs) plus very high efficient horn super tweeter (STs). Furthermore, in my multichannel multi-amplifier setup, I already eliminated the passive LC-network, i.e. eliminated all the coils and capacitors, and the crossover is done upstream by software crossover EKIO.
View attachment 83675

The crossover configurations by software crossover EKIO are;

15--50 Hz for sub-woofers (SWs): -12dB/Oct LR filters: driven by active SWs YAMAHA YST-SW1000
45--500 Hz for woofers (WOs): -12dB/Oct LR filters: to be driven by ACCUPHASE E-460 or Benchmark AHB2 or A-36 (or A-48??)
500--6,000 Hz for Be-SQs: -12dB/Oct LR filters: to be driven by A-36
6,000-- Hz for Be-TWs and STs : -12db/Oct LR filter (Lo-cut at 6,000 Hz)+1.5 microF -6dB/Oct Lo-cut CAP for STs: to be driven by A-36

(note: I already have E-460 as my reference amplifier)

I believe there should be no power problem with A-36; one for BE-SQs, one for Be-TWs+STs, even though I do not have the actual efficiency values for these high efficient drivers.

I am now searching for the efficiency value for the WOs of YAMAHA NS-1000 which is YAMAHA JA3058(A) cone woofer to assess whether A-36 would be powerful enough to drive it or not. I do hope Coach_Kaarlo who is driving YAMAHA NS-2000 with two of Benchmark AHB2 would post his thoughts on this issue. Please note that I will dedicatedly and directly drive the WOs by one A-36 with no LC-network, i.e. no coils nor capacitors for WOs.

Of course, if I would have one A-36 (purchase or demo unit), I could easily check all of the above.

From "save the forests" and "bill for electricity" points of view, I think it would be better for me to use E-460 or AHB2 for WOs though.....


Thank you for your contributions to this forum / thread, and for your methodical and thorough process. I am enjoying following your progress.


I can only offer limited information - my opinion is based on limited experience, and as yet I have been unable to measure the low frequency response discrepancy accurately. However it is easy to reproduce with simple tests.



When I was auditioning the NS-2000 speakers I was able to spend 5-6 hours each session comparing different amplifiers with the same speakers. The Benchmark AHB2, the Accuphase P300, and the Harmon Kardon Citation XX. We performed many tests - mainly focusing on small sections of music (10 sec max) and repeating until we could hear or not hear a difference. We also played at very high levels to generate woofer driver excursion so we could also measure / observe the differences.

The AHB2 was comparable to the Harmon Kardon Citation XX in many areas - but not the attack or kick. The AHB2 was open, transparent, spacious, detailed, clear, maybe missing the last 10-15% but you are working not listening to find these small differences. But play a track with strong base or big kick drum or EDM type of kick and there is a huge difference. For a while I was thinking that the AHB2 amp is actually more accurately reproducing what was recorded, but then I put my headphones on. The HD-650's matched the other amps, and highlighted what the AHB2 was missing.

I still have my 2 x AHB2's which I use in mono, because despite never driving the speakers into clipping - the mono power increase delivers a much better transient response to my ears, particularly with big old Yamaha drivers I think.

Some members on here have sent me private messages regarding this AHB2 issue, and speculated why it might be. And some have also posted publicly about possible causes of the differences.


Some more detailed comments here, I hope you find them interesting and maybe even helpful;


https://www.audiosciencereview.com/...ements-of-benchmark-ahb2-amp.7628/post-383627

https://www.audiosciencereview.com/...ements-of-benchmark-ahb2-amp.7628/post-384407

https://www.audiosciencereview.com/...ements-of-benchmark-ahb2-amp.7628/post-406146

https://www.audiosciencereview.com/...ements-of-benchmark-ahb2-amp.7628/post-410227

https://www.audiosciencereview.com/...ements-of-benchmark-ahb2-amp.7628/post-412332



https://www.audiosciencereview.com/...state-of-the-art-amplifiers.14257/post-490087

https://www.audiosciencereview.com/...state-of-the-art-amplifiers.14257/post-491091

https://www.audiosciencereview.com/...state-of-the-art-amplifiers.14257/post-490993



I will be very interested to read how the AHB2 performs in your testing, particularly as you have something specific to look for or observe.

Of course one must be aware of confirmation bias, but a 50% drop in cone extension at the same levels is ALWAYS observable.
 
Thank you for your contributions to this forum / thread, and for your methodical and thorough process. I am enjoying following your progress.

I can only offer limited information - my opinion is based on limited experience, and as yet I have been unable to measure the low frequency response discrepancy accurately. However it is easy to reproduce with simple tests.

When I was auditioning the NS-2000 speakers I was able to spend 5-6 hours each session comparing different amplifiers with the same speakers. The Benchmark AHB2, the Accuphase P300, and the Harmon Kardon Citation XX. We performed many tests - mainly focusing on small sections of music (10 sec max) and repeating until we could hear or not hear a difference. We also played at very high levels to generate woofer driver excursion so we could also measure / observe the differences.

The AHB2 was comparable to the Harmon Kardon Citation XX in many areas - but not the attack or kick. The AHB2 was open, transparent, spacious, detailed, clear, maybe missing the last 10-15% but you are working not listening to find these small differences. But play a track with strong base or big kick drum or EDM type of kick and there is a huge difference. For a while I was thinking that the AHB2 amp is actually more accurately reproducing what was recorded, but then I put my headphones on. The HD-650's matched the other amps, and highlighted what the AHB2 was missing.

I still have my 2 x AHB2's which I use in mono, because despite never driving the speakers into clipping - the mono power increase delivers a much better transient response to my ears, particularly with big old Yamaha drivers I think.

Some members on here have sent me private messages regarding this AHB2 issue, and speculated why it might be. And some have also posted publicly about possible causes of the differences.

Some more detailed comments here, I hope you find them interesting and maybe even helpful;
https://www.audiosciencereview.com/...ements-of-benchmark-ahb2-amp.7628/post-383627
https://www.audiosciencereview.com/...ements-of-benchmark-ahb2-amp.7628/post-384407
https://www.audiosciencereview.com/...ements-of-benchmark-ahb2-amp.7628/post-406146
https://www.audiosciencereview.com/...ements-of-benchmark-ahb2-amp.7628/post-410227
https://www.audiosciencereview.com/...ements-of-benchmark-ahb2-amp.7628/post-412332

https://www.audiosciencereview.com/...state-of-the-art-amplifiers.14257/post-490087
https://www.audiosciencereview.com/...state-of-the-art-amplifiers.14257/post-491091
https://www.audiosciencereview.com/...state-of-the-art-amplifiers.14257/post-490993

I will be very interested to read how the AHB2 performs in your testing, particularly as you have something specific to look for or observe.
Of course one must be aware of confirmation bias, but a 50% drop in cone extension at the same levels is ALWAYS observable.

Hello Coach_Kaarlo,

I thank you indeed for your very informative post and the related links on your amplifier exploration for wonderful YAMAHA NS-2000. I believe your information and list of links are highly valuable for not only my project but also for many people visiting here having interests and even working plan for similar multichannel multi-amplifier system building.

I have actually already visited some of your listed links (and pushed "Like"), but not yet for the others. I am carefully reading and learning the contents of these links; thank you so much again.

So far, I have tested ACCUPHASE E-460, YAMAHA MX-A5200, DENTEC DP-NC400-4-EXP, Benchmark AHB2, YAMAHA A-S301(B) (only for tentative use) and ROTEL RB-1582 MkII in my single-amplifier + LC-Network full range system and also in my multichannel multi-amplifier system (without LC-network) using my very much flexible setup of;
WS000993.JPG

During the coming one week period, I decided to have my interlude or intermezzo calm-down days in my project to carefully consider and plan the next way-forward in my amplifier exploration.

In this interlude/intermezzo, all of;

your above wonderful post #271

tpaxadpom's nice perspectives and suggestions in his post #266

my post #267 on "high range phase shift depending on the NFB feedback"

Balle Clorin's kind post #269 on Class-A amplifiers

should be taken into serious investigation (in my brain) and consideration, I believe.

First of all, let me emphasize again about the merits and advantages of full elimination of LC-network (network coils and capacitors) in multichannle multi-amplifier system.

For example, in the power train of driving the woofers (WOs), the elimination of 5.0 mH coil (and 94 microF capacitor) from the circuit has effectively and considerably reduced the workload for the amplifier, i.e. looking from the amplifier, the WOs suddenly became much more efficient and easy to be driven. I could easily and audibly confirm it by letting software crossover EKIO to output only the 45-500 Hz LO range L & R channels through DAC8PRO into the amplifier to drive WOs with and without LC-network. I could confirm this with E-460 and also with AHB2.

I could also confirm and "hear" the same improvements in driving Be-SQ and Be-TW by eliminating the coils and capacitors from the circuits.

Yes, as generally well known , we can drive the SP drivers considerably more efficiently by eliminating the LC-network coils and capacitors to give better sound quality in all of cleanliness, resolution, speed, sonority, 3D perspective, and even in phase shift characteristics. Theoretically, we can use less powerful Hi-Fi amplifiers in multichannel multi-amplifier system without LC-network coils and capacitors. (Of course, I understand well that this does not always mean that I can use rather small class-A amp, like A-36, to effectively drive the woofers.)

My second but rather serious consideration is related to tpaxadpom's nice perspectives and suggestions in his post #266. As I replied to him in my post #268, and I also wrote at the end of my post #265, I now feel the same as expressed by tpaxadpom; he wrote "In my opinion chasing the ultimate resolution, soundstage and other "audiophile" aspects we have lost the sound presentation as a whole.". I still love "the sound presentation as a whole" given by ACCUPAHSE E-460, my reference amp, in my system, even though E-460 would be "measured" inferior to the recent hi-end class-D amplifiers. I agree with tpaxadpom that we are missing "something unknown" in using class-D amps.

I assume this may be related to the "secrets" for the many tube amplifier enthusiasts and hi-end pre-amplifier enthusiasts.

From the extremely subjective point of view, my multichannel multi-amplifier project should achieve "improved sound presentation as a whole" for my ears and brain, even if the objective measurement would fail to validate it, or even if the objective measurement would find worse.

As shared in my post #264, in multichannel multi-amplifier system, ROTEL RB-1582 MKII gave considerably better sound presentation in driving Be-tweeters (Be-TWs) and super tweeters (STs) in comparison with YAMAHA A-S301(B) probably due to RB-1582 MKII's much better power reserve and also much better "phase shift characteristics in high range", I assume. This finding and also through our careful listening sessions on rather quiet but wonderful recording of piano solo and period string instruments, I came to the idea of possible use of Hi-Fi class-A amplifiers, and looked on various information on ACCUPHASE A-36 and A-48.

It was a kind of surprise for me, therefore, finding the description in the A-36 Brochure on "high range phase shift depending on the NFB feedback" as I posted #267. Very interestingly, and I do not know why, there is no such description in the A-48 Brochure. 

Edited to add: I also found similar description on A-250 Brochure, so it looks all of their class-A amplifiers are featured with "Current feedback principle assures excellent phase characteristics in high range".  

I am still expecting and looking forward to hearing any comment and/or technical background on this issue in this thread from some of you, experts in "phase shift and NFB feedback".

Consequently, I am now rather tentatively reluctant to put my priority on the "pure class-D Path" in amplifier exploration.

Now, after having above day-1 interlude/intermezzo consideration, I assume that it would be really nice for me if I could test and evaluate E-460, AHB2 and A-36 simultaneously in my multichannel multi-amplifier system as amplifier-X, -Y, and Z, respectively, and more interestingly randomly assigned to amplifier-X, -Y and -Z.

Here, I mean that I would like to try in multichannel multi-amplifier system;

Case-1: E-460 for WOs, AHB2 for Be-SQs, A-36 for Be-TWs+STs

Case-2: AHB2 for WOs, E-460 for Be-SQs, A-36 for Be-TWs+STs

and even,
Case-3: AHB2 for WOs, A-36 for Be-SQs, E-460 for Be-TWs+STs

It would not be feasible, but I may even try,
Case-4: A-36 for WOs, AHB2 for Be-SQs, E-460 for Be-TWs+STs

If I could test and evaluate two of A-36 with E-460, then how about this?
Case-5: E-460 for WOs, A-36 for Be-SQs, A-36 for Be-TWs+STs

Hopefully, during next week, I would like to start communication with "the" audio pro-shop for AHB2 demo unit again, and with ACCUPHASE (through their dealers?) for possible demo unit of A-36.

Just Edited: typo of Be-SWs has been corrected to Be-TWs.
 
Last edited:
Hi Quality Small Power amps seem to be getting rarer. Connecting a 150 watt per channel power amp directly to a beryllium mid range or tweeter seems a little dangerous to me. I think dual coupling capacitors and fuses or other kind of speaker protection should be left in place permanently.
To protect against DC on the outputs. Be drivers are getting too hard to get and too expensive to risk them imo.

https://neurochrome.com/products/guardian-686
 
Last edited:
Hello gene,

Thank you for your nice suggestions.
Yes, I fully agree with you, and will keep the protection caps for Be-SQs, Be-TWs and STs all the way;
WS000992.JPG


I am also considering to have back-up Be-SQ, Be-TW, two of each, and hopefully the woofers JA-3058(A) or JA-3301, since very fortunately we still can find many of them in the auction or second market in Japan;
https://auctions.yahoo.co.jp/search...q=-1&oq=&sc_i=&fr=auc_top&p=NS-1000&x=18&y=23

And as a reminder again for myself and all of you, we should avoid any of magnetic susceptible metals in this type of DIY circuit including the metal plates and screws of the connection terminal blocks; please refer to my post #250 for the details.
 
Last edited:
Note: I am an end user audio enthusiast and I have no conflict of interest at all with any of the manufacturers, import companies, distributors and audio shops relevant to this post. This post is not intending to intensively share objective and/or subjective evaluations of the candidate amplifiers, but I would like to share about how I would test and try candidate amplifiers in this project.

Hello friends,

In the English Brochure for ACCUPHASE A-36, a pure class-A power amplifier, they described;
"Current feedback principle assures excellent phase characteristics in high range. As shown in the illustration, the A-36 uses the output signal current rather than voltage for feedback. Since the impedance at the current feedback point is very low, there is almost no phase shift. A minimal amount of NFB therefore results in maximum improvement of circuit parameters."
View attachment 83438


In my comparison between YAMAHA A-S301(B) and ROTEL RB-1582 MkII in dedicated drive of high range with Be-TWs and Horn STs, I found and heard better high range sound with RB-1582 MkII probably and mainly due to "less phase shift".

As I am not an expert on "high range phase shift depending on the NFB feedback", I would appreciate having your thoughts and technical input on this issue, especially whether I can expect further better audible high range phase features/stabilities with using A-36 in my multichannel multi-amplifier system, or not.

I found these web papers;
https://www.eeeguide.com/parallel-current-negative-feedback-circuit/
https://www.eeeguide.com/effects-of-negative-feedback-in-amplifiers/
would be the technical background for Accuphase's description of "Current feedback principle assures excellent phase characteristics in high range. As shown in the illustration, the A-36 uses the output signal current rather than voltage for feedback. Since the impedance at the current feedback point is very low, there is almost no phase shift. A minimal amount of NFB therefore results in maximum improvement of circuit parameters."

I also found this textbook;
https://www.amazon.com/Feedback-Amp...92376439/ref=mt_other?_encoding=UTF8&me=&qid=

Although these deep (for myself) technical/engineering background descriptions are far beyond my knowledge and understandings, it looks that the "Current NFB" approach featured in Accuphase class-A amplifiers is based on the established technologies and methodologies.
 
I believe there should be no power problem with A-36; one for BE-SQs, one for Be-TWs+STs, even though I do not have the actual efficiency values for these high efficient drivers.

I am now searching for the efficiency value for the WOs of YAMAHA NS-1000 which is YAMAHA JA3058(A) cone woofer to assess whether A-36 would be powerful enough to drive it or not. I do hope Coach_Kaarlo who is driving YAMAHA NS-2000 with two of Benchmark AHB2 would post his thoughts on this issue. Please note that I will dedicatedly and directly drive the WOs by one A-36 with no LC-network, i.e. no coils nor capacitors for WOs.

Of course, if I would have one A-36 (purchase or demo unit), I could easily check all of the above.

From "save the forests" and "bill for electricity" points of view, I think it would be better for me to use E-460 or AHB2 for WOs though.....
Troels Graversen (speaker designer from Europe) have measured Yamaha NS1000M and reworked the passive crossover. He provided a number of measurements that should be of interest to you. See here. Tweeter and Midrange drivers are ~93dB efficient. I won't tell you what the woofer sensitivity is, just read the entire article (ignore his comments about caps). He also switched crossover network and applied 3rd order low pass for the woofer driver. It should be easy for you to experiment with different settings given that you are doing crossovers on PC. BTW since you have NS1000 speakers you should have YAMAHA JA3058 woofers (assuming they are original). YAMAHA JA3058(A) woofer was introduced in NS1000M.
I was planning to bring NS1000M and have them measured in anechoic chamber, with pandemia I am not sure when this will happen.
 
Last edited:
Troels Graversen (speaker designer from Europe) have measured Yamaha NS1000M and reworked the passive crossover. He provided a number of measurements that should be of interest to you. See here. Tweeter and Midrange drivers are ~93dB efficient. I won't tell you what the woofer sensitivity is, just read the entire article (ignore his comments about caps). He also switched crossover network and applied 3rd order low pass for the woofer driver. It should be easy for you to experiment with different settings given that you are doing crossovers on PC. BTW since you have NS1000 speakers you should have YAMAHA JA3058 woofers (assuming they are original). YAMAHA JA3058(A) woofer was introduced in NS1000M.
I was planning to bring NS1000M and have them measured in anechoic chamber, with pandemia I am not sure when this will happen.

Thank you for your kind post again,

As shared in my post #84, I have been aware of the nice web page of Troels Graverson. According to your nice suggestion, I will carefully read it again.

Yes, you are right that the woofer driver of my NS-1000 is YAMAHA JA3058, it is not JA3058(A). I hope I could find healthy backup JA3058 in auction or second market in Japan. In case if I cannot find good JA3058, much more available JA3058(A) would be OK for my backup purpose.

I am really interested in your plan to first measure NS1000M in anechoic chamber. I do hope that your plan will happen in the very near future when the COVID19 pandemic calming down.

BTW, very fortunately, today I could arrange one demo unit of ACCUPHASE A-36 coming to my home this weekend Saturday. I am very much looking forward to testing and evaluating A-36 soon in my single-amplifier + LC-network full range system, and also in multichannel multi-amplifier system together with E-460 (and the tentative YAMAHA A-S301). Quite regretfully, Benchmark AHB2 will not be available together with A-36, though. I have firmly memorized in my brain the sound with AHB2, as you may assume.

As I emphasized in the above post #272, the SP drivers, i.e. WOs, Be-SQs and Be-TWs, are rather easily/efficiently driven by an amplifier in the multichannel multi-amplifier system because of the complete elimination of all the coils, capacitors and attenuators. I would like, therefore, even to try A-36 to dedicatedly and directly drive WOs.

I would highly appreciate having any of your suggestion regarding my possible tests and evaluations of A-36 in my very flexible setup. The demo unit of A-36 may stay my home until around October 5, this means it will stay here for 8 full days.
 
Hello friends,

Just for your interests and reference, today I had interesting conversation with @chebum at here and here regarding the crossover filters, IIR, FIR or linear phase. Interestingly, he wrote, "I also tried both. Linear phase filters had more interesting sound, but IIR filters were more comfortable during longer sessions."

In this multichannel multi-amplifier project with software crossover, I am eventually using software crossover EKIO, and "EKIO uses IIR filters. The processing is done using a cascade of second order transposed direct form II biquad sections. Every calculation is done using 64 bit floating point numbers." Please refer to here and here for the details.

Also in my post here, I confirmed "EKIO gives no audible post-ringing nor pre-ringing at all in my rather sensitive test environment with -12 dB/Oct LR filters." I agree with @chebum that IIR XO filters by EKIO are really comfortable during longer listening session, although I have not yet done any intensive comparison between EKIO and other software crossover using FIR or linear phase filters.
 
Last edited:
@dualazmak , I just finished reading your entire journey, I have been planning something similar and you sharing with such discipline, consistency and rigorous methodology makes this thread the foundation for creating a scalable multi amping system or multichannel for what matters. Truly golden project, my respects to you sir, well done.
 
@dualazmak , I just finished reading your entire journey, I have been planning something similar and you sharing with such discipline, consistency and rigorous methodology makes this thread the foundation for creating a scalable multi amping system or multichannel for what matters. Truly golden project, my respects to you sir, well done.

Hello luisma,

Welcome to this thread, and I thank you so much for your kind and encouraging words.

As I wrote in my above post #277, I will soon test and evaluate pure Class-A amplifier in my project, both in single-amplifier+LC-network system and in multichannel multi-amplifier system.

Your continuing kind attention and participation on this thread/project will be much appreciated.
 
Back
Top Bottom