PMC + ATC mod and tri-amping

Innocent Bystander posted:

Allante93 posted:

 

{(Ain't that)}

Bad in Passive Mode, but can't wait to take that XO out of the Equation! .......

 

REPEAT, TEMPORARY UNTIL I PURCHASE A SNAXO 362 AND SC WITH THE BURNDY GOING FROM THE SC TO THE SNAXO, THEN THE SNAXO WILL DELIVER THE  SIGNALS TO THE REAR OF THE BRIKS THREE WAY SPEAKER. 

HENCE, ELIMINATING THAT CROSSOVER SHOWN ABOVE! 

Well, I'm having a bit of a play with tri-amping myself - but of the active variety. My initial steps are complicated by other variables, in fact those being the prime subject of study at this point  (posted separately thread https://forums.naimaudio.com/to...c-mod-and-tri-amping), but as I noted in that post it will lead to a trial of straight triamping of my PMCs in the not too far distant future.

My active XO at this point is the ATC EC23, and amps are all Musical Fidelity of the same family (P270 & 2x P170), so not a direct answer to the OP question,

{{however the question of active tri-amping vs passive mono-amping perhaps does have relevance. Perhaps I should compare passive tri-amp as well, but play time is not unlimited!}}

I also have a Behringer DCX2496, which could make for another comparison, relatively budget digital XO  vs expensive analogue. Maybe these are projects for when I retire, ears permitting!) One noticeable thing is that the heat output of a bank of 3 amps is considerable! (though perhaps because these have quite a strong bias into class A.) 

Quite an experiment you've going on!

This is the second time I've heard mention of that Behringer XO.

The First time, I think it was Active Naim with Kudos, due to the fact, the Snaxo was not applicable.

But now we know, Naim is collaborating with Kudos, as it relates to the Snaxo.

{{Now that I'm on your thread, I feel comfortable!}}

I haven't done any experiments with Active Systems!

But Scraping up 3K in the early 90's, to purchase my Briks, I've learned a lot.

1st, Passive Briks with Single LK 280

2nd, Passive bi-amped Briks with the addition of an Linn Spark.

3rd, Full Blown Active Briks, Linn XO with Bingo Card, LP 12/Ghenki

4th, dusted off the XO, which was configured for Bi-amping, and introduced the Briks to Naim!

282>HCDR>200>250.2

5th, Reconfigured the XO, for Tri-Amping, sold the 200 and here I am today:

Cdx2>282>HCDR>3 x 250.2>Briks

What did I Learn?

1st, Naim Amps sound Better than Linn Amps!

2nd, the 250.2, being an regulated amp didn't slip on the job, and went into protection mode, when handling the duties of driving the woofers and midranges, in Bi Amp Mode!

3rd, the 250.2 has no problem when it's task is limited to independent drivers.

However, the 250.2 responsible for the woofer, does get warm when I'm kicking the Briks!

SQ wise, only incremental improvements going from Bi Amp to Tri amp mode, the Main Advantages being more Headroom, and no more heating problems!

That's all I know, Innocent!

Nice Experiment!

Allante93!

Brief update: I've now removed the crossovers from the PMCs, adding protective capacitors, and done the necessary measurements to set the active crossover appropriately for the two different active setups. I've also arranged the wiring so that I can readily swap between passive (using the original crossover externally) and active driving of the PMCs, and active using the PMCs' bass with the ATC based satellites on top.

Next will come some listening sessions, so hopefully I will be able to report back in a couple of weeks or so.

Interesting post. I'll be keen to know your findings. Some pictures would be cool as well!

Updates 1- primary test: ATC SM75-150 mid + Scanspeak D3004/66000 tweeter vs PMC EB1i's native mid (Vifa D75MX-41) and tweeter (Seas H0881-06 27TFFC), in both cases tri-amped with Musical Fidelity P270 power amp (bass) +2x P170 (mid+treble). The active crossover was an old ATC EC23 analog device, with optimum crossover settings predetermined (including phase) and reset between the different speaker setups.

As well as listening myself, my visiting brother-in-law joined in (a keen audiophile who, until hearing my system with Hugo a couple of years ago, used to insist that vinyl sounds better.) At each changeover - which included additional tests of EB1 active vs EB1 passive - he had no idea what he was listening to. Our findings were in agreement, and wholly consistent with my first impression posted above.

The mid and top end with the ATC/Scanspeak is a distinct improvement, the sound being better defined and more natural, the latter very noticeable with, for example, intricate acoustic guitar and with female vocals, both of which also sounded more "in the room".

So, this mid and tweeter are definitely to feature in my system end-game: I just have to determine how to implement that:

A) Speaker options

i) Retain as tested: not the prettiest of setups, though easy. Probably my short term answer - if long term I would need to finish off the satellite cabs.

Ii) Modify the EB1 cabinets to replace the original drivers. Neatest solution but I will probably wait to see how I get on with iii):

Iii) Carry on with my transmission line bass cabinet design experimentation until I equal or better the EB1's sound, then simply replace the PMCs with my new home built speakers. A rather long term project as it is quite time-consuming.

B) Crossover/amplification options

i) Keep tri-amped as tested. Disadvantages: amp size, electricity consumption and heat (the relatively high class-A bias consuming 275W quiescent). And no way can I contemplate affording any major triple amp upgrade (or even double, utilising my existing Bryston). If I do stay with tri-amping, I will also try with a Behringer DCX2496 digital active crossover.

Ii) Once speaker set up finalised, design suitable passive crossovers (not necessarily easy to get right), and abandon tri-amping. This decision will be led by my conclusions on passive vs active driving of the EB1s, to be reported here. (The EB1's crossover looks to me to be of very good quality, consistent with the manufacturer's claims, and I would not expect to be able to make something that bettered it in quality.

Pics of the unfinished ATC based satellite (300x200x115mm)

Update 2: secondary test, PMC EB1i active vs passive.

Having removed the crossovers from inside the EB1s I wired them externally so that I could change between active powering using the three MF power amps and ATC active crossover detailed above, and passive powering with the MF P270 (for consistency with the active setup). The active powering of the midrange and tweeter includes protective capacitors (-3dB points 1.5 octaves below the respective 24dB/octave crossover frequencies); bass is direct drive.

Whilst active and passive did sound different, the difference was far from huge, and rather less obvious than I had been expecting (expectancy based on reading over the years). In particular I was expecting something really tangible at the bass end due to active being capable of better control of the driver, however the only specific thing I could put a finger on was that actively bass drums seemed a fraction fuller in sound, maybe a bit like the difference between hearing a firework explode from a bit closer (though sound levels unchanged). At the top end, cymbals on some tracks that passively sounded almost just like jabs of white noise revealed a hint of a metallic start to the sound. Overall I felt active sounded better, maybe a touch more open, but given that differences were rather subtle and that the fiddly changeover process made blind testing impossible it is hard to say that expectation bias was completely absent.

I also wonder if and to what extent my somewhat difficult listening room may be limiting my ability to hear differences of this nature: Whilst I have speaker and listening positions optimised for a reasonably level response, avoiding major bass peaks and troughs, the room does have rather a long decay time, to optimise which needs substantial room treatment (with similarly substantial practical issues and cost). But with a possible home move on the distant horizon my willingness to commit to the room treatment is inhibited - it will await a decision to stay indefinitely, otherwise see what a new room might bring. I will have to revisit this active vs passive test after either moving or treating this room.

Another factor of possible significance in the test was the 'wonky' amplification, P270 on bass and P170s mid and top, compared with passive P270. Differences in bass are definitely valid, but elsewhere could be due to any differences between P170 and P270. (P170 is P270's smaller sibling, I believe essentially similar design and so character, but 70wpc compared to 150). It is of course possible to retest with the other two wonky arrangements, and if I get around to doing that I'll update accordingly.

Based on my listening tests reported above, in my present room and with the amplification used, I do not feel that with my EB1i loudspeakers active is justified, with its increased cost and complexity. So, unless something changes, continuing with these speakers would be in passive mode for the forseeable future.

However, the ATC based satellite on which my last post reported, for now at least necessarily active, is another matter, that makes me question whether I will indeed continue with unmodified EB1s...

Also still to come, aside from the aforementioned possible follow-ups to my evaluation of EB1i active, is a comparison of the ATC EC23 analog and Behringer DCX2496 digital active crossovers. Additionally, I'll report my comparison of Musical Fidelity P270 power amp used in the above tests with Bryston 4Bsst2, which could be significant to choice of amps if I were to go active - maybe it might lead to a comparison of the EB1s passive with the old MF amps vs passive with Bryston...

Innocent Bystander posted:

 

Pics of the unfinished ATC based satellite (300x200x115mm)

 

 

Nice work IB - a very interesting thread. Your project reminds me somewhat of the System-1 from Code Acoustics - worth a google 

Update 3: Tertiary test - power amp comparison

Done alongside the testing reported in my last post, this comparison is a spin-off - it is not directly relevant to the main theme of this thread, except insofar as it confirms or denies the current bank of active amps as likely to get the best out of the active project. The power amps I compared are Bryston 4Bsst2 and Musical Fidelity P270, used to drive PMC EB1i speakers (passively).

First, a brief description of the amps: the Bryston's reputation is built on neutrality and grip on speakers (as well as build quality and reliability), the MF on its musicality, at least partly due to its strong class A bias. Despite similar sized boxes, the Bryston is actually more powerful at twice the MF's rated 150wpc/8?. My EB1i speakers are nominally 4?, into which Bryston claims a massive 500w - yet extremely enthusiastic playing has been known to cause their built-in overload warning lights to flash on transients, though no clipping was audible. I don't know what the MF's rating is into 4?, presumably about half of the Bryston but at similar levels I haven't noticed audible clipping (no warning light on the MF). They are both convention cooled with big heatsinks at each side, though the MF's fins get too hot to touch and remain that way when on, regardless of music level, whereas the Bryston is only mildly warm to the touch when idling, somewhat warmer after loud playing but never too hot to touch. Bryston amps are a natural pairing for PMC speakers as apparently they were the amps used in original test and set up of the speakers.

When I bought the Bryston some 20 months ago, it was a spur of the moment thing when I saw one advertised at a particularly attractive price (a mis-spelling on fleabay meant a dearth of bidders!) I had to import from Canada, yet even with the seller taking it first to the manufacturer for voltage conversion and shipping to Britain it was good value. It arrived just as I was going away for a holiday, so immediately on my return I plugged it in and thought this sounds great - I intended to do a direct comparison with the MF it supplanted, but have only just got around to it! So here goes:

Knowing Bryston's reputation for grip, I was expecting something noticeable at the bottom end even though the MF hadn't seemed weak. But actually I struggled to pick out anything. It was further up the spectrum that the differences were apparent, and quite noticeable. The MF had always sounded great to me, but going to it straight after the Bryston was as if the sound had been slightly blunted in some way, minor, but clearly noticeable. It was not that the MF lacked clarity compared to the Bryston, but rather that the Bryston added a sparkle that the MF lacked. I don't mean a 'zing' or 'tizz', but, perhaps, like looking at a highly polished surface - then replacing it with a mirror. And it is the Bryston that sounds better, rather than the MF. For reference, my Bryston is 13 years old (still under warranty), and the MF at least 25. The latter was serviced by MF 2 or 3 years ago, though not not much found needing to be done, even the main caps found to be in spec, while the Bryston was upgraded to sst2 spec by the manufacturer early in 2016.

So, my conclusion is clear: the Bryston 4Bsst2 is the better of the two amps, with a positive difference to music played through it compared to the MF P270.

That of course leads to questions relating to active options:
i) How does the MF P270/P170/P170 active triamping of EB1i compare against the same speakers passive with the Bryston?
ii) How does the ATC based satellite + EB1 bass with active MF triamping compare against EB1i passive with Bryston?
And iii) what would the ATC based satellite + EB1 bass sound like triamped with Bryston amps?
The first two of these I can and will evaluate. The last I can't without the considerable expense of buying, which would add quite a different dimension to this project, so presently at least, not on the cards.

Thanks to Richard for re-opening this thread.

I have a couple of updates, firstly on the amplifier front: The chance appearance of a well-priced Bryston amp on eBay (which went cheaply possibly because of a fortunate miss-spelling of the name!) gave me the opportunity to swap out the MF amps. It is model 9Bst: a 5-channel version of their 3Bst same spec and basic circuitry as the 4Bst  though with lower maximum power output, and the ST an older version than my 4Bsst² incarnation. With lower power requirement in the upper frequencies, amps that are closely matched apart from power capability are ideal partners for the mid and treble amps to partner the 4B on bass. I have removed the 5th channel module - a useful spare - making a neat system.

Maybe not as good as sst² versions all round, and maybe not as good as, say, 4B on mid nd high with something even more powerful on bass, the performance does beat the MF P270/P170/P170 setup, with greater clarity and authority.

Now to the nitty gritty: Indid say I’d come back with my assessment of the ATC EC23 analogue active crossover vs the much cheaper Behringer DCX2496 digital one. The ATC sounded better, the Behringer somewhat marred by its background noise level which was faintly audible during very quiet music, and in some ways I can’t define just didn’t sound quite as good - but it did have a distinct advantage of greater flexibility, including allowing a modicum of adjustable correction of imperfections in frequency response, whether caused by the speaker s themselves or the room. After some deliberation I decided to take a risk and upgrade the DCX. This was a DIY job, completely replacing the analog circuitry, changing the clock, replacing voltage regulators and adding a ferrite filter to the output of the power supply - cost in parts about £450, so allowing for my ‘free’ labour a significant multiplication of the purchase cost of the stock unit.

 The result is worth the cost and effort - now quieter than the ATC, the Behringer effectively disappears and just does its job. The ability to tailor the response means Incan save two configurations, one with the speakers at their optimum room position, with some but minimal ‘tweaking’, and the other with the speakers in a more general-room-purpose position where they live a major part of the time, the adjustability greatly compensating for the negatives of that position, something the ATC cannot do.

Whether there is any disbenefit of the extra DSP with the Behringer I can’t tell, as if there is it is outweighed audibly by the benefit. So it is a keeper, and the ATC crossover will go to a new home.

How it compares with other digital crossovers I have no idea as the Behringer is the only one I have heard. Ditto how it compares with other analogue  crossovers like SNAXO. But it is clear to me that the additional DSP a digital one can offer can be very advantageous, though that will vary between different systems and rooms (and room layouts.)

A primary advantage of the digital crossover is easy tailoring to any speaker for which the passive crossover can be removed or bypassed, removing the need for a specific factory set up device: All one really needs to know about a speaker one is trying to bi- or tri-amp is a reasonable idea of suitable starting crossover frequencies. Together with monitoring software like REW and a measuring microphone, suitable settings can easily be found (and the microphone can also be used with the XO’s inbuilt time delay/phase correction assessment tool). Some people have indicated frustration with absence of a SNAXO tailored to their choice of speaker to allow bi- or tri-amping:  a flexible and easily set up digital crossover like the Behringer presents a possible solution. The modifications I made are not difficult for anyone capable of wielding a soldering iron, however there is at least one supplier offering a similar upgrade ready-made, though of course somewhat more expensive, while there may be other brands of digital crossover equally good or better.

And of course a crossover like this not only can be used with a variety of speakers, but also virtually any power amp, which can include Naim. It may or may not be the best possible, but for me the modified DCX2496 does a very good job and is remarkable value for money, with its advantages in imperfect room situations likely to beat any analogue crossover.

I'm surprised you did not give Bryston's analogue crossover 10B-STD a go. Though you need 2 of them in mono operation for 3-way speakers. That doesn't require factory setup and comes in 2 versions. One with dials and the other (which I would prefer) where you fit Linkwitz-Riley filter cards. 

If I ever decide I need to go active, rather than force myself to choose from a limited number of options that work with SNAXO, I'd more than likely fit a Bryston analogue crossover into my Naim system.

I do get what you say about the flexibility and cost performance of digital ones though. While it will always be a lossy conversion to and from digital, I suppose that is a moot point is the losses are less than the losses incurred by a passive crossover. 

Is this becoming a PMC 20.26?

Richieroo posted:

Is this becoming a PMC 20.26?

It was above the 20.26 in original stock form: The EB1 is an extended bass version of the IB1, an earlier incarnation of the current IB2. This is  bringing it up to being the EB2 PMC never produced, an extended bass version of the IB2 - but also with the advantage of being actively driven.

Any chance of a pic of the whole setup?

feeling_zen posted:

I'm surprised you did not give Bryston's analogue crossover 10B-STD a go. Though you need 2 of them in mono operation for 3-way speakers. That doesn't require factory setup and comes in 2 versions. One with dials and the other (which I would prefer) where you fit Linkwitz-Riley filter cards. 

If I ever decide I need to go active, rather than force myself to choose from a limited number of options that work with SNAXO, I'd more than likely fit a Bryston analogue crossover into my Naim system.

I do get what you say about the flexibility and cost performance of digital ones though. While it will always be a lossy conversion to and from digital, I suppose that is a moot point is the losses are less than the losses incurred by a passive crossover. 

Re Bryston 10B, although of course I can’t be certain of performance as  I haven’t compared, I’m not sure that it has any advantage over the ATC EC23 - in fact may not be as good because I don’t think it has adjustable phase between the channels, somethingbthe ATC has (likewise the DCX).

feeling_zen posted:

Any chance of a pic of the whole setup?

I’m away at the moment- I’ll see whatvI can do in due course

I mentioned that In my room I have two listening/speaker positions, and that  the Behringer active crossover's additional DSP capability allowed setting up multiple configurations allowing some degree of tailoring to the two positions. To expand on that, one position, the 'serious listening' position, has the speakers about 9 ft apart, in front of wall/window (baffle at bottom ~65cm from wall, top ~95cm from glass), and listening position about equilateral, an armchair I pull in front of a sofa.  The other is the 'resting' one, where the listening position is further back, on the aforementioned sofa, and the speakers are further apart, either side of the 12ft  wide window blind that is actually a projector screen - in other words the same positioning as when watching a movie. To change I walk the speakers between the two positions using tape measure  to finalise, and carry the armchair.

In that 'resting' position position the sound is compromised because the speakers are right up against the wall in a corner (though 135º not 90º corner so not as bad), and the speakers are a tad far apart relative to the listening position.  OK for background music, but otherwise not great. However, since setting up the crossover's additional DSP capability to even off most of the negative effect of the speakers' corner proximity, making it quite acceptable, albeit not quite as good as in the optimum position, I have found myself hefting the speakers and chair less often, frequently content just to sit on the sofa as is, especially when not just focussed on the music. It is also better for when friends are round, for which the optimised position is awkward. I was wary of using the the additional DSP function, but overall any negative it has (nothing I have noticed) is outwighed by the positive.

Pics still to follow - will have to wait till the new year now.