Keuze voor units: Een ontwerper aan het woord

[markvl]

Dit topic is geopend naar aanleiding van posts in een andere thread over keuze van units en crossovers.

Hierbij een quote vam Mark L. Schifter (eigenaar van AV123) uit een thread op hun forum: http://forum.av123.com/showthread.php?s=&threadid=735&highlight=ref+crossover

De vraag was:
"given your expanding line of speakers and w/o revealing any secrets would you care to speak to how various mechanical properties (some measured others not), e.g # and type of speaker, crossover properties, frequency response, imaging, etc. along with the actual sound of a male, female or favorite instrument are combined (or avoided) to produce voicing commonality, assuming that's a goal across your expanding line of speakers?"

Part One…

This question posed is so important… I think I will answer it in two parts… The first with just some general discussion points… and the second, some actual things we do to yield “our sound”…

Hello Theron…

First, please allow me to apologize to you for this extremely tardy reply. The simple truth of it is these days I’m concentrating more and more on product… and less and less on e-mail and the forums. I tend to get focused on something… and just go go go until it’s done.

On the product side of things… there are many projects working at the same time right now. This and the new factory are taking up the lion’s share of my time. There are so many planning and execution issues to deal with… especially from this far away. I’m heading out this coming Tuesday to get an in person update in China on many issues…

Your question is an interesting one… and it would take many pages (and lots of time) to reply in full. I will offer a few tidbits that I think are important to the “voicing commonality” in our line… and a few of the tricks the design team has used and what *I* personally look for. I’m going to answer this in two parts… and some of this might be more than what you’re actually asking… but I hope it will all make some sense in the end.

First… I do want to say that the voicing within our line is purposely “tweaked”. Our RSS 300 uses a tweeter that is less “weighty” and more “airy” than the VIFA Ring Radiator. This was on purpose… We wanted a tad more “air” from our surround speaker… and hence we use aluminum (in concert with silk) dome to produce a nice brighter-and-lighter sound.

We decided to use this same tweeter for the RS150 and RSC100, as first, we had it designed and we were pleased with the result… and (b) we liked the way it integrated with the 5.25” woofer we designed. The XO became easy to deal with… and the result was one we really liked.

I want to also say here and now that the “voicing” of the REF Series was (also) purposely different in texture and character from that of the Rocket line. There were many reasons for this… and I’ll get into more of that later…

As for the crossover in general… I tend to like 3rd order. This is just me, as many designers like the newer (and trendier) 4th order types (in their many variants). I feel we can enjoy the benefit of phase order, and power handling when 3rd order is utilized. Power Handling is a big issue when one considers the demands of Reference Level playback from these small(er) sized speakers.

I’d like to also add some additions to this discussion. Drivers are in fact very important. Tom Roper recently had some very good things to say in several threads here and on avs about drivers… and the plain truth of it is… he’s right. I will say however that the “art” (if you will) comes in with driver utilization WITHIN the constraints of XO and the enclosure as well. While the drivers are the “critical element”… the box and the XO are not far behind.

There are tricks that one can use to “mask” problem areas of the design. I have to apply some simple formulas to what we are trying to accomplish here in an overall product-to-market sense and cost realities. I cannot use a $50 dollar tweeter or a $100 dollar woofer in a design I was to price at $399 a pair. It just does not work… (perhaps this is more information than you or others might like to hear… but I feel it’s important for everyone to understand)…

We design in everything we can --- given the price target and our overall margin needs. We then give it our ALL. Please do not discount these statements. Our products are unique within the industry as I feel we ARE designing in more BECAUSE of the margin calculations we’ve decided to work within. This is critical. If I had to deal with dealers, training these dealers… bad debt, larger staff requirements etc. etc. we could NEVER hit these price targets. Then, taking it one important step further we can use better parts and much nicer enclosures, again, because of the margins we’ve decided that are acceptable for our business as juxtaposed to the prices we’ve decided to charge for these wonderful products. It ALL adds up…

Now I know as Part One is ending here I haven’t really answered your question at all… but I AM getting there… All of these parts and enclosure tweaks we use are all part of the package --- again, in the context of the prices we are charging for these products (using our business methodology) it’s really amazing we can do what we are able to do at these numbers. I’m using design and manufacturing techniques that are found in speakers costing MANY MANY times the price of our Rockets or REF’s…

It all gets back to what we have put in… and what we wanted to accomplish (at XXX price) within the acceptable business margins we have decided to operate our business under. It all really adds up rather nicely, and makes it easy for me to understand why we are giving other speaker companies pause.

More to follow… Part Two will be more “tweaky” for sure…

All the best…

mls



Part Two

Let me begin by saying that due to my usual time constraints, and some bandwidth issues I’ve faced on this trip… I will now actually have to do this in three parts… I’ve only and hour or so this morning as I’m getting ready to move onward.

Part One was sort of the appetizer for me (and hopefully for you also)… I felt some general things needed to get sorted before going on. I will not necessarily reveal any tricks or magic pixy dust methods… just stuff that some of you may already know… and perhaps a practical tip or two. I’m going to speak a lot in general terms here… and use the RS 750 as my main subject for this part of the reply.

The crossover (or diving network as some like to call it) is where a lot of the art resides. There are several different methods that can be used to divide the signals out to the various drivers. We’ve used some rather interesting techniques here, but as this relates to the “voicing” question(s) let me talk about the following as it relates to Rocket, and in this case… particularly the RS 750.

We have found that in general very low DCR (DC Resistance) inductors are best. We use a “heavy enough” gauge of wire to do the trick (although not too heavy for other reasons), as we are very careful about saturation. I actually changed the overall gauge of wire on the first legs of the 750 XO’s several times before we got the value right (not the actual value in mH DP spec’d… but the DCR vis-à-vis the necessary value of the component).

I listened a lot to heavier gauge coils using what’s called an “air core” type of inductor, and finally I found that the sound was not punchy enough for my taste (there are a number of reasons for this… but Arnie Nudell always told me to “listen to the damn network also” --- and he was, as expected… always right about this)… so I switched to an “iron core” type on woofer one (RS 750)… and BINGO… everything clicked into place. In all other legs of the network we use air core types, again with low DCR…

When needed, we do use large value electrolytic capacitors (this is not unusual in any way)… but we are careful to use particular brands which we know to be “good parts”, even taking to stacking smaller values to get the necessary value we are looking for. I tend to stick with things that I’ve used in the past at Genesis etc. Arnie is a vast source of practical knowledge and I’ve had a chance for myself to listen to different types of parts in speakers costing 10’s of thousands of dollars… and also doing the same for speakers costing quite a bit less than that.

Please note that in the early days of Infinity, Arnie was building some incredible $500 to $1000 dollar speakers (the Monitor Junior comes to mind… as does the 1001A and POS I)… and later all of the EOSONE models which came about as a result of a partnership between Polk Audio and Genesis. These were very nice speakers for the money, and I had good practical knowledge of the circuits and the parts used… and more importantly… how to improve even these lower end speakers with the judicious use of the “right parts”…

To continue… many designers like to “bypass” these “lytics” with Mylar type or Film and Foil type capacitors. As a rule… in the bottom end… we have not done this. I know many designers swear by this… but DP’s holistic approach was so good… and so right on… I mainly just had to tweak a bit. All actual parts were spec’d by me… but EVERY actual value (except for the tweeter circuit) came from DP. Why mess with the master… We also used a very high grade of circuit board material… and not the norm for China which is sort of a “pressed cardboard-like material” with point to point wiring. These XO’s will not stand the test of time (in many cases) as they are mechanically unstable. We use a very high quality material (in fact all of the materials are of extra high quality) and a much better grade of solder than I normally see overseas. All of this adds up.

We did decide to spend the extra money and use “high end” Mylar, or in some places Film and Foil capacitors for the high frequency part of the dividing network… To my ears we were able to achieve the right sound by using the multi-pole network we chose and the topology… but also… equally important… the parts we finally settled on.

Each component in the XO makes an impact on the overall sound. Also the hook up wire we use (internally between the drivers and the XO… and the XO to the input connectors) are of the highest quality. I’m not going try to argue it out with anyone… but all of this DOES seem to very much matter. Also… please note that the connectors we use (ONIX branded) are the very best available in this genre of product (at least that I’m aware of)…

Let me also comment that we changed the dampening materials used in the 750 several times before we got “it” right. This is a critical part of the equation… and as the early beta testers can attest (our now famous Rocketeers)… I asked them to take out and replace much of the dampening materials a few times.

The early stuff was just flat out wrong… as we did not have enough material in there… Then we tested several types to get the snap I wanted… but not loose the “damped” effect (proper dampening) we also had to have (physics aside) for the drivers and the array configuration we choose. And then there was the fastening to the wall(s) of this material… and in some cases being careful not to run afoul with the bracing. (More about bracing next time… I changed this twice to get it right, cut holes and all)… It all adds up…

I know that I *still* have not really answered the questions posed… but I felt you needed these appetizers first before the entrée is to be served. There is so much that goes into a well designed system outside of just the drivers (although Tom Roper DOES have it mostly right, IMHO in what he has said about this)… and the proper use of the “right” components all the way through counts for much of the result…

Next time I will talk more about the actual voicing part… and how I go about deriving the sound I am looking for. We’ll examine the drivers themselves as a major part of the process and how I set up entire listening sessions between seemingly similar products to prove out what “voice” will win in the end.

Fun stuff…

Wish I could add more now… but I’m pressed as usual.

Thanks again for all of the support…

All the best…

mls

Groeten,
Mark

[markvl]

De specifieke questie in het vorige topic was de noodzaak voor een supertweeter.

In de Reference reeks van AV123 zijn 3 speakers leverbaar.
- Ref1 (1 Vifa XT Concentric Ring Radiator, 1 Atohm 5 1/4")
- Ref2 (1 Vifa XT Concentric Ring Radiator, 2 Atohm 5 1/4")
- Ref3 (1 Vifa XT Concentric Ring Radiator, 1 Magnetostat Super Tweeter, 2 Atohm 7")

Gezien het bereik van de Vifa lijkt het op zich niet nodig een super tweeter toe te voegen.
Heeft iemand ervaring met het verschil van zo'n ding?

Groeten,
Mark