Cheaper path to speaker quality?

I would like to open a discussion on amplifier damping, and driver cost.

It is my belief that the biggest problem with passive crossovers, even those with expensive caps and silver coils, is that the amp's control of the driver goes out the window at the crossover point! It is designed to do that. The inductors in the woofer circuit and caps in the tweeter circuit, (let's keep it a simple 2 way system) decouple the drivers from the amplifiers at the critical crossover point (loss of dampening). Check Qes and Qms, speakers are mainly controlled electrically, (and this is the way it should be). This means that at the crossover point the driver are going out of control. I suspect that is why many high-end manufactures stick to first order crossovers, (as well as other advantages). In fact the dampening is compromised through out the driver's range by a passive crossover.

People pay fortunes for low impedance speaker cables, if they only knew what their crossover was doing to their amp-speaker connection.

Some argue that an amplifier can have too high dampening,( low impedance). I have tried high impedance connections between some of my amplifiers and compared them with direct connections and I would have to say that some recordings sound more musical. On tight, dynamic and well-recorded pieces though, the dynamism was lost, probably no surprise here. The conclusion that I have come to is that high impedance amps can mask some harsh recordings by allowing more low order speaker harmonics. It's not impossible is it?

How can multi amping be cheaper? I can brew good quality 100w amps for around $100 Aust. each. I feel that directly connecting a low impedance amplifier to each driver will get me better control of the speaker than buying expensive drivers and feeding them through passive crossover. I'm betting that a $200 woofer direct amped will sound tighter, more controlled than a $1500 woofer fed through a passive xo. We often find our selves in the regions of diminishing returns when purchasing drivers as the manufacturers need to go to great expense to design and produce drivers with fractionally better behaviour. Behaviour that might be achieved cheaper by better amp-speaker coupling.

Of course well-behaved, expensive drivers, direct amped would then be even better.

There are real dangers to the drivers of direct amping, especially the tweeters! I have never toasted a tweeter but I have replaced many in other peoples systems, (including active systems).

Maybe these are just the musings of an ageing mind going soft. Your thoughts gentleperson's?
Regards WALKER
 
There is no question that multi-amping provides better control, higher volume levels, and better sound. But I disagree with your comments on how affordable it is.
An active xo will be many many times more expensive then a passive xo. Amplifiers are almost always more expensive then the drivers, unless you buy extremely expensive drivers and power them with a class-D HT amp.
Keep in mind unless you're willing to go for high-quality DSP an active xo will also cause some distortion.

And in the end, amplifiers make much lest difference in quality than the loudspeakers themselves. If I bought the cheapest workable amp and used it to drive high quality drivers it would definitely sound better than a $25,000 tube amp driving speakers I took from an intercom.
 
Ignite, do you think that a well designed active crossover costs more than a well designed passive ox? I have to say that I usually roll my own I active crossovers for less than $100. Passive crossovers get very expensive very quick if you use the same quality components.

With the price of drivers, you get into the law of diminishing returns very quickly. In the repair industry I have used bass drivers ranging from $20 to >$1000Aus, (12"). It has been my experience that once you get to $200 the improvements are usually lost in the enclosure, room acoustics and I believe the crossover.

If cost weren't an issue I would buy the expensive drivers and multi-amp then, but I feel that I'll get the best bang for buck if I buy quality, value for money drivers and multi-amp them. The amps that I'm planning to use are Doug Self type design, 100W and I can build them for less than $100 Aus. I'm planning on using 16 of them, one for each driver, which is still cheaper than most commercial power amps.

If I mate a $200 driver with a $100 amp directly connected, I think that it should out perform a $1000 driver with passive crossover and a $1,000 amp, (or a $25 000 amp).

Your thoughts, what am I missing?
Regards WALKER
 
Again, we have to consider that some amps work better with some drivers than others. If all amps produced a "miracle" sound completely true to the source material, then yes, amping directly would be theoretically "perfect". However, the sound coming from a driver will vary greatly depending on the quality of an amp. A passive design with few components in the signal path being driven by a Pass or Mark Levinson amp may outperform the same speaker being driven directly by a $200 home theater receiver.

(I'm still all for active systems, don't get me wrong :))
 
Super thanks for the input. You said " A passive design with few components in the signal path" do you think that this is why many top shelf speaker manufacturers opt for first order crossovers? That has been my suspicion.

I suspect that even a cheap amp direct connected will end up doing a better job, than a Pass amp with a passive crossover in its way.

The cheap little 100W amps I'm considering will outperform most $2000 HT amps.

Regards WALKER
 
Kelticwizard, yup, and damn good ones too. My junk box fills two sheds, (just ask "she who must be obeyed"). The enclosures aren't going to cost me because I'm making them from scrap.
Transformers and heatsinks are cheap if you’re not in a hurry and know where to look.
I'll be surprised if they end up costing me any more, (but I've been surprised before).
Even if I didn't already have the parts I can buy the 100W D Self style amp kit (PCB and components) for ~$80, 500VA transformer $50, 10 000uF 100V caps $10. Those are the big-ticket items.

How does that compare with US prices?

Regards WALKER
 
kelticwizard said:
Just thought I would throw in that one Australian dollar equals 52 cents American. So Walker is saying that you can build a 100 watt single channel amplifier for $52 American.


That is really cheap. Wow. Perhaps I should put more thought into learning how to build an amp.

Edit: Can somebody point me in the direction of some material on this stuff? I'd love to see how hard it'd be.

[Edited by Ignite on 11-22-2001 at 10:14 PM]
 

kelticwizard

diyAudio Moderator Emeritus
2001-09-18 2:33 am
Connecticut, The Nutmeg State
Walker:

Wasn't doubting you, just breaking it down for my fellow Americans.

Your thread saves me the effort of starting my own on a similar subject. I have not built a full scale amp, but I do design little filters, etc. Home theater receivers are five channel. My question was: wouldn't it be better to buy a pair of home theater receivers and put active filters on each channel to go to separate drivers in your loudspeakers? The components of active filters are pretty cheap, it would seem to be a case of merely hooking them up.

I realize that people rave about the advantages of a really good amp over receiver sound, but wouldn't the advantages of active crossovers more than compensate? This is assuming that 100 wats per channel is more than sufficient to meet your needs-I realize that really good amps go up to 1,000 watts per channel and more. I realize that sudden peaks can require the use of a lot of power suddenly, but 100 watts, for many people, will be sufficient to cover those peaks, I would think.

Just a thought.

[Edited by kelticwizard on 11-22-2001 at 10:28 PM]
 
Ignite, try some of the ESP (Elliott Sound Products) or Leo Simpson/Silicon Chip magazine, (the name may be dumb but the designs aren’t) designs and let me know what you think.

Kelticwizard, I like your idea and it has merit, unfortunately I have not been impressed by ANY of the HT amps that I have seen. Most specify 100W x 5ch and use a 200VA transformer but that’s not the only issue. It has been very disappointing even from manufacturers that build good HIFI gear. Having said that, I think that you would often be better off doing as you suggested than the normal setup of amp/passive crossover/speaker.

I like inefficient speakers, (heavily damped, low Qes) so I like low impedance amps that can live up to their specifications, and that sound clean.

By using 5 x 100W amps you will get greater and cleaner output than some 1000W amps because of the efficiencies of multi amping and dang if it aint often cheaper:)

Regards WALKER
 
Xovers

I just bought an active cross but as a toy #1 and as a tool #2 to find that perfect xover point. As a simplistic view I like a SE tube amp with a gain pot hooked up to a 2 way with a 6db high pass on the tweet. 1 source, 1 interconnect to mono block, 1 cable to speaker which is high efficiancy 2 way with as simple a cross as possible. Carefull selection of a full range and a tweeter in a TL cabinet. Complicated xovers are power suckers and can rob a great SE of it's best qualities before the signal reaches the drivers. An isolation transformer for the analog gear is helfull in quieting background. If you want to isolate everything get another for any dig gear. I will let you know how I do in next year in the Texas DIY Speaker Shootout. As far as cost goes there is no reason why $USD 300/channel shouldn't get a DIYer the same or better than $50K for a system in a "High End" store. $10K for cables? Bite me!
 
Hope this helps

My preference has always been to have the individual speaker drivers connected to their own amplifier. That means that a stereo MTM setup would need 6 power amps. If the output devices are directly connected to the drivers without crossover components, the control that the amp has over the speaker is increased.
With this setup the crossover is placed between the control/pre-amp and the power amps, which many believe is the proper place for them. It works extremely well in theory and practice. It is the preferred method for sound reinforcement / band / stage etc.

The reason that it isn't used domestically is because speaker manufacturers don't make amps as a rule.
:)
 

Brett

Member
2002-01-07 6:02 pm
<b>My preference has always been to have the individual speaker drivers connected to their own amplifier. </b>

Me too, but cost, heat, complexity and voicing issues can sometimes make it more of a hassle than it's worth. Besides there are simple ways of implementing crossovers on some speakers. My horns have conveniently placed acoustic roll offs and so don't need complex crossovers.

It's also imperative that all the amplifiers used are of the same topology and have a similar voice (identical amps are great), or you have thrown away much of the benefit. It will sound more hi-fi, and less of a whole.

<b>That means that a stereo MTM setup would need 6 power amps.</b>

It should only need 4, as the M drivers will be in parallel.

<b>If the output devices are directly connected to the drivers without crossover components, the control that the amp has over the speaker is increased.</b>

Yep. Usually.

<b>With this setup the crossover is placed between the control/pre-amp and the power amps, which many believe is the proper place for them. It works extremely well in theory and practice.</b>

Yes it can, but I've seen a few implementations where the signal is sent through an unwieldy and over complex crossover stuffed full of 5534s, or tube CF's and sounds better with a simple passive (speaker level xover) in it's place.

Passive line level xovers are the best of all IME

<b>It is the preferred method for sound reinforcement / band / stage etc.</b>

Yes but what is used for PA, and the reasoning behind it, may not transpose ito a domestic environment. PA drivers, esp in larger rigs, are often driven very hard, so that heat becomes a significant issue in the driver. As drivers get hotter, their T/S parameters shift around a lot and a passive xover optimised at the 1W level will not still be optimised at 400W+, esp after a couple of hours at this level. With sensibly efficient drivers, ie >92 real dB, pref more, then the power and heat levels will not vary significantly in a domestic environment, so it's easier to set the xover up.

Most larger PA's use horns, and often controllers like SMAART to equalise the path length in the horns and arrays which can't be done well in the analogue domain.

<b>The reason that it isn't used domestically is because speaker manufacturers don't make amps as a rule.</b>

No, I would argue that it is for the issues I listed at the beginning of the post, as well as restricting choice as to what the customer prefers in their 'sound'. Almost every active speker I've seen has the amps built in to the enclosure somewhere, probably the most stupid location imaginable. The Linn and Linn/Naim systems are exceptions.

Cheers
 
Ignite said:
If I bought the cheapest workable amp and used it to drive high quality drivers it would definitely sound better than a $25,000 tube amp driving speakers I took from an intercom.

I think that in this case you would be wrong most of the time. I am a firm believer that in a system where every component is important, the speaker is the least important. As Thatch-ear says it is amazing what can be had from a mostly full-range system. My next speakers will have drivers that cost me less than $20 CAD ($1 CAD is about 65 cents USD) and i know they will sound better than my already decent sounding, more expensive (but still Frugal-phile (tm) priced) speaker system.

dave
 
Brett said:
<b>That means that a stereo MTM setup would need 6 power amps.</b>

It should only need 4, as the M drivers will be in parallel.

It could be done that way. One reason for more amps is simply to squeeze a bit more dynamics out of the system. In my current system the bipolar speakers are driven by a single SE EL84 amp which, because the speakers aren't overly efficient, doesnèt play quite as loud as i'd like. Using 2 amps, one per speaker, would give me a modest 3dB more level. I can also put a different filter on the back driver if i decide that i'd like to use it only for baffle-step compensation.

And if i had just the two amps per speaker, i would think that using one for the back driver and one for the front driver + a cap XOed tweeter (at 10k mind you) would generally be a better application than driving the FRs with one & the tweeter with the other.

You have pointed out some very important things for consideration thou... i think that it can be summarized pretty much as execution is of vital importance.

dave