Difficult to drive speakers - AVS Forum
Forum Jump: 
 
Thread Tools
Old 12-19-2013, 02:34 AM - Thread Starter
 
Heinrich S's Avatar
 
Join Date: Aug 2007
Posts: 974
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 30
Why do some manufactures design speakers that have impedances that dip down to 1 ohm, in some cases? I've seen floor standing speakers that have impedance dips down very low.

Is this a design fault? Was it intentional? I don't understand why this is so, when there are also many well engineered speakers out there that do not have these abnormal impedance dips.

One good example is the Infinity Kappa range. Notorious for being extremely difficult to drive. Impedance dips down below 2 ohms. There are other exotic speakers out there that have abnormal impedance dips.

To the speaker designers, what are your thoughts?
Heinrich S is offline  
Sponsored Links
Advertisement
 
Old 12-19-2013, 04:55 AM - Thread Starter
 
Heinrich S's Avatar
 
Join Date: Aug 2007
Posts: 974
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 30
Just a sidequest? Dali Megaline, Genesis etc speakers with 20+ drivers per side? How on earth do you keep the impedance withing limits? Serial/Parallel wiring? Obvious they will need current and lots of it, but I gather quite difficult to drive ?
Heinrich S is offline  
Old 12-19-2013, 05:34 AM
 
Shaun B's Avatar
 
Join Date: Oct 2013
Posts: 117
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 14

Just my 2c.  The only real reason a speaker's impedance would dip below a given number (let's use 1 ohm for this argument) is because the designer chose for some reason not to resolve it. Doesn't mean the speaker is bad, it just means it was not done.

Shaun B is offline  
Old 12-19-2013, 05:40 AM
AVS Addicted Member
 
arnyk's Avatar
 
Join Date: Oct 2002
Location: Grosse Pointe Woods, MI
Posts: 14,387
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 763 Post(s)
Liked: 1178
Quote:
Originally Posted by Heinrich S View Post

Why do some manufactures design speakers that have impedances that dip down to 1 ohm, in some cases? I've seen floor standing speakers that have impedance dips down very low.

Back in the 1980s there were a few speakers that were stress tests of the amplifiers of the day. As I have documented in other posts, SS amp output devices are now 3-5 times more capable of delivering peak current even into reactive loads, then the first few generations of SS amps. And that is just the spec sheets, Reserves that used to be shaved are now overkilled.. Enough money has been made or lost on this issue that few are willing to build amplifiers that can't live with speakers or speakers that can't live with amplifiers.

There are now very few speakers that dip down to 1 or 2 ohms, but quite a few that dip down to 3.5 to 4.5 ohms. The only speaker I am aware of that dips down to 1 or 2 ohms does so at the extreme far end of the audio band where most recordings don't have much energy. I posted a detailed analysis of this a few months back.
Quote:
Is this a design fault?

Yes. As rule not many people actually did it, it wasn't done in mainstream speakers very often, and just about everybody has learned to stop doing it in order to optimize their market position.
Quote:
Was it intentional?

I can't speak in general, but I have had an over-the-web fairly close relationship with a consultant who designed a speaker like this for Infinity. He knew what he was doing but it was felt that people would pay for beefy amps to drive the speakers. This is one reason why amplifiers like the Pass Threshold SA4e came into existence. It was hard to short them with something that wouldn't disappear if you drove the amp hard. Basically, an audio spot-welder. ;-)
Quote:
I don't understand why this is so, when there are also many well engineered speakers out there that do not have these abnormal impedance dips.

It is now very rare.
Quote:
One good example is the Infinity Kappa range. Notorious for being extremely difficult to drive. Impedance dips down below 2 ohms. There are other exotic speakers out there that have abnormal impedance dips.

The Kappas have been off the market how many decades?
Quote:
To the speaker designers, what are your thoughts?

Designing speakers that eat amplifiers or amplifiers that are easily damaged by speakers is a form of economic suicide.

One example of a power amp that was designed rather suicidally was the original Ampzilla. It had half as many nearly identical output devices as the original Dyna 400, and later on the Dyna 400 was upgraded to have twice as many output devices and sold as the Dyna 416. So the Dyna 416 had 4 times as many output devices of a similar type in a similar series-connected configuration.


Ampzilla 2 had 3 times as many output devices as the original Ampzilla. I can't tell you how the output devices in Ampzilla 2 compared to those in Ampzilla because the Ampzilla 2 used house numbers on those devices but they were later and therefore probably beefier devices.

Also, we don't know what was actually screwed into these legacy amps as built throughout the production runs, as we pretty much only have schematics for them. But they were probably initially built with the devices on the blue print. Same story explains why the later Phase Linear Amps were less likely to live up to the nickname "Flame Linear", not that some of them didn't blow., too.

See the pattern? Speakers got easier, amps got stronger, reliability ensued. We now hear about amps that shut themselves down, but not so many funeral pyres.
arnyk is offline  
Old 12-19-2013, 05:48 AM
AVS Addicted Member
 
arnyk's Avatar
 
Join Date: Oct 2002
Location: Grosse Pointe Woods, MI
Posts: 14,387
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 763 Post(s)
Liked: 1178
Quote:
Originally Posted by Heinrich S View Post

Just a sidequest? Dali Megaline, Genesis etc speakers with 20+ drivers per side? How on earth do you keep the impedance withing limits? Serial/Parallel wiring?

Yes, series/parallel voice coils. Take 4 drivers in series parallel and you get about the same impedance curve as just one of them. Basic electronics.
Quote:
Obvious they will need current and lots of it,

Wrong. A speaker with a certain impedance and a certain efficiency is the same load and gets as loud with the same amount of power no matter how it is implementated. 4 drivers, 1 driver its pretty much the same. Actually 4 drivers might be a tad easier because the drivers work together acoustically and in some physical configurations become more efficient.

Intuition says you need a large amp for a large speaker, but the laws of physics say that all other things being equal the large speaker is more efficient and takes less power to make the same amount of sound. As always, the laws of physics win!

However, on AVS I often see the laws of intuition trump the laws of physics and then people unknowingly but defensively end up with big boat anchor amps that barely turn over in actual use. Not many of those HiFi boat anchors have reliable power output/clipping indicators, or nobody watches them and reaches the correct conclusion which is that they over-configured their system. Well, its a hobby and bragging rights are often part of the experience.
Quote:
but I gather quite difficult to drive ?

I don't know. Show me some efficiency numbers and an impedance curve!
arnyk is offline  
Old 12-19-2013, 06:56 AM - Thread Starter
 
Heinrich S's Avatar
 
Join Date: Aug 2007
Posts: 974
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 30
Quote:
Originally Posted by Shaun B View Post

Just my 2c.  The only real reason a speaker's impedance would dip below a given number (let's use 1 ohm for this argument) is because the designer chose for some reason not to resolve it. Doesn't mean the speaker is bad, it just means it was not done.

If a designer could fix it but chose not to then I don't know, that tells me that perhaps that designer was incompetent.
Heinrich S is offline  
Old 12-19-2013, 07:26 AM - Thread Starter
 
Heinrich S's Avatar
 
Join Date: Aug 2007
Posts: 974
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 30
Quote:
Originally Posted by arnyk View Post

Yes, series/parallel voice coils. Take 4 drivers in series parallel and you get about the same impedance curve as just one of them. Basic electronics.
Wrong. A speaker with a certain impedance and a certain efficiency is the same load and gets as loud with the same amount of power no matter how it is implementated. 4 drivers, 1 driver its pretty much the same. Actually 4 drivers might be a tad easier because the drivers work together acoustically and in some physical configurations become more efficient.

Intuition says you need a large amp for a large speaker, but the laws of physics say that all other things being equal the large speaker is more efficient and takes less power to make the same amount of sound. As always, the laws of physics win!

The Tannoy Kingdom 18 and Infinity Kappa 9 would contradict that. The Kingdom has a sensitivity of 96 dB 1 watt, if memory serves. My good friend has a pair of these speakers. Massive in size. However he has gone through several amps, many of them just shut down because they couldn't cope.

He now uses a Krell 600 watt amp which doubles down as impedances halves. No problems since. Even the very high sensitivity wasn't helping. Otherwise only 1-2 watts would be needed. In his case that most certainly wasn't the case. In fact, I was there when his amps shut down.

He tested the flagship Onkyo NR5010. Shut down. I saw it shut down with my own eyes. He put on a 200 watt Rotel power amp. Shut down.

He then used a Krell FPB600, the music kept on going without a hitch. No problems, no shutdowns.
Heinrich S is offline  
Old 12-19-2013, 07:27 AM
AVS Addicted Member
 
arnyk's Avatar
 
Join Date: Oct 2002
Location: Grosse Pointe Woods, MI
Posts: 14,387
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 763 Post(s)
Liked: 1178
Quote:
Originally Posted by Heinrich S View Post


If a designer could fix it but chose not to then I don't know, that tells me that perhaps that designer was incompetent.

Depends on what the boss says... ;-)

Different designers work under different constraints. Smaller speaker houses rely on stock drivers. Larger houses custom design their drivers. That covers impedance variations due to the drivers. Impedance is also due to the crossovers and thatis more under the control of the designer, but has to be optimized for the drivers.
arnyk is offline  
Old 12-19-2013, 07:36 AM
AVS Addicted Member
 
arnyk's Avatar
 
Join Date: Oct 2002
Location: Grosse Pointe Woods, MI
Posts: 14,387
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 763 Post(s)
Liked: 1178
Quote:
Originally Posted by Heinrich S View Post

Quote:
Originally Posted by arnyk View Post

Yes, series/parallel voice coils. Take 4 drivers in series parallel and you get about the same impedance curve as just one of them. Basic electronics.
Wrong. A speaker with a certain impedance and a certain efficiency is the same load and gets as loud with the same amount of power no matter how it is implementated. 4 drivers, 1 driver its pretty much the same. Actually 4 drivers might be a tad easier because the drivers work together acoustically and in some physical configurations become more efficient.

Intuition says you need a large amp for a large speaker, but the laws of physics say that all other things being equal the large speaker is more efficient and takes less power to make the same amount of sound. As always, the laws of physics win!

The Tannoy Kingdom 18 and Infinity Kappa 9 would contradict that. The Kingdom has a sensitivity of 96 dB 1 watt, if memory serves. My good friend has a pair of these speakers. Massive in size. However he has gone through several amps, many of them just shut down because they couldn't cope.

He now uses a Krell 600 watt amp which doubles down as impedances halves. No problems since.

You've got me confused with someone who tries to speak in perfect generalities. Of course there are some exceptions given the thousands of speaker systems out there.

When I say "...all other things being equal..." I mean just that. Obviously with these two speakers all other things weren't exactly equal. ;-) In the case of the Kappa 9s I think they were the ones with a dual voice coil woofer where the two voice coils did not have the usual same impedance. This was done to obtain a certain bass tuning. Very unusual birds.
arnyk is offline  
Old 12-20-2013, 03:06 AM - Thread Starter
 
Heinrich S's Avatar
 
Join Date: Aug 2007
Posts: 974
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 30
Quote:
Originally Posted by arnyk View Post

Depends on what the boss says... ;-)

Different designers work under different constraints. Smaller speaker houses rely on stock drivers. Larger houses custom design their drivers. That covers impedance variations due to the drivers. Impedance is also due to the crossovers and thatis more under the control of the designer, but has to be optimized for the drivers.
Quote:
Originally Posted by arnyk View Post

You've got me confused with someone who tries to speak in perfect generalities. Of course there are some exceptions given the thousands of speaker systems out there.

When I say "...all other things being equal..." I mean just that. Obviously with these two speakers all other things weren't exactly equal. ;-) In the case of the Kappa 9s I think they were the ones with a dual voice coil woofer where the two voice coils did not have the usual same impedance. This was done to obtain a certain bass tuning. Very unusual birds.

But if you look at the Infinity Primus range, with their very smooth on-axis and off-axis, you don't see 1 ohm impedance dips. So it's more than possible to design and build a speaker that has a flat response without resorting to abnormal impedances.
Heinrich S is offline  
Old 12-20-2013, 05:23 AM
AVS Addicted Member
 
arnyk's Avatar
 
Join Date: Oct 2002
Location: Grosse Pointe Woods, MI
Posts: 14,387
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 763 Post(s)
Liked: 1178
Quote:
Originally Posted by Heinrich S View Post


But if you look at the Infinity Primus range, with their very smooth on-axis and off-axis, you don't see 1 ohm impedance dips. So it's more than possible to design and build a speaker that has a flat response without resorting to abnormal impedances.

Totally agreed. It can be done. The counterpoint is that the flagship Primus P363 does dip below 4 ohms in the mid-bass and upper treble. The people who call it a 4 ohm speaker will get no arguments from me. That said, I know for sure that driving it does not give mid-grade and low end amps and AVRs any sort of tummy aches.



The Harman speaker design teams have the dual advantages of a very experienced driver design team and their own production facilities.
arnyk is offline  
Old 12-25-2013, 08:02 AM - Thread Starter
 
Heinrich S's Avatar
 
Join Date: Aug 2007
Posts: 974
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 30
If there is, and has been for a long time, affordable modelling software that can accurately predict combined driver/crossover impedance, why should we put up with poorly designed loudspeakers?
Heinrich S is offline  
Old 12-25-2013, 11:44 AM
AVS Addicted Member
 
arnyk's Avatar
 
Join Date: Oct 2002
Location: Grosse Pointe Woods, MI
Posts: 14,387
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 763 Post(s)
Liked: 1178
Quote:
Originally Posted by Heinrich S View Post

If there is, and has been for a long time, affordable modelling software that can accurately predict combined driver/crossover impedance, why should we put up with poorly designed loudspeakers?

The question you ask probably relates back to when we had the ability to measure impedance curves quickly and accurately, which would be no later than the 1980s for people who were in the business.

The answer is that impedance curve variations can be traded off for other things including frequency response and efficiency. That's one of the big things in engineering - trading one thing off to improve another. After that it is all about priorities and judgments.

Right now I don't see that many speakers with difficult impedance curves. Back before the 1980s, we did see some real amp-busters and amps that were easy to bust.
arnyk is offline  
Old 12-25-2013, 12:52 PM - Thread Starter
 
Heinrich S's Avatar
 
Join Date: Aug 2007
Posts: 974
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 30
Please correct me if I'm wrong but the impedance of the speaker varies quite considerably with enclosure size?
Heinrich S is offline  
Old 12-25-2013, 01:14 PM
AVS Special Member
 
A9X-308's Avatar
 
Join Date: Mar 2008
Location: Australia; now run by adults.
Posts: 5,458
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 175 Post(s)
Liked: 115
Quote:
Originally Posted by Heinrich S View Post

Please correct me if I'm wrong but the impedance of the speaker varies quite considerably with enclosure size?
For a given driver, it's nominal impedance won't vary with enclosure size, but there will be differences in it's LF impedance depending upon the tuning of the enclosure. Add a xover and it could change quite a lot.
A9X-308 is offline  
Old 12-25-2013, 05:35 PM
AVS Addicted Member
 
arnyk's Avatar
 
Join Date: Oct 2002
Location: Grosse Pointe Woods, MI
Posts: 14,387
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 763 Post(s)
Liked: 1178
Quote:
Originally Posted by Heinrich S View Post

Please correct me if I'm wrong but the impedance of the speaker varies quite considerably with enclosure size?

If the speaker driver has an open back (woofers have open back, midranges and tweeters generally are closed back) and we are around the range where the driver resonates, then there are major variations in its impedance curve. Both box size and box design (sealed, vented, or horn) can cause big variations. Even the stuffing can make a difference.
arnyk is offline  
Old 12-25-2013, 08:28 PM
AVS Special Member
 
A9X-308's Avatar
 
Join Date: Mar 2008
Location: Australia; now run by adults.
Posts: 5,458
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 175 Post(s)
Liked: 115
Quote:
Originally Posted by A9X-308 View Post

For a given driver, it's nominal impedance won't vary with enclosure size, but there will be differences in it's LF impedance depending upon the tuning of the enclosure. Add a xover and it could change quite a lot.

To expand on this, I'll use one of my favourite drivers, the JBL 2225 (2226 is near identical).

Here are the basic T/S.

Fs 40.00 Hz
Re 6.30 Ohm
Qms 2.50
Qes 0.31
Sd 890.0 cm2
Vas 170.0 l
Xmax peak 5.00 mm
Le 1.10 mH
Le2 0.00 mH
Re2 0.00 Ohm
Nominal Power 200.0 W

Note the Re is 6.3Ω which makes it a nominal 8Ω driver. With no xover the minimum impedance of the driver will be it's Re but around the tuning of the driver/enclosure, whether OB, ported or sealed, there will be at least one resonant peak at the bottom of the passband.

Here is the impedance curve for an open baffle.



Here is the same driver in a 60L sealed enclosure. Note the enclosure tuning has risen (peak is at higher frequency) due to the air spring of the box.



Same driver and enclosure, this time tuned to 40Hz. 40Hz is the lowest impedance between the two peaks.



This driver is a 15"midbass driver, but a midrange would show a similar pattern, as would a subwoofer, just the actual impedance numbers would vary as would the frequency range.

If you now add a xover, depending upon the driver above it, and the choices made by the designer in how the xover was implemented, the resultant impedance curve could vary wildly from this (see Arny's Infinity example above).
FWIW, this is a very easy driver to drive: high efficiency and moderate impedance.
A9X-308 is offline  
 
Thread Tools


Forum Jump: 

Posting Rules  
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off