Speaker impendance and bi amping

[IvoCRO]

Hi all,

I would like to hear your opinion about speaker impedance and connection to an power amplifier with different impedance taps like many McIntosh amps has. If we have 4 ohm speaker it should be connected to amplifiers 4 ohm tap, that's clear. This speaker also has bi-wire option and therefore bi-amping option. Impedance of this speaker is 4 ohm when lo and hi inputs are connected in parallel, using formula 1/Z(total)=1/Z(hi input) + 1/Z(lo input).

If speaker is connected in bi-amping, then each amplifier has load of only Z(hi input) or Z(lo input) which should be different (and higher) than total impedance.

So which tap should be used in bi-amping case?

On many amps there is warning for using both A + B speakers saying that minimum impedance of each speaker should be 8 ohm (which gives 4 ohm when using in parallel A + B mode).

Ideally, 4 ohm is achieved when you have two 8 ohm resistors in parallel. But there is more than only resistance in speaker and crossover.

According to that logic, for bi-amping, should 8 ohm tap be used when only one section (lo or hi) of 4 ohm speaker is connected, or stay with 4 ohm tap?

I would say to stay with 4 ohm tap, but I would like to hear what you think.

Ivo

 

[c_dk]

There are so many different paths to multiple amping a speaker system, it is important to specify which path you are exploring.

- The professional audio path, often with multi driver arrays with horns "aimed" at various locations in the vast seating area, possible time delays added, special equalization for each driver array, external crossover to facilitate the arrays, used in almost all pro audio systems since the beginning of electronic audio.

- Amateur budget build up a audiophile system path. A amp is purchased smaller than what the user really wants but is all that can be afforded at that time. A second amp is added at a later date and at that time the user has a choice. One branch would be to switch the amps to mono therefore most likely doubling the power so the system might sound better and be able to play louder. Another branch might be that instead of purchasing say, another 50 watt amp the owner is a bit more flush now and purchases a 100 watt amp. Uses the 100 watt amp for the bass,the 50 for the mid and treble and might find it sounds better at normal listening levels than the brute force path.

There are derivitives of these paths......satelite sub systems as used in home theater follow much of the first path, for example.

Those who have purchased pride and joy amplifers......like a new MC2505 as a 18 year old, might want to find a way to keep using that prized unit.......been there, done that.

And then there was the B&W matrix 800 series speakers as well as some other speakers designed by masochist speaker designers who had no regard for the amps the poor user might have to buy to use their gem of a speaker.

Which path would you like to explore?

 

[62caddy]

Hi all,

I would like to hear your opinion about speaker impedance and connection to an power amplifier with different impedance taps like many McIntosh amps has. If we have 4 ohm speaker it should be connected to amplifiers 4 ohm tap, that's clear. This speaker also has bi-wire option and therefore bi-amping option. Impedance of this speaker is 4 ohm when lo and hi inputs are connected in parallel, using formula 1/Z(total)=1/Z(hi input) + 1/Z(lo input).

If speaker is connected in bi-amping, then each amplifier has load of only Z(hi input) or Z(lo input) which should be different (and higher) than total impedance.

So which tap should be used in bi-amping case?

On many amps there is warning for using both A + B speakers saying that minimum impedance of each speaker should be 8 ohm (which gives 4 ohm when using in parallel A + B mode).

Ideally, 4 ohm is achieved when you have two 8 ohm resistors in parallel. But there is more than only resistance in speaker and crossover.

According to that logic, for bi-amping, should 8 ohm tap be used when only one section (lo or hi) of 4 ohm speaker is connected, or stay with 4 ohm tap?

I would say to stay with 4 ohm tap, but I would like to hear what you think.

Ivo

The same impedance applies to each speaker input individually as when combined and the 4 ohm taps should be used.

 

[c_dk]

The same impedance applies to each speaker input individually as when combined and the 4 ohm taps should be used.

Not nessesarily.......in pro applications it would not be uncommon to find different impedences from different driver assemblies.......if you observe the impedence characteristics of certain home loudspeakers they too can have wild impedence differences throughout the frequency ranges which are created by the drivers handling that range.

The budget audiophile path opened up home users to the more sofisticated path that the pros were already working with.

 

[IvoCRO]

Which path would you like to explore?

path would definitely be amateur one, staying with stereo only.

Amps are MC2200 and MC2205. So, two exactly the same amps. MC2205 would drive left and right low section and MC2200 would drive left and right high section of speaker.

Of course, I could bridge each amp in mono and drive each channel with it's own amp. But let leave bridge option aside for now.

 

[62caddy]

path would definitely be amateur one, staying with stereo only.

Amps are MC2200 and MC2205. So, two exactly the same amps. MC2205 would drive left and right low section and MC2200 would drive left and right high section of speaker.

Of course, I could bridge each amp in mono and drive each channel with it's own amp. But let leave bridge option aside for now.

You would have twice the effective output capability with the latter.

 

[IvoCRO]

You would have twice the effective output capability with the latter.

But that would take me to the dark side of bridging and never even thinking of going back. I have read many stories about that. Side effect is that meters would be only on one channel. And you know owner feelings about meters.......

 

[c_dk]

This might be complicated.....

It was explained to me by Ken Kessler over a adult lunch during a break from CES, that a British speaker manufacturer " discovered" biwiring speakers as a positive upgrade to crossover design.

So designing a speaker crossover that brought out the connections for the bandwidth limiting circuitry to the outside of the speaker added possible sound improvements and then opened the door to dedicating amplifiers to each seperate speaker as a group of sub system speakers.

Is it suprising then that audiophiles on tight budgets might have discovered better sound by employing one amp to drive say a 4 ohm mid tweeter subsystem with a bigger amp driving the 8 ohm woofer subsystem.

Twii posts often about his woofer sections not liking autoformer amps, I do not recall if his horn drivers might like a transformered coupling.

I do believe there is a speaker company that routinely had a woofer section that had a extremely low impedence, I know B&W had a less than three ohm midrange load.

In these cases a experimenting user might discover that driving a woofer section off of the 2 ohm taps and the midrange off the 8 ohm taps might be beneficial. Or vica versa.

These were experiments that we started doing back in the late 80s, early 90s and as other companies being the copy cats that they all are started including these crossover design changes and I think now all manufacturers do it.

It always intersted me that Jim Thiel never subscribed to these concepts even though his end users were usually well tuned in to such "tweeks" and his particular designs did often impose some challenging impedence loading.

 

[62caddy]

I'm not sold on the benefits of passive bi/tri amplification either.

 

[damacman]

It is interesting that I talked to a British loudspeaker designer, Joe Ackroyd (Goodmans, Wharfedale, and Royd Audio designer) when I called his speaker factory to ask some questions about his speakers (Royd Audio at the time). He said he put in bi-wiring capability because customers were asking for it. He did not feel bi-wiring did anything to change the sound of speakers, so he always recommended not to bi-wire. He also said he tested it and found no differences.

As did Roger Russell, then chief loudspeaker engineer of McIntosh - an outspoken critic of both biwiring and biamping, even today.

However c_dk's points are all valid ones which can be verified by anyone who chooses.

 

[c_dk]

Ken was referencing I believe a situation were Celestion found a sound difference with their breadboarded experimental crossover outside of the cabinet and the completed crossover mounted inside the cabinet.

He joked that they did the "American" thing of testing all the wrong parameters before finally finding the right thing that was isolating the speaker wire runs from amp to crossover/drivers.

We were experimenting with any number of B&W and Vandersteen speakers with different amps and different wires run.

The most demonstrative was basic zip cord,16 gauge doubled up with a 20 watt Arcam amp and small B&Ws.

 

[Mike Gibson]

I don't have an audiophile's ear but when I went to bi-amping my speakers I noticed 2 main differences. One was the speakers sounded fuller and had better bass (it seemed to me). The other was the convenience of switching crossover frequencies at will and on the fly. Something I had wanted to do with my Altecs for quite a while. After I swicthed to a bi-amped, electronically crossed setup I have never gone back. That's not to say bi-amping is the answer to audio nirvana for everyone though. YMMV.

 

[IvoCRO]

I would say that benefits of bi-amping and bi-wiring depends on all components in chain - amp, speaker cable and speaker itself. Like you sad, there could be improvement, but not always. If you go for bi-amping and remove crossover from speaker, using active crossover, advantage should be noticed too. But this requires some measuring equipment to tune crossover cuts and gains, and speaker itself when it is driven without crossover.

 

[SaturationPt]

I think that bi-wiring is similar to people's opinions on boutique speaker cables, where there certainly can be a difference measured (regardless of how minute) but is it audible? Some say certainly and others claim "snake oil" so it's not something you'll get conclusive opinion-based direction on a discussion forum.

Bi-amping is much more complicated because it's not a simple matter of removing jumpers and adding a second set of wires. In some cases (probably most) the jumpers just connect two halves of the crossover, and you can run full-range into both because each set of terminals still goes through its half of the crossover (high-pass / low-pass), so although there is the benefit of having separate amps not receiving artifacts from the other drivers (for example the HF section not getting potential rail dips from a hard-low note) and you can run tubes on top, SS on the bottom if that's what you feel will sound best. However, you're still using the speaker manufacturer's crossover and the amplifiers will see some open-circuit potenitally (for example the high-pass will pass the HF to the HF section of the speaker, but output of the connected amplifier could see frequencies significantly below the HF cutoff as an open-circuit). Many amplifiers won't care (I did this for my first time with a set of ADCOM amplifiers that didn't care), other amplifiers could have problems with it.

Bi-amping into speakers with a more complex crossover where you hook directly to the speaker driver via a set of jumpers for each driver (example might be the JBL 250Ti, search images of the terminal strip on the back) is more complicated and REQUIRES a proper external crossover to tie directly to the drivers or you can damage the speaker. A simple electronic crossover such as used in pro-gear will have switches to set crossover frequencies, and this will suffice for the protection of the speakers. A "low-level" unit will make the crossover before the amplifiers so that the amplifiers don't "see" the frequencies it's not required to drive into the speakers so it too is protected. However if you were to look into the design of crossovers, the better ones are not simply high-pass, band-pass, low-pass filters but complicated circuits with "shaping" involved where the speaker designer (example again being the JBL 250Ti designed by Greg Timbers) worked some complicated magic in the crossover creating peaks and valleys in the signal to the drivers to offset imperfections in the drivers' response and the resonance of the driver in the designed cabinet. This little bit of magic cannot be reproduced by a simple high/band/low-pass external filter, it requires a proper active crossover (or DSP) with a calibrated microphone and testing in the room to tune in/out these same peaks and valleys.

So experimenting with bi-amping is a lot of fun (IME), and can produce some excellent results that the stock crossover cannot (if you can go direct) and I feel that there is a real benefit if set up properly. Without the intrumentation and higher-level crossover options it is still fun and can provide some benefits, provided you understand your system components' limitations and do a little research first on the speakers' internal connections to the crossover/terminals/drivers. Have fun and don't take too seriously the opinions on whether you should or shouldn't, this is a hobby as well as a science so enjoy both aspects.

 

[TSmith8605]

Hi all,

I would like to hear your opinion about speaker impedance and connection to an power amplifier with different impedance taps like many McIntosh amps has. If we have 4 ohm speaker it should be connected to amplifiers 4 ohm tap, that's clear. This speaker also has bi-wire option and therefore bi-amping option. Impedance of this speaker is 4 ohm when lo and hi inputs are connected in parallel, using formula 1/Z(total)=1/Z(hi input) + 1/Z(lo input).

If speaker is connected in bi-amping, then each amplifier has load of only Z(hi input) or Z(lo input) which should be different (and higher) than total impedance.

So which tap should be used in bi-amping case?

On many amps there is warning for using both A + B speakers saying that minimum impedance of each speaker should be 8 ohm (which gives 4 ohm when using in parallel A + B mode).

Ideally, 4 ohm is achieved when you have two 8 ohm resistors in parallel. But there is more than only resistance in speaker and crossover.

According to that logic, for bi-amping, should 8 ohm tap be used when only one section (lo or hi) of 4 ohm speaker is connected, or stay with 4 ohm tap?

I would say to stay with 4 ohm tap, but I would like to hear what you think.

Ivo

You are thinking of speaker impedance as if it were simple resistance: one number for which Ohm's law applies.

But speaker impedance is more complicated than that in that it varies with frequency and also is characterized by two numbers (magnitude and phase) at any given frequency. Magnitude is similar to resistance and phase refers to the time delay between the current flow and voltage caused by the reactive elements in the speaker's crossover and the speaker itself. Both vary with frequency. The reactive elements store and release energy and so can shift the timing of the current relative to the voltage.

I am a very big fan of John Atkinson at Stereophile Magazine (I have no affiliation) who does a great job of measuring speaker performance in conjunction with speaker reviews. To illustrate my point, here is a speaker impedance measurement picked at random from such a review: https://www.stereophile.com/content/jbl-stage-a170-loudspeaker-measurements

Note that the magnitude of the impedance, solid line varies pretty dramatically with frequency as does the phase.



Just for fun, here's another impedance measurement from a completely different type of speaker:
https://www.stereophile.com/content/magnepan-lrs-loudspeaker-measurements



In the case of bi-amping passively using the bi-wire inputs on your speaker, you are dividing these curves into two pieces (which don't show up in the above measurements), one for the low-pass (woofers) and one for the high-pass (tweeters). For example, if you look at the impedance of the tweeter connection, you will see the effect of the crossover network's job to keep bass energy out of the tweeter. That means the crossover will raise the (magnitude of the) impedance at that input towards infinity (like an open circuit) at frequencies below the crossover frequency so that little lower-frequency energy flows to the tweeter. Then using your (magnitude of) Z equation in the woofer frequency range ( 1/Z(total)=1/Z(hi input) + 1/Z(lo input) ) where Zlo is getting to be infinity at lower frequencies and 1/Zlo then becomes zero and so 1/Z(total)=1/Z(hi input) + 0 or 1/Z(total)=1/Z(hi input) measured at the tweeter input. The same idea applies at in woofer input operating in the tweeter frequency range.

I hope this make sense and I welcome corrections to anything wrong here.

What speakers do you have? If there is a Stereophile review, maybe we can find measurements. Regardless you should confirm from the operator's manual and / or the manufacturer that they are designed for such passive bi-amp operation.

With this bit of background, I would refer you back to post #16 above.

You asked for an opinion but my opinion would dependent on what speakers you have.

 

[damacman]

Well that could at least explain why Magnepan has such a cult following!

 

[TSmith8605]

Well that could at least explain why Magnepan has such a cult following!

I just ordered a pair of Magneplaner 1.7i speakers. Hope I like them.

 

[62caddy]

I just ordered a pair of Magneplaner 1.7i speakers. Hope I like them.

I was thoroughly impressed with the .7 that heard up at Audio Classics a while back so the 1.7 should be even better. Only small complaint was the LF was a bit lacking. Wall/corner placement would help in this department but unfortunately isn't an option due to acoustic radiation from the rear panels of this design. For this reason a subwoofer might be advisable.

 

[c_dk]

Too close to the back wall intensifies the remodulation of the Mylar sheet from the rear wall reflection....the bass is what it is.

At the very least with all the different amp/speaker interactions, even not considering the truly way out there troublesome designs, I believe it should be apparent that many experimenters have found improvement without a external crossover.

 

[TSmith8605]

I was thoroughly impressed with the .7 that heard up at Audio Classics a while back so the 1.7 should be even better. Only small complaint was the LF was a bit lacking. Wall/corner placement would help in this department but unfortunately isn't an option due to acoustic radiation from the rear panels of this design. For this reason a subwoofer might be advisable.

I am aware of these issues. If I like the speakers, I may think about subs. If not, there's Barter Town.

Too close to the back wall intensifies the remodulation of the Mylar sheet from the rear wall reflection....the bass is what it is.

At the very least with all the different amp/speaker interactions, even not considering the truly way out there troublesome designs, I believe it should be apparent that many experimenters have found improvement without a external crossover.

I have been using flat electrostatics for 30 years and Magneplaner Tympani 1D before that so I am aware of these issues. The area behind my 'stats is very heavily damped with sound absorbers since they are just over two feet from the walls. The new 'planers will go in a different room where they can be well away from the wall. We'll see how that works. At some point, I will experiment with the new 'planers in place of the 'stats and see what happens. Intend to have some fun.

But I did not want to hijack OP's thread. Will say that I passively bi-amp the 'stats and their satellite woofer and the new 'planers (I think) have no provision for bi-amping. 40 years ago, I bi-amped the older T 1D with a custom, passive crossover between the preamp and amps.