Impedance of a crossover as seen by an amplifier

[Scotty .V]

Howdy all!

New guy here. (both to this forum and Speaker Building)
I have a question that is keeping me from pulling the trigger on a DIY 3-way speaker design.

I have loved all things audio since I was a little kid and just recently got into system building. However, I am confused at how 2-way or 3-way Speakers from a manufacture have an impedance rating of lets say 6Ohm's. How do they get 6Ohm's? I know it's not exactly 6Ohm's and that the impedance increases or decreases with changes in frequency.

I only ask this because I want to build my own 3-way calculated crossover. How do I make sure that the impedance load seen by my amplifier is within a safe range? Is there a way to measure the impedance load of a crossover? I will also note that I did "successfully" build a 2-way crossover a couple of months ago for a really simple cabinet design. It works okay! But I feel it could have been better.

Anyway, enough rambling. I just want to make sure that I'm not going to blow up any amps anytime soon with a custom crossover. Any help would be greatly appreciated!

Thanks!
-Scotty

P.S. If I have got something wrong while explaining my question, please feel free to correct me. I love this stuff and I want to learn as much as I can.

 

[Tim D]

Howdy all!

New guy here. (both to this forum and Speaker Building)
I have a question that is keeping me from pulling the trigger on a DIY 3-way speaker design.

I have loved all things audio since I was a little kid and just recently got into system building. However, I am confused at how 2-way or 3-way Speakers from a manufacture have an impedance rating of lets say 6Ohm's. How do they get 6Ohm's? I know it's not exactly 6Ohm's and that the impedance increases or decreases with changes in frequency.

I only ask this because I want to build my own 3-way calculated crossover. How do I make sure that the impedance load seen by my amplifier is within a safe range? Is there a way to measure the impedance load of a crossover? I will also note that I did "successfully" build a 2-way crossover a couple of months ago for a really simple cabinet design. It works okay! But I feel it could have been better.

Anyway, enough rambling. I just want to make sure that I'm not going to blow up any amps anytime soon with a custom crossover. Any help would be greatly appreciated!

Thanks!
-Scotty

P.S. If I have got something wrong while explaining my question, please feel free to correct me. I love this stuff and I want to learn as much as I can.

Hi Scotty. Welcome to AK!

You are on the right track. The impedance varies based on frequency. It has to do with the storage nature of the inductors (time varying magnetic fields) and the storage nature of capacitors (time varying electric charge fields). The inductors and capacitors store and release energy at the frequencies of the sound your amplifier is producing. The act of storing and releasing can act to push back against the amplifier (it resists/pushes back against the output of the amplifier) and that result is different for each frequency. In fact the speaker drivers have a voice coil that also has inductance as part of its makeup too. It is a bit of a complex science to figure out the impedance (a bunch of math) of a speaker system (all the speakers in a cabinet with the crossover). Typically the manufacturer will rate the impedance of a speaker system it sells in ohms (an impedance measurement) and that is just a guess based on an approximate average over the frequency range (typically from 20Hz to 20,000Hz or thereabouts). I recently wanted to measure this so I bought a package called "DATS" from Parts Express. It will show you the frequency varying nature of the impedance on you speakers or speaker systems. There are 3 typical values for speaker impedances. Those are 4 ohms, 8 ohms and 16 ohms (You will sometimes find in-between values such as 6 ohms, but those are less common). The 16 ohms are from the bygone era of classic tube amplifiers and are not generally produced in that range anymore. Most amplifiers will work with 16 ohms, but usually at reduced efficiency/power output. The 8 ohms is the most common and as such almost every amplifier will drive these efficiently. The 4 ohms speakers are a bit less common and there are some amplifiers that struggle or have problems driving 4 ohms speakers (sometimes older equipment out of spec will have trouble). In general the lower the impedance - the more power can be sent to the speakers, but this has the effect of making the amplifier work harder and they will typically get hotter (shorter life, higher failure rate, etc). In fact, a lot of solid state amplifiers are setup with A and B speaker connections. In general, the users should avoid paralleling two sets of 4 ohms speakers because the end result is that the amplifier will see half of the impedance and this will very often cause overheating.

I'll let others correct me where I'm wrong or not clear. Let me know if you don't understand any of this and I'll try another way.

 

[Tim D]

Here is a plot of the impedance of a single woofer (from one of my Realistic Nova 7B speakers). It shows a bump in the lower frequency range at about 25Hz. That is where the speaker is naturally tuned to vibrate called the resonant frequency. That frequency is important in speaker design. The point though is that you can see that the impedance depends on the frequency. (Sorry if it is hard to see).

 

[spkrdtr]

Those black background plots are almost worthless when posted in forums. Ever print one out? Lot's of black ink wasted.

 

[ConradH]

There are many different crossover designs, not to mention driver impedance curves, so you have to work each one out separately. Usually, the crossover will just present each driver to the amplifier over the drivers frequency range, not reducing the impedance excessively on its own. We hope. For a quick look, model the crossover in LTSpice, using 6 ohms for the drivers. For a more accurate look, use a driver model instead of the resistor. Naturally there's a learning curve for LTSpice, but there's a good Yahoo forum and some threads over at DIYaudio. For the final word, you have to build it and measure the impedance, not terribly difficult if you have a signal generator and meter with good AC response.

 

[whoaru99]

Howdy all!

New guy here. (both to this forum and Speaker Building)
I have a question that is keeping me from pulling the trigger on a DIY 3-way speaker design.

I have loved all things audio since I was a little kid and just recently got into system building. However, I am confused at how 2-way or 3-way Speakers from a manufacture have an impedance rating of lets say 6Ohm's. How do they get 6Ohm's? I know it's not exactly 6Ohm's and that the impedance increases or decreases with changes in frequency.

It's a nominal rating, typically taken at or around the lowest point after the low frequency resonance peak. However, some play loose with this number.

There is an IEC method for rating impedance but some (many?) mfg apparently don't use it. I would not assume an impedance rating is by IEC method unless specifically listed that way, or by consultation with the mfg.

 

[bowtie427ss]

One source defines nominal impedance as the approximate designed impedance of a circuit or device. I kind of like this definition.

Another defines it as the average impedance across the speaker system's usable range.

If you search around you can find a dozen more definitions, all different.

When it comes right down to it, for the sake of accuracy the term needs to be defined by the party/s claiming the rating as there is not an agreed industry standard for the specification.

 

[elnaldo]

I measure impedance with a computer and "LIMP " software, part of the ARTA software suite.

 

[spkrdtr]

If the design the OP wants is a kit or fully explained in this or another forum or web site, the OP should contact the designer about the Z vs Freq. curve 10-20kHz if he can't or doesn't want to use a Dats or WT2.

 

[leesonic]

If you select an amplifier that is known to be good at driving low impedance loads, you won't have to worry about it.

Lee.

 

[whoaru99]

One source defines nominal impedance as the approximate designed impedance of a circuit or device. I kind of like this definition.

Another defines it as the average impedance across the speaker system's usable range.

If you search around you can find a dozen more definitions, all different.

When it comes right down to it, for the sake of accuracy the term needs to be defined by the party/s claiming the rating as there is not an agreed industry standard for the specification.

In this context nominal is basically the equivalent of named, or in name only. An 8 ohm speaker is 8 ohms in name only, 8 ohms nominal.

And, there is a standard available. Making all the horses drink the water is another matter.

IEC IEC60268-5. Basically, if rated by this method, the actual impedance of the speaker at any frequency (including at DC) cannot be less than 80% of the rated value. So, if by IEC rating a speaker was called an "8 ohm" speaker, no place on the curve could the impedance be less than 6.4 ohms and still be called an "8 ohm" speaker.

Those "amp killer" Kappa 9s or whatever they are with the 0.9 ohms impedance? If rated by IEC method they would have to be listed at (rounded) 1 ohm speakers.

 

[toxcrusadr]

"Is there a way to measure the impedance load of a crossover?"

Hopefully it has become clear that the impedance of a loudspeaker is as much or more about the DRIVERS as it is the crossover. If it's a fairly simple conventional crossover and the drivers are being operated within their linear range, it's LIKELY the impedance of the entire system will be within a safe range. But the only way to know exactly what it looks like is to model the crossover using the impedance curves for the specific drivers in question (as already stated earlier in the thread).

If you're into DIY which you clearly are, a driver testing gadget and software + speaker modeling software are in order. The latter can be had as freeware I think, although I haven't done it myself.