Yes it does. And that's why basically every audio company quotes a 2.83V reference for their speaker products. It takes literally seconds with a voltmeter to check the output voltage from an amp to confirm that it's 2.83V. And amplifier gain is adjusted by Volts, NOT Watts. I don't have to sit down and work out what voltage I have to set the amplifier to in order to meet your "1 Watt" reference, which apparently has now been changed, by your subsequent note below, LOL. So which should I use now - Re, or measured minimum impedance in the passband?
Amp gain is set with a volume knob, not by volts or watts. Measuring the voltage is simple but the better way if you want to compare equivalent power is to adjust voltage based on minimum impedance in the passband.
Re is almost always very close to the minimum impedance in the passband, so as a shortcut you can just adjust voltage in reference to Re.
So what if you have to measure impedance? If you want to make an equivalent power comparison you have to do what is required to do that.
This is more sensible as a reference because it's much closer to equal power input.
How does it make any sense to do an equal power comparison of two very different speakers at 2.83V?
If I have a speaker that has 1.4 ohms Re, 1 watt referenced to Re would be 1.1V.
If I have a speaker that has 16 ohms Re, 1 watt referenced to Re would be 4V.
MEASURING THEM BOTH AT 2.83V IS NOT EQUAL POWER, NOT EVEN CLOSE. It doesn't tell you much of anything at all about the speaker to measure everything at 2.83V.
Last edited:
I've already shown you that "equal power" is a myth. It is a made-up parameter. Made up by you. The impedance of any subwoofer does not remain constant through its operating frequency range and your "equal power" assumption is based on a fixed impedance, which does not reflect any subwoofer alignment that I know about. I haven't even mentioned yet how your "minimum impedance" adjustment (which you had to do "on the fly" in this discussion, seeing that it was readily apparent that your previous suggestion to base it on a driver's Re made no sense) also makes no sense because how would you account for sealed subwoofer systems which usually show their minimum impedance OUTSIDE of their usual operating range? You're basing your whole argument on a made-up parameter, which negates the entire argument.
Interesting design Lord S, look forward to seeing you progress!
JaG, Brian is correct, Re is only applicable to one frequency, and a constant current source frequency sweep will give a single resonance peak at that frequency. For "real world" comparisons, it is much more appropriate to measure sensitivity, and a constant voltage source is used to do this. Hence the standardization of 2.83V, an arbitrary value that reflects 1W into an ideal frequency independent constant 8ohm load, (that no driver ever is!).
If you want to explore this further, I recommend a copy of "Testing Loudspeakers" by Joseph D'Appolito, a great book everyone should have on their shelf.
JaG, Brian is correct, Re is only applicable to one frequency, and a constant current source frequency sweep will give a single resonance peak at that frequency. For "real world" comparisons, it is much more appropriate to measure sensitivity, and a constant voltage source is used to do this. Hence the standardization of 2.83V, an arbitrary value that reflects 1W into an ideal frequency independent constant 8ohm load, (that no driver ever is!).
If you want to explore this further, I recommend a copy of "Testing Loudspeakers" by Joseph D'Appolito, a great book everyone should have on their shelf.
I have to say i'm kind of with JAG on this one, comparing systems is more "fair" using minimum impedance in the passband...
5 ohm minimum impedance and 8 ohm minimum impedance would both be called "8 ohms nominal" this can make it look like the 5 ohm cab 2dB's more sensitive. While in fact the other cab is just drawing 2dB's less power. On a 2ohms stable amplifier you could connect 3 of the 5 ohms cabs, or 4 of the 8 ohms. So it's not equal power, but it is equal power in minimum impedance! There is no reason to compare 2 different systems with equal power, across the entire band.
If you'd have the same cab, same driver, just different Re (with BL and Le adjusted as well, same motor), you'll find the total impedance will scale with the Re. So it's not just the load @ minimum impedance that is lower, it's the load across the entire band.
I really love the comparisons of the Prosound Shootouts of 2007 http://audioroundtable.com/ProSpeakers/messages/486.html
Power based on minimum impedance.
5 ohm minimum impedance and 8 ohm minimum impedance would both be called "8 ohms nominal" this can make it look like the 5 ohm cab 2dB's more sensitive. While in fact the other cab is just drawing 2dB's less power. On a 2ohms stable amplifier you could connect 3 of the 5 ohms cabs, or 4 of the 8 ohms. So it's not equal power, but it is equal power in minimum impedance! There is no reason to compare 2 different systems with equal power, across the entire band.
If you'd have the same cab, same driver, just different Re (with BL and Le adjusted as well, same motor), you'll find the total impedance will scale with the Re. So it's not just the load @ minimum impedance that is lower, it's the load across the entire band.
I really love the comparisons of the Prosound Shootouts of 2007 http://audioroundtable.com/ProSpeakers/messages/486.html
Power based on minimum impedance.
Last edited:
Fortunately hornresp (and others) can simulate impedance. So you can check the minimum impedance, adjust Eg accordingly and compare it to other simulated systems
And to get back to the Vtc Atc stuff, just add it to the sim and you'll have an accurate simulation of your current design. Compare it to you current sim (so with compensation, thus without Vtc) and see if it even makes a difference.
Based on that you can add your pieces of wood if necessary. Or change the drawings as xoc1 suggested in post 9 and get a bit extra extension, or shave a few cm from you cabinet size
And to get back to the Vtc Atc stuff, just add it to the sim and you'll have an accurate simulation of your current design. Compare it to you current sim (so with compensation, thus without Vtc) and see if it even makes a difference.
Based on that you can add your pieces of wood if necessary. Or change the drawings as xoc1 suggested in post 9 and get a bit extra extension, or shave a few cm from you cabinet size
Fortunately hornresp (and others) can simulate impedance. So you can check the minimum impedance, adjust Eg accordingly and compare it to other simulated systems
Indeed, I think we're actually agreeing here. Re is of no import in itself. System minimum impedance within passband is. The best way of getting that is with a constant voltage source simulation. Yes?
I'd say:
Amplifier is a voltage and current source, the amount of power you can have from a system is amplifier dependent. If you plug two subs at the same amplifier one of them will play loud and is the one that has lower impedance and this is the reason that all amplifier gives more power with 4 ohm then 8 ohm. Simple as that.
The driver are divided basically at two classes 4ohm and 8ohm, so if you are comparing two system at the same class it really make sense to me to compare at specific voltage and see what shows higher SPL, if one company can reach lower Re for a given driver then others so it has some credits, they can play with all parameter and Re is one of then.
I think we should never forget that the power will always be limited by the amplifier. And efficiency is produce more SPL at the same power. Comparing the same class.
In the past we were limited in terms of electronic components so to have more power it was easy to rise the voltage. Now you can increase power increasing current source.
Power = Voltage x current (simple way)
This doesn't make sense, 1,4 ohm and 16 ohm are part of different classes, if it was 4,7 ohm against 5,6 ohm, I see no problem to compare both @ 2,83V and 4,7 will have some advantage regarding electric power but it doesn't means that the SPL curve will be better.
Example using car subject:
Wouldn't you compare two engine sold as 2L ? in fact one can be 1,84L and the other 2,1L.
If one of then is normal aspirated and the other is turbo, will you still compare them?
What do think is more efficient?
And the others technologies like variable valve train ans so on, all of then will impact the engine efficiency.
Compare things at the same class is the simple and easiest way to go;
Amplifier is a voltage and current source, the amount of power you can have from a system is amplifier dependent. If you plug two subs at the same amplifier one of them will play loud and is the one that has lower impedance and this is the reason that all amplifier gives more power with 4 ohm then 8 ohm. Simple as that.
The driver are divided basically at two classes 4ohm and 8ohm, so if you are comparing two system at the same class it really make sense to me to compare at specific voltage and see what shows higher SPL, if one company can reach lower Re for a given driver then others so it has some credits, they can play with all parameter and Re is one of then.
I think we should never forget that the power will always be limited by the amplifier. And efficiency is produce more SPL at the same power. Comparing the same class.
In the past we were limited in terms of electronic components so to have more power it was easy to rise the voltage. Now you can increase power increasing current source.
Power = Voltage x current (simple way)
This doesn't make sense, 1,4 ohm and 16 ohm are part of different classes, if it was 4,7 ohm against 5,6 ohm, I see no problem to compare both @ 2,83V and 4,7 will have some advantage regarding electric power but it doesn't means that the SPL curve will be better.
Example using car subject:
Wouldn't you compare two engine sold as 2L ? in fact one can be 1,84L and the other 2,1L.
If one of then is normal aspirated and the other is turbo, will you still compare them?
What do think is more efficient?
And the others technologies like variable valve train ans so on, all of then will impact the engine efficiency.
Compare things at the same class is the simple and easiest way to go;
@pinkmouse: Yes
@Lordsansui: An amplifier is a voltage amplifier with a certain voltage limit, certain current limit and amplifiers like the "Lab.Gruppen fp14000" for example have a power limit.
If I look at my most recent subs, FLH's with 18SW115, I can use both the 8 ohm and 4 ohm version, practically same result, different impedance. Comparing sensitivity at 2,83V will give the 4 ohm driver a clear advantage.
While there is no actual difference, it's only a different impedance... The only reason to select a different impedance is because it matches your amplifier-situation better, not sensitivity.
@Lordsansui: An amplifier is a voltage amplifier with a certain voltage limit, certain current limit and amplifiers like the "Lab.Gruppen fp14000" for example have a power limit.
If I look at my most recent subs, FLH's with 18SW115, I can use both the 8 ohm and 4 ohm version, practically same result, different impedance. Comparing sensitivity at 2,83V will give the 4 ohm driver a clear advantage.
While there is no actual difference, it's only a different impedance... The only reason to select a different impedance is because it matches your amplifier-situation better, not sensitivity.
Hello Brian Steele
I have specific questions for you and your feedback will help me a lot. Thank you in advanced.
I went to your site, again, and collect data from all your "Proof of Concept" tests, so lets see some numbers:
POC#2
POC#3
POC#4
1 - Regarding compression ratio used, did you observed/measured high distortion?
2 - Regarding diaphragm displacement, did you tested all cab at max power? what you observed the distortion increases?
3 - From POC#2 to POC#3 you change the TH design and increase compression ratio, what you found as a result? is there any note specific for compression ratio?
I have specific questions for you and your feedback will help me a lot. Thank you in advanced.
I went to your site, again, and collect data from all your "Proof of Concept" tests, so lets see some numbers:
POC#2
Driver = Dayton PA310-8
Pe=450W
Xmax=5mm
Sd=530,9sqcm
S1=270sqcm
Compression ratio = 530,9/270=1,96
Diaphragm displacement @450W(48,37V) = 10,9mm (5,9mm more then Xmax)
Pe=450W
Xmax=5mm
Sd=530,9sqcm
S1=270sqcm
Compression ratio = 530,9/270=1,96
Diaphragm displacement @450W(48,37V) = 10,9mm (5,9mm more then Xmax)
POC#3
Driver = Dayton PA310-8 (The same from POC#2)
Pe=450W
Xmax=5mm
Sd=530,9sqcm
S1=232sqcm (Reduced S1)
Compression ratio = 530,9/232=2,28
Diaphragm displacement @450W(48,37V) = 10,8mm (5,8mm more then Xmax)
Pe=450W
Xmax=5mm
Sd=530,9sqcm
S1=232sqcm (Reduced S1)
Compression ratio = 530,9/232=2,28
Diaphragm displacement @450W(48,37V) = 10,8mm (5,8mm more then Xmax)
POC#4
Driver = B&C 18TBX100
Pe=1200W
Xmax=9mm
Sd=1210sqcm
S1=492,9sqcm
Compression ratio = 1210/492,9=2,28
Diaphragm displacement @1200W(78,23V) = 13,7mm (4,7mm more then Xmax)
Pe=1200W
Xmax=9mm
Sd=1210sqcm
S1=492,9sqcm
Compression ratio = 1210/492,9=2,28
Diaphragm displacement @1200W(78,23V) = 13,7mm (4,7mm more then Xmax)
1 - Regarding compression ratio used, did you observed/measured high distortion?
2 - Regarding diaphragm displacement, did you tested all cab at max power? what you observed the distortion increases?
3 - From POC#2 to POC#3 you change the TH design and increase compression ratio, what you found as a result? is there any note specific for compression ratio?
Fortunately hornresp (and others) can simulate impedance. So you can check the minimum impedance, adjust Eg accordingly and compare it to other simulated systems
And to get back to the Vtc Atc stuff, just add it to the sim and you'll have an accurate simulation of your current design. Compare it to you current sim (so with compensation, thus without Vtc) and see if it even makes a difference.
Based on that you can add your pieces of wood if necessary. Or change the drawings as xoc1 suggested in post 9 and get a bit extra extension, or shave a few cm from you cabinet size
Oh, and how about these situations:
1. Sealed subwoofer systems - minimum impedance appears OUTSIDE the passband. What do you use?
2. All systems - minimum impedance changes as power input increases, and in different ways too (vented system will usually change faster than TH). Do you base it at minimum impedance @1W, or minimum impedance at expected average drive level?
See? It gets more and more complicated. Which is why no manufacturer that I know of uses the 1W/"minimum impedance" method and they use a standard 2.83 drive voltage instead and leave it up to the owner of the system to intelligently match the rated impedance of the system (or systems) with the pro audio amp being used to drive them.
Yes, you're absolutely right. There's always compromise
I personally would like to see 100W into minimum impedance at 10 meters and additional graphes for higher powers, to see power compression. To me that's the only way to determine the best subwoofer system for PA use.
And additionally how they perform in stacks. That's info practically no manufacturer supply's. While this is more important than performance of 1 subwoofer, for those that need more than one ofcourse... As we al can agree doubling subwoofers isn't simply 6dB extra. It's more complicated than that...
So:
2. Impedance rising differently would show itself in the powercompression charts, because rising of impedance itself is not the only important factor, how the system behaves after this increase is.
1. You're right, i'd say: pick the lowest impedance your intended bandwith. Not perfect...
I personally would like to see 100W into minimum impedance at 10 meters and additional graphes for higher powers, to see power compression. To me that's the only way to determine the best subwoofer system for PA use.
And additionally how they perform in stacks. That's info practically no manufacturer supply's. While this is more important than performance of 1 subwoofer, for those that need more than one ofcourse... As we al can agree doubling subwoofers isn't simply 6dB extra. It's more complicated than that...
So:
2. Impedance rising differently would show itself in the powercompression charts, because rising of impedance itself is not the only important factor, how the system behaves after this increase is.
1. You're right, i'd say: pick the lowest impedance your intended bandwith. Not perfect...
Standards are a beautiful thing.
Hi Y'all,
Here is one more vote for using the sensitivity secification based on 2.83Vac/1m as a comparative reference.
The alternative might be to compare systems based on output @Xmax, not always easy to determine in the real world, but easily dialed in in Hornresp.
1W/1m is still of interest, but not for comparing systems. There is just too much variation in the impedance curves...
I like Danley's method of measuring @10m/28.3Vac, but even they have not been consistent, I remember at least one box they claim their measurement @10/20V. So even there, more than one "Standard".
Regards,
Hi Y'all,
Here is one more vote for using the sensitivity secification based on 2.83Vac/1m as a comparative reference.
The alternative might be to compare systems based on output @Xmax, not always easy to determine in the real world, but easily dialed in in Hornresp.
1W/1m is still of interest, but not for comparing systems. There is just too much variation in the impedance curves...
I like Danley's method of measuring @10m/28.3Vac, but even they have not been consistent, I remember at least one box they claim their measurement @10/20V. So even there, more than one "Standard".
Regards,
I thought Danley's reasons are obvious.
20 volts is 100 watts into 4 ohms, as 2.83 volts is for 8 ohms. So the idea is not only to measure and compare at 100 watts but also dovetails into the next reason; which is determining the 1 metre distance for a TH is way too difficult; where lies the Point Source? Hence, 10 metres makes this less of a blur. It also neatly fits into the 100 watt drive level.
I didn't do anything on the fly, you asked a specific question and I gave a specific answer.
In most cases the minimum impedance in the passband is very close to Re, so that's an acceptable figure to use. In some cases, like full size horns the minimum impedance in the passband can be more than 2x Re, so technically it's always best to use minimum impedance in the passband instead of Re, even if it's not necessary in the vast majority of cases.
Giving spl figures at a given voltage vs a given wattage serves different purposes, both of which are valid given the context of any given situation.
In MOST situations what people are trying to do is make equal power comparisons, or show the spl level at a certain amount of power defined by watts. In both those cases the voltage referenced to minimum impedance in the passband is more accurate than specifying a voltage based on nominal impedance or just a flat rate 2.83V for everything.
I fully realize that the impedance curve is not flat so "equal power" is not exactly equal, but it's a LOT MORE equal than showing everything at 2.83V.
No, I have both drivers and amps that are rated for 2, 6 and 16 ohms, so which category do they fit in? These "classes" of nominal impedance are a dumbed down construct of the concept of a complex impedance curve.
It makes sense to Danley, they rate everything at 2.83V regardless of actual or nominal impedance. That makes perfect sense for a company that produces a large amount of products (probably most of their products) that have Re well under 8 ohms and even under 4 ohms. It's gives really big numbers that look fantastic on a spec sheet and since the average consumer doesn't know what these numbers mean they like the big numbers. Most consumers are expecting the sensitivity spec to be an "equal power" spec.
Comparing them at the closest meaningful point on the impedance curve is even more accurate than to break it down into classes that allow up to 2x variation in the same class, and in Danley's case an almost infinite amount of variation because there's only one class.
Take a look at any Danley spec sheet. Why is sensitivity referenced to 2.83V always but max power handling is specified as watts, specifically not as voltage? Based on Brian's argument everything should be specified as voltage because the amps are a voltage source.
It allows both numbers to be as high as possible.
If you have a 3 ohm minimum impedance in the passband it allows a very high sensitivity number if you spec it at 2.83V. But it would be a very low number (compared to a cab with 8 ohms minimum impedance in the passband) if you specified max power as voltage instead of watts.
So why is the data sheet switching between a static voltage for sensitivity to a wattage for max power? It looks fantastic on a spec sheet and it can fool people that don't know how these numbers are derived or what they mean.
It allows both numbers to be as high as possible.
If you have a 3 ohm minimum impedance in the passband it allows a very high sensitivity number if you spec it at 2.83V. But it would be a very low number (compared to a cab with 8 ohms minimum impedance in the passband) if you specified max power as voltage instead of watts.
So why is the data sheet switching between a static voltage for sensitivity to a wattage for max power? It looks fantastic on a spec sheet and it can fool people that don't know how these numbers are derived or what they mean.
To expand on this, let's assume the average customer is dumb as a bag of rocks, which in most cases is true.
There are two subwoofer models for sale, the ONLY difference between these two models is how the dvc driver is wired - one is in parallel and one is in series. Let's call these models Sub P (for parallel) and Sub S (for series). All else is equal between these models except the dvc wiring. Minimum impedance in the passband for Sub P is 2 ohms and for Sub S it's 8 ohms. Both subs use the same driver so they can each handle 1000 watts max. Both subs use the same driver so they have the same frequency response curve shape.
So here's what the spec sheet says for each model.
Sub P - 100 db @ 2.83V, 1000 watts max power
Sub S - 94 db @ 2.83V, 1000 watts max power
So your average dumb as a bag of rocks consumer sees these numbers and says - HEY! Both subs have the same max power handling and Sub P is 6 db more sensitive so Sub P can go 6 db louder!
Your average dumb as a bag of rocks consumer doesn't realize that the efficiency is the same and the max spl is the same.
Your average dumb as a bag of rocks consumer doesn't realize that Sub P is taking 4 watts at 2.83V while Sub P is only drawing 1 watt at 2.83V.
And Brian, this one is for you. The impedance curve of Sub P at any given frequency is exactly 4x lower than Sub S. This means that if you simulate them both at 1 actual watt instead of simulating them both at 2.83V, they will literally be exactly equal power sims at every frequency at every point regardless of the fact that the impedance curve is not a flat line.
The ONLY information the average consumer is going to take away from the two spec spec sheet is that Sub P can go 6 db louder, the uninformed perception of which is ironically literally the only thing that's untrue wrt those two specs. Sub S will require more voltage to hit the same level (but exactly the same power in watts) due to it's 4x lower impedance but in every other way Sub S and Sub P are identical.
If the spec sheet intent was to compare fairly it would either be showing voltage based on a baseline equivalent power level at minimum impedance in the passband or the spec sheet would show efficiency instead of sensitivity. Using a static 2.83V to compare everything is a deliberate trick intended to fool ignorant consumers. How many average consumers know the difference between sensitivity and efficiency or how many volts their amplifier can put out? For the average consumer EVERYTHING is about watts, not volts. What amp manufacturer advertises their products' voltage rails? None, they advertise the power in watts. Even here where we are discussing this issue, how many of us know the voltage rails of their amps? I would guess a few people, but not many. What sim user shows a sim and says this design can take X amount of volts? Never, they always state the power in watts.
When choosing to reference voltage or watts you have to take into consideration the intent of the information presented, a company showing sensitivity of all products at 2.83V is deliberately intending to mislead the uninformed section of their customer base, which is most people.
There are two subwoofer models for sale, the ONLY difference between these two models is how the dvc driver is wired - one is in parallel and one is in series. Let's call these models Sub P (for parallel) and Sub S (for series). All else is equal between these models except the dvc wiring. Minimum impedance in the passband for Sub P is 2 ohms and for Sub S it's 8 ohms. Both subs use the same driver so they can each handle 1000 watts max. Both subs use the same driver so they have the same frequency response curve shape.
So here's what the spec sheet says for each model.
Sub P - 100 db @ 2.83V, 1000 watts max power
Sub S - 94 db @ 2.83V, 1000 watts max power
So your average dumb as a bag of rocks consumer sees these numbers and says - HEY! Both subs have the same max power handling and Sub P is 6 db more sensitive so Sub P can go 6 db louder!
Your average dumb as a bag of rocks consumer doesn't realize that the efficiency is the same and the max spl is the same.
Your average dumb as a bag of rocks consumer doesn't realize that Sub P is taking 4 watts at 2.83V while Sub P is only drawing 1 watt at 2.83V.
And Brian, this one is for you. The impedance curve of Sub P at any given frequency is exactly 4x lower than Sub S. This means that if you simulate them both at 1 actual watt instead of simulating them both at 2.83V, they will literally be exactly equal power sims at every frequency at every point regardless of the fact that the impedance curve is not a flat line.
The ONLY information the average consumer is going to take away from the two spec spec sheet is that Sub P can go 6 db louder, the uninformed perception of which is ironically literally the only thing that's untrue wrt those two specs. Sub S will require more voltage to hit the same level (but exactly the same power in watts) due to it's 4x lower impedance but in every other way Sub S and Sub P are identical.
If the spec sheet intent was to compare fairly it would either be showing voltage based on a baseline equivalent power level at minimum impedance in the passband or the spec sheet would show efficiency instead of sensitivity. Using a static 2.83V to compare everything is a deliberate trick intended to fool ignorant consumers. How many average consumers know the difference between sensitivity and efficiency or how many volts their amplifier can put out? For the average consumer EVERYTHING is about watts, not volts. What amp manufacturer advertises their products' voltage rails? None, they advertise the power in watts. Even here where we are discussing this issue, how many of us know the voltage rails of their amps? I would guess a few people, but not many. What sim user shows a sim and says this design can take X amount of volts? Never, they always state the power in watts.
When choosing to reference voltage or watts you have to take into consideration the intent of the information presented, a company showing sensitivity of all products at 2.83V is deliberately intending to mislead the uninformed section of their customer base, which is most people.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.