Speaker Impedance

On Thu, 21 Dec 2017 01:14:51 -0600, "Tim Williams"
<tiwill@seventransistorlabs.com> wrote:

><krw@notreal.com> wrote in message 
>news:ck4m3dl14n550qbo1o60u8gh5hj5ktokja@4ax.com...
>> The impedance of any device must be (equal to or) higher than it's DC
>> resistance or it's a battery.
>
>Unless there's an impedance in parallel with the DCR, like capacitance. 

Good point.

>Fortunately, that doesn't happen to voice coils until well into the MHz (and 
>even then, the valley is probably much higher than the DCR, due to other 
>losses).

Yes, I was thinking more in these terms.

Remember, it's the real part of the impedance that makes noise (and
heat).

On Thu, 21 Dec 2017 13:04:16 +0200, upsidedown@downunder.com wrote:

>On Wed, 20 Dec 2017 21:01:49 -0500, krw@notreal.com wrote:
>
>>On Wed, 20 Dec 2017 20:51:46 +0200, upsidedown@downunder.com wrote:
>>
>>>On Wed, 20 Dec 2017 10:34:32 -0500, krw@notreal.com wrote:
>>>
>>>>On Wed, 20 Dec 2017 06:41:30 -0800 (PST), George Herold
>>>><gherold@teachspin.com> wrote:
>>>>
>>>>>On Wednesday, December 20, 2017 at 3:21:29 AM UTC-5, Robert Baer wrote:
>>>>>> OGee wrote:
>>>>>> > I have one test instruments to test with, a VOM.
>>>>>> >
>>>>>> > I have several old speakers that I want to hook up to a amp I am trying
>>>>>> > to build.
>>>>>> >
>>>>>> > How do I determine the impedance of the speakers with only a VOM ?
>>>>>> >
>>>>>> > What value of resistance would I measure across the speaker coil for
>>>>>> > different speaker impedance of say 4 or 8 or ? and other ohm speaker
>>>>>> > impedance ?
>>>>>> >
>>>>>> > Is there a table someplace ?
>>>>>> >
>>>>>> > TIA
>>>>>>    Resistance will always be a few ohms or less, WAY below "impedance" 
>>>>>
>>>>>I've got two speakers in my lab.  
>>>>>small '32 ohm' Resistance = 30 ohms.
>>>>>Car size speaker, '8 ohms', R = 7.3 ohms.
>>>>
>>>>What is "car size"?  4000lbs?  ;-)  Car speakers aren't 8-ohms, rather
>>>>2-ohm (that V^2/R thing), usually.
>>>
>>>With a "12 V" car battery voltage you can get about 4 Vrns output
>>>voltage from a half bridge, i.e.  0.5 W into 32 ohms or 8 W into 2
>>>ohms. Some full bridge ICs claim 22 W into 4 ohms, requiring 14.4 V
>>>battery voltage (alternator maximum output voltage).
>>
>>OK, you got the arithmetic right(ish).  The typical car amplifier is
>>more like 50W into a 2-ohm speaker.
>
>That sounds like some kind of PMPO power :-).

No.
>
>50 W sine into 2 ohms is 10 Vrms, 14,1 Vpk or 28.3 Vpp. Theoretically,
>this would require 28.3 Vdc half bridge or 14.1 Vdc full bridge power
>supply.

Assuming 0% distortion.  Automotive amplifiers are typically rated to
10% distortion (most claim something around 49W).  BTW, they're all
full bridge.  Half-bridge makes no sense.

>For bipolars in (darlington)emiter follower configuration, add two Vbe
>drops for a half bridge  or four Vbe drops in a full H-brudge
>configuration.  At those currents Vbe might be in the order of 1 V,
>requiring 30 V resp. 18 V supply voltage.

They have the new fanged transistors called Metal Oxide
Semiconductors.

> With a common emitter voltage amplifier stage at output, there will
>be one resp. two Vce(sat) drops. With Vce(sat) of 0.7 V, the power
>supply needs to be at least 29 V resp. 15.5 Vdc. These voltage levels
>are clearly out of the safe area for "24 V" resp "12 V" lead
>batteries.

If you didn't have a dick, you'd be a woman.  Hmmm.
>>>
>>>With a DC/DC inverter power supply, the voltage swing and speaker
>>>impedance can be more freely selected.
>>
>>Obviously.  What do you think a 400W boost regulator costs?  What does
>>it add to the cost of the car.  What's the market?  ...just to "freely
>>select" speaker impedance.  
>
>A 400 W DC/DC converter doesn't cost a lot (at least outside car audio
>outlets). Since that converter needs about 40 A current it should be
>installed into the engine compartment close to the car battery. This
>might add some cost due to the environment requirements.

Clueless.

>BTW, with the increased use of hybrids and full electric cars with
>high battery voltages for the electric motors, why not use the motor
>battery for also driving the audio amplifiers :-)

Because high-voltage semiconductors are expensive, too.

On Thu, 21 Dec 2017 03:46:46 -0800 (PST), jurb6006@gmail.com wrote:

>>"BTW, with the increased use of hybrids and full electric cars with
>high battery voltages for the electric motors, why not use the motor
>battery for also driving the audio amplifiers :-) "
>
>That would make them incompatible. 
>
>Also in 24 volt systems there is usually a battery balancer if there is to be any 12 equipment run off it. 
>
>In a nutshell, a nominal 12 volt supply will yield about 8.9 watts into 8 ohms. Divide and multiply as needed, like 17.9 into 4 ohms, and so forth. 

Double that, again, for a full bridge output.

>Looking at the DC convertor in a standard cat amp, it doesn't add that much cost on a manufacturing level. The transformer is cheap, you need a uPc type chip for the drive/control. The pound of silicon is cheap these days. Just a bunch of MOSFETs in  parallel. the heatsinks cost more. heatsinks you got a tradeoff, either alot of aluminum or cheat with a high efficiency design, which is a more expensive solution, or maybe even one of those berylium heat pipes. (which doesn't really lend itself well to this application)

Wrong.  

k...@notreal.com wrote:

-----------------------
> > 
> Remember, it's the real part of the impedance that makes noise (and
> heat).
>


** The "add 20% to the DC value " rule for nominal impedance has a breakdown. 

If you find the exact impedance minimum frequency, then the impedance value is perfectly resistive and it follows that components making up the value are also resistive - the reactive ones have self cancelled.

We all know the biggest component, voice coil resistance which accounts for about 80%.

The next is friction losses in the spider and surround, which accounts for about 10% closely followed by eddy current losses in the magnet structure. 

Last and least is actual sound radiation at under 1%. 
 
Move the test frequency a little and the impedance rises due to motional back emf OR voice coil inductance. Both these the effect of reducing current flow in the voice coil and so drive force which must de taken into account when designing the enclosure, crossover points etc. 



....   Phil

On Thursday, December 21, 2017 at 8:26:53 PM UTC-5, Phil Allison wrote:
> k...@notreal.com wrote:
> 
> -----------------------
> > > 
> > Remember, it's the real part of the impedance that makes noise (and
> > heat).
> >
> 
> 
> ** The "add 20% to the DC value " rule for nominal impedance has a breakdown. 
> 
> If you find the exact impedance minimum frequency, then the impedance value is perfectly resistive and it follows that components making up the value are also resistive - the reactive ones have self cancelled.
> 
> We all know the biggest component, voice coil resistance which accounts for about 80%.
> 
> The next is friction losses in the spider and surround, which accounts for about 10% closely followed by eddy current losses in the magnet structure.

Nice, Thanks.  

George H. 
 
> 
> Last and least is actual sound radiation at under 1%. 
>  
> Move the test frequency a little and the impedance rises due to motional back emf OR voice coil inductance. Both these the effect of reducing current flow in the voice coil and so drive force which must de taken into account when designing the enclosure, crossover points etc. 
> 
> 
> 
> ....   Phil

On Thu, 21 Dec 2017 19:18:21 -0500, krw@notreal.com wrote:

>On Thu, 21 Dec 2017 03:46:46 -0800 (PST), jurb6006@gmail.com wrote:
>
>>>"BTW, with the increased use of hybrids and full electric cars with
>>high battery voltages for the electric motors, why not use the motor
>>battery for also driving the audio amplifiers :-) "
>>
>>That would make them incompatible. 
>>
>>Also in 24 volt systems there is usually a battery balancer if there is to be any 12 equipment run off it. 
>>
>>In a nutshell, a nominal 12 volt supply will yield about 8.9 watts into 8 ohms. Divide and multiply as needed, like 17.9 into 4 ohms, and so forth. 

Do the math. 8.9 W into 8  ohms is 8.5 Vrms, 12 Vpk and 24 Vpp, thus
8.9 W is for a bridged amplifier.

>Double that, again, for a full bridge output.

Your figures are _already_ for a bridged amplifier.

For example TDA2005 in bridged configuration produces 20 W into 4 ohms
with 14.4 Vdc and 10 % THD and 22 W into 3.2 ohms.

Such high THD allows a quite significant clipping i.e. non-sinus
waveform.


>>Looking at the DC convertor in a standard cat amp, it doesn't add that much cost on a manufacturing level. The transformer is cheap, you need a uPc type chip for the drive/control. The pound of silicon is cheap these days. Just a bunch of MOSFETs in  parallel. the heatsinks cost more. heatsinks you got a tradeoff, either alot of aluminum or cheat with a high efficiency design, which is a more expensive solution, or maybe even one of those berylium heat pipes. (which doesn't really lend itself well to this application)
>
>Wrong.  

A simple inverter would cost much less than a few  high quality
amplifiers it is feeding.

On Fri, 22 Dec 2017 09:50:27 +0200, upsidedown@downunder.com wrote:

>On Thu, 21 Dec 2017 19:18:21 -0500, krw@notreal.com wrote:
>
>>On Thu, 21 Dec 2017 03:46:46 -0800 (PST), jurb6006@gmail.com wrote:
>>
>>>>"BTW, with the increased use of hybrids and full electric cars with
>>>high battery voltages for the electric motors, why not use the motor
>>>battery for also driving the audio amplifiers :-) "
>>>
>>>That would make them incompatible. 
>>>
>>>Also in 24 volt systems there is usually a battery balancer if there is to be any 12 equipment run off it. 
>>>
>>>In a nutshell, a nominal 12 volt supply will yield about 8.9 watts into 8 ohms. Divide and multiply as needed, like 17.9 into 4 ohms, and so forth. 
>
>Do the math. 8.9 W into 8  ohms is 8.5 Vrms, 12 Vpk and 24 Vpp, thus
>8.9 W is for a bridged amplifier.

You assume a sine wave (i.e. 0% distortion).  
>
>>Double that, again, for a full bridge output.
>
>Your figures are _already_ for a bridged amplifier.

>For example TDA2005 in bridged configuration produces 20 W into 4 ohms
>with 14.4 Vdc and 10 % THD and 22 W into 3.2 ohms.
>
>Such high THD allows a quite significant clipping i.e. non-sinus
>waveform.

Of course but that's reality.
>
>>>Looking at the DC convertor in a standard cat amp, it doesn't add that much cost on a manufacturing level. The transformer is cheap, you need a uPc type chip for the drive/control. The pound of silicon is cheap these days. Just a bunch of MOSFETs in  parallel. the heatsinks cost more. heatsinks you got a tradeoff, either alot of aluminum or cheat with a high efficiency design, which is a more expensive solution, or maybe even one of those berylium heat pipes. (which doesn't really lend itself well to this application)
>>
>>Wrong.  
>
>A simple inverter would cost much less than a few  high quality
>amplifiers it is feeding.

Wrong.

On Fri, 22 Dec 2017 08:56:05 -0500, krw@notreal.com wrote:

>On Fri, 22 Dec 2017 09:50:27 +0200, upsidedown@downunder.com wrote:
>
>>On Thu, 21 Dec 2017 19:18:21 -0500, krw@notreal.com wrote:
>>
>>>On Thu, 21 Dec 2017 03:46:46 -0800 (PST), jurb6006@gmail.com wrote:
>>>
>>>>>"BTW, with the increased use of hybrids and full electric cars with
>>>>high battery voltages for the electric motors, why not use the motor
>>>>battery for also driving the audio amplifiers :-) "
>>>>
>>>>That would make them incompatible. 
>>>>
>>>>Also in 24 volt systems there is usually a battery balancer if there is to be any 12 equipment run off it. 
>>>>
>>>>In a nutshell, a nominal 12 volt supply will yield about 8.9 watts into 8 ohms. Divide and multiply as needed, like 17.9 into 4 ohms, and so forth. 
>>
>>Do the math. 8.9 W into 8  ohms is 8.5 Vrms, 12 Vpk and 24 Vpp, thus
>>8.9 W is for a bridged amplifier.
>
>You assume a sine wave (i.e. 0% distortion).  


Of course, that has been the standard practice for nearly a century:
Connect a known resistor to the amplifier and an oscilloscope across
the resistor. Turn up the sine signal until the waveform starts
visually  to distort. Read peak to peak voltage and calculate the
power. It is to visually detect a few percent distortion (far before
reaching 10 % THD).

Now the  fucking PMPO people try to inflate the power readings for
marketing purposes. Why don't the use a square wave, to get bigger
numbers :-).

A clipping full range amplifier is really bad for music signals due to
various intermodulation products. A high distortion on a bass guitar
or subwoofer amplifier is less harmful with less simultaneous tones.

On Fri, 22 Dec 2017 18:24:01 +0200, upsidedown@downunder.com wrote:

>On Fri, 22 Dec 2017 08:56:05 -0500, krw@notreal.com wrote:
>
>>On Fri, 22 Dec 2017 09:50:27 +0200, upsidedown@downunder.com wrote:
>>
>>>On Thu, 21 Dec 2017 19:18:21 -0500, krw@notreal.com wrote:
>>>
>>>>On Thu, 21 Dec 2017 03:46:46 -0800 (PST), jurb6006@gmail.com wrote:
>>>>
>>>>>>"BTW, with the increased use of hybrids and full electric cars with
>>>>>high battery voltages for the electric motors, why not use the motor
>>>>>battery for also driving the audio amplifiers :-) "
>>>>>
>>>>>That would make them incompatible. 
>>>>>
>>>>>Also in 24 volt systems there is usually a battery balancer if there is to be any 12 equipment run off it. 
>>>>>
>>>>>In a nutshell, a nominal 12 volt supply will yield about 8.9 watts into 8 ohms. Divide and multiply as needed, like 17.9 into 4 ohms, and so forth. 
>>>
>>>Do the math. 8.9 W into 8  ohms is 8.5 Vrms, 12 Vpk and 24 Vpp, thus
>>>8.9 W is for a bridged amplifier.
>>
>>You assume a sine wave (i.e. 0% distortion).  
>
>
>Of course, that has been the standard practice for nearly a century:

No, it's not.

>Connect a known resistor to the amplifier and an oscilloscope across
>the resistor. Turn up the sine signal until the waveform starts
>visually  to distort. Read peak to peak voltage and calculate the
>power. It is to visually detect a few percent distortion (far before
>reaching 10 % THD).
>
>Now the  fucking PMPO people try to inflate the power readings for
>marketing purposes. Why don't the use a square wave, to get bigger
>numbers :-).

Wrong.
>
>A clipping full range amplifier is really bad for music signals due to
>various intermodulation products. A high distortion on a bass guitar
>or subwoofer amplifier is less harmful with less simultaneous tones.
>
Wrong again.

>">In a nutshell, a nominal 12 volt supply will yield about 8.9 watts into 8 ohms. Divide and multiply as needed, like 17.9 into 4 ohms, and so forth.

Double that, again, for a full bridge output. "

No, that is for a full bridge. That's right, half bridge only puts out 2.25 watts/8 ohms.

>">Looking at the DC convertor in a standard cat amp, it doesn't add that much cost on a manufacturing level. The transformer is cheap, you need a uPc type chip for the drive/control. The pound of silicon is cheap these days. Just a bunch of MOSFETs in  parallel. the heatsinks cost more. heatsinks you got a tradeoff, either alot of aluminum or cheat with a high efficiency design, which is a more expensive solution, or maybe even one of those berylium heat pipes. (which doesn't really lend itself well to this application)"

"Wrong."

Which part ?