78 noise

3 minute wonder
Post Reply
Coffee Phil
vinyl addict
vinyl addict
United States of America
Posts: 5747
Joined: 20 Sep 2008 08:22
Location: California

Re: 78 noise

Post by Coffee Phil » 17 Aug 2019 05:03

Hi H. Callahan,

OK, I guess I’ll have to wait for someone to translate your stuff from discotheque physics to Newtonian Physics so I can understand it.

Pity that the EE program at my university required that clearly outmoded type of physics instead of letting us dance our bell bottoms off, learning your clearly superior science.

Phil
Last edited by Coffee Phil on 17 Aug 2019 05:38, edited 1 time in total.

josephazannieri
long player
long player
United States of America
Posts: 2906
Joined: 20 Apr 2008 06:01
Location: Norwalk,Ohio,USA

Re: 78 noise

Post by josephazannieri » 17 Aug 2019 05:28

Yo Phil and others discussing speaker characteristics:

I am going to have to get a little nitpickey with your statements here:
Coffee Phil wrote:
15 Aug 2019 22:46

These speakers are highly inefficient. The voice coil impedance (above resonance) is largely the DC resistance, hence the current through the coil and the force it exerts on the cone is proportional to the applied voltage. I think the bulk of the remaining force acting on the coil is the mass of the cone.
Actually, the speaker's voice coil impedance is not exclusively or even primarily resistive. It is a combination of resistance and inductance, and, at resonance, the self-generating back EMF caused by the speaker running amok. A given speaker with a given voice coil will have a constant inductance which is always there, much as any coil of wire will have a constant inductance. A speaker is a reactive load whose measured impedance varies according to the frequency applied to the voice coil.

A typical cone speaker will have an impedance peak at the resonance point. This occurs because at resonance, the speaker generates a back EMF caused by the motion of the cone. The proof is this, again from my demonstrations in Clyde, Ohio. If you take two similar cone speakers and hook the voice coils in parallel and push on the cone of one of the speakers, the cone of the other speaker will move about the same amount. This is because the voice coil of the speaker being pushed generates a current from being moved within the magnetic field of that speaker. At resonance, the speaker cone will vibrate not only because of the signal applied, but also because of the self-generated motion caused by the resonance. As the signal frequency rises above the resonant point, the measured impedance drops gradually until it reaches a low point, for most speakers somewhere around 400 Hz. This is because the speaker is no longer generating a back EMF, and the inductance of the voice coil, which may be as much as 2.2 mH for a typical 15 inch woofer, is not yet affecting the passage of current through the voice coil. For the example given, the woofer's impedance will start to rise at about 600 Hz, and may go as high as 25 or 30 ohms, reducing the current passing through the voice coil.

Now of course a smaller speaker with a smaller voice coil will not have as much inductance, and its impedance will not rise as much as a woofer's impedance will. This means that the voice coil with less inductance will not have as much of an impedance rise at high frequencies, and will operate better at high frequencies. While the current through the coil may be constant, the amount of motion will be affected at the low end by resonance, and at the high end by the impedance rise caused by voice coli inductance, which will reduce the current through the voice coil and reduce the motion of the cone. I am talking about speakers in free air, which behave differently from enclosed speakers. But that's another story.

Now an acoustic cutting head does not have a voice coil, but it does have a resonant point, so voice coil inductance will not affect its response to a tone of a certain frequency. But, the resonant point will give a bulge in the head's frequency response. Given the relatively small size of an acoustic cutting head, it is quite possible that the head's resonant point may occur in the frequency range that is being recorded.

Of course these rules apply in Clyde, Ohio, but I am pretty sure that they also apply in California. As a lawyer, I am glad to have the chance to criticize, and I thank you for your charitable indulgence of my pontificating.

And good luck from the old topic drifter, critic, and pontificator,

Joe Z.

Coffee Phil
vinyl addict
vinyl addict
United States of America
Posts: 5747
Joined: 20 Sep 2008 08:22
Location: California

Re: 78 noise

Post by Coffee Phil » 17 Aug 2019 06:58

Hi Joe,

Nit pit away. Sometimes I need to be taken to task. I was a bit sloppy there. Indeed as we approach the upper range of the driver inductive reactance does become significant and even predominates. My experience is that a driver which is called 4 Ohms typically measures ~3.2 to 3.5 Ohms DC resistance. I did say above resonance but I neglected to speak of the higher frequencies where inductive reactance comes into play.

No doubt the back EMF comes into play in controlling the cone motion however recently I have read stuff about driving speakers with a current source as opposed to the voltage source which most of our amplifiers are today. I’m not sure I buy into that as I remember the triode (low output resistance) vs pentode (current source) amp thing. This is of coarse from literature as I’m old but not THAT old. That impedance thing was of coarse before negative feedback was common. My Citation 5 with beam pentodes and feedback is pretty much a voltage source.

Remember I am an amateur speaker tinkerer.

Oh yes, I’m pretty sure we have the same laws of physics in California as you have in Clyde, Ohio.

Phil


josephazannieri wrote:
17 Aug 2019 05:28
Yo Phil and others discussing speaker characteristics:

I am going to have to get a little nitpickey with your statements here:
Coffee Phil wrote:
15 Aug 2019 22:46

These speakers are highly inefficient. The voice coil impedance (above resonance) is largely the DC resistance, hence the current through the coil and the force it exerts on the cone is proportional to the applied voltage. I think the bulk of the remaining force acting on the coil is the mass of the cone.
Actually, the speaker's voice coil impedance is not exclusively or even primarily resistive. It is a combination of resistance and inductance, and, at resonance, the self-generating back EMF caused by the speaker running amok. A given speaker with a given voice coil will have a constant inductance which is always there, much as any coil of wire will have a constant inductance. A speaker is a reactive load whose measured impedance varies according to the frequency applied to the voice coil.

A typical cone speaker will have an impedance peak at the resonance point. This occurs because at resonance, the speaker generates a back EMF caused by the motion of the cone. The proof is this, again from my demonstrations in Clyde, Ohio. If you take two similar cone speakers and hook the voice coils in parallel and push on the cone of one of the speakers, the cone of the other speaker will move about the same amount. This is because the voice coil of the speaker being pushed generates a current from being moved within the magnetic field of that speaker. At resonance, the speaker cone will vibrate not only because of the signal applied, but also because of the self-generated motion caused by the resonance. As the signal frequency rises above the resonant point, the measured impedance drops gradually until it reaches a low point, for most speakers somewhere around 400 Hz. This is because the speaker is no longer generating a back EMF, and the inductance of the voice coil, which may be as much as 2.2 mH for a typical 15 inch woofer, is not yet affecting the passage of current through the voice coil. For the example given, the woofer's impedance will start to rise at about 600 Hz, and may go as high as 25 or 30 ohms, reducing the current passing through the voice coil.

Now of course a smaller speaker with a smaller voice coil will not have as much inductance, and its impedance will not rise as much as a woofer's impedance will. This means that the voice coil with less inductance will not have as much of an impedance rise at high frequencies, and will operate better at high frequencies. While the current through the coil may be constant, the amount of motion will be affected at the low end by resonance, and at the high end by the impedance rise caused by voice coli inductance, which will reduce the current through the voice coil and reduce the motion of the cone. I am talking about speakers in free air, which behave differently from enclosed speakers. But that's another story.

Now an acoustic cutting head does not have a voice coil, but it does have a resonant point, so voice coil inductance will not affect its response to a tone of a certain frequency. But, the resonant point will give a bulge in the head's frequency response. Given the relatively small size of an acoustic cutting head, it is quite possible that the head's resonant point may occur in the frequency range that is being recorded.

Of course these rules apply in Clyde, Ohio, but I am pretty sure that they also apply in California. As a lawyer, I am glad to have the chance to criticize, and I thank you for your charitable indulgence of my pontificating.

And good luck from the old topic drifter, critic, and pontificator,

Joe Z.

Coffee Phil
vinyl addict
vinyl addict
United States of America
Posts: 5747
Joined: 20 Sep 2008 08:22
Location: California

Re: 78 noise

Post by Coffee Phil » 17 Aug 2019 18:40

Hi Joe,

Just a bit more on this subject creep to moving coil speakers. Just to be fairly sure that I’m not talking through my tailpipe or using disco-physics, I figured I should check out my assertion that sound pressure from a cone speaker is proportional to cone acceleration. Among what I found is this:

https://physics.stackexchange.com/quest ... -excursion


This gives some credibility to the notion of driving a speaker voice coil with a current source vs the very low output impedance amplifiers we now love.

It could be that the notion that amplifiers driving moving coil speakers should be voltage sources came during the time when power tetrodes were appearing in the early ‘30s. Tetrodes and pentodes offered more power sensitivity and were more efficient yet triodes sounded better. This was before negative feedback was typical in audio power amplifiers and voltage source drive to speakers had to be “honestly” gotten from triodes rather that those nasty pentodes.

Could it be that voltage source drive in and of itself was not the big deal but rather the pentodes were going into voltage saturation (with the resulting distortion) on the impedance peaks of the speaker?

Phil

Coffee Phil wrote:
17 Aug 2019 06:58
Hi Joe,

Nit pit away. Sometimes I need to be taken to task. I was a bit sloppy there. Indeed as we approach the upper range of the driver inductive reactance does become significant and even predominates. My experience is that a driver which is called 4 Ohms typically measures ~3.2 to 3.5 Ohms DC resistance. I did say above resonance but I neglected to speak of the higher frequencies where inductive reactance comes into play.

No doubt the back EMF comes into play in controlling the cone motion however recently I have read stuff about driving speakers with a current source as opposed to the voltage source which most of our amplifiers are today. I’m not sure I buy into that as I remember the triode (low output resistance) vs pentode (current source) amp thing. This is of coarse from literature as I’m old but not THAT old. That impedance thing was of coarse before negative feedback was common. My Citation 5 with beam pentodes and feedback is pretty much a voltage source.

Remember I am an amateur speaker tinkerer.

Oh yes, I’m pretty sure we have the same laws of physics in California as you have in Clyde, Ohio.

Phil


josephazannieri wrote:
17 Aug 2019 05:28
Yo Phil and others discussing speaker characteristics:

I am going to have to get a little nitpickey with your statements here:
Coffee Phil wrote:
15 Aug 2019 22:46

These speakers are highly inefficient. The voice coil impedance (above resonance) is largely the DC resistance, hence the current through the coil and the force it exerts on the cone is proportional to the applied voltage. I think the bulk of the remaining force acting on the coil is the mass of the cone.
Actually, the speaker's voice coil impedance is not exclusively or even primarily resistive. It is a combination of resistance and inductance, and, at resonance, the self-generating back EMF caused by the speaker running amok. A given speaker with a given voice coil will have a constant inductance which is always there, much as any coil of wire will have a constant inductance. A speaker is a reactive load whose measured impedance varies according to the frequency applied to the voice coil.

A typical cone speaker will have an impedance peak at the resonance point. This occurs because at resonance, the speaker generates a back EMF caused by the motion of the cone. The proof is this, again from my demonstrations in Clyde, Ohio. If you take two similar cone speakers and hook the voice coils in parallel and push on the cone of one of the speakers, the cone of the other speaker will move about the same amount. This is because the voice coil of the speaker being pushed generates a current from being moved within the magnetic field of that speaker. At resonance, the speaker cone will vibrate not only because of the signal applied, but also because of the self-generated motion caused by the resonance. As the signal frequency rises above the resonant point, the measured impedance drops gradually until it reaches a low point, for most speakers somewhere around 400 Hz. This is because the speaker is no longer generating a back EMF, and the inductance of the voice coil, which may be as much as 2.2 mH for a typical 15 inch woofer, is not yet affecting the passage of current through the voice coil. For the example given, the woofer's impedance will start to rise at about 600 Hz, and may go as high as 25 or 30 ohms, reducing the current passing through the voice coil.

Now of course a smaller speaker with a smaller voice coil will not have as much inductance, and its impedance will not rise as much as a woofer's impedance will. This means that the voice coil with less inductance will not have as much of an impedance rise at high frequencies, and will operate better at high frequencies. While the current through the coil may be constant, the amount of motion will be affected at the low end by resonance, and at the high end by the impedance rise caused by voice coli inductance, which will reduce the current through the voice coil and reduce the motion of the cone. I am talking about speakers in free air, which behave differently from enclosed speakers. But that's another story.

Now an acoustic cutting head does not have a voice coil, but it does have a resonant point, so voice coil inductance will not affect its response to a tone of a certain frequency. But, the resonant point will give a bulge in the head's frequency response. Given the relatively small size of an acoustic cutting head, it is quite possible that the head's resonant point may occur in the frequency range that is being recorded.

Of course these rules apply in Clyde, Ohio, but I am pretty sure that they also apply in California. As a lawyer, I am glad to have the chance to criticize, and I thank you for your charitable indulgence of my pontificating.

And good luck from the old topic drifter, critic, and pontificator,

Joe Z.

H. callahan
senior member
senior member
Posts: 840
Joined: 25 Feb 2013 17:59

Re: 78 noise

Post by H. callahan » 18 Aug 2019 05:28

Coffee Phil wrote:
17 Aug 2019 05:03
Hi H. Callahan,

OK, I guess I’ll have to wait for someone to translate your stuff from discotheque physics to Newtonian Physics so I can understand it.

Pity that the EE program at my university required that clearly outmoded type of physics instead of letting us dance our bell bottoms off, learning your clearly superior science.

Phil
Nice to see they also taught you how not to be arrogant.

I never asked someone to dance his buttom off nor claimed my "physics" being superior.

You know what Newton did when he wanted to create his law of gravity? He went on the interent, into a forum and did demand physical formulas. He did not observe nature, he didn´t do experiments and did not conclude from experiments how gravity must be - and then did not make a formula describing his findings.

All you have is theories. You have the theory that its ok to play shellacs with diamond styli but you can´t back that up with physics. You just assume that all the people who did play shellacs with steel needles for decades were wrong.
Also you just have the theory that acoustics are constant amplitude, though you state yourself that there is contradicting literature.

But unless someone is able to provide you physical proof and even papers from that period, you´re not willing to reconsider your theories - but make fun of those who cannot explain it the physical way.

You may be good at theorizing - i can´t judge as i´m not that capable of physics - but that seems to be about it.

My patience is at its end -

Have a nice day.

josephazannieri
long player
long player
United States of America
Posts: 2906
Joined: 20 Apr 2008 06:01
Location: Norwalk,Ohio,USA

Re: 78 noise

Post by josephazannieri » 18 Aug 2019 06:29

Yo Phil and other loudspeaker analysts:

The damping factor of an amplifier is controlled by the impedance of the output stage. Not the impedance of the speaker, but the actual impedance of the output stage, as seen from the viewpoint of the speaker. When you hook a speaker across the secondary of an output transformer, the speaker "sees" the impedance of the secondary. Assuming the impedance of the secondary of the output transformer is 1 ohm, and the speaker is rated at 8 ohms, the damping factor is 8. Transistor amplifiers have much higher damping factors because the impedance of their output stages can be way less than 1 ohm. The higher the damping factor, the better the amp controls the speaker, particularly at resonance where the back EMF coming off the speaker is shorted out by the low impedance of the output stage. When measuring speaker impedance curves using a meter, you need to put a resistor in series with the speaker voice coil to make sure that your signal generator is not shorting the speaker voice coil. Then you measure impedance just across the voice coil. There are additional items that control the damping of an amplifier, such as resistance of the wires hooking the speaker to the amp, and if there is a crossover net, the resistance of any inductors in series with the speaker.

Of course you are not measuring impedance directly, you are measuring AC volts across the voice coil, but you can set the voltage so that it will read directly in ohms. There is a procedure for measuring impedance which is found in Testing Loudspeakers By Joseph D'Appolito, Ph D. (Audio Amatuer Press, 1998) at pp. 9-35. D'Appolito describes an experiment where you push on the cone of a woofer with the voice coil open. The woofer will move easily. Then you short the speaker terminals with a piece of wire. Now the speaker will give great resistance to the push that previously was so easy. This shows that the closer the impedance the speaker "sees" is to a dead short, the better the amp will control the loudspeaker.

But returning for a moment to your discussions with H Callahan, he does have a point. Sound always moves at a certain speed, so all sonic endeavors are constant velocity. Even in my experience years ago acting as sound man for Passion Play (a Jethro Tull tribute band), all sounds high and low, even the pants-flapping 40 Hz low E on the Fender bass, left the bandstand at about 1130 feet per second, or a perhaps a little faster when we were playing in Cleveland (nearer to sea level than the local clubs). But the acceleration of a cone will still be determined by the magnetic and mechanical characteristics of the loudspeaker.

And getting back to the mechanical characteristics of an acoustic recording diaphragm, the speed of its acceleration from a sound hitting it will be determined its mechanical characteristics. So I guess you and I are somewhat agreed, though expressing ourselves differently.

So, good luck to all the topic drifters who are still with us, from the old speaker cobbler,

Joe Z.

Coffee Phil
vinyl addict
vinyl addict
United States of America
Posts: 5747
Joined: 20 Sep 2008 08:22
Location: California

Re: 78 noise

Post by Coffee Phil » 18 Aug 2019 08:32

Hi Joe,

Damping factor is indeed a parameter of the amplifier. It is as you said the intended speaker impedance divided by the source impedance of the amplifier. The source impedance of the amplifier can be made quite low by the use of negative VOLTAGE feedback. In solid state amps this number can be very low as greater negative feedback can be applied with stability since you don’t have the phase shift of the output transformer eating up phase margin.

Neglecting feedback for the moment the output transformer within its useful range reflects the plate resistance of the output tube(s) by the turns ratio squared. Triodes have fairly low plate resistance while pentodes have very high plate resistance. Back in the day before feedback was common what we call damping factor today had to be gotten “honestly” by using triodes in the output stage.

Now even more on this: The model of the speaker has a fair bit of resistance in series. Like I said in the other post ~ 3.2 Ohms in a four Ohm speaker. That resistance is series with the amplifier’s output resistance so the “effective” damping factor is going to be scarce more than unity even if the amplifier boasts a damping factor of 1000. This suggests that for amplifiers with damping factors in excess of ten, boasting of huge damping factors is silly.

The notion that huge damping factor is a good thing does not seem to be universally accepted. I have recently read articles suggested that driving speakers with a current source might be a good thing. I’ll give my thoughts on that in a bit. You may also remember some Fisher tube amps having a pot which could mix voltage feedback and current feedback. Current negative feedback will raise the source impedance of the amp. I’m not sure that the term damping factor was used in the ‘50s but those Fishers were boasting variable damping factor. Clearly in the fifties the folks at Fisher were questioning the notion that the more damping factor, the better.

Getting back to the notion of current source drive for speakers. I’m still pretty agnostic on this but consider the triode / pentode thing in the thirties. I get the idea that most folks preferred the sound of triode amps. Is this because triodes were inherently better or did the current source output of pentode amps give huge voltage swings at the impedance peaks of the speaker resulting in distortion. If this current source amplifier thing goes anywhere I think that sufficient voltage swing capability will have to be assured to prevent clipping at the speaker impedance peaks.

Phil
josephazannieri wrote:
18 Aug 2019 06:29
Yo Phil and other loudspeaker analysts:

The damping factor of an amplifier is controlled by the impedance of the output stage. Not the impedance of the speaker, but the actual impedance of the output stage, as seen from the viewpoint of the speaker. When you hook a speaker across the secondary of an output transformer, the speaker "sees" the impedance of the secondary. Assuming the impedance of the secondary of the output transformer is 1 ohm, and the speaker is rated at 8 ohms, the damping factor is 8. Transistor amplifiers have much higher damping factors because the impedance of their output stages can be way less than 1 ohm. The higher the damping factor, the better the amp controls the speaker, particularly at resonance where the back EMF coming off the speaker is shorted out by the low impedance of the output stage. When measuring speaker impedance curves using a meter, you need to put a resistor in series with the speaker voice coil to make sure that your signal generator is not shorting the speaker voice coil. Then you measure impedance just across the voice coil. There are additional items that control the damping of an amplifier, such as resistance of the wires hooking the speaker to the amp, and if there is a crossover net, the resistance of any inductors in series with the speaker.

Of course you are not measuring impedance directly, you are measuring AC volts across the voice coil, but you can set the voltage so that it will read directly in ohms. There is a procedure for measuring impedance which is found in Testing Loudspeakers By Joseph D'Appolito, Ph D. (Audio Amatuer Press, 1998) at pp. 9-35. D'Appolito describes an experiment where you push on the cone of a woofer with the voice coil open. The woofer will move easily. Then you short the speaker terminals with a piece of wire. Now the speaker will give great resistance to the push that previously was so easy. This shows that the closer the impedance the speaker "sees" is to a dead short, the better the amp will control the loudspeaker.

But returning for a moment to your discussions with H Callahan, he does have a point. Sound always moves at a certain speed, so all sonic endeavors are constant velocity. Even in my experience years ago acting as sound man for Passion Play (a Jethro Tull tribute band), all sounds high and low, even the pants-flapping 40 Hz low E on the Fender bass, left the bandstand at about 1130 feet per second, or a perhaps a little faster when we were playing in Cleveland (nearer to sea level than the local clubs). But the acceleration of a cone will still be determined by the magnetic and mechanical characteristics of the loudspeaker.

And getting back to the mechanical characteristics of an acoustic recording diaphragm, the speed of its acceleration from a sound hitting it will be determined its mechanical characteristics. So I guess you and I are somewhat agreed, though expressing ourselves differently.

So, good luck to all the topic drifters who are still with us, from the old speaker cobbler,

Joe Z.

Coffee Phil
vinyl addict
vinyl addict
United States of America
Posts: 5747
Joined: 20 Sep 2008 08:22
Location: California

Re: 78 noise

Post by Coffee Phil » 19 Aug 2019 00:16

Sorry to stay off topic so long, but as I like to maintain credibility, I figured I should find the article which I read. It was in EDN and the link is below:

https://www.edn.com/design/consumer/442 ... tage-drive

There were a couple reader comments which I’ll comment on here:

One comment mentioned that if the speaker became disconnected the output voltage from the amp would get scary and might be hazardous. No doubt, but as a sometimes live on the edge guy I did not think of that. I was more concerned that the amp would voltage rail on the speaker impedance peaks and distort like the pentodes of the thirties.

Another commenter suggested current drive did not become the standard because Analog EEs could not figure out how to do it. As such a person I found that insulting and I would welcome the challenge. I believe the reason may be that when negative voltage feedback became common it seemed to mitigate the pentode nastiness and it was assumed that the reason was the more voltage source like output of the amplifiers was the reason. I now speculate that the pentodes could have been clipping on the speaker impedance peaks. That is why I said these proposed amps would need to have sufficient voltage headroom to prevent that. Of course this huge voltage headroom would make them more scary if the speaker is disconnected.

Phil



Coffee Phil wrote:
18 Aug 2019 08:32
Hi Joe,

Damping factor is indeed a parameter of the amplifier. It is as you said the intended speaker impedance divided by the source impedance of the amplifier. The source impedance of the amplifier can be made quite low by the use of negative VOLTAGE feedback. In solid state amps this number can be very low as greater negative feedback can be applied with stability since you don’t have the phase shift of the output transformer eating up phase margin.

Neglecting feedback for the moment the output transformer within its useful range reflects the plate resistance of the output tube(s) by the turns ratio squared. Triodes have fairly low plate resistance while pentodes have very high plate resistance. Back in the day before feedback was common what we call damping factor today had to be gotten “honestly” by using triodes in the output stage.

Now even more on this: The model of the speaker has a fair bit of resistance in series. Like I said in the other post ~ 3.2 Ohms in a four Ohm speaker. That resistance is series with the amplifier’s output resistance so the “effective” damping factor is going to be scarce more than unity even if the amplifier boasts a damping factor of 1000. This suggests that for amplifiers with damping factors in excess of ten, boasting of huge damping factors is silly.

The notion that huge damping factor is a good thing does not seem to be universally accepted. I have recently read articles suggested that driving speakers with a current source might be a good thing. I’ll give my thoughts on that in a bit. You may also remember some Fisher tube amps having a pot which could mix voltage feedback and current feedback. Current negative feedback will raise the source impedance of the amp. I’m not sure that the term damping factor was used in the ‘50s but those Fishers were boasting variable damping factor. Clearly in the fifties the folks at Fisher were questioning the notion that the more damping factor, the better.

Getting back to the notion of current source drive for speakers. I’m still pretty agnostic on this but consider the triode / pentode thing in the thirties. I get the idea that most folks preferred the sound of triode amps. Is this because triodes were inherently better or did the current source output of pentode amps give huge voltage swings at the impedance peaks of the speaker resulting in distortion. If this current source amplifier thing goes anywhere I think that sufficient voltage swing capability will have to be assured to prevent clipping at the speaker impedance peaks.

Phil
josephazannieri wrote:
18 Aug 2019 06:29
Yo Phil and other loudspeaker analysts:

The damping factor of an amplifier is controlled by the impedance of the output stage. Not the impedance of the speaker, but the actual impedance of the output stage, as seen from the viewpoint of the speaker. When you hook a speaker across the secondary of an output transformer, the speaker "sees" the impedance of the secondary. Assuming the impedance of the secondary of the output transformer is 1 ohm, and the speaker is rated at 8 ohms, the damping factor is 8. Transistor amplifiers have much higher damping factors because the impedance of their output stages can be way less than 1 ohm. The higher the damping factor, the better the amp controls the speaker, particularly at resonance where the back EMF coming off the speaker is shorted out by the low impedance of the output stage. When measuring speaker impedance curves using a meter, you need to put a resistor in series with the speaker voice coil to make sure that your signal generator is not shorting the speaker voice coil. Then you measure impedance just across the voice coil. There are additional items that control the damping of an amplifier, such as resistance of the wires hooking the speaker to the amp, and if there is a crossover net, the resistance of any inductors in series with the speaker.

Of course you are not measuring impedance directly, you are measuring AC volts across the voice coil, but you can set the voltage so that it will read directly in ohms. There is a procedure for measuring impedance which is found in Testing Loudspeakers By Joseph D'Appolito, Ph D. (Audio Amatuer Press, 1998) at pp. 9-35. D'Appolito describes an experiment where you push on the cone of a woofer with the voice coil open. The woofer will move easily. Then you short the speaker terminals with a piece of wire. Now the speaker will give great resistance to the push that previously was so easy. This shows that the closer the impedance the speaker "sees" is to a dead short, the better the amp will control the loudspeaker.

But returning for a moment to your discussions with H Callahan, he does have a point. Sound always moves at a certain speed, so all sonic endeavors are constant velocity. Even in my experience years ago acting as sound man for Passion Play (a Jethro Tull tribute band), all sounds high and low, even the pants-flapping 40 Hz low E on the Fender bass, left the bandstand at about 1130 feet per second, or a perhaps a little faster when we were playing in Cleveland (nearer to sea level than the local clubs). But the acceleration of a cone will still be determined by the magnetic and mechanical characteristics of the loudspeaker.

And getting back to the mechanical characteristics of an acoustic recording diaphragm, the speed of its acceleration from a sound hitting it will be determined its mechanical characteristics. So I guess you and I are somewhat agreed, though expressing ourselves differently.

So, good luck to all the topic drifters who are still with us, from the old speaker cobbler,

Joe Z.

Coffee Phil
vinyl addict
vinyl addict
United States of America
Posts: 5747
Joined: 20 Sep 2008 08:22
Location: California

Re: 78 noise

Post by Coffee Phil » 19 Aug 2019 17:47

Hi H. callahan,

Sorry about the arrogance! Back in my working career days I worked with a lot of people who were much smarter than me and it served to keep my arrogance in check.

Now to the diamond stylus thing: I have never said that it is "wrong" to play a shellac record with a diamond needle. I played many records that way myself when I was fortunate enough to possess such a player. Now about my Admiral 78 changer to which you referred. It is recommended that steel needles be changed after one play or at the most two plays. It would be a good trick to swap needles mid change cycle. Clearly a steel needle is not the best choice for this machine. For that reason I use a jewel tipped stylus in that machine.

Your suggestion that I am unwilling to change my beliefs with the acquisition of new credible information is simply not so. I try to be a life long learner and I do read and even perform experiments to gain understanding of the world around me.

Now knowledge of if acoustic records are cut "constant amplitude", "constant velocity", or some combination of both is not likely to change the world, but I find it interesting and have performed experiments to get a clue about the subject. Now admittedly my experiments are not as precise as I would like but I believe they are better than none at all. Ideally I would secure a period record cutter and period wax blank and put them into an anechoic chamber with a sound source which could produce a frequency sweep with constant sound pressure. The resulting wax disc would then be viewed under a microscope to observe the change in modulation over the frequency sweep. Clearly if I would go to that effort and expense for that pursuit my wife would have to consider having me committed to some institution. I have performed admittedly less rigorous experiments which at least partially satisfies my need for knowledge on this subject. I have an assortment of acoustically recorded recorded recorded both vertical and lateral and the ability to play them back amplitude responding, velocity responding, or with an assortment of different EQ curves. I have posted the result of those tests. Make of them what you will.

I am now weary of this subject and have no more to say about it unless someone has additional credible historical documents, more experimental results, or some sort of modeling of early 20 century acoustic disc or cylinder recorders.

Phil



H. callahan wrote:
18 Aug 2019 05:28
Coffee Phil wrote:
17 Aug 2019 05:03
Hi H. Callahan,

OK, I guess I’ll have to wait for someone to translate your stuff from discotheque physics to Newtonian Physics so I can understand it.

Pity that the EE program at my university required that clearly outmoded type of physics instead of letting us dance our bell bottoms off, learning your clearly superior science.

Phil
Nice to see they also taught you how not to be arrogant.

I never asked someone to dance his buttom off nor claimed my "physics" being superior.

You know what Newton did when he wanted to create his law of gravity? He went on the interent, into a forum and did demand physical formulas. He did not observe nature, he didn´t do experiments and did not conclude from experiments how gravity must be - and then did not make a formula describing his findings.

All you have is theories. You have the theory that its ok to play shellacs with diamond styli but you can´t back that up with physics. You just assume that all the people who did play shellacs with steel needles for decades were wrong.
Also you just have the theory that acoustics are constant amplitude, though you state yourself that there is contradicting literature.

But unless someone is able to provide you physical proof and even papers from that period, you´re not willing to reconsider your theories - but make fun of those who cannot explain it the physical way.

You may be good at theorizing - i can´t judge as i´m not that capable of physics - but that seems to be about it.

My patience is at its end -

Have a nice day.

Coffee Phil
vinyl addict
vinyl addict
United States of America
Posts: 5747
Joined: 20 Sep 2008 08:22
Location: California

Re: 78 noise

Post by Coffee Phil » 23 Aug 2019 02:29

Hi Bob,

This book is defiantly worth a read! I want a copy but a search gave me “sticker shock”. North of $100 is the best I could find and on Amazon it is more than $600.

I may convince myself to pop for the ~ $100 copy, but for now I’ll struggle on line printing pages where I have to go rapidly from text to illustrations.

Thanks again for finding this.

Phil
Bob Dillon wrote:
14 Aug 2019 21:42
Modern Gramophones and Electrical Reproducers (pub. 1929) - page 33

In discussing the recorder, claims are for :

Frequencies below 300 hz - constant amplitude
300 hz to 5000 hz - constant velocity
Beyond 5000 hz - constant acceleration

Table of contents : https://archive.org/details/ModernGramo ... reproducer The entire chapter 1 at least is worth a read
Jump to page 24 & 25 (in the book itself).

Bob Dillon
senior member
senior member
United States of America
Posts: 566
Joined: 03 Mar 2019 20:22

Re: 78 noise

Post by Bob Dillon » 23 Aug 2019 05:35

If speaking on behalf of the authors off the book, you're welcome !

Coffee Phil
vinyl addict
vinyl addict
United States of America
Posts: 5747
Joined: 20 Sep 2008 08:22
Location: California

Re: 78 noise

Post by Coffee Phil » 26 Aug 2019 03:15

I have been spending some time in:

Modern Gramophones and Electrical Reproducers by P. Wilson, MA and G.W. Webb

The above book was linked by Bob.

I now have some thoughts on this constant amplitude, vs. constant velocity, vs. constant amplitude, vs. constant acceleration thing for acoustic cut records.

On page 33 there is a graph which illustrates the response of a contemporary electrical cutting head as equalized in cutting records in the late ‘20s and the response of a typical acoustic cutting head. The graph is below:

https://www.vinylengine.com/turntable_f ... 488/medium

Curve b is the 1920's electrical setup. It is described as approximately constant amplitude below 300 Hz. From 300 Hz to 4000 Hz is described as constant velocity. Above 4000 Hz the curve is described as constant acceleration. They do not say how the data was measured but it looks like what we would expect from a non–equalized velocity responding pickup.

Curve a is of a typical acoustic cutter. I would assume the disc is read with the same pickup as used for curve b. There is little which can be described as constant amplitude, constant velocity, or constant acceleration. The low frequency slope is ~ double the slope of the electrical cutter. From ~400 Hz to 700 Hz it is close to constant velocity. From 700 Hz to ~ 1500 Hz it is again a greater slope than constant amplitude. After the peak at ~1500 Hz the negative slope looks like -18 db /octave. That is way beyond the -6 dB octave which we would expect for constant acceleration.

I find that playing acoustic records back velocity responding lacks bass and sounds overly bright. If I play them back amplitude responding they sound “believable” but lack intelligibility. Using a bass turn of 300 Hz and a treble cut of 3400 Hz is about the best compromise which I have found so far. Looking at the graph above would suggest arranging for a 12 dB / octave bass turn would be better but I think rumble would get out of hand. I can live with the low frequency response I have achieved so far. Looking at the resonant peak at ~ 1500 Hz makes me think following my phono stage with a graphic equalizer which can knock that down and bump up the response at ~ 3000 Hz may give me some improvement. A made to purpose equalizer for that acoustic curve may not be as useful. I believe that the book did say the curve was "typical" and that in the day different soundboxes could be in the user’s kit to optimize different records. I may start looking for an old graphic equalizer on eBay to use for this purpose.

Phil

PS: You may note the vertical axis of the graph is T.U. That stands for transmission unit. It is 10 log of the power ratio. Sound familiar? It should, as in 1928 AT&T renamed the T.U. decibel.

Bob Dillon
senior member
senior member
United States of America
Posts: 566
Joined: 03 Mar 2019 20:22

Re: 78 noise

Post by Bob Dillon » 26 Aug 2019 20:07

Interesting.

In making my own digital transfers of acoustical records, I've sometimes found it pleasing to give a very small nudge to the area around 3,000 Hz too, using audio software. These days, with my Puffin phono stage that I'm currently using, I like the 500Hz turnover rather than 300, which often sounds too thin to me.

I can't remember if I linked to this already, but in case I haven't, this is an excellent article on the differing properties of various reproducers/soundboxes. https://www.antiquephono.org/sound-subs ... j-wakeman/

Coffee Phil
vinyl addict
vinyl addict
United States of America
Posts: 5747
Joined: 20 Sep 2008 08:22
Location: California

Re: 78 noise

Post by Coffee Phil » 27 Aug 2019 00:11

Hi Bob,

I had not seen that article before. Thanks for linking it. It will provide hours of interesting reading for those of us who like playing acoustic records.

Phil
Bob Dillon wrote:
26 Aug 2019 20:07
Interesting.

In making my own digital transfers of acoustical records, I've sometimes found it pleasing to give a very small nudge to the area around 3,000 Hz too, using audio software. These days, with my Puffin phono stage that I'm currently using, I like the 500Hz turnover rather than 300, which often sounds too thin to me.

I can't remember if I linked to this already, but in case I haven't, this is an excellent article on the differing properties of various reproducers/soundboxes. https://www.antiquephono.org/sound-subs ... j-wakeman/

Coffee Phil
vinyl addict
vinyl addict
United States of America
Posts: 5747
Joined: 20 Sep 2008 08:22
Location: California

Re: 78 noise

Post by Coffee Phil » 28 Aug 2019 23:03

Here is a link to a copy of a record in my collection: http://www.loc.gov/jukebox/recordings/d ... 4052/[url][/url]

It is a "politically incorrect" version of Dvorak's Humoresqe #7 as a lullaby. It is kind of special to me as my mother used to sing a version of the same work of Dvorak to me as a lullaby which I'm sure was "politically correct" when I was a baby. So far this is the only version of the work which I can find on record as a lullaby.

It dates from 1915 and is of coarse an acoustic recording. The Library Of Congress link sounds similar to my record played back velocity responding. I prefer it with a bass turn of 300 Hz and a treble cut of 3180 Hz, but I still think there is more that can be had. I had mentioned that attempting to make a dedicated playback EQ to work with the curve which I posted above may not be the best as the curve is "typical" and the lumps and bumps may vary from record to record. I thought that a graphic equalizer following my mono stage may be more useful. I have recently found other mentions of that concept on the internet. I am now watching some 1/3 octave equalizers on eBay. It will have to be cheap as other than this I'm not even a big fan of tone controls so if it doesn't do magic for these acoustic records it will go back on eBay.

Phil

Post Reply