Real world loudspeaker current demands
AES E-Library: Peak Current Requirement of Commercial Loudspeaker Systems
Thanks, that's interesting enough without paying the $20 fee. I see it only applies to multiway systems - a good enough reason to be using chip amps in active set-ups.
AES E-Library: Peak Current Requirement of Commercial Loudspeaker Systems
Thanks, that's interesting enough without paying the $20 fee. I see it only applies to multiway systems - a good enough reason to be using chip amps in active set-ups.
It does not only apply to multi-way speakers. It is worse in multi-way speakers, because there are more components adding up. That document talks about 6,6 times for a commercial (multi-way) system. AndrewT mentioned 3 times, which is a rule-of thumb figure for inductive motors in general, not especially for speakers.
All inductive motors have an inrush current, when they start or change their speed or direction. A speaker membrane does those things all the time, so the inrush current becomes effective pretty often. The impedance curve of a speaker shows the average current consumption per voltage. It swamps those inrush currents.
Those current peaks bring up two questions. Whether it is worthwhile to oversize the power supply accordingly. And when it comes to chipamps, whether those short peaks will trigger the SPiKE system and it takes more ICs in parallel than the current derived from the nominal speaker impedance suggests to keep SPiKe from triggering.
All inductive motors have an inrush current, when they start or change their speed or direction. A speaker membrane does those things all the time, so the inrush current becomes effective pretty often. The impedance curve of a speaker shows the average current consumption per voltage. It swamps those inrush currents.
Those current peaks bring up two questions. Whether it is worthwhile to oversize the power supply accordingly. And when it comes to chipamps, whether those short peaks will trigger the SPiKE system and it takes more ICs in parallel than the current derived from the nominal speaker impedance suggests to keep SPiKe from triggering.
Do inductive motors have inrush currents?
Sorry, I didn't make my writing clear - the 'it' in my sentence was the 6.6X multiplier in the paper you cited. I don't recall the synopsis talking about active systems - if you've read the paper perhaps you could confirm whether it does or not?
Do either you or AndrewT have a reference for this 3X figure? John Borwick cites a 3X figure for a multiway speaker in his 'Loudspeaker and Headphone Handbook' but doesn't back this up with any references or experiments.
I'm taking it that the term 'inductive motor' means a motor with an inductive impedance when looking into its power terminals. In which case it cannot have an inrush current and also be inductive. Inrush currents belong only in the realm of capacitors, not inductors. By definition an inductor can't have its current changed instantaneously. Applying a step voltage waveform to anything inductive results in a current ramp, not an inrush current.
Which makes it unhelpful to call it an 'inrush current' at all. That term applies to things subject to a step change in voltage which is impossible in a hi-fi system (except perhaps at switch on/switch off).
Basic misunderstanding - there are no inrush currents under these conditions, since when measuring the impedance of a speaker, only a sinewave is used, which always has a very gradual rate of change, never a step change.
Oh dear, it looks as though you've got yourself a mini-tutorial
Hint - if you don't like tutorials, keep your misunderstandings of circuit theory to yourself
Sorry, I didn't make my writing clear - the 'it' in my sentence was the 6.6X multiplier in the paper you cited. I don't recall the synopsis talking about active systems - if you've read the paper perhaps you could confirm whether it does or not?
Do either you or AndrewT have a reference for this 3X figure? John Borwick cites a 3X figure for a multiway speaker in his 'Loudspeaker and Headphone Handbook' but doesn't back this up with any references or experiments.
I'm taking it that the term 'inductive motor' means a motor with an inductive impedance when looking into its power terminals. In which case it cannot have an inrush current and also be inductive. Inrush currents belong only in the realm of capacitors, not inductors. By definition an inductor can't have its current changed instantaneously. Applying a step voltage waveform to anything inductive results in a current ramp, not an inrush current.
Which makes it unhelpful to call it an 'inrush current' at all. That term applies to things subject to a step change in voltage which is impossible in a hi-fi system (except perhaps at switch on/switch off).
Basic misunderstanding - there are no inrush currents under these conditions, since when measuring the impedance of a speaker, only a sinewave is used, which always has a very gradual rate of change, never a step change.
Oh dear, it looks as though you've got yourself a mini-tutorial
Hint - if you don't like tutorials, keep your misunderstandings of circuit theory to yourself
AN-1192 'Good results'
pacificblue has pointed out to us on this thread that the writers of AN-1192 say the following:
1% gain setting resistors (Ri and Rf) will give good results but it is recommended 0.1% tolerance resistors be used for setting the gain of each op amp for closer matched gain and equal output current and power dissipation.
Since I'd already simulated the circuit with the 0.1% Ri and Rf, I decided to take a quick look at what happens when, in place of the recommended 0.1% tolerance resistors, we try 1% types. After all, they do say they'll give 'good results'. So perhaps somebody will be tempted to try with 1% types. As a result of my simulation, the best I can say is don't bother.
If you recall, the earlier simulation gave high frequency (>1kHz) currents from the two amps as 3.4A and 2.3A, corresponding to equivalent impedances of 6.7 and 9.8ohms - a fair match.
With 1% types at worst case the currents turn out to be 7.2A and 1.6A. At first I thought there must be some mistake as these don't add up to the output current (5.7A). The truth is, the current from the lower gain amplifier has reversed its phase - its now pulling current out of the load rather than pushing it in. So now the matching is so bad that the parallel combination is considerably worse than running with a single amplifier.
One possible conclusion is that the 'good results' the writers were speaking about is the financial results of their company - they get to sell two amplifier chips in place of one. No, that's being too cynical.... If anyone would like to see what the plots look like, drop me a message.
pacificblue has pointed out to us on this thread that the writers of AN-1192 say the following:
1% gain setting resistors (Ri and Rf) will give good results but it is recommended 0.1% tolerance resistors be used for setting the gain of each op amp for closer matched gain and equal output current and power dissipation.
Since I'd already simulated the circuit with the 0.1% Ri and Rf, I decided to take a quick look at what happens when, in place of the recommended 0.1% tolerance resistors, we try 1% types. After all, they do say they'll give 'good results'. So perhaps somebody will be tempted to try with 1% types. As a result of my simulation, the best I can say is don't bother.
If you recall, the earlier simulation gave high frequency (>1kHz) currents from the two amps as 3.4A and 2.3A, corresponding to equivalent impedances of 6.7 and 9.8ohms - a fair match.
With 1% types at worst case the currents turn out to be 7.2A and 1.6A. At first I thought there must be some mistake as these don't add up to the output current (5.7A). The truth is, the current from the lower gain amplifier has reversed its phase - its now pulling current out of the load rather than pushing it in. So now the matching is so bad that the parallel combination is considerably worse than running with a single amplifier.
One possible conclusion is that the 'good results' the writers were speaking about is the financial results of their company - they get to sell two amplifier chips in place of one. No, that's being too cynical....
If anyone would like to see what the plots look like, drop me a message.Well, I looked up the translation for the German word Einschaltstrom. The dictionary gave a choice of 'inrush current' and 'starting current', both from the realm of electrical engineering without mentioning that inrush current was restricted to capacitors. You may not have noticed the German flag at the left of my posts that indicates I am not a native speaker of your language and may not have the grasp of every last subtlety in English the way you have. With average or higher intelligence any reader should be able to get, what I was explaining. And if you think that inductive motors do not draw higher currents, when they accelerate, than when they idle or decelerate, you may not have half as much reason for your arrogant behaviour as you think.
Oh, and your above post in its entirety fits perfectly into that dictionary's description of 'nit-picking'.
Inrush currents and start up currents.
Well 'starting current' I've never heard applied to speakers either - the term simply does not apply in this realm. Of course the cone, voice coil and other moving parts store energy - perhaps that's what you're getting at?
This has little or nothing to do with your grasp of English, which is perfectly adequate for you to communicate your misunderstandings - whether in motor circuits or opamp circuits. It has everything to do with your tenuous grasp of electronic and electrical engineering. I had indeed noticed the German flag on your posts.
Cute So your (unfounded) argument is then that because I didn't get what you were explaining, I'm therefore below average intelligence
Of course they draw higher currents during acceleration than when they're rotating at constant angular velocity otherwise they'd be violating energy conservation. How does this apply though to a loudspeaker cone - which, when oscillating sinusoidally - is constantly either accelerating or decelerating?
Do enlighten the rest of us of which dictionary and what its description actually says.
Well 'starting current' I've never heard applied to speakers either - the term simply does not apply in this realm. Of course the cone, voice coil and other moving parts store energy - perhaps that's what you're getting at?
This has little or nothing to do with your grasp of English, which is perfectly adequate for you to communicate your misunderstandings - whether in motor circuits or opamp circuits. It has everything to do with your tenuous grasp of electronic and electrical engineering. I had indeed noticed the German flag on your posts.
Cute
So your (unfounded) argument is then that because I didn't get what you were explaining, I'm therefore below average intelligence Of course they draw higher currents during acceleration than when they're rotating at constant angular velocity otherwise they'd be violating energy conservation. How does this apply though to a loudspeaker cone - which, when oscillating sinusoidally - is constantly either accelerating or decelerating?
Do enlighten the rest of us of which dictionary and what its description actually says.
Remedy for LF gain imbalance
For those still following the AN-1192 saga, the simplest solution to the LF gain imbalance in AN-1192 figure6 is to increase the value of Ci - from 68uF to 470uF. The big lump in the plot at 4Hz then goes away - see the new output plots. The vertical scales have been kept the same as before. There's a barely noticeable rise now, peaking just below 2Hz, nothing much to worry about.
For those still following the AN-1192 saga, the simplest solution to the LF gain imbalance in AN-1192 figure6 is to increase the value of Ci - from 68uF to 470uF. The big lump in the plot at 4Hz then goes away - see the new output plots. The vertical scales have been kept the same as before. There's a barely noticeable rise now, peaking just below 2Hz, nothing much to worry about.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.