If you have a DVC woofer and use one VC to drive the woofer(amp signal) and the other to sense the back-EMF you are correct, and are good to go for velocity based MFB. Well, with the understanding that there will be an erroneous signal added to the back-EMF from mutual coupling with the driven coil which increases in amplitude with increasing frequency.
But, if you are sensing current thru the same VC you are driving with the amplifier, you can’t ignore the amp signal because it is larger in magnitude than the back-EMF signal you are trying to sense. In fact, it becomes orders of magnitude larger than the back-EMF as you go down in frequency below resonance. For a woofer in a sealed box, the back-EMF magnitude falls -6dB/octave below resonance. Look at your plot of current sensing voltage in Post#292, it is flat below resonance because the voltage from the amplifier is the dominate current flowing thru the woofer and it is constant in amplitude. The only way to extract the back-EMF from a single VC woofer is to use a bridge circuit to null out the dominate amplifier contribution to the sensed current. In theory, what you are left with is the back-EMF. In practice, the lossy inductance contribution to the woofer impedance makes it impossible to completely balance the bridge and extract a clean back-EMF signal above resonance. Below resonance, a simple resistive bridge should suffice.
No. The example in the patent is for a sealed box.
One set of measurements is in room response, the other is from an anechoic chamber.
Last edited:
1. Correct.
2. Partial credit. There is no acceleration, but don't stop there. Go ahead and attach an accelerometer to the VC to provide the acceleration feedback. Now, as in 1., corrective feedback will be provided to try to get the VC moving.
The unanswered bonus question:
In this thought experiment we have...
1) a velocity sensor providing corrective feedback
2) an accelerometer sensor providing corrective feedback
What is the current sensor sensing as far as VC motion and what sort of corrective feedback is it providing if any?
Ben, the back-EMF is there absolutely, as you state, but it is very small compared to the current sensed by the resistor.. So this configuration mainly reacts to changes in impedance of the speaker connected, for they are dominant.
So, any kind of velocity sensing seems to have a use limited to sub-woofers.
It is still implemented today in Yamaha YST products at impressive low cost.
It is a commercial success since 25 years.
The peak of impedance at resonance results from back-EMF and does not reflect the existence of the reactive components which are used in modelisation.
Isn't that how feedback works? Was Stahl and all the others using VC feedback confused? Somebody (maybe me) is confused.
While the bridge approach magnifies the error component of the feedback signal, I think the main purpose of using it (aside from Werner's patent lawyer at RCA, the easiest mention I know of) is to balance those other pesky impedance gremlins by modelling them in the contra-leg.
It isn't much trouble to use a Wheatstone Bridge and I did so. The series resistor in the leg needs to be properly sized but the components in the contra-legs are the usual quarter-watt pieces (hint: use a pot to achieve the balance you like).
Anybody who wants to experiment, just have a peek at #292 and give a series resistor a try.
(forr - thanks for your helpful posts.)
Ben
Last edited:
Yes, Yamaha YST loudspeakers gives good performance, but I think they use something different from Stahl method. AFAIK Yamaha didn't bought Stahl's patent.
Katsuo Nagi from Yamaha refers twice to Stahl's papers in his US patent # 4,908,870 (March 1990).
Back-EMF is a voltage due to the voice coil velocity in the magnetic gap. It reflects the mechanical phenomenon of resonance and is translated by a peak into the impedance curve. Physically speaking, this peak is not due to an electrical resonance of passive components.
No it's not... The mechanical phenomenon causes a steep rise in impedance, that's what you measure. That's not the back-EMF. The back-EMF is related to the movement of the VC and a very small portion of the measured (current)signal.
Thank you for valuable information.
It seems that YST is just negative impedance output amplifier, lowering only the driver Qts - still good for cheap drivers! With Stahl method you can manipulate more of the TS parameters.
No, that's not correct. The driver's Qts stays the same, only with (minor) negative (output)impedance, the amplifier is capable of controlling the load better than an amplifier with positive (output)impedance.
The mechanical phenomenon could not be the cause of the steep rise because it is a mechanical, not electrical ! The mechanical phenomenon is first translated into voltage (not current) to have an effect on the impedance curve. That's what does the driver which is a two-ways transducer, converting electrical energy to mechanical energy and mechanical energy to electrical energy.
The Qes of the driver does not exist on its own. It depends on the whole electrical circuit where it is inserted. If the amp has a negative resistance output, the Qes is lowered, the effect is like removing some Ohm from the value of the voice coil resistance. As Qts is related to Qes, it does certainly not remain the same. Does more damping mean "better controlled" ?
Then explain why in the TS parameters nowhere the amplifier is included in the Qes calculation...
Don't give me this crap!
There is a circuit called A.N.I.C. (Advanced Negative Impedance Converter) mentonned in the Yamaha manual.
Negative outupt amplifiers are known since the fifties, maybe earlier, the first reference being to mind is Werner and Carell (JAES, October1958) and Thiele.
Before the calculations, you need some data. To get them you have to include the driver in a circuit where there is current generator, a voltage generator or a known resistive output impedance.
The parameters given by the manufacturers always suppose the driving amp has a zero Ohm output impedance.
A skilled user knows how to manipulate the output impedance.
A bit of search will lead you to documentation on all that.
I don't see material arguments among the respected previous posters if you add the words, "...acts like..." or "... in a manner of speaking..." or "... in proportion to..." or "...maybe you could say..." here and there.
While there are differences, in practice all MF amps act "... as if..." they had negative output impedance. Otherwise, how would they pump lower power into faster-moving speakers? Or you could apply those "... as if.." words to speakers. In the end, all the MF approaches are pretty similar electronically.
Patents - and here including IPAL once again - are just tricks to show you own a piece of property trivially different from somebody else's. Werner said it all in 1957.
B.
While there are differences, in practice all MF amps act "... as if..." they had negative output impedance. Otherwise, how would they pump lower power into faster-moving speakers? Or you could apply those "... as if.." words to speakers. In the end, all the MF approaches are pretty similar electronically.
Patents - and here including IPAL once again - are just tricks to show you own a piece of property trivially different from somebody else's. Werner said it all in 1957.
B.
Last edited: