Return to Home Page - www.GMweb2.net
The 'T-bass' for open baffles.
A dynamic driver resonance equaliser.
By Graham Maynard graham.maynard1@virgin.net
Did you ever hear the expression 'Really Pumping' as applied to high SPL LF reproduction ?
If yes, then that was likely when enclosure energy had been tuned to reinforce the LF output from a driver; ie. when the resultant LF output took on a character of its own in order to become impressive, and where the overall reproduction no longer matched original audio waveforms driving the system due to delay between initial drive cone motion and cabinet returned energy becoming radiated.
The 'T'-bass is a passive circuit for use between the low impedance output of a SS power amplifier and a conventional low frequency cone driver on an open baffle. It may be used to partly assist in compensating for the reduction of LF output when a driver is used on a plain baffle with shallow side stiffening wings (not a U-frame), where a reduction of radiated SPL arises when compared to the same driver being used in an enclosure. An open baffle is disadvantaged in relation to low frequency SPL via the driver's maximum displacement capabilities when there is front to back cancellation around a baffle edge, but open baffles do not return enclosure stored 'air spring' energy in 'Loudspeaker System Time'. Open baffle mounted LF drivers however are still not immune to basic resonance characteristics modifying the driver cone amplitude response with frequency during 'Music Time'.
The first L and C (6.4mH and 1,000uF) component values in this circuit may be optimised to counter driver-baffle stored energy around the system resonant frequency. These may be trimmed to reduce the amplitude distortion which develops during Music Time between t=0 and t=1/Fs as the normally resonant driver-baffle combination otherwise stores and returns energy to the radiating cone following sustained or percussive low frequency transients. (Yes, store-return effects do arise on open baffles as well as within enclosures.) The 1 ohm resistors control the Q and absorption balance of the series tuned but parallel acting LC circuit, such that it can reduce *loudspeaker* current flow after and around t=1/Fs and at frequencies close to Fs in a manner which counters the driver store-return increase in cone radiation that would otherwise arise via direct driver to amplifier connection after t=0. The transformer is shown as 500VA and 2x 40V, though for light duties it may be as small as 250VA and 2x 20V. The 1,000uF capacitor may be 2x low ESR 470uF 100Vdc symmetrically connected in '+' to '-' parallel.
Thus this transformer based circuit may be used to generate a reduction in electrical drive after t=0 which counters the increased driver-baffle output, with both responses acting simultaneously but inversely in real *Music Time*. Crossovers and equalisers do not do this, because no matter how theoretically correct or smooth the resulting amplitude/ frequency response might be as measured in time isolation, those other circuits still cannot prevent the driver-baffle combination subsequently storing and returning energy with respect to the amplifier, and thus still 'voicing' reproduction in direct relation to their own Driver-Baffle-Time characteristics.
Once the 'T'-bass circuit has been implemented however, additional line level EQ may indeed be applied at LF because this cannot any longer exaggerate the baffle-driver Fs based amplitude distortion effects which otherwise would normally arise during Music Time!
As with the modus operandii of my GEM amplifier, there are aspects of dynamic current flow interaction through active NFB - reactive load amplifier~loudspeaker interfaces during live Music Time which are not always understood due to modern testing methods suggesting conclusions and explanations related to, but also limited by, the circumstances associated with those time isolated sine-wave or impulse measurement findings themselves. Thought patterns and hypotheses based upon such testing might suggest reasons to deny possibilities for other types of resultant behaviour, but those same methods do not necessarily provide adequate proof to indicate that different approaches are not possible and thus cannot be implemented. There is always more than one way of viewing and resolving a problem, so to anyone who rigidly doubts the effectiveness of my circuits, I suggest that you are likely to change your mind after you have checked them out 'hands-on' and listened to the audible evidence for yourself.
The 5mH and 4.7uF are not really part of this 'T'-bass circuit, they could be 2mH and 10uF depending on the chosen driver; these components are merely shown as being one possible way to reduce mid frequency output, maybe with say 4.7R in series with the capacitor, whilst the 22uF and 4.7R form the Zobel network which no dynamic loudspeaker should be run without across its terminals. It is essential that all component values are optimised for most satisfactory reproduction via individual driver-baffle-room systems, though this does NOT mean that you should then attempt to achieve the flattest sinewave amplitude response at your normal listening position because that too will lead to a distortion of the driver's transduction response in Music Time!
Return to Home Page - www.GMweb2.net