Graham,
I've been considering the T-Bass system. I haven't built one yet, because I'm one of those cautious people who prefer to understand how something works before trying it. After all, if it works, then there must be valid physical reasons why. To that end, I've been working through your explanations. I'm having trouble with a couple of things, which I hope you can clear up.
I agree that the beginning of a sine wave burst contains other frequencies. It has to, because it is changing in time, and indeed if you spectrum analyse the change from "no signal" to "first cycle" you see other frequencies. And I agree that those higher frequencies are reproduced by the higher frequency drivers in a multi-way system, or by the one driver in a full-range.
But I have always understood that the sum of the outputs (addition of all the harmonics) will add up to the original waveform. You say that it doesn't. Do I understand you correctly?
You also point out that the acoustic output of a driver driven below its resonance is less during the first cycle than in subsequent cycles.
You say that this is demonstrated in the graphs in Linkwitz's discussion of how the Linkwitz Transform works. I see it in the graphs, and in the actual results obtained by StigErik and others. There are two points that arise from this that I hope you can clarify:
1: Referring to the Linkwitz graph, look at the phase relationship between the input signal and the acoustic output. It's established beyond doubt that the (steady state sine wave) output signal phase of a speaker below resonance leads the input phase by 90 degrees. But during the first 1/4 cycle, the input and output are more or less in phase. They would have to be, because the speaker cannot respond to a signal that has not arrived yet.
However, the output phase very quickly (within the first cycle) changes to 90 degrees advanced. In order to do so, it has to take a "short cut", namely the reduced amplitude of the first cycle. Another way of looking at it would be to say that the first output cycle takes only 3/4 of the time of the first cycle of the input, and the waveform of that cycle is non-sinusoidal. Therefore, there are other frequencies present during that cycle. Does this make sense, in reference to your own understanding of the effect?
2: You say that the reduced acoustic output during the first cycle is because the speaker is storing the input energy within itself. What mechanism is used to store that energy? And when is this energy released? What effect does the release of this stored energy have on the music?
Your T-bass circuit is designed to increase the drive to the speaker during the first cycle, so that the acoustic output waveform more closely follows the applied waveform. This will store even more energy in the driver. What effect will the release of this extra energy have on the music?
Signed,
Perplexed.
I've been considering the T-Bass system. I haven't built one yet, because I'm one of those cautious people who prefer to understand how something works before trying it. After all, if it works, then there must be valid physical reasons why. To that end, I've been working through your explanations. I'm having trouble with a couple of things, which I hope you can clear up.
Regarding those loudspeaker traces you have developed; imagine the harmonics of those waves as would already be developed within that first cycle; they will be coming through say the mid range.
Imagine how distorted the resultant propagating bass/ mid composite would be during that first cycle especially.
I agree that the beginning of a sine wave burst contains other frequencies. It has to, because it is changing in time, and indeed if you spectrum analyse the change from "no signal" to "first cycle" you see other frequencies. And I agree that those higher frequencies are reproduced by the higher frequency drivers in a multi-way system, or by the one driver in a full-range.
But I have always understood that the sum of the outputs (addition of all the harmonics) will add up to the original waveform. You say that it doesn't. Do I understand you correctly?
You also point out that the acoustic output of a driver driven below its resonance is less during the first cycle than in subsequent cycles.
You say that this is demonstrated in the graphs in Linkwitz's discussion of how the Linkwitz Transform works. I see it in the graphs, and in the actual results obtained by StigErik and others. There are two points that arise from this that I hope you can clarify:
1: Referring to the Linkwitz graph, look at the phase relationship between the input signal and the acoustic output. It's established beyond doubt that the (steady state sine wave) output signal phase of a speaker below resonance leads the input phase by 90 degrees. But during the first 1/4 cycle, the input and output are more or less in phase. They would have to be, because the speaker cannot respond to a signal that has not arrived yet.
However, the output phase very quickly (within the first cycle) changes to 90 degrees advanced. In order to do so, it has to take a "short cut", namely the reduced amplitude of the first cycle. Another way of looking at it would be to say that the first output cycle takes only 3/4 of the time of the first cycle of the input, and the waveform of that cycle is non-sinusoidal. Therefore, there are other frequencies present during that cycle. Does this make sense, in reference to your own understanding of the effect?
2: You say that the reduced acoustic output during the first cycle is because the speaker is storing the input energy within itself. What mechanism is used to store that energy? And when is this energy released? What effect does the release of this stored energy have on the music?
Your T-bass circuit is designed to increase the drive to the speaker during the first cycle, so that the acoustic output waveform more closely follows the applied waveform. This will store even more energy in the driver. What effect will the release of this extra energy have on the music?
Signed,
Perplexed.
Regarding 2: I will try to do recordings of a single sine wave period only. The energy apparently lost in the first period is expected to "ring" somewhat after the input signal is stopped, since the "stored energy" must be released. This is indeed interesting!
Last edited:
Are there any of you that has build a T-bass circuit that can perform the same acoustic measurements that I showed here earlier? I'm interested in seeing what your waveforms looks like, with and without T-bass. Thanks in advance. 🙂
The ringing after the input stops would be a logical place for it to be released, and that is supported by experiment. The question is, does the T-bass circuit make the "ringing" worse?
I note that your experiment does not show significant improvement in the first cycle waveform. About the only thing I have seen which does, is the Linkwitz Transform. The first cycle being smaller is apparently due to the speaker acting as a high pass filter. The LT lowers the corner frequency of this filter, so the waveform improves. From this, I wonder if a speaker with true response down to DC would have an accurate first cycle?
I have the components but not the time and space to build a T-bass and try it at the moment. All I can do is think about it. Pretty frustrating, but it gives me an idea what life must be like for Steven Hawking.
I note that your experiment does not show significant improvement in the first cycle waveform. About the only thing I have seen which does, is the Linkwitz Transform. The first cycle being smaller is apparently due to the speaker acting as a high pass filter. The LT lowers the corner frequency of this filter, so the waveform improves. From this, I wonder if a speaker with true response down to DC would have an accurate first cycle?
I have the components but not the time and space to build a T-bass and try it at the moment. All I can do is think about it. Pretty frustrating, but it gives me an idea what life must be like for Steven Hawking.
Hi Don,
You asked >> But I have always understood that the sum of the outputs (addition of all the harmonics) will add up to the original waveform. You say that it doesn't. Do I understand you correctly? <<
How can they add correctly during the first low frequency cycle when the LF driver's transduction is so badly amplitude distorted ?
And as you say, a loudspeaker cannot respond to a signal before it arrives, so when it does arrive, the transduction is indeed distorted in time.
T-bass is to increase the leading edge via transformer action - requires low Z amplifier and low R windings. Then the series tuned resonant circuit reduces LS drive to counter driver resonance.
(Not store more energy - but damp it.)
I do not know detail of the resistances in StigErik equipment/ leads/ components, or whether his C+L match the driver, or what the driver/ loading is.
A loudspeaker directly connected to a NFB amplifier cannot fail to store energy and resonate or transduce with back-EMF induced angle, whether that amplifier is Linkwitz EQed or not.
Re the storing of energy as in your (2), the decreased first half cycle amplitude shows that the driver has stored some energy which could not generate air motion.
You asked >> But I have always understood that the sum of the outputs (addition of all the harmonics) will add up to the original waveform. You say that it doesn't. Do I understand you correctly? <<
How can they add correctly during the first low frequency cycle when the LF driver's transduction is so badly amplitude distorted ?
And as you say, a loudspeaker cannot respond to a signal before it arrives, so when it does arrive, the transduction is indeed distorted in time.
T-bass is to increase the leading edge via transformer action - requires low Z amplifier and low R windings. Then the series tuned resonant circuit reduces LS drive to counter driver resonance.
(Not store more energy - but damp it.)
I do not know detail of the resistances in StigErik equipment/ leads/ components, or whether his C+L match the driver, or what the driver/ loading is.
A loudspeaker directly connected to a NFB amplifier cannot fail to store energy and resonate or transduce with back-EMF induced angle, whether that amplifier is Linkwitz EQed or not.
Re the storing of energy as in your (2), the decreased first half cycle amplitude shows that the driver has stored some energy which could not generate air motion.
Last edited:
Stig Erik,
Have you checked with a 20Hz signal? My circuit cut the output at 20Hz by at least half.
Graham, more stupid questions.
What happens if the parallell coil is bigger? I suppose that the boost/impedance matching of fs moves down in frequency.
Peter
Have you checked with a 20Hz signal? My circuit cut the output at 20Hz by at least half.
Graham, more stupid questions.
What happens if the parallell coil is bigger? I suppose that the boost/impedance matching of fs moves down in frequency.
Peter
Hi Graham, thanks for responding.
Before I continue, I would like to make it clear that I am not skeptical about the T-bass circuit. I believe it works. By "works", I mean that it is constructed from real components in a practical circuit and has a measurable effect. I'm just having difficulty reconciling the behaviour of the circuit (as simulated and measured by experiment) with your description of how it works. It's probably just a difference in language, rather like the difference between the English language and what is spoken in America. 🙂
Do they add correctly or not? I have always assumed they do, but in this specific scenario I haven't seen any evidence either way yet. Not that you have to provide it, it just means I have to revisit the papers where I saw it derived and explained and relate them to this scenario.
Your second point first, the series tuned resonant circuit: A series tuned circuit has a low impedance at resonance. That is, current flow through it is at a maximum. I would have thought that this would increase LS drive, not decrease it. And a resonant circuit stores energy. How can it damp the driver resonance? Does it absorb energy from the speaker?
Maybe I'm not reading the circuit correctly. Which components constitute the series resonant circuit?
Back to the first point:
OK, we agree the first cycle is distorted in both amplitude and frequency. You propose T-bass to reduce the amplitude distortion, by increasing the level of the first leading edge (1/4 to 1/2 cycle).
I will now describe how I understand T-bass does this. Please correct me where I get it wrong.
T-bass applies a step-up that approximately doubles the voltage applied to the driver during the first cycle, followed by a resonant circuit that reduces the drive for following cycles.
There's more to it for BSC etc, but do I have the core function correct?
What I cannot yet see is how this differs in effect from a standard low-pass filter applied before the driver. When you look at the spectrum of the first leading edge, it contains components down to DC.
You can hear this in practice, if you have a power supply with a very low resistance primary power transformer. When you switch it on, you hear a "grunt" or damped hum/ buzz from the transformer. This is the DC component of the first cycle causing the transformer core to saturate. It dies away as the DC component is dissipated in the coil /mains supply resistance. (Note, this noise is not due to the inrush charging current of the filter capacitors. You will hear it in a transformer with a purely resistive load too.)
So what I see T-bass does, just like a low pass filter, is boost the lowest frequency components. What I have yet to quantify is how much the driver's impedance and back-EMF, as seen at the driver terminals, affects and is affected by the T-bass circuit. That shouldn't be any more difficult than designing a crossover, and it may be solvable using the same methods.
"Aha", you say, "but a low pass filter causes a phase change and the transformer applies the boost without causing a phase change." But in any circuit where the frequency response is not constant there must also be a phase change. Plotting the frequency / phase response of the T-bass circuit shows that it definitely changes the phase with frequency. The transformer has inductance, so it must also add a change in phase, but its main job is to boost the input to compensate for the losses in the resonant circuit - if I understand the circuit operation to be as you describe it.
Did it store it, or did it not accept it? Intuitively, I feel it stores it, because the energy dissipated after the input stops has to come from somewhere. But I have no proof either way. I will have to "look it up" - find an authoritative proof either way.
So my current understanding remains this:
The first cycle of a sine burst contains frequencies down to (in theory) DC. A driver with a flat response to DC will reproduce the input with no distortion of the first cycle. Real drivers are high-pass filters, so will not pass the very low frequency components of the first cycle. Adding a low-pass filter ahead
of the driver will boost the level of the very low frequency components.
The above is, I believe, the frequency domain equivalent description of your time-domain description of the T-bass operation.
Just a couple more thoughts on phase: There's proof everywhere that the phase of the acoustic output of a dynamic driver, with respect to the input waveform, is about -90 degrees for frequencies above the driver fundamental resonance, passes through zero at driver resonance, and approaches +90 degrees below resonance. The phase difference below resonance is quite clear in the graphs showing the driver response to the start of a tone burst. Now consider the mythical driver (perhaps excepting the TRW) with flat response down to DC. By definition, its resonant frequency is 0 Hz. Its phase will never go positive, hence there will be no time distortion of the first cycle. Will there still be amplitude distortion of the first cycle? I don't know. I'll try a simulation. This is one area where simulation works better than a physical model. (And there are simulators that can accurately model the T-bass circuit. I expect you weren't aware of them - for example, I hadn't heard of Tina before this thread.)
Regards,
Don.
Before I continue, I would like to make it clear that I am not skeptical about the T-bass circuit. I believe it works. By "works", I mean that it is constructed from real components in a practical circuit and has a measurable effect. I'm just having difficulty reconciling the behaviour of the circuit (as simulated and measured by experiment) with your description of how it works. It's probably just a difference in language, rather like the difference between the English language and what is spoken in America. 🙂
How can they add correctly during the first low frequency cycle when the LF driver's transduction is so badly amplitude distorted ?
Do they add correctly or not? I have always assumed they do, but in this specific scenario I haven't seen any evidence either way yet. Not that you have to provide it, it just means I have to revisit the papers where I saw it derived and explained and relate them to this scenario.
And as you say, a loudspeaker cannot respond to a signal before it arrives, so when it does arrive, the transduction is indeed distorted in time.
T-bass is to increase the leading edge via transformer action - requires low Z amplifier and low R windings. Then the series tuned resonant circuit reduces LS drive to counter driver resonance.
(Not store more energy - but damp it.)
Your second point first, the series tuned resonant circuit: A series tuned circuit has a low impedance at resonance. That is, current flow through it is at a maximum. I would have thought that this would increase LS drive, not decrease it. And a resonant circuit stores energy. How can it damp the driver resonance? Does it absorb energy from the speaker?
Maybe I'm not reading the circuit correctly. Which components constitute the series resonant circuit?
Back to the first point:
OK, we agree the first cycle is distorted in both amplitude and frequency. You propose T-bass to reduce the amplitude distortion, by increasing the level of the first leading edge (1/4 to 1/2 cycle).
I will now describe how I understand T-bass does this. Please correct me where I get it wrong.
T-bass applies a step-up that approximately doubles the voltage applied to the driver during the first cycle, followed by a resonant circuit that reduces the drive for following cycles.
There's more to it for BSC etc, but do I have the core function correct?
What I cannot yet see is how this differs in effect from a standard low-pass filter applied before the driver. When you look at the spectrum of the first leading edge, it contains components down to DC.
You can hear this in practice, if you have a power supply with a very low resistance primary power transformer. When you switch it on, you hear a "grunt" or damped hum/ buzz from the transformer. This is the DC component of the first cycle causing the transformer core to saturate. It dies away as the DC component is dissipated in the coil /mains supply resistance. (Note, this noise is not due to the inrush charging current of the filter capacitors. You will hear it in a transformer with a purely resistive load too.)
So what I see T-bass does, just like a low pass filter, is boost the lowest frequency components. What I have yet to quantify is how much the driver's impedance and back-EMF, as seen at the driver terminals, affects and is affected by the T-bass circuit. That shouldn't be any more difficult than designing a crossover, and it may be solvable using the same methods.
"Aha", you say, "but a low pass filter causes a phase change and the transformer applies the boost without causing a phase change." But in any circuit where the frequency response is not constant there must also be a phase change. Plotting the frequency / phase response of the T-bass circuit shows that it definitely changes the phase with frequency. The transformer has inductance, so it must also add a change in phase, but its main job is to boost the input to compensate for the losses in the resonant circuit - if I understand the circuit operation to be as you describe it.
Re the storing of energy as in your (2), the decreased first half cycle amplitude shows that the driver has stored some energy which could not generate air motion.
Did it store it, or did it not accept it? Intuitively, I feel it stores it, because the energy dissipated after the input stops has to come from somewhere. But I have no proof either way. I will have to "look it up" - find an authoritative proof either way.
So my current understanding remains this:
The first cycle of a sine burst contains frequencies down to (in theory) DC. A driver with a flat response to DC will reproduce the input with no distortion of the first cycle. Real drivers are high-pass filters, so will not pass the very low frequency components of the first cycle. Adding a low-pass filter ahead
of the driver will boost the level of the very low frequency components.
The above is, I believe, the frequency domain equivalent description of your time-domain description of the T-bass operation.
Just a couple more thoughts on phase: There's proof everywhere that the phase of the acoustic output of a dynamic driver, with respect to the input waveform, is about -90 degrees for frequencies above the driver fundamental resonance, passes through zero at driver resonance, and approaches +90 degrees below resonance. The phase difference below resonance is quite clear in the graphs showing the driver response to the start of a tone burst. Now consider the mythical driver (perhaps excepting the TRW) with flat response down to DC. By definition, its resonant frequency is 0 Hz. Its phase will never go positive, hence there will be no time distortion of the first cycle. Will there still be amplitude distortion of the first cycle? I don't know. I'll try a simulation. This is one area where simulation works better than a physical model. (And there are simulators that can accurately model the T-bass circuit. I expect you weren't aware of them - for example, I hadn't heard of Tina before this thread.)
Regards,
Don.
Regarding 2: .... The energy apparently lost in the first period is expected to "ring" somewhat after the input signal is stopped, since the "stored energy" must be released. This is indeed interesting!





😉
Michael
Last edited:
I follow this thread to help constructors out, not armchair challengers.




😀
Michael
Hi Don,
I like the way you are now thinking - "out of the box" !
Not enough folk have been doing this.
You cannot see the difference between T-bass and say crossover type circuit because you have not real world analysed each separately, and then compared both with direct.
Go back to the waveform via a simple choke and capacitor filter and look not only at the reduced first cycle, but the first cycle *delay* (phase change not apparent before signal arrived) during music time.
So many folk fail to augment full range with a LF driver because of impulse delay introduced by LF crossover.
>> What I have yet to quantify is how much the driver's impedance and back-EMF, as seen at the driver terminals, affects and is affected by the T-bass circuit. <<
The amp is step up loaded by the LS through the transformer for the first low frequency half cycle, then due to the choke grounded centre transformer limb the impedance seen by the driver increases so that stored resonant input is reduced, whist the series C+L shunt the transformer and the load amplifier where LS resonance would have caused high impedance.
The transformer core must not suffer 'switch on surge' saturation because that would cause amplifier clipping or damage.
>> There's proof everywhere that the phase of the acoustic output of a dynamic driver, with respect to the input waveform, is about -90 degrees for frequencies above the driver fundamental resonance, passes through zero at driver resonance, and approaches +90 degrees below resonance. <<
Yes, and that develops within the LF driver during the first cycle after an impulses arrives.
So tune T-bass to increase leading waveform but counter driver energy storage !!!!!
If there are better simulators then go for it.
I had this circuit *real world* running BEFORE I ever examined it with TINA.
TINA then indicated ways in which I might improve - but TINA was wrong !!!!!
Hi StigEric,
I think you are looking for more than T-bass can provide, and then asking me about what it *cannot* do.
The highest Fs driver I used this circuit on was circa 40Hz. My final baffle had 18" with Fs 23Hz, so the system worked just fine at 20Hz.
Hi Michael,
Nice simulations, but good only as far as equivalence goes.
( The transformer winding resistances are way too high ! )
If you've got a Beyma 21" why not dig out a PSU trafo and knock the circuit up ?
However, the Beyma appears to have a high Qms so possibly not the best for OB running anyway. ( I'm not saying it cannot be tried however. )
Do those links relate to real-world hands-on LF audio investigation, observation and experience ? If yes I will check. If not I'm too busy with other things !!!!!
Cheers ........... Graham.
I like the way you are now thinking - "out of the box" !
Not enough folk have been doing this.
You cannot see the difference between T-bass and say crossover type circuit because you have not real world analysed each separately, and then compared both with direct.
Go back to the waveform via a simple choke and capacitor filter and look not only at the reduced first cycle, but the first cycle *delay* (phase change not apparent before signal arrived) during music time.
So many folk fail to augment full range with a LF driver because of impulse delay introduced by LF crossover.
>> What I have yet to quantify is how much the driver's impedance and back-EMF, as seen at the driver terminals, affects and is affected by the T-bass circuit. <<
The amp is step up loaded by the LS through the transformer for the first low frequency half cycle, then due to the choke grounded centre transformer limb the impedance seen by the driver increases so that stored resonant input is reduced, whist the series C+L shunt the transformer and the load amplifier where LS resonance would have caused high impedance.
The transformer core must not suffer 'switch on surge' saturation because that would cause amplifier clipping or damage.
>> There's proof everywhere that the phase of the acoustic output of a dynamic driver, with respect to the input waveform, is about -90 degrees for frequencies above the driver fundamental resonance, passes through zero at driver resonance, and approaches +90 degrees below resonance. <<
Yes, and that develops within the LF driver during the first cycle after an impulses arrives.
So tune T-bass to increase leading waveform but counter driver energy storage !!!!!
If there are better simulators then go for it.
I had this circuit *real world* running BEFORE I ever examined it with TINA.
TINA then indicated ways in which I might improve - but TINA was wrong !!!!!
Hi StigEric,
I think you are looking for more than T-bass can provide, and then asking me about what it *cannot* do.
The highest Fs driver I used this circuit on was circa 40Hz. My final baffle had 18" with Fs 23Hz, so the system worked just fine at 20Hz.
Hi Michael,
Nice simulations, but good only as far as equivalence goes.
( The transformer winding resistances are way too high ! )
If you've got a Beyma 21" why not dig out a PSU trafo and knock the circuit up ?
However, the Beyma appears to have a high Qms so possibly not the best for OB running anyway. ( I'm not saying it cannot be tried however. )
Do those links relate to real-world hands-on LF audio investigation, observation and experience ? If yes I will check. If not I'm too busy with other things !!!!!
Cheers ........... Graham.
Last edited:
My tests was performed using a Seas W22EX001 by the way, forgot to mention that... sorry. Fs = 25 Hz.
My tests was performed using a Seas W22EX001 by the way, forgot to mention that... sorry. Fs = 25 Hz.
So have you tuned the C, the L and both series Rs for best LF SYSTEM reproduction ?
If that Seas driver is enclosure mounted then the driver amplitude response is bound to fall more sharply below the *higher system resonance* (unless tuned or t-lined) than would be the case with open baffle mounting due to enclosures raising driver resonant frequency and loading them below that new frequency.
Natural driver Fs CANNOT be the sole determining factor.
Hi StigEric,
So 'output' might be reduced at 20Hz - as T-bass is supposed to with continuous LF sines.
However, if the L,C and Rs are properly adjusted and your trafo and choke are low R with sufficient inductance, then you should still have leading edge boost.
If you are looking for boost at 20Hz I could imagine the choke needing to be 15 to 22mH for an 8 ohm driver, and yet it must have winding R less than 1 ohm, so ferrite likely necessary.
First off you could try removing the R in series with the inductor.
Cheers ..........Graham.
So 'output' might be reduced at 20Hz - as T-bass is supposed to with continuous LF sines.
However, if the L,C and Rs are properly adjusted and your trafo and choke are low R with sufficient inductance, then you should still have leading edge boost.
If you are looking for boost at 20Hz I could imagine the choke needing to be 15 to 22mH for an 8 ohm driver, and yet it must have winding R less than 1 ohm, so ferrite likely necessary.
First off you could try removing the R in series with the inductor.
Cheers ..........Graham.
Last edited:
Hi Don,
I like the way you are now thinking - "out of the box" !
Not enough folk have been doing this.
You cannot see the difference between T-bass and say crossover type circuit because you have not real world analysed each separately, and then compared both with direct.
Yes, it's very easy to say. "it can't work" and pass on. I'm trying to think it through and work out *why* it does (or equally, does not) work as you specify. I see a set of components, connected in a particular way, and I know they must interact in a certain way. They must obey certain well understood and proven laws of physics.
Now if that interaction does not result in the function that you specified, then either I have made an error - in component value, for example - or you have made an error in the specification of its function. Saying that I should actually build one, and then I will see it works, is not a valid explanation. The circuit can be accurately simulated. The simulation tools have been proven accurate many times over - if you build it in simulation, its simulated behaviour will be very close to its actual behaviour. If you disagree with that, then I'm done here - I will not consider something where I am being asked to suspend my belief in the laws of physics.
(Note that I do not include my ears as accurate tools, the human auditory system is one of the least trustworthy senses we have, especially in repeatability and objective accuracy.)
You can say that the physics and simulators are wrong, that they do not accurately model a T-bass behaviour. But in that case, the onus is on you to show how, why and where they are wrong.
Go back to the waveform via a simple choke and capacitor filter and look not only at the reduced first cycle, but the first cycle *delay* (phase change not apparent before signal arrived) during music time.
So many folk fail to augment full range with a LF driver because of impulse delay introduced by LF crossover.
By this, I assume you mean a "standard" choke and capacitor low pass filter before a LF driver that provides bass augmentation for a fullrange driver. So many folk get that right, too. It's not rocket science any more, there are well designed and accurate tools that allow you to get close to the correct values and topology before you lift a soldering iron. I very well remember the way it was done 40 years ago, with slide rule and graph paper. The process took so long it's a wonder any good speakers got built at all, and there were more bad ones than there are today. But I won't argue the point further until I try it.
Speaking of which, do you have a real-world example of a driver and T-bass values that you deem to perform well? It would be a lot more accurate if we had a "worked example" so we are all working from the same source. I assume the one you are using now has been optimised for best performance?
The amp is step up loaded by the LS through the transformer for the first low frequency half cycle, then due to the choke grounded centre transformer limb the impedance seen by the driver increases so that stored resonant input is reduced, whist the series C+L shunt the transformer and the load amplifier where LS resonance would have caused high impedance.
The transformer core must not suffer 'switch on surge' saturation because that would cause amplifier clipping or damage.
We agree that the circuit needs to boost the voltage to the driver during the first half cycle, and not boost it for subsequent cycles. Correct?
You say that the transformer provides a step up (boosts the voltage) during the first half cycle, so the first part is satisfied.
There appears to be an inconsistency at the next step - your schematic from earlier in the thread clearly shows the choke connected to one end of the winding, not the centre tap as you appear to be saying above.
But assuming for now that the schematic is correct, you say that the impedance of the choke then raises the impedance as seen by the driver, reducing the drive for subsequent cycles. The problem is that the choke specified (6.4mH) is too small to do this. Work out the impedance of the choke at 20 Hz, you will see what I mean. (Less than 1 ohm.)
You next say that the series C+L (I assume the 2x 470uF and the 6.4mH) "shunt the transformer and the load amplifier where LS resonance would have caused high impedance".
What's a "load amplifier"?
Also, the impedance of the capacitors at 20 Hz is over 8 ohms.
And the resonant frequency of the series LC resonant circuit formed by the L and C is about 65 Hz, well above any of the suggested operating frequencies of the circuit.
Do you see why I am having trouble understanding the circuit operation? The numbers just don't add up for it to work the way you say it does.
So tune T-bass to increase leading waveform but counter driver energy storage !!!!!
I would love to. But, and this is where I expect we will have to agree to differ, I will need to simulate it (or do it the hard way, with slide rule and paper) before I build it. I will not build it until I have results that indicate it will work as you describe.
If there are better simulators then go for it.
I had this circuit *real world* running BEFORE I ever examined it with TINA.
TINA then indicated ways in which I might improve - but TINA was wrong !!!!![/QUOTE}
I would be very interested in seeing the results of your examination, and the values which led you to believe that TINA was wrong. They may explain my inability to grasp the T-bass function.
Regards,
Don.
Last edited:
Nice simulations, but good only as far as equivalence goes.
😉
( The transformer winding resistances are way too high ! )
.
???
The value you see noted is the assumed transformer inductance of 1 Henry at each branch.
Do those links relate to real-world hands-on LF audio investigation, observation and experience ?.
???
Plots shown are visualisations of what T-bass can or can not do - actually it's intended as an eye opener only.
#############
If *you* provide :
- correct TSP values
- correct T-bass circuit values
- measurements for validation
of a tuned version of yours - then *I* will prove, that you are wrong in several core aspects of your explanation regarding T-bass circuit:
1.) T-bass actually *is* equalisation / response shaping
2.) T-bass actually can be mimic'ed by a active filter too (have not tried to be honest – but is a simple consequence of point 1.) )
Should you (or someone else) ever be interested in such a "acid test" – just let me know...
This is not to say T-bass does not *sound* as you and others (me included) have found it to *sound* - just your explanations are off – biased to some degree that is.
You will have noticed that your core aim – restoring some of the amplitude loss at the first cycles – is easily demonstrated to be simple HP filter behaviour treatment as published some postings back – meaning you can :
1. change F res (as Linkwitz does)
2. change order of HP-filter (as you do at least in the F-res area)
3. change type of HP-filter (as you do as well)
The "restoration" of some of the "amplitude loss at first cycles" has its penalty in causing loooooong "ringing" – clearly seen on CSD's I've published – simply caused by increasing HP filter order / and or filter type.
No magic in here ...
IMO this is the core audible "improvement" regarding restoration of "box sound" – if you are after such.
On the pro side there is some shorter decay visible above F-res. We would have to check back with some other simus to get a feeling if its not only "math fake" (artefacts) though (from auditioning I'd tend to say its a "real" positive effect)...
🙂
Michael
Last edited:
Michael - Your simulation is rubbish and therefore proves nothing !!!!!
Go measure the resistance of a real-world 500W-2x 40V transformer !!!!!
I am not writing about about increasing first cycles, so you are incorrect here too.
T-bass boosts the first *half cycle* amplitude - that is when correctly constructed and tuned to match the driver system - it then reduces continuing amplitude.
Also, is your 21" the correct type of LS to use on OB with this circuit, seems far too underdamped to be used without cabinet loading ?
Stop making so many assumptions and chiding me for your errors !
Because of ill health (and others with ill health in our family) all my audio is set aside, and I have no intention of getting it out to spoon feed you or Don.
Also box sound is detestable so stop making assumptions there about my thoughts.
It is clear you are still unable to understand this - but that is neither my fault nor my responsibility.
The best way to begin to understand this it is to build, tune and listen - not play with your software !
Seems to me that you think everyone else who has had success with this circuit must be wrong because your incorrectly inputted computer says so.
.
Go measure the resistance of a real-world 500W-2x 40V transformer !!!!!
I am not writing about about increasing first cycles, so you are incorrect here too.
T-bass boosts the first *half cycle* amplitude - that is when correctly constructed and tuned to match the driver system - it then reduces continuing amplitude.
Also, is your 21" the correct type of LS to use on OB with this circuit, seems far too underdamped to be used without cabinet loading ?
Stop making so many assumptions and chiding me for your errors !
Because of ill health (and others with ill health in our family) all my audio is set aside, and I have no intention of getting it out to spoon feed you or Don.
Also box sound is detestable so stop making assumptions there about my thoughts.
It is clear you are still unable to understand this - but that is neither my fault nor my responsibility.
The best way to begin to understand this it is to build, tune and listen - not play with your software !
Seems to me that you think everyone else who has had success with this circuit must be wrong because your incorrectly inputted computer says so.
.
Last edited:
Because of ill health (and others with ill health in our family) all my audio is set aside...
.
Sad to hear about....
Anything else, well ....
Michael
Michael - Your simulation is rubbish and therefore proves nothing !!!!!
Go measure the resistance of a real-world 500W-2x 40V transformer !!!!!
Graham,
What, in your opinion, is a typical resistance of a real-world 500W-2x 40V transformer? You must know, otherwise you can't say that Michael's value is inaccurate - not and expect to be believed, at any rate.
Because of ill health (and others with ill health in our family) all my audio is set aside, and I have no intention of getting it out to spoon feed you or Don.
I'm sorry to hear that. I understand how it makes you feel, I have alluded elsewhere to the circumstances that likewise prevent me from doing much more than think at the moment. Still, it's an ill wind, etc... if I didn't have time to think, I wouldn't have started pondering T-bass.
The best way to begin to understand this it is to build, tune and listen - not play with your software !
Anyone can construct the T-bass circuit and experiment with different values. They will certainly hear a difference in the sound as they do so. But this does not lead to an understanding of how it works. Understanding comes from knowing how components work at a basic, physical/electrical level, and how they interact with each other. We can do this with sliderule and paper, or we can use a simulator. All a simulator does is to allow us to automate the tedious calculations of the component interactions. I, for one, would not care to replace my current scientific calculator with the slide rule and log tables that I used before it.
Seems to me that you think everyone else who has had success with this circuit must be wrong because your incorrectly inputted computer says so.
.
Here we have the crucial point.
There are people who have built this circuit and are happy with the results it produces. I have no problem with that. It does produce an audible and measurable effect. But the results they are getting do not appear to be due to the effects you describe. That is, it does not significantly boost the first half cycle. Both simulated and real measurements show that. So whatever they are hearing is due to some other effect, and the simulations and real world measurements - performed in your "music time", the time/waveform domain, not just the frequency domain - indicate what that effect is.
So you say we're "doing it all wrong", and you're not going to "spoon feed" us. That's your choice. All we can do is to prove, by peer reviewable means such as simulation and real world measurements, what the T-bass circuit actually does.
One parting thought:
We've all been using (at your suggestion) a sine wave tone burst to illustrate the problem. Real music transients - plucked bass, for example - don't look anything like this at their leading transient. As resonators, they exhibit their own phase behaviour on the leading half cycle. They don't exhibit the behaviour of a gated sine wave. So if they don't have the problem, then the driver doesn't have a problem - it is never asked to reproduce something that it can't. By trying to make the driver reproduce something that doesn't exist in music, the T-bass could actually be making the music less life-like. But the audible effect of this is getting into psycho-acoustics, how we perceive and emotionally react to music, and that is outside the scope of this technical discussion.
Regards,
Don.
Last edited:
- Home
- Loudspeakers
- Full Range
- 'T'-bass drive for OB LF drivers.