Hi CLS,
I'm so pleased you are enjoying the T-bass.
It is like a revelation, and there is no way you tell anyone what they are not enjoying until somehow they can hear it for themselves.
EQ and crossovers destroy cherence. If suddenly starting sine waveforms are viewed via a simulator their lesser leading edge and EQ or filter introduced delay become obvious.
YET from that leading edge into the rest of the first cycle the waveform becomes compressed into a leading LF phase sine characteristic: this before the first half cycle managed to achieve full amplitude: so music waveforms become badly distorted on an on-going basis in a manner which sine investigation cannot reveal !!!
Yes the transformer provides the initial boost to counter that which is missing via real world transduction, and then after half a cycle the L and C increase the impedance at the ground end of the transformer so that continuing drive is reduced before the driver can store too much energy which would otherwise muddy reproduction.
EQ and LP do simply CANNOT do that, and the T-bass does it with reasonable phase coherence too !
Those who study with sines (including amplifier designers) will never see the start-up distortion because it has passed before they have a *steady* waveform they can measure; that is unless they measure the first cycle from t=0 which does show the distortion, but exponents will state "Oh you can't do that!"
Of course you can; why not for dear sake.
Class-AB amplifiers have similar micro sudden start-ups at every half cycle crossover, and these too cannot be adequately viewed via steady sine investigation.
Class-A is smooth - but inefficient.
That is why I went on to develop my class-A//AB amplifier which is class A throughout crossovers and class-AB into power. It is different to other types which purport to do the same, and it is as good to hear as an amplifier as T-bass is for LF OB.
However I do not 'thread' my class-A//AB amplifier (which took me 30 years to get right!) on diyAudio because of all the noise and personally aimed bad attitude which developed from posters I suffered from here for years.
My amplifier -
http://www.gmweb2.net/Documentation/GEM document authored by Graham Stephen Maynard.htm
The write-up needs to be re-done, but I've not had time recently.
Hey someone just told me its Christmas next week !
Cheers ......... Graham.
_________________________________________________
Hi bzfcocon,
You posted whilst I was writing.
DO PLEASE set up your investigation.
Have you ever listened to the test CD reproduction of a distant canon firing ?
That is the best way you can hear that we acutely sensitive to AF coherence at LF. If the bass LS is not *in phase* as well as having the correct amplitude then the sound we listen to no longer comes from open space !!!
EQ and crossovers do not, because it is entirely impossible for them to do so, maintain phase and amplitude linearity in real time, so their AF arriving at a driver is distorted before the LS system distorts it some more.
T-bass works with the driver it is in series with !
Cheers ............ Graham.
I'm so pleased you are enjoying the T-bass.
It is like a revelation, and there is no way you tell anyone what they are not enjoying until somehow they can hear it for themselves.
EQ and crossovers destroy cherence. If suddenly starting sine waveforms are viewed via a simulator their lesser leading edge and EQ or filter introduced delay become obvious.
YET from that leading edge into the rest of the first cycle the waveform becomes compressed into a leading LF phase sine characteristic: this before the first half cycle managed to achieve full amplitude: so music waveforms become badly distorted on an on-going basis in a manner which sine investigation cannot reveal !!!
Yes the transformer provides the initial boost to counter that which is missing via real world transduction, and then after half a cycle the L and C increase the impedance at the ground end of the transformer so that continuing drive is reduced before the driver can store too much energy which would otherwise muddy reproduction.
EQ and LP do simply CANNOT do that, and the T-bass does it with reasonable phase coherence too !
Those who study with sines (including amplifier designers) will never see the start-up distortion because it has passed before they have a *steady* waveform they can measure; that is unless they measure the first cycle from t=0 which does show the distortion, but exponents will state "Oh you can't do that!"
Of course you can; why not for dear sake.
Class-AB amplifiers have similar micro sudden start-ups at every half cycle crossover, and these too cannot be adequately viewed via steady sine investigation.
Class-A is smooth - but inefficient.
That is why I went on to develop my class-A//AB amplifier which is class A throughout crossovers and class-AB into power. It is different to other types which purport to do the same, and it is as good to hear as an amplifier as T-bass is for LF OB.
However I do not 'thread' my class-A//AB amplifier (which took me 30 years to get right!) on diyAudio because of all the noise and personally aimed bad attitude which developed from posters I suffered from here for years.
My amplifier -
http://www.gmweb2.net/Documentation/GEM document authored by Graham Stephen Maynard.htm
The write-up needs to be re-done, but I've not had time recently.
Hey someone just told me its Christmas next week !
Cheers ......... Graham.
_________________________________________________
Hi bzfcocon,
You posted whilst I was writing.
DO PLEASE set up your investigation.
Have you ever listened to the test CD reproduction of a distant canon firing ?
That is the best way you can hear that we acutely sensitive to AF coherence at LF. If the bass LS is not *in phase* as well as having the correct amplitude then the sound we listen to no longer comes from open space !!!
EQ and crossovers do not, because it is entirely impossible for them to do so, maintain phase and amplitude linearity in real time, so their AF arriving at a driver is distorted before the LS system distorts it some more.
T-bass works with the driver it is in series with !
Cheers ............ Graham.
peterbrorsson said:Hi sreten,
Doh, the first post showing the circuit got me confused
http://www.diyaudio.com/forums/attachment.php?s=&postid=1622325&stamp=1222862646
Thanks,
Peter
That post tells that the primary is left unconnected - which is, of course very important. You could just as well get a toroid with only the secondaries - but it will probably require custom manufacture and would be therefore more expensive than a standard one you buy in an electronic store (in Europe : Conrad, for example).
bzfcocon said:
I'm definitely no expert on this, but I don't think so. (of course I maybe wrong)
I don't know if anybody did this- touching the cone when it's playing. I did it several times... OK, many times... I do it a lot.
I mean, when a bass note hit, the cone motion and the sound come out simultaneosly. (OK, maybe both the touching feel and hearing senses are simultaneosly delayed, or delayed by different degree. Or, I don't have that kind of 'resolution' in my senses. I don't know... )
OTOH, if we do sense the bass after several cycles, we should have it come out in advance (even earlier) to "catch up" the mid-high, don't we? Otherwise, how can the LF portion "fill" the area under the elevated line of a step response (or a rised square wave) ? If we can not sense the very beginning of a bass note, how would an 'ideal' square wave response be like? That's really beyond me.
Back to the T-bass, it's not only excellent on those jumpy punchy things. It also grabs and shakes you on those sustained bass -- double bass played by bow, pipe organ... etc. They are no longer flat and lifeless like a calm river, but full of dense and clear textures. More like a huge heavy roller crunching along a gravel ground, making it shake at EVERY little bumps and dips.
If we do need several cycles to sense the bass, then this kind of sustained bass notes in music would never give us that sense of 'textures". They are not really sustained flat only, but with continuously changed intensity. And we do sense those changes.
I've only got this kind of performance from a pair of big bass horns in a very large space before, now I have it with T-bass & OB.
FWIW my modest speakers (bass = 4 x 7" woofers on 30cm + 15cm side wings) produce bass in my TINY ROOM that is the most pleasing in terms of texture, pace, depth, coherence and listenability I've heard in a home speaker system.
Others have had more impact or extension (and some perhaps more detail on certain types of music - due to driver quality!), but none have been as coherent and able to deliver the deep notes at the same time as not possessing overhang on heavy bass notes... and this delivered at pretty high output levels. With bigger woofers the effect must be spellbinding as distortion will be even lower and output even deeper.
Simon
Others have had more impact or extension (and some perhaps more detail on certain types of music - due to driver quality!), but none have been as coherent and able to deliver the deep notes at the same time as not possessing overhang on heavy bass notes... and this delivered at pretty high output levels. With bigger woofers the effect must be spellbinding as distortion will be even lower and output even deeper.
Simon
All right, all right, I am ordering the coils 
Some more questions:
- currently, my bass drivers (1 Alpha15 per channel) are crossed over actively LR4 to the mids at about 140 Hz. Am I supposed to give up the LP crossover entirely or can I keep it like that? Currently, the active LR4 is avery easy way to achieve "perfect", in-phase crossover - apart from the group delay, of course.
- if I got this right, the part after the transformer IS a sort of LP filter in itself. I believe this could be ommited if the circuit is used with an existing LP filter ?
If I understand correctly, the main advantage of the circuit should be in the "boost" domain just under the dipole/reflection peak.
Some more questions:
- currently, my bass drivers (1 Alpha15 per channel) are crossed over actively LR4 to the mids at about 140 Hz. Am I supposed to give up the LP crossover entirely or can I keep it like that? Currently, the active LR4 is avery easy way to achieve "perfect", in-phase crossover - apart from the group delay, of course.
- if I got this right, the part after the transformer IS a sort of LP filter in itself. I believe this could be ommited if the circuit is used with an existing LP filter ?
If I understand correctly, the main advantage of the circuit should be in the "boost" domain just under the dipole/reflection peak.
bzfcocon said:
Hi,
It is a notch filter (+zobel) that could be rearranged to be a 2nd order
electrical low pass filter, it is optional in terms of whatever T-bass is.
Regarding your skepticism, it should be healthy. There is lot claimed
in this thread that AFAICT will not stand up to rigorous examination.
/sreten.
Hi BZ,
Yes the second choke is a partial filter, but not like a crosover which introduces phase lag.
You should be able to balance the lead of the mid with the lag of the choke/LF driver at your chosen frequency thereby optimising coherence.
The Alpha has a slight peak before roll-off. That peak can be completely tuned out with a capacitor in parallel with the output choke.
A Zobel across the driver can further cut mid frequencies by holding the load impedance low as choke impedance rises, but overdoing this will affect coherence.
Everything needs to be adjusted bit by bit in order to achieve satisfaction because all components interact and an effect to one part of the spectrum also has impact upon another.
There is a learning curve, and it needs a 'hands/ears-on' to get it right, though when it is right the detail in the bass is quite amazing.
As CLS says, the bowed basses are as alike real life in a way I have never heard an enclosured LS/system ever match.
So you could try your LR4 arrangement, but it might prove to be inferior.
Cheers ........... Graham.
Yes the second choke is a partial filter, but not like a crosover which introduces phase lag.
You should be able to balance the lead of the mid with the lag of the choke/LF driver at your chosen frequency thereby optimising coherence.
The Alpha has a slight peak before roll-off. That peak can be completely tuned out with a capacitor in parallel with the output choke.
A Zobel across the driver can further cut mid frequencies by holding the load impedance low as choke impedance rises, but overdoing this will affect coherence.
Everything needs to be adjusted bit by bit in order to achieve satisfaction because all components interact and an effect to one part of the spectrum also has impact upon another.
There is a learning curve, and it needs a 'hands/ears-on' to get it right, though when it is right the detail in the bass is quite amazing.
As CLS says, the bowed basses are as alike real life in a way I have never heard an enclosured LS/system ever match.
So you could try your LR4 arrangement, but it might prove to be inferior.
Cheers ........... Graham.
T-bass for Eminence Beta15
Hi Graham,
I have some Eminence Beta15s I'd love to try your T-bass on.
http://eminence.com/proaudio_speake...il_link=BETA-15A&speaker_size=15&SUB_CAT_ID=2
Fs=35Hz, Re=6.3 ohm, Le=1.1mH, Qts=0.58, Vas=11.2 cuft, and have a nasty 10db peak at 2kHz
They are on small U frames (1 per side for now but I have 4) about 2 ft out from the corners in a small room.
I have 150VA, 2 x 45V transformers (a bit larger than your suggested 2 x 40V) and pretty much all the parts except for the 6.4mH coils. I have some 10mH inductors but they may be too far out of your spec.
Would you suggest any changes to the basic schematic for my situation (especially the 6.4mH coils as they are $$ and I don't want to buy more values than I have to)?
Thanks for your help and I can hardly wait to try out T-bass!
Cheers,
Jeff
Hi Graham,
I have some Eminence Beta15s I'd love to try your T-bass on.
http://eminence.com/proaudio_speake...il_link=BETA-15A&speaker_size=15&SUB_CAT_ID=2
Fs=35Hz, Re=6.3 ohm, Le=1.1mH, Qts=0.58, Vas=11.2 cuft, and have a nasty 10db peak at 2kHz
They are on small U frames (1 per side for now but I have 4) about 2 ft out from the corners in a small room.
I have 150VA, 2 x 45V transformers (a bit larger than your suggested 2 x 40V) and pretty much all the parts except for the 6.4mH coils. I have some 10mH inductors but they may be too far out of your spec.
Would you suggest any changes to the basic schematic for my situation (especially the 6.4mH coils as they are $$ and I don't want to buy more values than I have to)?
Thanks for your help and I can hardly wait to try out T-bass!
Cheers,
Jeff
Hi Jeff,
All I can do is encourage you to try it.
The Beta-15A is a naturally tighter LF reproducer than the Alpha-15A.
A series output choke with // capacitor will take out energisation at the Beta's peak.
The L+C will help with your corner OB position.
The transformers will do to start, but will have too high a winding resistance to do your LSs justice
If your amp is good you could go for two Betas per side with proper transformer.
Yes the choke is a little high in value, but you can still make best use of it by trying different capacitor values, maybe closer to 470uF.
You will be enjoying some nice sound over Christmas if you get this done.
Cheers .......... Graham.
PS. Of those who have tried it, *not one* has said that it does not work.
All I can do is encourage you to try it.
The Beta-15A is a naturally tighter LF reproducer than the Alpha-15A.
A series output choke with // capacitor will take out energisation at the Beta's peak.
The L+C will help with your corner OB position.
The transformers will do to start, but will have too high a winding resistance to do your LSs justice
If your amp is good you could go for two Betas per side with proper transformer.
Yes the choke is a little high in value, but you can still make best use of it by trying different capacitor values, maybe closer to 470uF.
You will be enjoying some nice sound over Christmas if you get this done.
Cheers .......... Graham.
PS. Of those who have tried it, *not one* has said that it does not work.
Keep on tuning the T-bass, I put a 0.5 Ohm resistor in series with the choke the other day.
At first it sounded OK, seemed not much changed. Listened more and more, I found the magic was "diluted". Shorting the resistor brought back the "energy" right away. So I'd better keep it without. I guess the DCR of the choke and transformer I use are high enough, so I don't need any more series R to suppress the overshoot.
At first it sounded OK, seemed not much changed. Listened more and more, I found the magic was "diluted". Shorting the resistor brought back the "energy" right away. So I'd better keep it without. I guess the DCR of the choke and transformer I use are high enough, so I don't need any more series R to suppress the overshoot.
If it even worked, wouldn't it just give you excessive bass output? This circuit it used mainly to boost the bass after dipole roll-off... now as headphones don't really roll-off at music frequencies there's nothing to compensate for.
Just work on improving the amp or (just as likely at fault) your source.
Simon
Just work on improving the amp or (just as likely at fault) your source.
Simon
Plz excuse me if my questions sound stupid
Is it needed to use this circuit with Alpha on H baffle ?
I can't decide whether to use dual Alpha's on H baffle without T-bass, or use one alpha on U (or H on I) baffle with circuit ? I am trying to reach ~93 db sensitivity...
Sry if this was mentioned before, but i am newb and i would like to know if this circuit is crossover at the same time ?
Is it needed to use this circuit with Alpha on H baffle ?
I can't decide whether to use dual Alpha's on H baffle without T-bass, or use one alpha on U (or H on I) baffle with circuit ? I am trying to reach ~93 db sensitivity...
Sry if this was mentioned before, but i am newb and i would like to know if this circuit is crossover at the same time ?
Graham Maynard said:
Graham,
From your description, I actually understand that you suggest to have no "explicit" crossover between the bass and mid driver apart from the notch + Zobel on the output of T-Bass. For me, this raises a few questions:
- obviously, there has to be a roll-off of the bass driver to blend it with the mids. Do you suggest a Zobel+notch being sufficient ? Many large bass drivers have huge peaks in the mids and also a dipole peak that in my case is around 250Hz, while the current crossover point is at 150Hz, to me it does not look like the zobel+notch would be enough.
- how do you blend the mids in ? High passed or also let without any crossover - apart from the natural dipole roll-off which - AFAIK - also introduces phase shift ?
- to me, it looks like you actually advocate a very shallow crossover from bass to mids (would be something like 2nd order acoustic). You also mention "coherence" between bass and mids: however, it is not very clear what you mean by that, as it takes both the bass and the mids driver to define coherence.
The advantage of a LR4 crossover is that it is straightforward to do it right: of course, there would be a 360 deg phase difference in the crossover area, which as I understand you consider to be very detrimental
Initially, I have understood that the main benefit of the circuit would be in the low bass, but it looks like I have not got the whole picture.
Hi CLS,
Glad you determined that - so you could hear the difference introduced by 0.5ohm !!!!!
The other aspect to be wary of is cable resistance.
Between my amp and the T-bass circuit my lead resistance is likely to be circa 0.05 ohms, and I can hear the dynamic loss introduced by as little as another 0.1ohm.
I would suggest that the low frequency T-bass circuit is amplifier rather than loudspeaker sited if very thick interconnects are not being used.
Hi kristleifur,
The T-bass is in series with and interacts with a LF driver.
In the past I have not found good headphones to be 'wanting' of bass when they are properly air sealed to the ear.
Earphones do not have the same disadvantages as a driver on an OB, however it would indeed be possible to build a T-bass, load it resistively and potentially divide this to provide a feed, but I would suggest that the same results could be acheived without using a transformer.
What makes the T-bass different to EQ is the additional first half cycle kick which overcomes driver cone mass characteristics. Headphones should not have similar disadvantage, but I will sketch a suggested circuit and add it to this thread.
It would not be cheap to make though, and it could not be allowed to have too great an impact upon reproduction lest it would became tiring. Also each headphone type has its own optimum source impedance, which would then become modified by any series connected circuit - likely deleteriously.
Hi chakija,
>>dual Alpha's on H baffle without T-bass, or use one alpha on U (or H on I) baffle with circuit<<
Two drivers are used to increase LF displacement and (if in parallel) increase sensitivity. The T-bass has already worked just fine with Alpha-15A drivers, but you would of course need something like a 100W 4 ohm amplifier if you want to drive two Alphas in parallel.
However I must leave the choice of baffle type to yourself. Two Alpha-15As on one baffle with 6" 'U' support sides and T-bass drive are likely to be quite sufficient for home use.
T-bass is used to correct transduction characteristics no matter how many drivers are used. Thus it replaces an EQ or a crossover stage implemented LP circa 75Hz to flatten a LF *driver on OB* amplitude response.
Those crossover type circuits are what makes bass so hard to 'blend' with a mid or full-range driver response, because the crossover section rotates electrical phase; whereas the T-bass has minimal action in this regard.
Hi Sreten,
You are the first person I have ever put on an Ignore List, and you will remain so until you actually build this circuit so that you might have some *real world* experience to add relevence to the way in which you imagine you are accurately advising others.
Cheers ............ Graham.
PS PS PS
Hi BZ,
You posted whilst I was writing this and there are so many questions I'll have to come back later.
Glad you determined that - so you could hear the difference introduced by 0.5ohm !!!!!
The other aspect to be wary of is cable resistance.
Between my amp and the T-bass circuit my lead resistance is likely to be circa 0.05 ohms, and I can hear the dynamic loss introduced by as little as another 0.1ohm.
I would suggest that the low frequency T-bass circuit is amplifier rather than loudspeaker sited if very thick interconnects are not being used.
Hi kristleifur,
The T-bass is in series with and interacts with a LF driver.
In the past I have not found good headphones to be 'wanting' of bass when they are properly air sealed to the ear.
Earphones do not have the same disadvantages as a driver on an OB, however it would indeed be possible to build a T-bass, load it resistively and potentially divide this to provide a feed, but I would suggest that the same results could be acheived without using a transformer.
What makes the T-bass different to EQ is the additional first half cycle kick which overcomes driver cone mass characteristics. Headphones should not have similar disadvantage, but I will sketch a suggested circuit and add it to this thread.
It would not be cheap to make though, and it could not be allowed to have too great an impact upon reproduction lest it would became tiring. Also each headphone type has its own optimum source impedance, which would then become modified by any series connected circuit - likely deleteriously.
Hi chakija,
>>dual Alpha's on H baffle without T-bass, or use one alpha on U (or H on I) baffle with circuit<<
Two drivers are used to increase LF displacement and (if in parallel) increase sensitivity. The T-bass has already worked just fine with Alpha-15A drivers, but you would of course need something like a 100W 4 ohm amplifier if you want to drive two Alphas in parallel.
However I must leave the choice of baffle type to yourself. Two Alpha-15As on one baffle with 6" 'U' support sides and T-bass drive are likely to be quite sufficient for home use.
T-bass is used to correct transduction characteristics no matter how many drivers are used. Thus it replaces an EQ or a crossover stage implemented LP circa 75Hz to flatten a LF *driver on OB* amplitude response.
Those crossover type circuits are what makes bass so hard to 'blend' with a mid or full-range driver response, because the crossover section rotates electrical phase; whereas the T-bass has minimal action in this regard.
Hi Sreten,
You are the first person I have ever put on an Ignore List, and you will remain so until you actually build this circuit so that you might have some *real world* experience to add relevence to the way in which you imagine you are accurately advising others.
Cheers ............ Graham.
PS PS PS
Hi BZ,
You posted whilst I was writing this and there are so many questions I'll have to come back later.