here is the ac analysis.
as you can tell, it goes down to about 10hz, far beyond what my ears can hear and what my speakers can reproduce. and I don't have any Cd players that output in the 0-2hz range, nor do I have any analog amplifier that takes spdif output directly.
note: C7 was there for stability testing. V4 to test psrr. the circuit can run on dual rails without modification but I don't recommend that as it has poor DC output control and without an output capacitor it can be dangerous to your speaker.
if the bass isn't good enough for you, you can use bigger bootstrap capacitors, or increase R4/R12 but you would need to recalculate the rest of the circuitry. in theory, the use of a CCS here (in place of the bootstrap) should improve bass but if you can go down to 10hz, it should be more than enough for most applications.
as you can tell, it goes down to about 10hz, far beyond what my ears can hear and what my speakers can reproduce. and I don't have any Cd players that output in the 0-2hz range, nor do I have any analog amplifier that takes spdif output directly.
note: C7 was there for stability testing. V4 to test psrr. the circuit can run on dual rails without modification but I don't recommend that as it has poor DC output control and without an output capacitor it can be dangerous to your speaker.
if the bass isn't good enough for you, you can use bigger bootstrap capacitors, or increase R4/R12 but you would need to recalculate the rest of the circuitry. in theory, the use of a CCS here (in place of the bootstrap) should improve bass but if you can go down to 10hz, it should be more than enough for most applications.
Attachments
hi...
hi,tlf9999:
do you think the CD-walkman has bad bass output?
as I know that the most CD output only 1V(P-P),my CDwalkman is as same as normally...
why you all don't want to take more width sounding range's AMP of my design?perhaps,if you using it to drive your speaker,more deep and strong bass you will hear...
but first must have somebody to build one and sure it sounds good and worth to build...
have fun,good night!
WINCO
hi,tlf9999:
do you think the CD-walkman has bad bass output?
as I know that the most CD output only 1V(P-P),my CDwalkman is as same as normally...
why you all don't want to take more width sounding range's AMP of my design?perhaps,if you using it to drive your speaker,more deep and strong bass you will hear...
but first must have somebody to build one and sure it sounds good and worth to build...
have fun,good night!
WINCO
DarkHorse, I don't know anything about your walkman.
As to wide frequency response, well, if I cannot hear DC signals, and my speakers cannot reproduce DC signals, it does me no good for my amp to be able to reproduce DC signals,
.
The point is that wider frequency response is good only to the point of being useful. If I cannot hear it, it will be a waste of time and resources for me to extend my amp's range beyond it.
In this case, I think being able to do 40 - 50hz is more than adequate. and my experience is that the original jlh, with no modification, does that just fine.
I think there may be something wrong with your paritcular execution. the first would be the gain - 40x is way too much for your source -> at 20vpp, the amp is clipping at maximum input (especially with mosfet output devices) which generates the impression of dry bass.
I would dial the gain back to a more manageable level and see if it still has "weak" bass.
As to wide frequency response, well, if I cannot hear DC signals, and my speakers cannot reproduce DC signals, it does me no good for my amp to be able to reproduce DC signals,
.The point is that wider frequency response is good only to the point of being useful. If I cannot hear it, it will be a waste of time and resources for me to extend my amp's range beyond it.
In this case, I think being able to do 40 - 50hz is more than adequate. and my experience is that the original jlh, with no modification, does that just fine.
I think there may be something wrong with your paritcular execution. the first would be the gain - 40x is way too much for your source -> at 20vpp, the amp is clipping at maximum input (especially with mosfet output devices) which generates the impression of dry bass.
I would dial the gain back to a more manageable level and see if it still has "weak" bass.
Hi Darkhorse,
( I'm trying to improve the signal to noise ratio here ! )
I believe you said you were using 'line out' on the Sony.
Can you use 'headphone out' to feed your amplifier?
If yes, feed your amplifier with CD headphone output.
Then construct an output attenuator for your amplifier using say 120 ohms and 4.7 ohms.
Now move your (32ohm) headphones backwards and forwards between amplifier input and unloaded amplifier attenuator output.
Is the bass/mid/treble tonal balance the same between amplifier input and amplifier output ?
With your loudspeaker in another room connect its cable to amplifier output; is the headphone monitored output tonal balance altered by driver load/dynamics/back-EMF ?
Decent headphones should be capable of resolving any bass loss and phase shifts to the lowest limits of your source and hearing.
Indeed this would be an ideal way of directly observing the weakness of series output capacitor coupling, especially the values used in the original JLH class-A, and how increasing the value can reduce waveform distortion arising between amplifier input and output terminals.
Also do try the split feedback resistors, they will flatten that undesirable peak close to 1Hz.
Hi Mike,
Yes I quite agree with the need to minimise phase shifts for bass frequencies - right to the lower limit of human hearing - and this IS where a series output capacitor can be a problem.
Cheers ......... Graham.
( I'm trying to improve the signal to noise ratio here ! )
I believe you said you were using 'line out' on the Sony.
Can you use 'headphone out' to feed your amplifier?
If yes, feed your amplifier with CD headphone output.
Then construct an output attenuator for your amplifier using say 120 ohms and 4.7 ohms.
Now move your (32ohm) headphones backwards and forwards between amplifier input and unloaded amplifier attenuator output.
Is the bass/mid/treble tonal balance the same between amplifier input and amplifier output ?
With your loudspeaker in another room connect its cable to amplifier output; is the headphone monitored output tonal balance altered by driver load/dynamics/back-EMF ?
Decent headphones should be capable of resolving any bass loss and phase shifts to the lowest limits of your source and hearing.
Indeed this would be an ideal way of directly observing the weakness of series output capacitor coupling, especially the values used in the original JLH class-A, and how increasing the value can reduce waveform distortion arising between amplifier input and output terminals.
Also do try the split feedback resistors, they will flatten that undesirable peak close to 1Hz.
Hi Mike,
Yes I quite agree with the need to minimise phase shifts for bass frequencies - right to the lower limit of human hearing - and this IS where a series output capacitor can be a problem.
Cheers ......... Graham.
Graham Maynard said:
Yes, but why ? I only experienced that it behaves like that, the deep
bass experience improves by reducing phaseshift, even when amp already
has -3db at 10hz. This made me using up to 10uf at input and it
did definitely improve low bass...
The weird thing is, speakers already apply loads of phaseshift at
these frequencies, is this about phasecorrect damping ?
Mike
I can illustrate a point that Graham made earlier. I was experimenting with a motional feedback woofer, and the first version had a rising low frequency response, the gain increasing below the 61Hz resonant frequency. I had a Sony CDP470 cd player at the time, (before I knew any better), and played a disc which came free with Hi-Fi World. This CD was produced in conjunction with Meridian (as a promotion for the brand), and was live organ music recorded in the church. Well, probably some Cathedral. I thought that this would be a good workout for a sub, because organs can go very low All I noticed, though, was the cone flapping about due to the very low frequency wind noise which was faithfully recorded at the venue, and increased due to the rising LF response of the MFB sub. The movement was so slow it was visible, although I don't know the frequencies involved.
Graham is correct in my experience.
All I noticed, though, was the cone flapping about due to the very low frequency wind noise which was faithfully recorded at the venue, and increased due to the rising LF response of the MFB sub. The movement was so slow it was visible, although I don't know the frequencies involved.Graham is correct in my experience.
Hi Darkhorse!
My and my friends and listener's loudspeakers are "ordinary" 2- way, TML or BR speakers. I use the amp with Yamaha CDX-993 CD player directly connected. The amp is dual mono: 2*10.000uF LC filtered PSU/channel, 4700uF BC caps on the outputs, MJL15003 transistors. So, nothing special.
I don't care for simulatiors software, if the empiric shows good and nice sound!
I don't want argue with you, just to tell my experiences.
Greets:
Tyimo
My and my friends and listener's loudspeakers are "ordinary" 2- way, TML or BR speakers. I use the amp with Yamaha CDX-993 CD player directly connected. The amp is dual mono: 2*10.000uF LC filtered PSU/channel, 4700uF BC caps on the outputs, MJL15003 transistors. So, nothing special.
I don't care for simulatiors software, if the empiric shows good and nice sound!
I don't want argue with you, just to tell my experiences.
Greets:
Tyimo
tlf9999 said:
hi,tlf9999:
your are mistake from me,I got the CD's signal from line out,but I have add one 50Kohm VR to control the volume level,so from my speaker will not occur Clipping case.your JLH1969 make you satisfy,congratulation!but I don't feel any so I have to do something with it...now I know everybody's fortune is different.and do what that you like to do!
regards
WINCCO
Graham Maynard said:
hi,Mr.Garham:
thank you for your point,but my walkman's headphone output can't output enough signal voltage to drives my amp,because it can only offer current to drive my headphone...
I don't want to do more change with my PNP version,I trust there isn't serious thing with so low frequency(under 10Hz),I will plan to build one in future...
regards
WINCO
Tyimo said:Hi Darkhorse!
My and my friends and listener's loudspeakers are "ordinary" 2- way, TML or BR speakers. I use the amp with Yamaha CDX-993 CD player directly connected. The amp is dual mono: 2*10.000uF LC filtered PSU/channel, 4700uF BC caps on the outputs, MJL15003 transistors. So, nothing special.
I don't care for simulatiors software, if the empiric shows good and nice sound!
I don't want argue with you, just to tell my experiences.
Greets:
Tyimo
hi,Tyimo:
it doesn' matter!perhaps I met the wrong things and let my point is unsatisfy with JLH1969...
your say it can process well bass that is enough...do your want to do!
good day!
WINCO
johnnyx said:I can illustrate a point that Graham made earlier. I was experimenting with a motional feedback woofer, and the first version had a rising low frequency response, the gain increasing below the 61Hz resonant frequency. I had a Sony CDP470 cd player at the time, (before I knew any better), and played a disc which came free with Hi-Fi World. This CD was produced in conjunction with Meridian (as a promotion for the brand), and was live organ music recorded in the church. Well, probably some Cathedral. I thought that this would be a good workout for a sub, because organs can go very low
Graham is correct in my experience.
hi,johnnyx:
you say so much,but my english is poor...
so please tell me the output curve graph,low frequency(lower 100Hz) turn up is acceptable,is it?
greets!
WINCO
Hi Mike,
It is very easy to understand this low frequency phase induced waveform distortion.
An amplitude plot of the original JLH is -3dB down at about 13Hz, and in its day this was fair enough.
However the phase shift at this frequency is in excess of 70 degrees leading.
At 30Hz the lead is 34 degrees; for 90Hz, 10 degrees.
Draw a 30Hz fundamental; overdraw a third harmonic; add the component amplitudes, and re-draw the resultant.
Now do the same with the fundamental leading the harmonic by 25 degrees, as per the JLH capacitor coupled design.
All of a sudden the redrawn resultant waveform distortion is considerable, even though the sine distortion of each component measured in isolation would be slight !
This is what the loudspeakers reproduce - no matter how imperfect they might already be !
( Any wonder JLH went for direct output coupling with his update ! )
Hi Johnny,
Yes indeed that was the bass cone flapping/breathing I was reporting.
In a full range design it can cause wanted signal distortion, and can occasionally be introduced by pre-mastering analogue-digital induced error alone.
Some 'exponents' are not actually aware that this problem can arise; they have never observed it, and they don't know about it !
Also, whilst many listeners are entirely satisfied with the reproduction that results from a higher low frequency roll-off, not everybody is, and thus all different personal approaches to reproduction should be respected.
Hi Nelson,
Original JLH open loop can be circa 50dB, with device dependency.
Hi Darkhorse,
JLH stipulated that his class-A circuit should be driven by a low impedance pre-amp/source. In this regard he offered a circuit for a separate fet buffer for use with high impedance source inputs.
I am not aware of you mentioning the 50k pot before. This is a high resistance to common emitter bipolar circuitry and will seriously degrade amplifier capabilities.
It is the input components that set up the reference potential against which the NFB loop must dynamically act, and the JLH is a fine performer when driven by a pre-amplifier that has an integral output buffer stage.
Cheers ........ Graham.
It is very easy to understand this low frequency phase induced waveform distortion.
An amplitude plot of the original JLH is -3dB down at about 13Hz, and in its day this was fair enough.
However the phase shift at this frequency is in excess of 70 degrees leading.
At 30Hz the lead is 34 degrees; for 90Hz, 10 degrees.
Draw a 30Hz fundamental; overdraw a third harmonic; add the component amplitudes, and re-draw the resultant.
Now do the same with the fundamental leading the harmonic by 25 degrees, as per the JLH capacitor coupled design.
All of a sudden the redrawn resultant waveform distortion is considerable, even though the sine distortion of each component measured in isolation would be slight !
This is what the loudspeakers reproduce - no matter how imperfect they might already be !
( Any wonder JLH went for direct output coupling with his update ! )
Hi Johnny,
Yes indeed that was the bass cone flapping/breathing I was reporting.
In a full range design it can cause wanted signal distortion, and can occasionally be introduced by pre-mastering analogue-digital induced error alone.
Some 'exponents' are not actually aware that this problem can arise; they have never observed it, and they don't know about it !
Also, whilst many listeners are entirely satisfied with the reproduction that results from a higher low frequency roll-off, not everybody is, and thus all different personal approaches to reproduction should be respected.
Hi Nelson,
Original JLH open loop can be circa 50dB, with device dependency.
Hi Darkhorse,
JLH stipulated that his class-A circuit should be driven by a low impedance pre-amp/source. In this regard he offered a circuit for a separate fet buffer for use with high impedance source inputs.
I am not aware of you mentioning the 50k pot before. This is a high resistance to common emitter bipolar circuitry and will seriously degrade amplifier capabilities.
It is the input components that set up the reference potential against which the NFB loop must dynamically act, and the JLH is a fine performer when driven by a pre-amplifier that has an integral output buffer stage.
Cheers ........ Graham.
DarkHorse said:
it should have enough voltage output as well. give it a try and let's say what happens.
I also simulated yours (2n5401 input, 2n5551 vas, and irf510 output). they are biased similarly to yours (1.3ma input, 4.5ma vas on 9ma ccs, and 600ma output). the gain is set at 40x. As I don't have a model for 2sk170, I used a perfect ccs in my circuit.
It performed identical to my version of jlh posted earlier. I tried to dial back the gain, or to increase the output impedance of the source (to 100k) and the perform is unaltered.
So in theory yours should work. I suspect now that the problem is either in wiring or in component selection.
Source input impedance
Graham,
You mention that 50K is a high impedance for the bipolar input of the JLH. I am happy with the sound of the JLH using a 10K (approx) attenuator though I have never tried an input buffer. I'd therefore be interested in what you suggest is a maximum value here before degrading effects become audible, also how adverse effects caused by using a source with too high an output impedance might sound in practice.
Thanks,
Tim.
Graham,
You mention that 50K is a high impedance for the bipolar input of the JLH. I am happy with the sound of the JLH using a 10K (approx) attenuator though I have never tried an input buffer. I'd therefore be interested in what you suggest is a maximum value here before degrading effects become audible, also how adverse effects caused by using a source with too high an output impedance might sound in practice.
Thanks,
Tim.