| lineup |
It is a very original idea.
I have seen something like this, but WITH global feedback.
Distortion will be more like in some tube amplifiers. Not very low.
Maybe this amp can make good sound.
Very interesting idea.
:) |
|
|
| peranders |
| quote: | Originally posted by X.G.
the schematic is designed by my friend Locky_Z.he ask me to post here....
this is his post of the Hifidiy.net
http://www.hifidiy.net/dispbbs.asp?...ID=21315&skin=0
it is current output amp.the output current IL=Vin * Rc / (Rsc * R6)
any comment? |
Why not simulate it but remember that current sources are non-ideal and they are different in real life. Spontaniously I feel that the opamp is working a comparator => won't work as an amplifier. Which were the supposed benefits? Just different? |
|
|
| darkfenriz |
Intersting one and very unusual.
I think that current duty cycle may be an issue, namely both op-amp and output stage should have the same current duty cycle. Another words they should work in EXACTLY the same class. Otherwise differential amplifiers may get overdriven. |
|
|
| lumanauw |
| quote: | | Which were the supposed benefits? Just different? |
A friend of mine proposes similiar idea some time ago. The idea is to make an output stage that never turn off. Output stage is a current mirror+amplification with RC/RSC. And the current passing RC never turns off, it is the bias of the opamp at minimal point. So the current to the output RSC will also never turn off.
What is the value of R6 in the formula? Is it 470ohm? |
|
|
| Tube_Dude |
| quote: | Originally posted by peranders
Spontaniously I feel that the opamp is working a comparator => won't work as an amplifier. |
The OP amp work as a unity gain follower and the output signal is obtained at the power supply pins , because the output of the OP Amp is loaded with a low value resistor.
The current trough the power pins will be equal to the current in the output resistor, more the currents of the others internal stages of the OP amp.
As the amp is not closed in a overall feedback loop, this kind of OP amp driver , will introduce many nonlinearitys. |
|
|
| peranders |
| quote: | Originally posted by peranders
Spontaniously I feel that the opamp is working a comparator => won't work as an amplifier. |
...he, he, now I see the feedback of the opamp :up: I must have been :bfold: yesterday :nod: |
|
|
| Upupa Epops |
| Driving current is too low = low SR, high distortion... |
|
|
| jacco vermeulen |
| quote: | Originally posted by peranders
I must have been :bfold: yesterday |
Maybe you are today. :clown:
Can someone tell me which of the posted circuit is so surprising?
Output stages driven from opamp rails is 20 years old, especially MOSFET output stages.
Only difference i see is omiting the feedback, which does not look like a smart thing to me. |
|
|
| Tube_Dude |
| quote: | Originally posted by jacco vermeulen
Can someone tell me which of the posted circuit is so surprising?
Output stages driven from opamp rails is 20 years old, especially MOSFET output stages.
Only difference i see is omiting the feedback, which does not look like a smart thing to me. |
The output stage use feedback of current ,that increase the output impedance , via the 0,22 Ohms drain resistors of the output Mosfets. |
|
|
| jacco vermeulen |
Thanks Jorge,
i can see that the overall feedback has been replaced by the differentials and the drains supply feedback.
But what does it solve. Like Pavel said, distortion will be high, and it introduces more problems. |
|
|
| Tube_Dude |
| quote: | Originally posted by jacco vermeulen
Thanks Jorge,
i can see that the overall feedback has been replaced by the differentials and the drains supply feedback.
But what does it solve. Like Pavel said, distortion will be high, and it introduces more problems. |
Surely Jacco , but I never say that this is a good design... ;) |
|
|
| jacco vermeulen |
| True, obrigrado. |
|
|
| lineup |
I once build an amp using OP-amp this way.
The output was taken from collectors of output transistors.
Global Feedback back to OP-input,
was taken from a small resistor between load and ground.
The relationship between this resistor and load impedance set the gain.
You can call it current feedback, as same current in load as in small resistor.
I got problems with stability.
...................
Hi-Fi = High Fidelity, is to minimize all distortions.
Because distortion tells result is not like input.
And fidel = be true to
High-End is not always the same.
Because an amp is high-end it does not have to be Hi-Fi.
A low cost amp can be more Hi-Fi.
And because you amp is 0.001% distortion of all sorts distortion
you can not be sure listening to Hi-Fi.
Yes, you can be listening to Hi-Fi,
if BOTH amps and loudspeakers are very low distortion of all sorts.
We have to not confuse amplifier fidelity with total sound system fidelity.
Many people is not mainly in search for Ultra High Fidelity.
Some are - but not everybody.
Tube amplifiers, and other, shows us,
amplifiers does not have to be ultra Hi-Fi to sound good.
But to call them Hi-Fi amplifiers would not be correct.
One thing is however true:
You can not let total sound system distortion levels be TOO HIGH,
because at some point, it will NOT sound good.
===================================
As I said in my first post,
this circuit will have distortion in the magnitude of some tube amplifiers.
My guess is 0.5-2% at normal level.
Question is what this distortion will do to sound.
Can it sound good?
I wonder if this 'new idea' amp is stable?
And if not, how can it be stabilized. |
|
|
| Tube_Dude |
| quote: | Originally posted by lineup
I once build an amp using OP-amp this way.
The output was taken from collectors of output transistors.
Global Feedback back to OP-input,
was taken from a small resistor between load and ground.
The relationship between this resistor and load impedance set the gain.
You can call it current feedback, as same current in load as in small resistor.
I got problems with stability.
[/B] |
Have you put a zobel 10 Ohms / 220nF across the speaker load ? If not the inductance of the speaker will create another pole , and the stability becomes problematic. |
|
|
| Upupa Epops |
| Many people don't know, what they can... Many of them can by listening to take " impression " and for manufacturers is it good start point for washing brains... But it is all the same story... I'm so tired.... |
|
|
| Ultima Thule |
Voltage in, current out... peranders, you are partially right, without load the whole amplifier behaves like a comparator, it's a transconductance amplifier! :cool:
Cheers Michael |
|
|
| 81bas |
Hello!
did anyone at least simulate this schematic? =)
surprisings are:
1. automatic queiscent current regulation
2. output stage is never switching off
3. each output mosfet is linearized by negative current feedback
Thanks! |
|
|
| davidsrsb |
| As this is a transconductance amplifier, the gain will go very high into real speaker loads at resonances. Valve amps normally have voltage feedback to get the output impdedance down, just not as low as most solid state. This one will have an output impedance in the megaohm range. |
|
|
| X.G. |
| quote: | Originally posted by davidsrsb
As this is a transconductance amplifier, the gain will go very high into real speaker loads at resonances. Valve amps normally have voltage feedback to get the output impdedance down, just not as low as most solid state. This one will have an output impedance in the megaohm range. |
the type of this amp is not my taste :xeye:
my friend Locky_z did not build the amp in practice....
I re-designed a pre amp circuit which original designed by my another friend CWW11.its hardcore is super-pair stage.he did not view the Diyaudio and did not know the super-piar topol,but he re-invented the wheel ;) :D
I name the circuit: Non-NFB DC pre amp of Dual Converter. I had simulated,build and adjust for 4 months....because my design is for the Hifidiy.net which is DIYer-groups and the Hifidiy.net will sell the pre amp kits,I don't know that I post the detail of it on here will broke the rule of Diyaduio forum or not.
can I post here?
this is the test amp which my buddy build:
I continue to modify it two weeks ago:
 |
|
|
| jcx |
simulation will be difficult, most spice op amp models don't even try to sim power supply currents
your best bet is to use a complete device level model of a disctrete op amp like the jensen 990 |
|
|
| peranders |
| X.G. did you or your buddy make the pcb? In any case very nice looking. Do you possibly have any schematic? |
|
|
| X.G. |
hi, peranders.
the PCB made by the FHIFIDIY.net,and it drawn by my buddy AEON,its schematic is designed by me.
the post which we build and discuss is on here:
http://www.hifidiy.net/dispbbs.asp?...ID=19991&skin=0
sorry,I don't finish drawing of the whole schematic which all in my brain ;) ,but is the basic topology of it show below:
it have two JFET buffers,one for input and perform V/I convert which V/I resistor is R7,another for output and perform isolation.
the middle stage is the current transmission of super-pair and R13 is the volume controler which perform I/V converter.the gain of the pre amp=R13/R7
the pre-amp is Non NFB,in other words,it don't have loop feed back but have strong local NFB;) .the simulator tell me the distortion in ordinay play level is less than 0.01% and mainly from the JFET buffers.I don't have the distortioner to verified it:xeye: ,but I have check the BW of it in pratice....the BW of the pre amp is 700KHZ@-3dB!!! when R13(R10 in the pic below) is less than 1K.
the sound of it is very smooth, very detail, much much better than my OPA627+BUF634 pre which show here:
http://www.diyaudio.com/forums/show...5712&highlight=
I continue to modify it now:
|
|
|
| davidsrsb |
20 years ago Vellerman used to do a power amp kit with mosfets driven from common base transistors supplying an opamp.
Anyone remember a common emitter output design called the "Texan" back in the 70s |
|
|
| lumanauw |
Hi, XG,
I study your schematic, and it's actually a Folded Cascode, modified by placing Q11 and Q15 so the output is non-inverting (ordinary folded cascode gives inverting output).
Q1 and Q9 is for helping setting the voltage reference of the diodes?
The folded cascode is Q11 towards Q8.
But one thing is not clear. If the current passing the 220ohm resistor is 6.8mA, how many mA is passing through Q11, and how many mA is passing through Q8? I cannot calculate how many % of biasing current is passing to front path and end path. |
|
|
| X.G. |
hi,lumanauw| quote: | Originally posted by lumanauw
it's actually a Folded Cascode |
yes.....depend on what you think.maybe it can seen as a special diff stage+ special CB stage+buffer.
| quote: | Originally posted by lumanauw
Q1 and Q9 is for helping setting the voltage reference of the diodes?
The folded cascode is Q11 towards Q8.
|
Q1-Q8,Q2-Q9,Q6-Q11 andQ4-Q15 all are super-pair stages....you can see some posts of them in digital area,like this:
http://www.diyaudio.com/forums/show...5051&highlight=
http://www.diyaudio.com/forums/show...9750&highlight=
and more:
http://www.diyaudio.com/forums/sear...rder=descending
| quote: | Originally posted by lumanauw
But one thing is not clear. If the current passing the 220ohm resistor is 6.8mA, how many mA is passing through Q11, and how many mA is passing through Q8? I cannot calculate how many % of biasing current is passing to front path and end path. |
the current of Q8 is close to Q1, Q11 is close to Q6,and depend on the temp couple.the simulate show something for that,and my experience show the result.I think you can calculate all of them now.
PS:Are you a Chinese,lumanauw?
X.G. |
|
|
| lumanauw |
Hi, XG,
| quote: | | the current of Q8 is close to Q1, Q11 is close to Q6,and depend on the temp couple.the simulate show something for that,and my experience show the result.I think you can calculate all of them now. |
I don't get it. The current through Q1=Q6 (neglecting base currents), since they are in the same current path.
| quote: | | PS:Are you a Chinese,lumanauw? |
I'm Indonesian. Do you know where Indonesia is? |
|
|
| X.G. |
| quote: | Originally posted by lumanauw
I don't get it. The current through Q1=Q6 (neglecting base currents), since they are in the same current path.
|
yes,the current of Q1=Q6,and the current of Q1=(Vcc-Vref)/R18......oh! I made a mistake before:xeye: :eek: ,correct it :the current of Q8 is not(!) close to Q1,the current of Q11 is close to Q6.one sample of my experience,IIRC,when Q6=1.50mA,Q11 is 1.53mA,and Q8 can reach to 8mA
now,the total current of Q1,Q8 and Q11 equal to Vref/R1,so you can easily caculate the the current of Q8.
| quote: | Originally posted by lumanauw
I'm Indonesian. Do you know where Indonesia is? |
oh......I had viewed your post which about FM acoutis and saw your picture....
I know where Indonesia is,of course :cool: ;) ,but I don't know where Bandung is:cannotbe: |
|
|
| lumanauw |
Hi, XG,
Yes, it's funny. Most people don't know where Indonesia is, but know where Bali is. Bali is a small part of Indonesia :D |
|
|
| jacco vermeulen |
Overhere people go to a Chinese restaurant or take-away to eat Indonesian food.
Gila !!!
David,
you are too much of a busy person to come from Java. :clown: |
|
|
| lumanauw |
Hi, Jacco,
Do you like "Satay" or "Masakan Padang"? I think there are alot of those restaurants in Netherlands :D |
|
|
| Pingrs |
"Anyone remember a common emitter output design called the "Texan" back in the 70s"
My first amp. This was a "Practical Wireless" pre/power amp design by a TI engineer. Very popular because of ease of construction (741/748 op amps throughout + o/p stage). Combined it with a Skingley-Thompson (WW) FM tuner. Made three of them on home made pcb's. Still going strong.
I still have a spare FR4 pcb!
Brian. |
|
|
| davidsrsb |
Malaysian (Kajang) satay is sweeter than the Indonesian.
The Texan stood out for its better than average sound quality in its day, most competing (more conventional) kit amps were terrible. |
|
|
| jacco vermeulen |
Hi David,
i've lived in The Hague for 10 years, you'll find an Indonesian restaurant or Toko on every corner there.
They all have masakan padang, i am very fond of rendang.
Type "sate" in google, that should give you an idea how common satay is here.
Every french fries shop here has satay too, but not as good.
I'm very fond of Thai style satay: thin sliced, folded around the pin, grilled on charcoal.
I've been cooking Indonesian food since i was 10, also do my own sambals and atjars since then.
(and have become sambal badjak addicted, PEDIS !)
www.pasarmalambesar.nl/english/eng.htm
Anyone got a schematic of the Texan ? |
|
|
| Cortez |
A new idea, or only a question:
What if we would use a separated PS for the drivers in the output stage ?
(For the first transistors in the darlington)
Base concept: when the output transistors load the main PS, it
should give a lot of current in a short time, so only a less current
would remain to the drivers in the same time, when the PS is shared. |
|
|
| Cortez |
| Nothing ? Benefits / drawbacks, experiences ? |
|
|
|
