Re: Re: Re: Re: zero neg feedback
Of course it matters because we are not talking about subjective things; we are talking about objective things. We are not talking about whether you like chinese food or indian food. We are talking about how to design a good amp. Of course, different people have different opinions. But we don't agrue foolishly; We agrue with logical reasons. Whether Jesus resurrected on the third day can be settled by debate, by court evidence and by common sense. I don't think anyone can design any amp without knowing Ohm's law.
All I'm saying is I find it quite strange to see someone congrulating himself. It's against typical culture.
Please, I truely did not and do not mean to offend you. It's just a little comment .
Thank God for sending Jesus Christ to save me and everyone. May His name be on high forever. Amen.
James Yung
Praise! Audio
Tube_Dude said:
I don't know why you say that i was praising my self...what i was trying to say..is:
I have my opinions...and if some others don't agred with me ...doesn't matter!!
As Jonh Lennon said:
"I just believe ...in me!!!"
Of course it matters because we are not talking about subjective things; we are talking about objective things. We are not talking about whether you like chinese food or indian food. We are talking about how to design a good amp. Of course, different people have different opinions. But we don't agrue foolishly; We agrue with logical reasons. Whether Jesus resurrected on the third day can be settled by debate, by court evidence and by common sense. I don't think anyone can design any amp without knowing Ohm's law.
All I'm saying is I find it quite strange to see someone congrulating himself. It's against typical culture.
Please, I truely did not and do not mean to offend you. It's just a little comment .
Thank God for sending Jesus Christ to save me and everyone. May His name be on high forever. Amen.
James Yung
Praise! Audio
My apology to Tube_Dude.
I forgot I have asked him to pop me a email if he wanted to see some of my no FB designs. He has indeed sent me a email and I will show him one after Tues. (I've replied to him) Next time I'll try to remember not to promise so easily! ;-)
In fact, I think I will show that design here in diy.com.
James Yung
Praise! Audio
I forgot I have asked him to pop me a email if he wanted to see some of my no FB designs. He has indeed sent me a email and I will show him one after Tues. (I've replied to him) Next time I'll try to remember not to promise so easily! ;-)
In fact, I think I will show that design here in diy.com.
James Yung
Praise! Audio
Tube_Dude said:
Don't forget noise coming in from the psu - that can sound pretty rough - I guess that is why batteries or chokes are gaining popularity.
Also sound caused by a too small stability margin is worth avoiding
So apart from clipping, crossover, noise and poor stability margin distortion, basically distortion is OK !
what about 3rd harmonic ?
OK so apart from clipping, crossover, noise, stability margin and 3rd harmonic distortion, All distortion is fine by me....
or did I miss something.....
mike
whoops... forgot transient intermodulation... that's a big no no
About feedback : It's the same principle and follows the same rules, no matter if applied to a loop of one device or to a loop of five devices
About CFP : It's no longer 'local feedback' since there are two devices on the loop. CFPs may be unstable and oscillate if total phase shift is 180º for any frequency having negative feedback [It may happen with only 2 devices]. Anyway, the real flaw of the CFP is the slow turn-off of the high current device due to having a simple high value resistor as the only means to remove charge stored in the base. To quickly turn off a satuated bipolar device, a negative base current equal to half the collector current must be applied for one or two microseconds [lower currents provide longer charge removal periods]
About CFP : It's no longer 'local feedback' since there are two devices on the loop. CFPs may be unstable and oscillate if total phase shift is 180º for any frequency having negative feedback [It may happen with only 2 devices]. Anyway, the real flaw of the CFP is the slow turn-off of the high current device due to having a simple high value resistor as the only means to remove charge stored in the base. To quickly turn off a satuated bipolar device, a negative base current equal to half the collector current must be applied for one or two microseconds [lower currents provide longer charge removal periods]
Simple and stupid things may oscillate too...
The last stupid thing I've seen oscillating [on breadboard] was a simple BC550C used as a switch with some capacitive load on the collector and two 1N4148 as a baker clamp to prevent deep saturation. That thing oscillated slightly at about 14Mhz when the BC550C was turned on
Actually a baker clamp is nothing but a negative feedback circuit with three devices on the loop, so beware of 'local feedback'
The last stupid thing I've seen oscillating [on breadboard] was a simple BC550C used as a switch with some capacitive load on the collector and two 1N4148 as a baker clamp to prevent deep saturation. That thing oscillated slightly at about 14Mhz when the BC550C was turned on
Actually a baker clamp is nothing but a negative feedback circuit with three devices on the loop, so beware of 'local feedback'
Mankind needs gods as much as fish needs bycicles...
janneman :
The trick is that there is no need to apologize for crimes done 'in god's name'
Acting in the name of a supernatural being is a very clever excuse to kill or discriminate against all people you don't like with impunity [not to talk about the long tradition of pedophilia in the catholic church]
janneman :
The trick is that there is no need to apologize for crimes done 'in god's name'
Acting in the name of a supernatural being is a very clever excuse to kill or discriminate against all people you don't like with impunity [not to talk about the long tradition of pedophilia in the catholic church]
Very good explanation....
Hi Praise, Hi every body...
thx for your replies...
Sorry I wasn't not active at this forum for 3 days,,,
To Praise, don't worry about your promise, I'm not hard filling to you...I am already check my email and I've found your email..btw my design isn't finish yet and need to rebuild, but if you ask, I will send the schematic it to you.
To every body,
It's look too hard to built a good NFB amp... but I find some interest think there, but I don't know yet,,,maybe it called a natural distortion but it is sweet and I like it....I will keep encourage to develop it....My amp now just has low bandwidth now ( about 18 khz), and others technical performance is not so good, but it sound so sweet and slow, maybe like a tube amp..(just imagine, I don't have a tube amp right now, )
Best regards,,
Lukio
Hi Praise, Hi every body...
thx for your replies...
Sorry I wasn't not active at this forum for 3 days,,,
To Praise, don't worry about your promise, I'm not hard filling to you...I am already check my email and I've found your email..btw my design isn't finish yet and need to rebuild, but if you ask, I will send the schematic it to you.
To every body,
It's look too hard to built a good NFB amp... but I find some interest think there, but I don't know yet,,,maybe it called a natural distortion but it is sweet and I like it....I will keep encourage to develop it....My amp now just has low bandwidth now ( about 18 khz), and others technical performance is not so good, but it sound so sweet and slow, maybe like a tube amp..(just imagine, I don't have a tube amp right now, )
Best regards,,
Lukio
Global NFB closed around as linear stages as possible improves all of the technical and sonic characteristics. The wrong reputation of global NFB results from poor designs that try to cure hard non-linearities like class B crossover distortion and turn-on-off spikes only by means of global NFB, instead of sophisticated solution of individual stages before application of global NFB.
Together with a friend I designed and built in 1983 a class AB amplifier that is overall feedback free and non-switching. Below is the circuit of the amplifier and in the next post the servo that is used.
I removed the values and transistor types as I still have other plans with it. It will give you an idea.
It sounds very good with a very well controlled bass. In listening tests it was at par with a Stax DA-50M an definitely better than a Hafler DH220 and Bryston 2B.
The output impedance is adjusted to zero (DF is infinity) with the error cancellation adjustment, based on the Hawksford error feedback circuit. This circuit linearizes the output stage. Input stage and VAS stage use local feedback only. This topology was chosen for no interface intermodulation distortion and no transient intermodulation distortion.
Output power is 70W in 8 Ohm. Measured distortion at -20dB below clipping is 0.05% both at 1kHz and 10kHz. Measured distortion at 1dB below clipping is 0.1% at 1kHz and 0.13% at 10kHz. Low order harmonics only. Bandwidth 200kHz with input filter, 700kHz without input filter. SNR is 100dBA.
Steven
I removed the values and transistor types as I still have other plans with it. It will give you an idea.
It sounds very good with a very well controlled bass. In listening tests it was at par with a Stax DA-50M an definitely better than a Hafler DH220 and Bryston 2B.
The output impedance is adjusted to zero (DF is infinity) with the error cancellation adjustment, based on the Hawksford error feedback circuit. This circuit linearizes the output stage. Input stage and VAS stage use local feedback only. This topology was chosen for no interface intermodulation distortion and no transient intermodulation distortion.
Output power is 70W in 8 Ohm. Measured distortion at -20dB below clipping is 0.05% both at 1kHz and 10kHz. Measured distortion at 1dB below clipping is 0.1% at 1kHz and 0.13% at 10kHz. Low order harmonics only. Bandwidth 200kHz with input filter, 700kHz without input filter. SNR is 100dBA.
Steven
Attachments
Steven,
This is a clever circuit, brimming over with interesting ideas. Your performance figures look outstanding.
But it is complex, and a good pcb design would take quite some time to put together. Can you give it a single input diff pair? I suspect so......
If this gave you no perceived reduction in sonics, it would be much easier to build. In all my R&D I've never found any reason to favour two input diff pairs, at least from the sonic POV.
Thank you for sharing your ideas,
Cheers,
Hugh
This is a clever circuit, brimming over with interesting ideas. Your performance figures look outstanding.
But it is complex, and a good pcb design would take quite some time to put together. Can you give it a single input diff pair? I suspect so......
If this gave you no perceived reduction in sonics, it would be much easier to build. In all my R&D I've never found any reason to favour two input diff pairs, at least from the sonic POV.
Thank you for sharing your ideas,
Cheers,
Hugh
Thanks Hugh,
At that time complete symmetry was very popular and at least it allows for two times more local feedback for the same output. The input stage looks complex because of the input pairs using CFP's, but the difference in linearity is big compared to normal differential pairs. The input stage should deal with almost 2Vpp differential input voltage. Anyway, these are all small transistors, costing close to nothing. We made a single layer PCB, which is quite compact, has only one jumper wire and even contains the metal TO3 power transistors on an L-profile. Only at the moment I cannot find the PCB-film back. It's more than 20 years ago.
Steven
At that time complete symmetry was very popular and at least it allows for two times more local feedback for the same output. The input stage looks complex because of the input pairs using CFP's, but the difference in linearity is big compared to normal differential pairs. The input stage should deal with almost 2Vpp differential input voltage. Anyway, these are all small transistors, costing close to nothing. We made a single layer PCB, which is quite compact, has only one jumper wire and even contains the metal TO3 power transistors on an L-profile. Only at the moment I cannot find the PCB-film back. It's more than 20 years ago.
Steven
Steven, this is a very fine circuit able to give perfect results. CFP's in the input stages, the way how 2nd stage is coupled and crossover distortion compensation can lead to absolutely perfect sonic results. Borbely, Hawksford, Bruce Candy and Upupa Epops (Pavel Dudek) use similar topologies or parts of it. Yes, this is the way how to reach the top.
Hugh, I disagree. I know very well the limits of topologies like P3A or Project61 (you know what I speak about). They would never give the results and space for continuous improvement shown in Steven's circuit.
Hugh, I disagree. I know very well the limits of topologies like P3A or Project61 (you know what I speak about). They would never give the results and space for continuous improvement shown in Steven's circuit.
Interesting circuit, Steven. Yes, it is important to find novel ways to steer the output so that it has no DC offset. Personally, I believe in servos. BUT, it seems important to me that the servo is only significantly active at VERY LOW frequencies, and not as a low frequency roll-off. Years ago, I actually tried to make a servo also act as a high pass filter (ie 50Hz). This was not so good. It appears to be better to just use a quality cap to do the same thing.
Usually, it is best to minimize 'global' feedback. Probably, feedback pairs are better, but not perfect. In any case, the most linear circuit, (without feedback) is usually best.
Usually, it is best to minimize 'global' feedback. Probably, feedback pairs are better, but not perfect. In any case, the most linear circuit, (without feedback) is usually best.
Pavel,
No problem. I accept you and I disagree; it is the nature of technology. Neither did I mention my own designs; I am not self-serving, I believe you realise this.
I believe Steven has come up with something very impressive, and said so with sincerity. The circuit is inspired and I am studying it in detail right now. While I believe a fully complementary design is unnecessary, I would need to verify it in listening tests for this design. Hell, I could be wrong!! (It happened once before, in 1962, IIRC.)
Pavel, there is a huge difference between a commercial design and a full-on, academically inspired and researched conceptual design. I now realize the difference, and do not in any way denigrate this design, which uses the fully complementary philosophy of the Elektor Crescendo. However, I draw much of my inspiration from the conceptual circuits to be found in academia, and now from this one. Steven, Pavel, are you familiar with Peter Blomley's non-switching Class AB of the early seventies? There's another to consider.
Pavel, you are a touchy fellow, aren't you?
But we still love you....... A man who designs good power followers can't be all bad!
Cheers,
Hugh
No problem. I accept you and I disagree; it is the nature of technology. Neither did I mention my own designs; I am not self-serving, I believe you realise this.
I believe Steven has come up with something very impressive, and said so with sincerity. The circuit is inspired and I am studying it in detail right now. While I believe a fully complementary design is unnecessary, I would need to verify it in listening tests for this design. Hell, I could be wrong!! (It happened once before, in 1962, IIRC.)
Pavel, there is a huge difference between a commercial design and a full-on, academically inspired and researched conceptual design. I now realize the difference, and do not in any way denigrate this design, which uses the fully complementary philosophy of the Elektor Crescendo. However, I draw much of my inspiration from the conceptual circuits to be found in academia, and now from this one. Steven, Pavel, are you familiar with Peter Blomley's non-switching Class AB of the early seventies? There's another to consider.
Pavel, you are a touchy fellow, aren't you?
But we still love you....... A man who designs good power followers can't be all bad!
Cheers,
Hugh
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.