lumanauw said:
You are going to need several additional gain stages inserted between the drivers and the output devices in order to make error current negligible and circumvent all its ill effects. Also, the additional gain will make your input stage and VAS redundant since you would just be able to feed the bases directly with the input signal and the emitters with a fraction of the output signal.
This means that by optimizing your circuit we will end up with pretty mucho the same thing as Alexander amp, maybe with discrete input stage. Your optimized schematic is going to look much like an amplifier called 'Titan 2000' published in Elektor some time ago (well, maybe a bit les ostentatious).
Hi, PMA,
Thanks for the advice and the sim. I really appreciate it
Post #42 :
In what value of R5 the curve starting to show no "peaking"? And what is the "peaking" means?
I assume the blue one with peaking is the effect of R5 value, and the red one is the distortion figure.
What is the units in horizontal line. Is it frequency? The blue curve show acceptable below 100k. Is this below 100khz? But the red one starts non linear (between curves) at 10k, I think.
Post #43
I look at the distortion below 10khz. Below 10khz, the even orders are bigger than odd order. Above 8k, the blue one is everywhere. This one scares me The blue one dominated by higher frequency ones. If my interpretation is right, this won't result in nice sounding amp, too many higher order harmonics.
Do you have the same type of curve from your NP-PMA?
Thanks for the advice and the sim. I really appreciate it
Post #42 :
In what value of R5 the curve starting to show no "peaking"? And what is the "peaking" means?
I assume the blue one with peaking is the effect of R5 value, and the red one is the distortion figure.
What is the units in horizontal line. Is it frequency? The blue curve show acceptable below 100k. Is this below 100khz? But the red one starts non linear (between curves) at 10k, I think.
Post #43
I look at the distortion below 10khz. Below 10khz, the even orders are bigger than odd order. Above 8k, the blue one is everywhere. This one scares me
The blue one dominated by higher frequency ones. If my interpretation is right, this won't result in nice sounding amp, too many higher order harmonics.Do you have the same type of curve from your NP-PMA?
Hi, EVA,
They all takes the global feedback from the output speaker node.
(Infact, I've experienced Alexander amp using NE5532/TL082 as its brain and C5200,4 pairs for 1 channel, as the current generator. In that experiment, I took the feedback from speaker node. At first the sound is stunning, revealing every detail like never heard before. But it is fatiquing. I'm not blaming the topology, it may be caused by crossover distortion, because at that time I use EF darlington output with very low bias. Since then, I don't continue with this Alexander amp)
In this design I wanted to stress on the
-Separation of feedback node from speaker node by R5
-Having T6-T7 tries to do EC for small ripples caused by back EMF, so that this back EMF stops at A and don't enter the differential input.
-Using R5 for getting current drive effect.
The goals are different than building Alexander amp type Alexander amp stresses on very fast amp, very high slew rate and very high damping factor (very low output impedance).
Now, I wanted none of this 3 "good" property of Alexander amp
But I will still need input and VAS, dont I? Because the EC around T6-T7 is just a "voltage buffer", do not have voltage gain. Voltage gain will be still need to be generated by input+VAS.
But I don't want to build types of "current feedback" like Alexander or Titan
They all takes the global feedback from the output speaker node. (Infact, I've experienced Alexander amp using NE5532/TL082 as its brain and C5200,4 pairs for 1 channel, as the current generator. In that experiment, I took the feedback from speaker node. At first the sound is stunning, revealing every detail like never heard before. But it is fatiquing. I'm not blaming the topology, it may be caused by crossover distortion, because at that time I use EF darlington output with very low bias. Since then, I don't continue with this Alexander amp)
In this design I wanted to stress on the
-Separation of feedback node from speaker node by R5
-Having T6-T7 tries to do EC for small ripples caused by back EMF, so that this back EMF stops at A and don't enter the differential input.
-Using R5 for getting current drive effect.
The goals are different than building Alexander amp type
Alexander amp stresses on very fast amp, very high slew rate and very high damping factor (very low output impedance). Now, I wanted none of this 3 "good" property of Alexander amp
lumanauw said:
David,
blue curves are amplitude characteristics for R17 as a parameter. R17 = 0, 40, 80 Ohms ... etc. It is my CFP output stage according to http://www.diyaudio.com/forums/showthread.php?postid=700155#post700155 simulated, with R18 disconnected. Just an influence of R17.
Red curves are phase characteristics.
X axis is frequency. 100k = 100kHz.
lumanauw said:Hi, PMA,
Post #43
I look at the distortion below 10khz. Below 10khz, the even orders are bigger than odd order. Above 8k, the blue one is everywhere. This one scares me
Do you have the same type of curve from your NP-PMA?
Take into account that this is for the open-loop CFP output stage. And pretty high input voltage, 18V. For the values under 10V the distortion is very low. The distortion also depends on quiescent current and transistors used. The only aim was to show you that R17 (R5) increases the distortion, and that it is not good to use it.
Yes, I have measurements of my error correction amp, not only simulation.
Also, you can see measurements of the CFP class A output stage here:
http://web.telecom.cz/macura/pma1_measure_en.html
Nelson Pass said:
Nelson,
As far I have seen your circuit and patent, it is indeed a form of error correction but the specifics of Hawkfords idea is that it exactly feeds back the error to be corrected. At the same time, the correction ALSO increases the input voltage as a kind of boot strap circuit. This needs a precise ratio, if the feedback is either larger or smaller the results are worse. I don't think that is the case with your circuit. So, even if a circuit has the same topology but a different ratio it would still be different; but if I have missed something can you point me to a circuit of yours that has it? I would be most interested.
As far the patent stuff, it is always difficult to know who inspired whom when and how much. Most of us stand on the shoulders of giants. (You must feel the weight of all of us by now
).Jan Didden
Nowdays, in the era of internet and global world, it is very easy for one to combine a certain patent with this idea, with that idea, resulting another idea.
The ones that is really amazing is the ones appears from the older years, before the internet, and the idea has to come from 100% one's bright brain.
The ones that is really amazing is the ones appears from the older years, before the internet, and the idea has to come from 100% one's bright brain.
lumanauw said:
David, come on! The internet is just the last 2% of the technological times. I think it was EASIER before the net, you read the journals, and you were reasonably sure it was serious. Now with the net you have the big handicap that any fool can post as a scientist. And by Jove, they do!
Jan Didden
Hi, Janneman,
Maybe in your position, yes, you can obtain information easier before the internet era. But I think it is not the same in my place. Maybe you can get AES papers easily through your connections, but in here (Indonesia) it is very difficult to get/buy any information. For example, the DougSelf handbook, until this time, I never saw it in any bookstore here. I got it from buying from internet (thanks to the internet). And I learn alot more from internet. For me, information is very limited without the internet.
But one thing makes a difference. Before the internet, people (if they are interested to ) are learning more seriously, because the information is harder to get, and forces people to learn juice of the information to the last word available.
You are a good example. How you are good with Feedback theory, because you learn it before the internet era
Maybe in your position, yes, you can obtain information easier before the internet era. But I think it is not the same in my place. Maybe you can get AES papers easily through your connections, but in here (Indonesia) it is very difficult to get/buy any information. For example, the DougSelf handbook, until this time, I never saw it in any bookstore here. I got it from buying from internet (thanks to the internet
). And I learn alot more from internet. For me, information is very limited without the internet. But one thing makes a difference. Before the internet, people (if they are interested to ) are learning more seriously, because the information is harder to get, and forces people to learn juice of the information to the last word available.
You are a good example. How you are good with Feedback theory, because you learn it before the internet era
lumanauw said:[snip]But one thing makes a difference. Before the internet, people (if they are interested to ) are learning more seriously, because the information is harder to get, and forces people to learn juice of the information to the last word available.
You are a good example. How you are good with Feedback theory, because you learn it before the internet era
Yes. Yes. I remember copying a whole book in a French library in Paris for more money then the book cost, but it was and still is unavailable. (Sol Rosenstark, Principles of Feedback Theory or something like that).
Jan Didden
Hi Toino,
---Mikeks...I start to suspect that you are D Self…---
I once suspected Mikeks was Kevin Aylward because of this
http://www.kevinaylward.co.uk/ee/currentfeedbackmyth/currentfeedbackmyth.html
I was wrong, but there is a common point to all three, they have written in Electronics World when it was the great times.
---Mikeks...I start to suspect that you are D Self…---
I once suspected Mikeks was Kevin Aylward because of this
http://www.kevinaylward.co.uk/ee/currentfeedbackmyth/currentfeedbackmyth.html
I was wrong, but there is a common point to all three, they have written in Electronics World when it was the great times.
There's the real Douglas Self (the book writer) in here
http://www.diyaudio.com/forums/showthread.php?postid=1150322#post1150322
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.