Hi Kanwar,
Ahh, you are wise then. You can't afford knockoffs and I rather suspected you knew where to buy.
Mastertech,
I am sorry. I do not understand your last post.
As far as I know, amplifierguru has acquired a manufacturer for his amp sometime in the future. I wish him luck in this.
-Chris
Ahh, you are wise then. You can't afford knockoffs and I rather suspected you knew where to buy.
Mastertech,
I am sorry. I do not understand your last post.
As far as I know, amplifierguru has acquired a manufacturer for his amp sometime in the future. I wish him luck in this.
-Chris
Hello everybody,
Greg isn't gone somewhere...He is Still in the Forum watching down his threads and making appropriate decisions for posting some infos , when SINBIN is lifted from it...
So Greg if you are seeing this ...Dont worry in your absence we are taking great care of your threads......
K a n w a r
Greg isn't gone somewhere...He is Still in the Forum watching down his threads and making appropriate decisions for posting some infos , when SINBIN is lifted from it...
So Greg if you are seeing this ...Dont worry in your absence we are taking great care of your threads......
K a n w a r
Ipanema (would that be the boy from Ipanema?),
Any amplifier has a distortive forward transfer function. That is, the input waveform, as it passes through the amp, is distorted into something different by the time it gets to the output.
The normal correction system is to feed back a small portion of the output (that portion establishing the gain) to a differential amplifier which compares input to output. This system compares input to output, and drives the amp in such a way as to ensure that the output is much closer to the input.
This has many, many advantages, but the disadvantage is that the proportionate feedback correction is equal to the ratio of open loop (OLG) to closed loop (CLG) gain. That is, if the amp has gain of 25, but with no feedback can give a gain of say 2500, then the feedback factor is 2500/25, and this represents the reduction of distortion, namely one hundred times less than open loop.
So feedback reduces distortion by the ratio of OLG to CLG. If OLG distortion is 10%, then CLG will be 0.1%. Good, but not great. There are, however, other good reasons why negative feedback is useful; reduction of output impedance and extension of frequency response being the most important.
NFB gets a bad rap. Your body temperature, blood pressure, heatbeat, sugar levels and digestive processes are all controlled by negative feedback. In fact the absence of negative feedback in cell division is one way of defining cancer. It's incredibly ubiquitous and fundamental to nature.
Feedforward is very different. This principle accepts the inevitability of distortion and actually profiles it. It then reverses the distortion, producing a conjugate of the distortion at input levels, and mixes it with the input, 'pre-distorting' the signal. What comes out of this process is then applied to the amplifier, which distorts as expected, but since reverse distortion was applied at the input, the output theoretically comes out clean and free of distortion. This is the theory; but I don't believe this technique has enjoyed a sufficient market airing to show its effectiveness one way or the other.
The arguments are very mathematical, and redolent with complexity, but essentially it works by predistorting the input signal. The early work on feedforward was done in the thirties, but popularised by a very smart British researcher in the 70s and 80s name Malcolm Omar Hawksford, of the University of Essex.
Cheers,
Hugh
Any amplifier has a distortive forward transfer function. That is, the input waveform, as it passes through the amp, is distorted into something different by the time it gets to the output.
The normal correction system is to feed back a small portion of the output (that portion establishing the gain) to a differential amplifier which compares input to output. This system compares input to output, and drives the amp in such a way as to ensure that the output is much closer to the input.
This has many, many advantages, but the disadvantage is that the proportionate feedback correction is equal to the ratio of open loop (OLG) to closed loop (CLG) gain. That is, if the amp has gain of 25, but with no feedback can give a gain of say 2500, then the feedback factor is 2500/25, and this represents the reduction of distortion, namely one hundred times less than open loop.
So feedback reduces distortion by the ratio of OLG to CLG. If OLG distortion is 10%, then CLG will be 0.1%. Good, but not great. There are, however, other good reasons why negative feedback is useful; reduction of output impedance and extension of frequency response being the most important.
NFB gets a bad rap. Your body temperature, blood pressure, heatbeat, sugar levels and digestive processes are all controlled by negative feedback. In fact the absence of negative feedback in cell division is one way of defining cancer. It's incredibly ubiquitous and fundamental to nature.
Feedforward is very different. This principle accepts the inevitability of distortion and actually profiles it. It then reverses the distortion, producing a conjugate of the distortion at input levels, and mixes it with the input, 'pre-distorting' the signal. What comes out of this process is then applied to the amplifier, which distorts as expected, but since reverse distortion was applied at the input, the output theoretically comes out clean and free of distortion. This is the theory; but I don't believe this technique has enjoyed a sufficient market airing to show its effectiveness one way or the other.
The arguments are very mathematical, and redolent with complexity, but essentially it works by predistorting the input signal. The early work on feedforward was done in the thirties, but popularised by a very smart British researcher in the 70s and 80s name Malcolm Omar Hawksford, of the University of Essex.
Cheers,
Hugh
Feed Forward Error Correction
Terry,
Any chance you can exlain how FF EC works in very simple terms?
I don't see ( my ignorance) how control can be maintained without any form of FB.
If no FB and errors throughout amp, where is the ultimate level of error detected?
Seems to my naive idea like someone trying to trace something with an enlarging pantagon. i.e. what about magnified errors within the EC itself?
cheers
Terry,
Any chance you can exlain how FF EC works in very simple terms?
I don't see ( my ignorance) how control can be maintained without any form of FB.
If no FB and errors throughout amp, where is the ultimate level of error detected?
Seems to my naive idea like someone trying to trace something with an enlarging pantagon. i.e. what about magnified errors within the EC itself?
cheers
Re: Feed Forward Error Correction
What you have to understand is that the error is sampled in one
part of the circuit, ie; OP stage and the correction is applied before
the OP stage.
So they are two separate functions carried in two separate regions.
Let's say that in the unity gain OP section there is at a certain time
a 0.1V drop from the IP to the OP point (speaker drive point) due to
the finite and non linear impedances of the OP BJT's.
Well that 0.1V is sampled and applied as +0.1V before that stage
and theoretically you get a null.
Of course in real life you don't get a perfect null, because the error
correction itself is subject to the inherent non linearity of it's active
devices.
To make it work well requires a lot of fiddling with device bias points
etc.
Hope this all helps, please re-read Hughs explanation elsewhere
as I thought he did a pretty good job.
Best,
Terry
hifryer said:
What you have to understand is that the error is sampled in one
part of the circuit, ie; OP stage and the correction is applied before
the OP stage.
So they are two separate functions carried in two separate regions.
Let's say that in the unity gain OP section there is at a certain time
a 0.1V drop from the IP to the OP point (speaker drive point) due to
the finite and non linear impedances of the OP BJT's.
Well that 0.1V is sampled and applied as +0.1V before that stage
and theoretically you get a null.
Of course in real life you don't get a perfect null, because the error
correction itself is subject to the inherent non linearity of it's active
devices.
To make it work well requires a lot of fiddling with device bias points
etc.
Hope this all helps, please re-read Hughs explanation elsewhere
as I thought he did a pretty good job.
Best,
Terry
AKSA said:
hi Hugh,
West's technological edge over developing nations is overwhelming.we cant catch up with u in technology for 50yrs or so let alone beat you.we have very vast agri and processing industries that r no less in quality.here it is where the catch is.developed countries heavily subsidise their farms and other low tech industries to get their cost at levels of developing nations but when it comes to us protecting our high tech industries,its opposed by west.
('us' and 'we' includes China,India,Brazil and others)
the level playing feild is not mindless talk or nonsense.a poor farmer who wants to export bec local market is saturated cant do so bec his grain lost its cost advantage(TO SUBSIDIES),,he goes to work at factories but alas,the factory is at the verge of closing down!why?its tech is obsolete as compared to tech from west.what should he do ultimately?
Hi sagarverma,
The developing nations can produce excellent product, you have skilled engineers.
What bothers me about your post is that it implies that theft of designs and intellectual property is okay because you feel these countries are behind. Copies are never okay. The fact that they are usually inferior in quality deepens the crime.
Western countries had to develop technology from scratch. The mistakes and correct decisions are laid out for all to see, so the path is easier. They only problem is that the right choices and the easy choices are usually different. The human animal will usually take the easy route and cry when they are caught.
Build your country to be strong. Don't rape your own citizens. There are corporations that are only to glad to oblige you. Don't let them.
-Chris
The developing nations can produce excellent product, you have skilled engineers.
What bothers me about your post is that it implies that theft of designs and intellectual property is okay because you feel these countries are behind. Copies are never okay. The fact that they are usually inferior in quality deepens the crime.
Western countries had to develop technology from scratch. The mistakes and correct decisions are laid out for all to see, so the path is easier. They only problem is that the right choices and the easy choices are usually different. The human animal will usually take the easy route and cry when they are caught.
Build your country to be strong. Don't rape your own citizens. There are corporations that are only to glad to oblige you. Don't let them.
-Chris
anatech said:
Hi Chris,
PLEASE PLEASE tell me where i posted that copying is ok.
i Strongly condem copying.
my statement has been grossly misinterpreted.
Infact, i had posted a zip file in this thread itself that contains news peice bout copying of a Made in India bike by Chinese.you see the pics and see the shameless copying.+i also asked Greg to take care of his SKA design which Kanwar says,has been copied.
what i intent to say is that although we r developing fast,we have skilled workforce,but still at current pace,it will take many decades to catch up with west in terms of tech.
so our advantages,in terms of agri etc where our quality is at par but cost significantly low,must not be diluted by hefty subsidies.but if west insists with subsidies then it should not press our Govts. to leave our tech sector to face technological onslaught from western companies.
you cant clap with one hand,so if u demand us opening up our tech sector,it must be reciprocated by equal measure from your side too.
This is what 'level playing feild' is all about.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.