inquiries about LTP topology

[keantoken]

I have looked at the LTP topology and am wodering about it somewhat. I have some ideas that I am wondering about the ups and downs for. From what I know, here is a standard use of the LTP:
OUTPUTSTAGE is going to the output stage of the amp. OUT is the feedback output. Everythig else should be self-explanatory.

I have wondered about another idea I had and am showing in the next post.

[keantoken]

Here is my idea. The two bottommost transistors are complementary this time, making the bias flow through the emitters and collectors of the transisters. One downfall I can see to this is that the voltage on one side of the LTP is higher after this stage because the current flows one way. I also see that some complementary matching might be needed. Here it is:

[lumanauw]

Hi, Keantoken,

You got many interesting ideas
I'm not familiar with your LTP in post#1. Usually a differential pair only consist of Q1-Q2 (without Q3-Q4) with both emitors tied together. This simple (but very powerfull) configuration of 2 transistors tied together in the emitors indeed does the main job of a power amplifier.

I'm interested in your idea, but could you explain it with LTP that only has Q1-Q2 without Q3-Q4, with Q1-Q2's emitors tied together?

[keantoken]

By "You got many interesting ideas" I am supposing that is a good thing, thanks!

I am deriving this LTP variation from mikeB's symasym4 (in diagram).
I am not completely sure what you mean, as Q3-Q4 are the main points of my idea. Do you want me to explain how an LTP works so that you know I understand it?
I am simply wondering how this variation would work as supposed to the more used topology in my first post.
i am not really sure what Q3-Q4 are for (sorry... ), but I think that Q3 developes another comparing circuit with Q2 while using Q4 to amplify the input signal even more so that the output transistors can be biased at lower current therefore reducing noise generation (whew, long sentence!). Get what I mean? Am I right, mikeB? I think one of the ups of this idea is that instead of the bias current being split between the two bases, it runs through both bases at the same time, applying the full bias current to both transistors! (ohm's law might prove me wrong, though) Another intresting idea: I wonder what would happen if you fitted a resistor between the bases of Q3-Q4?

I severely dislike dissapointing people, so if I dissapointed you, tell me where I went wrong so that I can fix it next time!

[jan.didden]

Originally posted by keantoken
Here is my idea. The two bottommost transistors are complementary this time, making the bias flow through the emitters and collectors of the transisters. One downfall I can see to this is that the voltage on one side of the LTP is higher after this stage because the current flows one way. I also see that some complementary matching might be needed. Here it is:

OK, look at the voltage on the emitters of Q1 and Q2. Can you tell me what they are (approximately of course)? That will immediately show you whether this will work or not.

Hint: Assume that Vbe = 0.7V

Jan Didden

[lumanauw]

Hi, Keantoken,

If you know how a differential pair is working, you will understand other amp schematics (more complicated ones) more easily. Go to Nelson Pass' diy website, passdiy.com, read the article about Bride of Zen/BOZ, (or is it BOZOS?)

It is about 2 transistor of the same type, tied together in the emitors(sources for mosfets), and by making this 2 transistor differential pair with power devices, you already make an audio power amp that can move your speakers

Try to figure out how this 2 transistor tied together can make amplification of input in controlled way (certain dB constant gain factor with controlled DC offset, making linear amplification : out=K x input)

[ilimzn]

Originally posted by keantoken
I am not completely sure what you mean, as Q3-Q4 are the main points of my idea. Do you want me to explain how an LTP works so that you know I understand it?

Yes, that would actually be a good idea since your first idea is not a LTP, i.e. does not function as such. The second would sort-of work like a LTP assuming resistor values are well chosen - except that it would add one more Vbe into one side of the LTP, and one pretty much wasted transistor in the 'tail' of the LTP, which in turn would prevent the LTP in doing what it does best - compensating the nonlinearities of one transistor in the LTP with tghe other (to put it simplistically).

Once again, I admire your enthusiasm, but you seem to be lacking some of the very basic knowledge you need to set up sensible electronic circuits - standard or unusual. For instance the difference between transistor emitter and collector, that the Vbe under normal active conditions equals 0.5-0.7V and if it is not, there is something very wrong, and, if the transistor was ideal, that you could NOT make it any bigger no matter what forward current you tried, etc.
Let me put this simply: you cannot use a hammer if you don't know which end to hold. Electronics is based on many 'tools' and many rules as to which end you hold them by to do work - but there are some basic ones, perhaps maybe a dozen or less. You need to lern them. Why? Well, because if this was, say, chemistry, you would be an interesting color, burned, or perhaps even dead - not all ideas are good.
Fortunately, this is electronics, and there is such a thing as a simulator - which has one serious failing: Unlike the real world, it does not explode, smoke or even tell you in any way that you had a bad idea - in fact, you have to look for it in the results it gives you to see it.

You have to learn to crawl before you walk before you run before you jump, there's no way around it - a simulator will just let you do it with less of the electronics equivalent of scraped knees, or, god forbid, broken bones, but, sadly, it will not tell you that you have scraped your knee or broken your bone, unless you alrady know HOW to interpret the result. And, to do that, you would need already what you are trying to get from the simulator. It's a chicken and egg situation, which means you must search for the basic knowledge somewhere else - books, courses, web pages (many!).

I am not really sure what Q3-Q4 are for (sorry... ),

...This is the first sign that you really should not be using it other than on it's own until you figure out what it does.

...amplify the input signal even more so that the output transistors can be biased at lower current therefore reducing noise generation (whew, long sentence!)

But one which could even be proved to be dead wrong! As I said - learn to walk first. Noise generation is the least of your concerns right now. And, FYI lower current does not necesairly mean lower noise.

(ohm's law might prove me wrong, though)

Ohm's lay is the one that you really need to take as an absolute in electronics. It works every time, and if you don't know where to use it, it will, as it were, bite you in the a$$ when you least suspect it.

[/b]
I severely dislike dissapointing people, so if I dissapointed you, tell me where I went wrong so that I can fix it next time! [/B]

Not knowing something is hardly a reason for dissapointing people - but asking questions in order to learn, and not listening to the answers you don't necesairly like, that is. Think about it.
Homework: function of current sources, and function of current mirrors. If nothing else so you can understand which part of MikeB's Symasym is actually the LTP
Base bias current willalways be half the tail current ina ny LTP - if it's not, it isn't a LTP - you might want to consider the answer to the question as to why this is so.
Oh yes - and if my answer anoys you - maybe it gives you incentive to figure out how to beat me over the head with my own words. If so, I will be happy, because then you will have figured out for yourself what you want to know - and once you do, that knowledge will always stay with you, unlike answers I could give you 'on a platter'.

[keantoken]

Ah, yes... That last reply was kind of blunt, but I do think I might have just thrown the tightrope walker into the lion's den... I try not to be ignorant, but I try to learn the answers to all of my questions at first so that I don't have to run to a full-blown EE later just to figure something out. I think I'm just impatient. I don't just discard the answers that I don't like! That's not something a sensible person would do. I think it's lack of exposure/experience that gets me. I haven't really done anything with an LTP up till very recently, so I'm just trying to learn how it works. Sorry!

[lineup]

hi
here is an ordinary circuit using Long Tailed Pair input

it is a discrete op-amp

rather basic circuit
but nevertheless can produce HIFI
if put together and built in a good way:



more circuits from same DIY website
http://www.dibsplace.com/design/dcircats.html


lineup likes DIY Audio websites like this

[jan.didden]

Originally posted by keantoken
Ah, yes... That last reply was kind of blunt, but I do think I might have just thrown the tightrope walker into the lion's den... I try not to be ignorant, but I try to learn the answers to all of my questions at first so that I don't have to run to a full-blown EE later just to figure something out. I think I'm just impatient. I don't just discard the answers that I don't like! That's not something a sensible person would do. I think it's lack of exposure/experience that gets me. I haven't really done anything with an LTP up till very recently, so I'm just trying to learn how it works. Sorry!

Well, I think one of the problems is to try to learn some basic stuff without asking the basic questions. You throw up circuits that give the impression you know all about them and want to discuss improvements. Naturally, you get info that is way above your head, like the previous post.

You could consider instead of asking: "this is a modified LTP, what do you think of the improvement" you could ask: "I think this is an LTP, can somebody explain how this works". If you do THAT, you make progress fast. If you do what you did before, progres is zero.

Sorry to be so blunt, but that's the score. I have 30+ years of analog design under my belt. There are MANY things I don't know. If I want to know them, I ask for explanation. I make progress fast. We were all born ignorant. The only thing you can be blamed for is refusal to learn.

Jan Didden