Hello and greetings to all,
I am new to this forum (well, actually I got bored of lurking ;-) and this is my first post here.
As a student of (Industrial) Chemical Engineering I have had a course of basic electronics, but I am still in the learning process and what I am about to attempt here is not my area of expertise so I want to apologise in advance for any stupid mistakes I will be making here.
The problem is this: a guy I know is a graduated Electromechanics Engineer and while we were discussing amplifiers etc he said that he and a friend once wanted to design an amplifier but eventually gave up because they found out that it was cheaper to just buy one.
I said I didn't believe that and that this was a good opportunity to actually sit down and design one myself and prove they were wrong. I hope I'm not getting over my head into this: I've looked on the forum and there are some threads about Class A designs, but they all have multiple transistors or are push-pull when I would like to try Single Ended Pure class A. And I have the idea that the designs there are more complicated than the one I made -- dough I'm sure there will be a good reason for it
).It is a shame dough, that I can't find the thread about the most simple amplifier or something like that. The designs I saw there were along the lines of mine ... ah, I also would like to stress that I only intend to design the thing and then look up the price of the components. As a student I don't have the means (I think ... it will depend on the final design ;-) to buy and build it, but if anyone wants to do this, please do. Consider this an Open Source design or something.
For the last weeks I've been busy reading a lot about PSU's (www.tnt-audio.com) and general amplifier theory and now I have the design specification for my amplifier. Here it goes:
"The Amplifier's power supply, heatsinks, casing etc ... will be based and designed like the "Le Amp" from n0rh.
However, it will be a pure single-ended Class A Monobloc (for ease of design and least components) and of 10 watts peak power (or maybe less, depends on the heatsinks. I'd rather be scimpy on power as 1 watt RMS on my Kenwood KR-5600 sounds loud enough with my Bose 301 in my 12m^2 dorm room). Any modifications nessesairy to fullfill any of these specifications will be present in the final product as to allow the future modification (via PCB) of the 10 watt monobloc to the 100 watt monobloc (class A/B) that the Le Amp is."
As far as the PSU goes I already figured out that the 20.000µF of the n0rh should be sufficiënt for 10 watts Class A, but if I understood correctly
(http://www.tnt-audio.com/clinica/ssps1_e.html) then for Class A there should be around 30.000 µF. I would use 4 extra caps of 5.000µF for further smoothing of the ripple, and thus use something like in diagram 4 of that website. (give or take some different µF in the capacitors etc.) An added advantage is that then I'd use 2 bridge rectifiers who can devide the power surge to the capacitors as I feel the 35amp rectifier of the original Le Amp would become insufficiënt for the increased capacitor banks. The advantages of this approach would be of course the increased smoothness of the PSU voltage and the added dynamic headroom of the final amp - the final 100 watts amp that is, I doubt the difference would be noticeable in the 10 watt Class A. Disadvantage is the added price of the extra components, but then again it will be an improvement over the original Le Amp design.
And now for the real reason I'm posting here: the schematic of the Class A amp itself.
Note: I'll do the mathematics of the values of the components later. Right now I only want to know if I made huge mistakes in the design itself.
This moring I drew the first attempt of the amp on paper, and after installing PSpice (some Civil Engineers I know use it but I've never used it before) and getting lost in the different modules I found the one in where I could draw the schematic. Here is a screenshot of it.
http://studwww.ugent.be/~jsteeman/ClassA-Version1.JPG
(haven't figured out how to attach pictures. I can't see it when I preview my post.)
I want/have to keep things as simple as possible so I started from a simple transistor that accepts AC to amplify it, hence C2 and R5 to get rid of possible DC component from the préamplifier. If i'm correct these two will act as a highpassfilter so C2 should block everything under 15 or 10Hz (20hz seems as a too close margin to me).
Since a transistor can only conduct in one direction, it cannot take AC and therefore I put R4 there to deliver the bias current that will "lift" the AC untill it is fully transported thru the transistor. (This makes it easy to put a voltage-clipping indicator there: 0 Volts is considered clipping)
After the signal is amplified thru the transistor, the added DC bias must be removed and that is why I put C1 and R2 there.
So this is it. Any comments/improvements are verry much appreciated. I really want to learn this so if I went wrong somewhere, feel free to give me a nudge in the right direction.
thx,
Cordraconis
I am new to this forum (well, actually I got bored of lurking ;-) and this is my first post here.
As a student of (Industrial) Chemical Engineering I have had a course of basic electronics, but I am still in the learning process and what I am about to attempt here is not my area of expertise so I want to apologise in advance for any stupid mistakes I will be making here.
The problem is this: a guy I know is a graduated Electromechanics Engineer and while we were discussing amplifiers etc he said that he and a friend once wanted to design an amplifier but eventually gave up because they found out that it was cheaper to just buy one.
I said I didn't believe that and that this was a good opportunity to actually sit down and design one myself and prove they were wrong. I hope I'm not getting over my head into this: I've looked on the forum and there are some threads about Class A designs, but they all have multiple transistors or are push-pull when I would like to try Single Ended Pure class A. And I have the idea that the designs there are more complicated than the one I made -- dough I'm sure there will be a good reason for it
).It is a shame dough, that I can't find the thread about the most simple amplifier or something like that. The designs I saw there were along the lines of mine ... ah, I also would like to stress that I only intend to design the thing and then look up the price of the components. As a student I don't have the means (I think ... it will depend on the final design ;-) to buy and build it, but if anyone wants to do this, please do. Consider this an Open Source design or something.For the last weeks I've been busy reading a lot about PSU's (www.tnt-audio.com) and general amplifier theory and now I have the design specification for my amplifier. Here it goes:
"The Amplifier's power supply, heatsinks, casing etc ... will be based and designed like the "Le Amp" from n0rh.
However, it will be a pure single-ended Class A Monobloc (for ease of design and least components) and of 10 watts peak power (or maybe less, depends on the heatsinks. I'd rather be scimpy on power as 1 watt RMS on my Kenwood KR-5600 sounds loud enough with my Bose 301 in my 12m^2 dorm room). Any modifications nessesairy to fullfill any of these specifications will be present in the final product as to allow the future modification (via PCB) of the 10 watt monobloc to the 100 watt monobloc (class A/B) that the Le Amp is."
As far as the PSU goes I already figured out that the 20.000µF of the n0rh should be sufficiënt for 10 watts Class A, but if I understood correctly
(http://www.tnt-audio.com/clinica/ssps1_e.html) then for Class A there should be around 30.000 µF. I would use 4 extra caps of 5.000µF for further smoothing of the ripple, and thus use something like in diagram 4 of that website. (give or take some different µF in the capacitors etc.) An added advantage is that then I'd use 2 bridge rectifiers who can devide the power surge to the capacitors as I feel the 35amp rectifier of the original Le Amp would become insufficiënt for the increased capacitor banks. The advantages of this approach would be of course the increased smoothness of the PSU voltage and the added dynamic headroom of the final amp - the final 100 watts amp that is, I doubt the difference would be noticeable in the 10 watt Class A. Disadvantage is the added price of the extra components, but then again it will be an improvement over the original Le Amp design.
And now for the real reason I'm posting here: the schematic of the Class A amp itself.
Note: I'll do the mathematics of the values of the components later. Right now I only want to know if I made huge mistakes in the design itself.
This moring I drew the first attempt of the amp on paper, and after installing PSpice (some Civil Engineers I know use it but I've never used it before) and getting lost in the different modules I found the one in where I could draw the schematic. Here is a screenshot of it.
http://studwww.ugent.be/~jsteeman/ClassA-Version1.JPG
(haven't figured out how to attach pictures. I can't see it when I preview my post.)
I want/have to
keep things as simple as possible so I started from a simple transistor that accepts AC to amplify it, hence C2 and R5 to get rid of possible DC component from the préamplifier. If i'm correct these two will act as a highpassfilter so C2 should block everything under 15 or 10Hz (20hz seems as a too close margin to me).Since a transistor can only conduct in one direction, it cannot take AC and therefore I put R4 there to deliver the bias current that will "lift" the AC untill it is fully transported thru the transistor. (This makes it easy to put a voltage-clipping indicator there: 0 Volts is considered clipping)
After the signal is amplified thru the transistor, the added DC bias must be removed and that is why I put C1 and R2 there.
So this is it. Any comments/improvements are verry much appreciated. I really want to learn this so if I went wrong somewhere, feel free to give me a nudge in the right direction.
thx,
Cordraconis
Hi there. I think you can find that with a little creativity and salvaging, you can definitely DIY for cheaper than store bought if you're talking strictly power amps. Where store bought really wins out is in features like remote controls, fancy displays, and a whole slew of inputs. The cheapest way to go, bang for buck is with a chipamp (see chipamp forum). It's a bit harder to get your money's worth out of discrete amps without a lot of design experience invested. I'll pitch in though and say that it's way more fun to DIY and learn than it is to drop 5-bills into the hands of those l0s3rs at circuit city.
w/rt to your circuit. You've built a basic resistor biased emitter follower. Your main concern with a SE amp like this is that you have no voltage gain (so you rely on the preamp for that). But your secondary concern will be trading off your current output with damping factor. First of all, R4 doesn't need to be variable, you should do just fine with R5. Your bias current is going to be set by ( ( ( ( (R5/R4+R5) * VCC) - 0.7) / R2) got that?) which is to say, base voltage - Vbe drop divided by emitter resistor.
Now here's the trade off: You need a high bias current in order to supply a high output current. This means you want R2 as small as possible. However, your want your "current source" to be as high impedance as possible so that your AC signals go through the load instead of the current souce. This means you want R2 as big as possible.
I would say, determine what kind of load you're going to use (say 32-ohm headphones) and then choose an R2 of atleast 10x this value. So in this case, atleast 320 ohms or 330 ohms for a standard value. By the way, using another transistor as a current source would greatly help this issue.
Next, you're not going to get 10-watts out of a 2n2222, it's a pretty wimpy transistor. You could get a few milliwatts from it for headphones if that was your goal. If you're looking into something with power that is cheap, a 2N3905 might work.
Lastly, I would strongly suggest you look into MOSFETS and the Zen ampifiers. They are proven examples of quality simple class-A designs with minimal components.
Well that's long winded and quite enough. Hope this helps.
--
Danny
w/rt to your circuit. You've built a basic resistor biased emitter follower. Your main concern with a SE amp like this is that you have no voltage gain (so you rely on the preamp for that). But your secondary concern will be trading off your current output with damping factor. First of all, R4 doesn't need to be variable, you should do just fine with R5. Your bias current is going to be set by ( ( ( ( (R5/R4+R5) * VCC) - 0.7) / R2) got that?) which is to say, base voltage - Vbe drop divided by emitter resistor.
Now here's the trade off: You need a high bias current in order to supply a high output current. This means you want R2 as small as possible. However, your want your "current source" to be as high impedance as possible so that your AC signals go through the load instead of the current souce. This means you want R2 as big as possible.
I would say, determine what kind of load you're going to use (say 32-ohm headphones) and then choose an R2 of atleast 10x this value. So in this case, atleast 320 ohms or 330 ohms for a standard value. By the way, using another transistor as a current source would greatly help this issue.
Next, you're not going to get 10-watts out of a 2n2222, it's a pretty wimpy transistor. You could get a few milliwatts from it for headphones if that was your goal. If you're looking into something with power that is cheap, a 2N3905 might work.
Lastly, I would strongly suggest you look into MOSFETS and the Zen ampifiers. They are proven examples of quality simple class-A designs with minimal components.
Well that's long winded and quite enough. Hope this helps.
--
Danny
If you haven't looked at my class A design thread you should give it a look. My Design Project I have somewhat similar goals except that I actually intend to build it. My project is however also a learning experience.
I am designing to use as many parts that I already have as possible. Starting with post #9 you can see several attachments showing the growth of the design. My initial idea was to use a 2n3055 power transistor that I have on hand along with a few small signal transistors similar to your 2n2222.
I was well on the way when I realized that I did not have a high power resistor for the output circuit and that it would cost very little more to get another power transistor than to buy the resistor so I opted to go ahead and use an active load (a constant current source instead of the load resistor). For $1 more than the simpler design I get a significant increase in power and I expect I will get some drop in distortion. I am intentionally making the first design without interstage negative feedback to make the circuit simpler, less problamatic stability wise and potentially freer and more open sounding. feedback can easily be added later if necessary.
What I am finding is that "simple" is harder with transistors than with tubes (though the transistors will be cheaper). Of course if you don't actually intend to build it you could design a very simple tube amp that would be a useable circuit. Danny might have the right idea for you in using MOSFET output as these might be easier to design with (though the sense I get is that simple MOSFET circuits are not as linear as BJTs which aren't very linear either).
In either case you will need some voltage gain because 500mV only gives you 31mW. If you are going to use a single stage for both power and voltage gain you will need either a common emitter circuit (with load in the collector) or a common source (for MOSSFET) with load in the drain. I am not sure about the MOSFET (Nelson Pass made a single MOSFET amp that is quite famous which you should check out) but with the BJTs you are IMHO better off with at least a two stage design. Let the small signal device do the voltage gain so that the output device can put it's effort into current gain. You want a voltage gain in the first stage of at least 10 if you want to put out any power at all. Since power = E^2/R. To get 1W into 8ohms you need voltage of Sqrt(P*R)=sqrt(8)=2.83rms which is nearly 4 volts peak. Since your output voltage is going to be less than half of the power supply voltage it is clear that you need a power supply of at least 10 volts to get even a single watt of output and a gain approaching 10 for the average tape deck or CD output levels.
My project is however also a learning experience.I am designing to use as many parts that I already have as possible. Starting with post #9 you can see several attachments showing the growth of the design. My initial idea was to use a 2n3055 power transistor that I have on hand along with a few small signal transistors similar to your 2n2222.
I was well on the way when I realized that I did not have a high power resistor for the output circuit and that it would cost very little more to get another power transistor than to buy the resistor so I opted to go ahead and use an active load (a constant current source instead of the load resistor). For $1 more than the simpler design I get a significant increase in power and I expect I will get some drop in distortion. I am intentionally making the first design without interstage negative feedback to make the circuit simpler, less problamatic stability wise and potentially freer and more open sounding. feedback can easily be added later if necessary.
What I am finding is that "simple" is harder with transistors than with tubes (though the transistors will be cheaper). Of course if you don't actually intend to build it you could design a very simple tube amp that would be a useable circuit.
Danny might have the right idea for you in using MOSFET output as these might be easier to design with (though the sense I get is that simple MOSFET circuits are not as linear as BJTs which aren't very linear either).In either case you will need some voltage gain because 500mV only gives you 31mW. If you are going to use a single stage for both power and voltage gain you will need either a common emitter circuit (with load in the collector) or a common source (for MOSSFET) with load in the drain. I am not sure about the MOSFET (Nelson Pass made a single MOSFET amp that is quite famous which you should check out) but with the BJTs you are IMHO better off with at least a two stage design. Let the small signal device do the voltage gain so that the output device can put it's effort into current gain. You want a voltage gain in the first stage of at least 10 if you want to put out any power at all. Since power = E^2/R. To get 1W into 8ohms you need voltage of Sqrt(P*R)=sqrt(8)=2.83rms which is nearly 4 volts peak. Since your output voltage is going to be less than half of the power supply voltage it is clear that you need a power supply of at least 10 volts to get even a single watt of output and a gain approaching 10 for the average tape deck or CD output levels.
*Warning* Very long post ... again.
) was that the voltage would only be dependant of the PSU voltage. After some deeper thinking I see that this was wrong. Good news is that my worries about how to get the correct voltage from the 400VA-toroïd are gone since in my course I found the schematic for a voltage regulator!
I haven't decided yet what transistor to use. I looked at the pictures in the "Schematics" module and took the thing that looked like an Trans., so don't pay attention to what type it is. It's only to draw the schematic.
Better is the thread I mentioned in my first post, the one about the simplest class A amp. The thread I meant was "simple amplifier with BJT's" -- found by doing a forum search on "zen amplifier". I've noticed something new for me: in the output stage they sometimes use 2 transistors, but it's not push-pull. Why is that and could these Trans be an NPN and PNP as to recycle the output stage of the 100watt class A/B of my final design?
Maybe it is time to explain another problem about that 100watt amp I had in mind: I thought that the "Le Amp" I want to base everything on is a chipamp (thx for the tip!), but I misinterpreted the concept. After reading the chipamp forum for some hours I learned that it is actually the *whole* amp that's in an IC. My understanding was that a chipamp was the amp minus the PSU and the final gain stage, so there would be a PSU, a chip and 2 transistors on a heatsink inside such an amp.
You can understand my supprise when I looked more closely to the n0rh amp's screenshots ... byebye to the transistors I wanted to use in my Class A design.
On the other hand, it seems that the IC used in that amp costs around 30 euro's/piece. Talking about value for money indeed!
Anyway, I'll still want to go through with the designing of the Class A amp like I had the idea in my first post.
I wasted a lot of time behind the computer today, but I feel like i learned a lot by reading up on the things you suggested. Thx again for the chipamp tip.
--> woops, 2 o' clock in the moring. Definately need to get some sleep. Hope I didn't make too many typo's.
Cordraconis
Today I skimmed trough my basic-electronics coursematerial and I've build indeed a current gain stage. My reasoning (or actually the lack thereofazira said:
) was that the voltage would only be dependant of the PSU voltage. After some deeper thinking I see that this was wrong. Good news is that my worries about how to get the correct voltage from the 400VA-toroïd are gone since in my course I found the schematic for a voltage regulator! ah, I recognize this formula as a voltage divider, got it. It seems I unintentionally made one. In my next version I have moved R5 directly to the input stage of the preamp where it belongs -- Pspice has renamed it to R6.azira said:
Yes, as I said above, I could put a voltage gain after the inputstage, but I was thinking of avoiding an extra transistor - and suprise - in Mashaffer's thread someone also came up with the idea of a step-up transformer (only a very small one should be needed I guess). An added advantage is I don't need a DC-blocking capacitor. But what about the current? It will be less after the transformer, but will it be enough to drive the transistor?azira said:
azira said:
I haven't decided yet what transistor to use. I looked at the pictures in the "Schematics" module and took the thing that looked like an Trans., so don't pay attention to what type it is. It's only to draw the schematic.I have read the whole thread that mashaffer posted. Very interesting indeed. With my refreshed knowledge of voltage- and current gain stages I actually understand a lot of the schematics now.azira said:
Better is the thread I mentioned in my first post, the one about the simplest class A amp. The thread I meant was "simple amplifier with BJT's" -- found by doing a forum search on "zen amplifier". I've noticed something new for me: in the output stage they sometimes use 2 transistors, but it's not push-pull. Why is that and could these Trans be an NPN and PNP as to recycle the output stage of the 100watt class A/B of my final design?
Maybe it is time to explain another problem about that 100watt amp I had in mind: I thought that the "Le Amp" I want to base everything on is a chipamp (thx for the tip!), but I misinterpreted the concept. After reading the chipamp forum for some hours I learned that it is actually the *whole* amp that's in an IC.
My understanding was that a chipamp was the amp minus the PSU and the final gain stage, so there would be a PSU, a chip and 2 transistors on a heatsink inside such an amp. You can understand my supprise when I looked more closely to the n0rh amp's screenshots ... byebye to the transistors I wanted to use in my Class A design.
On the other hand, it seems that the IC used in that amp costs around 30 euro's/piece. Talking about value for money indeed!
Anyway, I'll still want to go through with the designing of the Class A amp like I had the idea in my first post.
Thank you very much for your time and help!azira said:
I wasted a lot of time behind the computer today, but I feel like i learned a lot by reading up on the things you suggested. Thx again for the chipamp tip.
I have read your thread, but I think the "simple BJT amp" thread is more my cup op thea for the moment. More simple designs.mashaffer said:If you haven't looked at my class A design thread you should give it a look. My Design Project I have somewhat similar goals except that I actually intend to build it.
yup. Most important thing I learned today.mashaffer said:
--> woops, 2 o' clock in the moring. Definately need to get some sleep. Hope I didn't make too many typo's.
Cordraconis
Attachments
Let's try to do simple DC design
Please, take in account that i am a Brazilian guy, my natural language is not English:
But some good intention will help.
First, to design a simple Class A circuit, you must decide if power or not.....and have to decide the output circuit current....this current is the one will flow continuosly from Colector to Emitter.... better to believe will move this way..... i know electrons goes to other side... forget it, if want to make simple things without a Enginnering degrée.
I decided 500mA colector current, because not to great, and 12 Volts voltage, because batteries can supply, a motorcicle battery can be used, or other kind of battery that can supply 500mA continuosly for long time without exaust.
So, decided Voltage and Current.... this way i can conclude the energy consumption of this circuit... this is Voltage, expressed in volts, multiplied by current expressed in amperes... this way:
12 X 0,5 = 6 Watts..... this is consumption, not the output power, the output power will be a small fraction of that, around 10 percent if you lucky.
I will decide also the Colector charge resistance, because i want to use one speaker that can support continuous 6 watts energy without burn in fire, or melt, or smell.... a power one can hold that .
I decide 8.2 volts colector resistance, and a NPN power transistor that i know the gain is around 20 when used with that condition.... my colector resistance will be a 8.2 speaker... around that.
But..... take a look at the scratch.....middle draw....a battery, and a colector to emitter circuit, the output circuit..... i decided 8 ohms resistance and i decided 500 miliampéres...but this energy is not flowing!..... i discover that this energy will flow if i put energy in the input circuit, the input circuit is from base to emitter circuit.... yes.... emitter is common to both energies, from output that normally bigger than input energy.... i am talking about bias, about iddle current.... without that, the transistor is off, nothing happens, energy do not travell from colector to emitter (please, believe that, fix that to be easy...of course Mr Exactus will came and say oposite)..... When a man go to a woman, the woman goes to the man.... who goes to where?.... depends where you are observing.... that's relative and theoretical, no one followed the electron and can have completely sure.... also no one knows the color of the electron, if is fat or not!....Blue.... that's OK!... lets fix it is blue.... all rigth!
But if i make a magic, putting energy in base to emitter circuit, and the transistor turn on, and energy will flow from positive battery side into colector, and will pass to emitter in direction of the ground....the earth..... and if the energy i decide will pass will be 500 mA, what will be the voltage that will appear in 8.2 ohms charge resistor, or speaker terminals.....the Ohms law said that Voltage is the result of current multiplied by resistance, and we have to take care, the voltage have to be expressed in volts and the current in amperes.... so:
V = 0,5 X 8,2 = 4.1 Volts will appear over resistance extremes.. but the supply have 12 volts!..... where's the rest voltage?
Will follow next post...because my telefone dial up line is unstable.
Carlos
Please, take in account that i am a Brazilian guy, my natural language is not English:
But some good intention will help.
First, to design a simple Class A circuit, you must decide if power or not.....and have to decide the output circuit current....this current is the one will flow continuosly from Colector to Emitter.... better to believe will move this way..... i know electrons goes to other side... forget it, if want to make simple things without a Enginnering degrée.
I decided 500mA colector current, because not to great, and 12 Volts voltage, because batteries can supply, a motorcicle battery can be used, or other kind of battery that can supply 500mA continuosly for long time without exaust.
So, decided Voltage and Current.... this way i can conclude the energy consumption of this circuit... this is Voltage, expressed in volts, multiplied by current expressed in amperes... this way:
12 X 0,5 = 6 Watts..... this is consumption, not the output power, the output power will be a small fraction of that, around 10 percent if you lucky.
I will decide also the Colector charge resistance, because i want to use one speaker that can support continuous 6 watts energy without burn in fire, or melt, or smell.... a power one can hold that .
I decide 8.2 volts colector resistance, and a NPN power transistor that i know the gain is around 20 when used with that condition.... my colector resistance will be a 8.2 speaker... around that.
But..... take a look at the scratch.....middle draw....a battery, and a colector to emitter circuit, the output circuit..... i decided 8 ohms resistance and i decided 500 miliampéres...but this energy is not flowing!..... i discover that this energy will flow if i put energy in the input circuit, the input circuit is from base to emitter circuit.... yes.... emitter is common to both energies, from output that normally bigger than input energy.... i am talking about bias, about iddle current.... without that, the transistor is off, nothing happens, energy do not travell from colector to emitter (please, believe that, fix that to be easy...of course Mr Exactus will came and say oposite)..... When a man go to a woman, the woman goes to the man.... who goes to where?.... depends where you are observing.... that's relative and theoretical, no one followed the electron and can have completely sure.... also no one knows the color of the electron, if is fat or not!....Blue.... that's OK!... lets fix it is blue.... all rigth!
But if i make a magic, putting energy in base to emitter circuit, and the transistor turn on, and energy will flow from positive battery side into colector, and will pass to emitter in direction of the ground....the earth..... and if the energy i decide will pass will be 500 mA, what will be the voltage that will appear in 8.2 ohms charge resistor, or speaker terminals.....the Ohms law said that Voltage is the result of current multiplied by resistance, and we have to take care, the voltage have to be expressed in volts and the current in amperes.... so:
V = 0,5 X 8,2 = 4.1 Volts will appear over resistance extremes.. but the supply have 12 volts!..... where's the rest voltage?
Will follow next post...because my telefone dial up line is unstable.
Carlos
Attachments
That tranformer is going to cost you a whole lot more than an extra transistor.
You should see in most schematics basically 3 variations of a pair of transistors in an output stage.
First: a pair with collectors looking at eachother T-junctioned with the load. This is a CCS loaded CE style, single ended class A, basically a BJT version of the Zen amp.
Second: a pair with emitters looking at eachother T-junctioned with the load. This is a push-pull output stage can be either class-A, class AB, or class B depending on bias.
Third: a pair of bjt's "stacked" emitter to collector with the load T-junctioned between the emitter-collector connection. This will be a CCS loaded emitter follower, also class A. It's basically a simple upgrade to what you've got in your schematic.
In either of these circuits, you win a LOT by adding that extra transistor.
Actually the way you drew the first circuit is very correct from a textbook point of view, proper bias circuitry and coupling capacitors. Minus the variable resistor I mentioned, it's what I would work with for studying a simple follower.
--
Danny
You should see in most schematics basically 3 variations of a pair of transistors in an output stage.
First: a pair with collectors looking at eachother T-junctioned with the load. This is a CCS loaded CE style, single ended class A, basically a BJT version of the Zen amp.
Second: a pair with emitters looking at eachother T-junctioned with the load. This is a push-pull output stage can be either class-A, class AB, or class B depending on bias.
Third: a pair of bjt's "stacked" emitter to collector with the load T-junctioned between the emitter-collector connection. This will be a CCS loaded emitter follower, also class A. It's basically a simple upgrade to what you've got in your schematic.
In either of these circuits, you win a LOT by adding that extra transistor.
Actually the way you drew the first circuit is very correct from a textbook point of view, proper bias circuitry and coupling capacitors. Minus the variable resistor I mentioned, it's what I would work with for studying a simple follower.
--
Danny
Cordraconis;
Glad to see you are thinking things through and applying yourself. If I might I would like to bring up a couple of points. First keep in mind that transistors are much cheaper than transformers. Here in the states I can buy an appropriate small signal transistor for the price of a can of Coca Cola. For the price of a transformer I could have lunch.
Here is an idea of what you face trying to design a single transistor amplifier. Attached is a simple common emitter stage using a common power transistor. Notice how small the bias resistor has to be to supply adequate bias current. Ignoring issues of bias stability we also have rather low input impedence. In my next posting I am attaching a simulation of the circuit using a ideal voltage source for a signal. Note that we have a pretty good voltage swing and could theoretically get some reasonable output from it.
However the second trace shows the result of even a relatively small output impedence for the preamp. Notice that we have lost all of the gain. The preamp simply can't drive the power transistor with enough current.
I am not saying that such a design is not possible but this shows a bit of what you have to overcome.
Glad to see you are thinking things through and applying yourself. If I might I would like to bring up a couple of points. First keep in mind that transistors are much cheaper than transformers.
Here in the states I can buy an appropriate small signal transistor for the price of a can of Coca Cola. For the price of a transformer I could have lunch. Here is an idea of what you face trying to design a single transistor amplifier. Attached is a simple common emitter stage using a common power transistor. Notice how small the bias resistor has to be to supply adequate bias current. Ignoring issues of bias stability we also have rather low input impedence. In my next posting I am attaching a simulation of the circuit using a ideal voltage source for a signal. Note that we have a pretty good voltage swing and could theoretically get some reasonable output from it.
However the second trace shows the result of even a relatively small output impedence for the preamp. Notice that we have lost all of the gain. The preamp simply can't drive the power transistor with enough current.
I am not saying that such a design is not possible but this shows a bit of what you have to overcome.
Attachments
Continuing....
But, where's the rest..... 4.1 volts will be at charge.... of course the 7.9 volts that must exist to complete 12 Volts total, will be measure from colector to emitter......7.9 volts over transistor, and 500mA crossing the unit.... what's the power there:
7.9 X 0,5 = 3.95 Watts
This way, normally, people use twice or more than twice for the transistor dissipation.... this way, one 10 watts power unit will be good, and must be attached to a 3 by 3 inches aluminium to dissipate heat, this aluminium can have the shape you want.
But this will happens, those voltages, if you put energy flowing in the base to emitter circuit... and this energy is smaller than collector current.... and the relation is exactly the DC gain value you measure in normal multimeters..... this transistor i already know it will be 20 when this current of 0,5 Amperes passes inside.
If the relation is 20, this informed us that the colector current divided by the gain will result the base current......and off course, if you have first the base current decided, the colector current will be this base current multiplied by gain.
Our case, is 0,5 A divided by 20.... and the result is 0,025 A.... we can read this as 25 miliampéres....if you put this current flowing from base to emitter.... guaranteed you will have 20 times more energy in colector to emitter.....20 times 0,025A is 0,5A!
But theres a rule, you cannot calculate only the exact energy, people use to force a 10 times bigger energy flow from points A to point b.... a two resistor divider network, this will fix voltages and current.
If 10 times current will pass througth this divider network, we have to multiply 0.025 by ten.... this way 0.250A, or 250 miliamperes will flow.....and whe know that transistor only start to conduct energy with around 600 milivolts measured from base to emitter.... in our case, from base to ground, because emitter is conected to negative bus... the ground bar.
So, we have to calculate the resistance that will keep this voltage when passing 250 miliampéres...better to calculate 700 milivolts, because in class A, a heavy current goes into base to emitter circuit... and to have this heavy current you need a big voltage to open completely the energy flow.....so, 10 times the needed energy will flow througth those two resistors.... you will need only 10 percent to flow from base to emitter.... so, if you want to be real precise.... 9 tenths will flow inside resistor in direction of the ground bar..... one tenth will flow inside base to emitter junction in direction the ground bar....but lets calculate with the real total amount of energy, the total amount will be 250 miliamperes.....
We need that 700 milivolts, 0.7Volts appear over lower resistance, because this one will be connected from base to emitter as you see right side..... this way, ohm law tell us that resistance is the result of the voltage divided by the current...our case:
0.7 : 0.250 = 2,8 ohms
Strange that....but i think it is correct..... if not, a lot of people will run to help me .........HEEEEELP!
And the 9.2 volts you need to complete 12 volts will be over the up resistor, lets calculate.
9,2 : 0.250 = 36,8 ohms....this way we gonna use 39 ohms.
And lets see the power over those resistors, the down one:
Power is the result of current multiplied by voltage, the current we know is 0.250A, and voltage also, this way:
0.250 X 9.2 = 2,3 Watts.... it is a wire resistor, or a big one to dissipate more than this power....better to use 5 watts unit.
The other resistor, the down one:
0.250 X 0.7 = 0.175 W this one is very small, every fourth watt will be more than enough.
And what will be the power... sorry, i do not know how to calculate exactly...but hundreds of friends can come and help....normally, the power will be around 10 percent the consumption in this kind of circuit....but the sound is wonderfull....and the circuit is sensitive..... only some milivolts will be enough to drive transistor to maximum undistorted swing.....you have to put one condenser in the input, to avoid the 0.7 volts we have in base goes to the audio source you connect to the amplifier..... this circuit will need something alike 20 milivolts of audio.... so, if you put the pre amplifier output, you have to make a resistor divider to reduce audio.....amplifier will produce 750 milivolts and all you need is 20.... the ratio of those values is 750 divided by 20, this result in 37,5 by one..... so, if your pre amplifier output impedance is 10K..... this way, both resistors constructed as dividers can have this value or more than that value...lets use 10K.... calculating two resistors, one to ground and other to pre amplifier output, in the midlle of those resistors the audio to go to our one stage class A amplifier.....the result is 270 as the down resistor and the rest, 9750 ohms to the amplifier..... as we have not this value, use 10 K.
In the reality, the low resistance value from base to emitter will be a low resistance that will drain a lot the energy from pre amplifier... i think it will not give 750 milivolts anymore.... so....no problem and be happy.
If work well, i started this thing.... if explode!......hummmmmm will be hard to find me.... i will change my Internet adress
In the reality, those base resistors, too low....hummmm smelling
s--t!.... better to increase them proportionally...lets see...hummm
28 ohms down and 360 up!.... i will check.
The problem with those circuits, is the quality, normally audio is very good, can also use condenser conected to colector, the positive side in the colector, value around 2200 microfarads/16V or a little more isolation.... the power, this way will be reduced... really small, but reducing input volume, quality will be good to headphones.... normally better than all your audio chain....Why?
Less components, less distortion,... and class A is the better we can have in quality.
If use power supply, that one must have electronic filtering.... heavy filtering.... a good size condenser in the regulator transistor base, a zenner diode from base to ground and a resistor from colector to base if the transistor used is NPN, the output you can collect in the emitter lead...put one condenser alike 1000 microfarads in emitter lead..... a zenner and 220 microfarads in base and use around 20 volts DC input with this regulator..... resistor can be 330 ohms - 2 watts and the zenner must be a 5 watts unit.... transistor 10 watts too.
Good luck!
Carlos
But, where's the rest..... 4.1 volts will be at charge.... of course the 7.9 volts that must exist to complete 12 Volts total, will be measure from colector to emitter......7.9 volts over transistor, and 500mA crossing the unit.... what's the power there:
7.9 X 0,5 = 3.95 Watts
This way, normally, people use twice or more than twice for the transistor dissipation.... this way, one 10 watts power unit will be good, and must be attached to a 3 by 3 inches aluminium to dissipate heat, this aluminium can have the shape you want.
But this will happens, those voltages, if you put energy flowing in the base to emitter circuit... and this energy is smaller than collector current.... and the relation is exactly the DC gain value you measure in normal multimeters..... this transistor i already know it will be 20 when this current of 0,5 Amperes passes inside.
If the relation is 20, this informed us that the colector current divided by the gain will result the base current......and off course, if you have first the base current decided, the colector current will be this base current multiplied by gain.
Our case, is 0,5 A divided by 20.... and the result is 0,025 A.... we can read this as 25 miliampéres....if you put this current flowing from base to emitter.... guaranteed you will have 20 times more energy in colector to emitter.....20 times 0,025A is 0,5A!
But theres a rule, you cannot calculate only the exact energy, people use to force a 10 times bigger energy flow from points A to point b.... a two resistor divider network, this will fix voltages and current.
If 10 times current will pass througth this divider network, we have to multiply 0.025 by ten.... this way 0.250A, or 250 miliamperes will flow.....and whe know that transistor only start to conduct energy with around 600 milivolts measured from base to emitter.... in our case, from base to ground, because emitter is conected to negative bus... the ground bar.
So, we have to calculate the resistance that will keep this voltage when passing 250 miliampéres...better to calculate 700 milivolts, because in class A, a heavy current goes into base to emitter circuit... and to have this heavy current you need a big voltage to open completely the energy flow.....so, 10 times the needed energy will flow througth those two resistors.... you will need only 10 percent to flow from base to emitter.... so, if you want to be real precise.... 9 tenths will flow inside resistor in direction of the ground bar..... one tenth will flow inside base to emitter junction in direction the ground bar....but lets calculate with the real total amount of energy, the total amount will be 250 miliamperes.....
We need that 700 milivolts, 0.7Volts appear over lower resistance, because this one will be connected from base to emitter as you see right side..... this way, ohm law tell us that resistance is the result of the voltage divided by the current...our case:
0.7 : 0.250 = 2,8 ohms
Strange that....but i think it is correct..... if not, a lot of people will run to help me .........HEEEEELP!
And the 9.2 volts you need to complete 12 volts will be over the up resistor, lets calculate.
9,2 : 0.250 = 36,8 ohms....this way we gonna use 39 ohms.
And lets see the power over those resistors, the down one:
Power is the result of current multiplied by voltage, the current we know is 0.250A, and voltage also, this way:
0.250 X 9.2 = 2,3 Watts.... it is a wire resistor, or a big one to dissipate more than this power....better to use 5 watts unit.
The other resistor, the down one:
0.250 X 0.7 = 0.175 W this one is very small, every fourth watt will be more than enough.
And what will be the power... sorry, i do not know how to calculate exactly...but hundreds of friends can come and help....normally, the power will be around 10 percent the consumption in this kind of circuit....but the sound is wonderfull....and the circuit is sensitive..... only some milivolts will be enough to drive transistor to maximum undistorted swing.....you have to put one condenser in the input, to avoid the 0.7 volts we have in base goes to the audio source you connect to the amplifier..... this circuit will need something alike 20 milivolts of audio.... so, if you put the pre amplifier output, you have to make a resistor divider to reduce audio.....amplifier will produce 750 milivolts and all you need is 20.... the ratio of those values is 750 divided by 20, this result in 37,5 by one..... so, if your pre amplifier output impedance is 10K..... this way, both resistors constructed as dividers can have this value or more than that value...lets use 10K.... calculating two resistors, one to ground and other to pre amplifier output, in the midlle of those resistors the audio to go to our one stage class A amplifier.....the result is 270 as the down resistor and the rest, 9750 ohms to the amplifier..... as we have not this value, use 10 K.
In the reality, the low resistance value from base to emitter will be a low resistance that will drain a lot the energy from pre amplifier... i think it will not give 750 milivolts anymore.... so....no problem and be happy.
If work well, i started this thing.... if explode!......hummmmmm will be hard to find me.... i will change my Internet adress
In the reality, those base resistors, too low....hummmm smelling
s--t!.... better to increase them proportionally...lets see...hummm
28 ohms down and 360 up!.... i will check.
The problem with those circuits, is the quality, normally audio is very good, can also use condenser conected to colector, the positive side in the colector, value around 2200 microfarads/16V or a little more isolation.... the power, this way will be reduced... really small, but reducing input volume, quality will be good to headphones.... normally better than all your audio chain....Why?
Less components, less distortion,... and class A is the better we can have in quality.
If use power supply, that one must have electronic filtering.... heavy filtering.... a good size condenser in the regulator transistor base, a zenner diode from base to ground and a resistor from colector to base if the transistor used is NPN, the output you can collect in the emitter lead...put one condenser alike 1000 microfarads in emitter lead..... a zenner and 220 microfarads in base and use around 20 volts DC input with this regulator..... resistor can be 330 ohms - 2 watts and the zenner must be a 5 watts unit.... transistor 10 watts too.
Good luck!
Carlos
Here is a sample of what you can do with about $2 more in hardware. I do not claim that this is optimal by any means just a quick spice run. Note that R1 would be an adjustable resistance to set the bias for 1/2 V+ at the output.
The left most transistor is providing voltage gain while the PNP transistor is providing the drive current for the output transistor. This way none of the devices is stressed and you can supply the voltage and current that you need for less money than a transformer to boot.
Notice that even with the 2K ohm source imp. we still get several volts of swing. The gain of the input stage is set by the ratio of the collector to emitter resistors. There are other bias arrangements also that can be used to increase the input impedence and resuce the loading of the preamp. If the preamp has less than 2k output impedence of course you will get more voltage gain.
I hope this helps you think through the issues involved. Please do not take any of these circuits as a complete design for they are far from it. just some examples of different approaches.
The left most transistor is providing voltage gain while the PNP transistor is providing the drive current for the output transistor. This way none of the devices is stressed and you can supply the voltage and current that you need for less money than a transformer to boot.
Notice that even with the 2K ohm source imp. we still get several volts of swing. The gain of the input stage is set by the ratio of the collector to emitter resistors. There are other bias arrangements also that can be used to increase the input impedence and resuce the loading of the preamp. If the preamp has less than 2k output impedence of course you will get more voltage gain.
I hope this helps you think through the issues involved. Please do not take any of these circuits as a complete design for they are far from it. just some examples of different approaches.
Attachments
stage three
DestroyerX:
I have read your calculations and found them very helpfull. I like your way of thinking and keeping things clear -- from a reasoning point of view. I will try to keep the same order of power>amount of ampére over trans>amount of ampére over base etc ... working backwards is the way to go.
I have to say that I had some troubles understanding your english (mainly the grammar), but it forced me to read things again and again so this way I learned the most. Never mind the form or language, it's the message that is the most important.
If some teachers are reading this I hope I didn't give them ideas for their courses.
Designing an amplifier (and probably any cirquit) is not dropping current- and voltage gain stages or "building blocks" in the right places and then troubleshooting and hoping it will work, but thinking of what actually happens with either the currents and the voltages. Again a lesson learned.
But one thing about your design: you put the speaker in "the AC+DC path", thereby not only heating it up unnessesairly but also because of the changeing resistance that varies with the frequency, I think your voltage-and-biascurrent network of resistors will be affected. Think of it as the 8Ohm resistor that is actually varying a few 100 times a second. I would not do that if it was possible.
Mashaffer:
Thanks for bringing the problem with the imput impendance to my attention.
Maybe it's indeed not such a bad idea to put an extra transistor in the design. After all, a Class AB amp has two of them in the output stage, so I might as well use one of them as a voltage gain, and the second one as a current gain. One of the two would be overkill but the point was to be able to reuse as much of it in a more powerfull amp anyway.
I took a quick look at your design and I was thinking of using a resistor to control the current that enters the base of the final transistor (so a 3rd one would not be nessesairy) but now I see that you don't have one and you even use the 3rd one to increase it! Hmmm ... I will try to look into this tomorrow.
Final toughts: the imput-transistor aka the voltage gain stage will help imput impendance, right? But maybe an op-amp as imput stage would be feasable, providing the incredible gain can be tamed? (op-amps are also in my course material, that's why.
)
I'll try to make a new schematic tomorrow.
goodnight!
Cordraconis
DestroyerX:
I have read your calculations and found them very helpfull. I like your way of thinking and keeping things clear -- from a reasoning point of view.
I will try to keep the same order of power>amount of ampére over trans>amount of ampére over base etc ... working backwards is the way to go.I have to say that I had some troubles understanding your english (mainly the grammar), but it forced me to read things again and again so this way I learned the most. Never mind the form or language, it's the message that is the most important.
If some teachers are reading this I hope I didn't give them ideas for their courses.
Designing an amplifier (and probably any cirquit) is not dropping current- and voltage gain stages or "building blocks" in the right places and then troubleshooting and hoping it will work, but thinking of what actually happens with either the currents and the voltages. Again a lesson learned.
But one thing about your design: you put the speaker in "the AC+DC path", thereby not only heating it up unnessesairly but also because of the changeing resistance that varies with the frequency, I think your voltage-and-biascurrent network of resistors will be affected. Think of it as the 8Ohm resistor that is actually varying a few 100 times a second. I would not do that if it was possible.
Mashaffer:
Thanks for bringing the problem with the imput impendance to my attention.
Maybe it's indeed not such a bad idea to put an extra transistor in the design. After all, a Class AB amp has two of them in the output stage, so I might as well use one of them as a voltage gain, and the second one as a current gain. One of the two would be overkill but the point was to be able to reuse as much of it in a more powerfull amp anyway.
I took a quick look at your design and I was thinking of using a resistor to control the current that enters the base of the final transistor (so a 3rd one would not be nessesairy) but now I see that you don't have one and you even use the 3rd one to increase it! Hmmm ... I will try to look into this tomorrow.
Final toughts: the imput-transistor aka the voltage gain stage will help imput impendance, right? But maybe an op-amp as imput stage would be feasable, providing the incredible gain can be tamed? (op-amps are also in my course material, that's why.
)I'll try to make a new schematic tomorrow.
goodnight!
Cordraconis
Good, me good, hahahahaha!, bad english, Yeeeaahhh!
I know that very well, never studied english...had one englishman that lived with my sister some years.... always nervous the guy....neurotic as a hell... and he teach me a lot of bad words i cannot use here..... i am an expert in insults!...hahahaha!...because the damn one.
In early school i had some lections, how to count to hundred, i am a boy, this is a book!..... but the movies, always in English...my country is full of Kelloggs, and Coke, and Palm Olive soap, and Anderson Clayton..... this is an extention of Uncle Sam.... i do not know if we have direct telephone line already in our Government...but will be in near future, i suppose.
This way, English awfull, i recognize...but wonderfull me indian, me fell god help you very much...me good!
I learned how to design, Dc calculation, entire amplifier, do it by myself, from back to front only....if you invert i will say... dah dah dah...no condition to me..... will broke my last neuronium working...cannot invert... i think to invert i will have to use the pen and papper upside down or backwards! hihihi!
I thank you to give atention, i was doing and no one looking..... i felt myself transparent..... good to be helpfull, i think is one the best feelings we can have, to help!
CARLOS
I know that very well, never studied english...had one englishman that lived with my sister some years.... always nervous the guy....neurotic as a hell... and he teach me a lot of bad words i cannot use here..... i am an expert in insults!...hahahaha!...because the damn one.
In early school i had some lections, how to count to hundred, i am a boy, this is a book!..... but the movies, always in English...my country is full of Kelloggs, and Coke, and Palm Olive soap, and Anderson Clayton..... this is an extention of Uncle Sam.... i do not know if we have direct telephone line already in our Government...but will be in near future, i suppose.
This way, English awfull, i recognize...but wonderfull me indian, me fell god help you very much...me good!
I learned how to design, Dc calculation, entire amplifier, do it by myself, from back to front only....if you invert i will say... dah dah dah...no condition to me..... will broke my last neuronium working...cannot invert... i think to invert i will have to use the pen and papper upside down or backwards! hihihi!
I thank you to give atention, i was doing and no one looking..... i felt myself transparent..... good to be helpfull, i think is one the best feelings we can have, to help!
CARLOS
You could try using a resistor biasing network for the base of the output transistor (I think this is what you mean) but keep in mind that the input transistor (voltage amplifier) will have to supply enough signal current to drive the output transistor which you may or may not be able to do. Also keep in mind that depending on how you intend to bias the output transistor you may need a coupling capacitor between the two stages to keep the bias circuits (DC levels) seperate.
What the third transistor (the PNP device) does is multiply the forward current gain of the output transistor so that it is a much smaller load for the voltage amplifier stage. If the output transistor has a forward current gain of 100 and the PNP paired with it has a forward current gain of 50 then the pair will have a combined current gain of 5000. So for 1A of signal current at the output the voltage stage will only have to provide 200uA instead of 10mA. If the current gain of the PNP is higher (it could be as much as 150) then the loading on the voltage amplifier is even less.
Yes it does provide some impedence buffering. There are other bias circuits that can provide quite a bit more input impedence if you like. Find a good text book and look up the "bootstrap bias" configuration for an example. In this configuration the base of the input transistor is not connected to the middle of the voltage divider bias resistors. Instead a capacitor is connected from the emitter of the transistor to the middle of the two resistors and a third resistor is used to connect that point to the base (and thus the input). See attached jpeg.
You could use an op amp if you use an appropriate power supply. Taming the gain is easy. Just use a standard inverting circuit and set the input resistor to feedback resistor ratio so that it gives you the exact gain that you desire. You will have to make sure that the op amp you choose can supply adequate current to drive the output transistor of course (or use the complementary pair again
).Just a note to be sure you understand one point. You may be perfectly aware of this and I don't mean to be insulting if I am staing the obvious. There are two different classes of transistor that we have been looking at.
The output transistor is a "power" transistor that is in a large (relatively) case which is designed to conduct heat to a heatsink. It will be in a metal case (like TO-3 case) or a plastic case with a metal tab (like TO-220 case). This transistor is designed to pass a lot of current and dissipate a relatively large amount of power (when properly heat sinked). It is not particularly optimized for voltage amplification though it can provide some.
The input transistor is a small signal type. This will be in a small plastic (or somtimes metal) case that is not really designed to be put on a heat sink or to handle a lot of current or power. It is however optimized for voltage amplification or small current amplification.
You would need to be sure to use the right type of transistor for each job. On my prototype class AB amp that I put together I used power transistors for both of the devices in the complementary pair (the NPN-PNP pair at the output) but I am pretty sure that you could use a small signal transistor for the PNP (driver) device as the current is fairly low. I used both as power transistors since I happened to have matched pairs on hand and I felt that the PNP power device would run cooler since it had a metal tab that would act as a small heat sink (it is not mounted on the main heatsink with the NPN output device).
Attachments
Cordraconis,
If you are looking to design a simple low power Class A amplifier, I can think of no better place to go for information than PassDIY. Nelson Pass has designed a series of low power one and two stage class A power amplifiers for the audio amatuer to build and learn from. The whole series of amplifiers named "Zen Amplifiers" were designed as an educational tool to teach some of the fundamentals of audio amplifier design. It is an outstanding website and well worth spending a few moments time at.
If you are looking to design a simple low power Class A amplifier, I can think of no better place to go for information than PassDIY. Nelson Pass has designed a series of low power one and two stage class A power amplifiers for the audio amatuer to build and learn from. The whole series of amplifiers named "Zen Amplifiers" were designed as an educational tool to teach some of the fundamentals of audio amplifier design. It is an outstanding website and well worth spending a few moments time at.
I took some time to read the PassDIY pages, and after googeling for bootstrap I ended up on the pages of http://sound.westhost.com/amp_design.htm which also took some time to read. BTW, in September I have to retake some exams that i didn't pass so I'm studying in between drawing schematics. 
The signal current of the VAS is dependant of R9, right? So if I keep the current high enough it shouldn't pose a problem I think.
I put the coupling capacitor C6 there. I saw the schematic of a guy who posted it to ask for comments on his class A design, and when I compared it to my own design I saw that the imput capacitor C5 could be omitted. However, if for some reason a DC would be introduced I think it could reduce the effeciëncy of the VAS and since efficiëncy is already low I desided to keep it.
Also, after reading up on the Aleph Current source of Neslon Pass I tought that coupling the output of the final transistor (witch has higher current than the base) to the base and therefore increasing the signal current could help driving that transistor more easy.
Now wait before you type something!
By proposing the feedback capacitor I realize that this is POSITIVE feedback, so the increased signal current would increase the output which increases again the signal current ...
This pops up 2 questions: Can it be tamed somehow? and how will the coupling capacitor C6 interact with the feedback capacitor C3?
I've taken a quick look at this site before, but since i don't have PDF-reader installed on this partition I didn't read it. After your suggestion i looked again and this time i noticed the links to the HTML pages.
I like the Aleph Current source and aspecially the Lightbulb-resistor idea! I like people who think like this.
(See version 3 of my schematic)mashaffer said:
The signal current of the VAS is dependant of R9, right? So if I keep the current high enough it shouldn't pose a problem I think.
I put the coupling capacitor C6 there. I saw the schematic of a guy who posted it to ask for comments on his class A design, and when I compared it to my own design I saw that the imput capacitor C5 could be omitted. However, if for some reason a DC would be introduced I think it could reduce the effeciëncy of the VAS and since efficiëncy is already low I desided to keep it.
Also, after reading up on the Aleph Current source of Neslon Pass I tought that coupling the output of the final transistor (witch has higher current than the base) to the base and therefore increasing the signal current could help driving that transistor more easy.
Now wait before you type something!
By proposing the feedback capacitor I realize that this is POSITIVE feedback, so the increased signal current would increase the output which increases again the signal current ...
This pops up 2 questions: Can it be tamed somehow? and how will the coupling capacitor C6 interact with the feedback capacitor C3?
Aha!
I've taken a quick look at this site before, but since i don't have PDF-reader installed on this partition I didn't read it. After your suggestion i looked again and this time i noticed the links to the HTML pages.
I like the Aleph Current source and aspecially the Lightbulb-resistor idea! I like people who think like this.
Attachments
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.