thanks a lot for this outstanding explanation......
now i'll take my time to analyze your answer and....i'll make to you some questions about this when i cannot understand, if i can!
Anyways....i was studying the loadline on that book(that one that you outlined) but it wasn't really clear at first and this is when i got stucked and open this 3d.
Now i will re-study it again to with your explanations.
just 2 questions:
- what do you mean when you say that we assume distortionless platecurrent? (the characteristic are not linear and so that introduce distortion)
- what is exactly the purpose of drawing ad loadline? what am i actually gonna visualize through this curve? what does the ellipse rapresent?
Thanks a lot for your eplanations!!
now i'll take my time to analyze your answer and....i'll make to you some questions about this when i cannot understand, if i can!
Anyways....i was studying the loadline on that book(that one that you outlined) but it wasn't really clear at first and this is when i got stucked and open this 3d.
Now i will re-study it again to with your explanations.
just 2 questions:
- what do you mean when you say that we assume distortionless platecurrent? (the characteristic are not linear and so that introduce distortion)
- what is exactly the purpose of drawing ad loadline? what am i actually gonna visualize through this curve? what does the ellipse rapresent?
Thanks a lot for your eplanations!!
Maybe it is easier to do some experiments. The current and voltage of the output tube is easily accessible. Then a oscilloscope will be able to draw the dynamic load line on the screen.
I think you hit the key point: AC loadlines are not often drawn because they're not often that useful.
Anyway, the ellipse represents the effect of reactance on the total load impedance. You can think of it as a perturbation of the pure resistance line.
The tube's distortion means (by definition) that the current going up and the current going down, for the same grid voltage swing, are not equal. Additionally, as the Ip/Vp characteristic tends toward the x-axis (lower current), the tube's characteristic shows curvature, which changes the up and down swing disproportionately differently.
Anyway, the ellipse represents the effect of reactance on the total load impedance. You can think of it as a perturbation of the pure resistance line.
The tube's distortion means (by definition) that the current going up and the current going down, for the same grid voltage swing, are not equal. Additionally, as the Ip/Vp characteristic tends toward the x-axis (lower current), the tube's characteristic shows curvature, which changes the up and down swing disproportionately differently.
Go to Steve Bench's site for a good "how to" on drawing load lines for tube amps.
http://www.members.aol.com/sbench101/
There are menu picks on the right side near the top, or a text in the lower center of the page that leads to load lines part 1 through 5.
This is as deep into the subject as you need to go to design a good tube amplifier. I also use TubeCad software (SE amp cad, and P-P calculator), which will draw the load lines. You can change parameters and watch the lines move. These are not free programs, but they are cheap.
I am sure that the non ideal parameters of speaker performance can be modeled, but to fully model the power and frequency dependent counter EMF effects of a speaker cone would require some serious advanced math and physics skills. Would the results be worth it, would you really want to design an amplifier taylored to a particular set of speakers, in a particular room?
I would rather use the math and computer models to get close, and build it. Then the fun part starts. I know several people who would rather model a circuit to the Nth degree first though. None of them are interested in vacuum tubes.
http://www.members.aol.com/sbench101/
There are menu picks on the right side near the top, or a text in the lower center of the page that leads to load lines part 1 through 5.
This is as deep into the subject as you need to go to design a good tube amplifier. I also use TubeCad software (SE amp cad, and P-P calculator), which will draw the load lines. You can change parameters and watch the lines move. These are not free programs, but they are cheap.
I am sure that the non ideal parameters of speaker performance can be modeled, but to fully model the power and frequency dependent counter EMF effects of a speaker cone would require some serious advanced math and physics skills. Would the results be worth it, would you really want to design an amplifier taylored to a particular set of speakers, in a particular room?
I would rather use the math and computer models to get close, and build it. Then the fun part starts. I know several people who would rather model a circuit to the Nth degree first though. None of them are interested in vacuum tubes.
thank you guys for your posts, they really helped me.
I've studyed on the book and i hope.....i've gotten the concept 🙂
I still need to absorbe those informations (i still don't apprechiate the concept of the trial dynamic poin and loadline).
Thanks 🙂
Best,
Stefano.
I've studyed on the book and i hope.....i've gotten the concept 🙂
I still need to absorbe those informations (i still don't apprechiate the concept of the trial dynamic poin and loadline).
Thanks 🙂
Best,
Stefano.
The A Number One problem with doing reactive loadlines is that it's pretty difficult, and may not reflect the actual operating conditions since different speeks behave differently, and, in the final analysis, all you really discover is the importance of not skimping when it comes to output xfmrs. The higher the primary inductance, the thinner and narrower the reactive line gets as it approaches a purely resistive situation.
You can help improve the situation by using good output xfmrs, and by selecting PA VTs that have the lowest r(p)'s. If that latter is not possible (i.e. beam tetrodes or high u RF power finals) you can always reduce the effective r(p) with local NFB. That'll improve speaker damping for enhanced bass performance, and by making the effective r(p) a smaller fraction of the output xfmr Xl(pri) you improve linearity across the board.
Otherwise, don't worry about it. 🙂
You can help improve the situation by using good output xfmrs, and by selecting PA VTs that have the lowest r(p)'s. If that latter is not possible (i.e. beam tetrodes or high u RF power finals) you can always reduce the effective r(p) with local NFB. That'll improve speaker damping for enhanced bass performance, and by making the effective r(p) a smaller fraction of the output xfmr Xl(pri) you improve linearity across the board.
Otherwise, don't worry about it. 🙂
Hi Stefanoo,
Simply speaking a DC load line is for setting up the DC or bias conditions for a tube alone or you may think it is just the plate loading resistor of the tube.
However, a AC load line for a tube, you should take the following stage or the next stage into consideration, it becomes a very complex loading indeed. But simply, you may consider a AC loading of the stage you considered as the plate loading with the second stage's grid resistor in parallel.
E.g Plate loading 100K and the next stage grid resistor is also 100K then the AC loading / load line should be 50K.
.........
Simply speaking a DC load line is for setting up the DC or bias conditions for a tube alone or you may think it is just the plate loading resistor of the tube.
However, a AC load line for a tube, you should take the following stage or the next stage into consideration, it becomes a very complex loading indeed. But simply, you may consider a AC loading of the stage you considered as the plate loading with the second stage's grid resistor in parallel.
E.g Plate loading 100K and the next stage grid resistor is also 100K then the AC loading / load line should be 50K.
.........
Perhaps it has not been said, but the elliptical loadline applies when multiple frequencies are being amplified. If only one frequency were amplified, the speaker would have only one impedance and the loadline would be straight.
The highest extent of the elipse indicates the lowest possible value of impedance of the speaker (basically at low bass and high treble frequencies). The lowest part of the elipse indicates the highest impedance of the speaker, (which will be at some mid frequencies). On average it is approximately constant.
Here's a page that might be of some use. It at least has a picture of the elipse! http://www.freewebs.com/valvewizard1/se.html
The highest extent of the elipse indicates the lowest possible value of impedance of the speaker (basically at low bass and high treble frequencies). The lowest part of the elipse indicates the highest impedance of the speaker, (which will be at some mid frequencies). On average it is approximately constant.
Here's a page that might be of some use. It at least has a picture of the elipse! http://www.freewebs.com/valvewizard1/se.html
Merlinb said:
Er, no. If the impedance has a reactive component then the loadline will be elliptical at a single frequency. Remember CIVIL? The current leads the voltage in a capacitor and lags in an inductor, so you have to consider what happens to current and voltage at 0, 90, 180, and 270, rather than just 0 and 180 as we do with a resistor, and that leads to the ellipse.
- Status
- Not open for further replies.