• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

Load lines. What?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I've read and read about load lines and I understand the graphs and how to calculate various things to put a dot in the graph, but, where do I WANT the dot? I hear about people placing the line to keep the tube in it's most linear operating point but all the lines looked curved to me. :confused:

I've been playing with this nifty calculator. 12AX7/ECC83 High-Mu Dual Triode Calculator

I guess the question is... How do I know what I want? Is it based on the B+ voltage I'm running? Or the gain I need? Or...?
 
They are curved... but it's the points where the said load line intersects them that define how the stage will behave. And the dot, you want it at a low current for this particular tube, but not too low because if you look closely then the points produced by the load line and the curves will be less equally spaced. If you want just a preamplifier stage you might even like it more with a little distortion (brings more "tube magic" if you ask me) but if it's part of an all-tube amp you want it to be "clean" as possible, since each stage will not only amplify that distortion but add it's own as well.
 
It's a hifi amplifier.

Do you mean like I've drawn the green lines?

attachment.php
 

Attachments

  • loadline.JPG
    loadline.JPG
    186.2 KB · Views: 616
Keep in mind that the published curves in the data sheets do not always represent what you will get in hand.... Even in the old days the curves were based on "Cherry-Picked" tubes... Also some of the curves were modified by hand to look a bit more linear....
The data sheets are only as good as how tight your operating conditions are...ie, the curves are based on having REGULATED voltages when the were sweeping...
So if you have an un-regulated power supply that sags 50 to 80 volts from idle to full power, then your fooling yourself to design with those curves... you can design the entire amplifier at full power output...
For Class A pre-amps you should be fine....
 
Crowhurst finally started making some sense to me after I read this:
How to "screw around" your tube load line

It's a funny read but it walks you through it pretty good. Sometimes it takes a few different approaches before things sink in. Between that article and guidance I received on this board I can predict my pre-voltage gain stage behavior much better.

In a class there would be lab experience and other students to draw from. It's difficult to learn from reading alone. -Fred
 
Keep in mind that the published curves in the data sheets do not always represent what you will get in hand.

That's true. However, load lines are usually accurate enough to predict the operating points of the tube and its most linear region. I always create a schematic before I build anything and work out all the voltages and component values theoretically before I start to solder. I am generally surprised and delighted to find that the circuit works as predicted from the math and the load lines. There is generally room for some tweaking and "dialing in", but my experience is that load lines are a critical tool for the tube audio engineer.
 
That's true. However, load lines are usually accurate enough to predict the operating points of the tube and its most linear region. I always create a schematic before I build anything and work out all the voltages and component values theoretically before I start to solder. I am generally surprised and delighted to find that the circuit works as predicted from the math and the load lines. There is generally room for some tweaking and "dialing in", but my experience is that load lines are a critical tool for the tube audio engineer.

Yes... I totally agree with you....
When I was a youngster I loved doing LOAD-LINE analysis ..especially the composite characteristics curves for Push-Pull load line analysis.. Would work the numbers to death and also be very happy with the build results...
As I got older I just throw things together and see what it sounds like without all the number crunching..sometimes more fun...
What I was trying to say in my last thread....is to be cautious of how you use the curves...
For example you have an all in one amplifier...integrated pre and power amp using the same supply.... And lets say the B+ on your front-end 12AX7 is at 300V at idle..... So you do your load line at 300V ... Now when the amp is playing music ...the B+ may dip to 200V for the front end B+ ...from the power tubes drawing current.... Now you have a moving target for your B+ ...ie, your load line is moving about the curves dynamically.... It gets worse if you are using a EF86 on the front end... The screen voltage will also move about and cause next order effect making the curves moving about...so you have moving load-line and curves moving around, which makes for more non-linear reproduction... A little regulator for the screens goes a long way...

Chris
 
Do you mean like I've drawn the green lines?

No. If you look at all the lines representing the grid voltage you will see that to the left side and at higher currents the lines are pretty evenly spaced. As the current drops and the plate voltage rises to the right side of the curves you will see the lines are compressed and get more closely spaced. The idea is to operate the tube in its most linear region and that region is where the grid lines are evenly spaced. The compression of the grid lines represents 2nd order distortion. In your drawing below you will see the two verticle green lines. I have chosen -1V as the bias point. You will see that 1 volt to the left to 0V and 1V to the right to 2V are the most equally spaced grid lines. Beyond 2V you see the lines start to compress. That is, as I said, 2nd harmonic distortion.
 

Attachments

  • 12AX7 curves.jpg
    12AX7 curves.jpg
    172.8 KB · Views: 326
Last edited:
Ok so you meant evenly spaced along the load line. That makes more sense.

How do plate and cathode resistors (cathode bias) effect the gain of the stage?

In that calculator, what is the difference between RT and Rk? What'd I'd like to do is adjust B+ voltage, plate resistor, and cathode resistor to see how it effects things. But which is which?

Crowhurst finally started making some sense to me after I read this:
How to "screw around" your tube load line

I read that one too. Thanks.

lets say the B+ on your front-end 12AX7 is at 300V at idle..... So you do your load line at 300V ... Now when the amp is playing music ...the B+ may dip to 200V for the front end B+ ...from the power tubes drawing current.... Now you have a moving target for your B+ ...ie, your load line is moving about the curves dynamically.
Chris

I was planning on an SMPS with voltage regulation so that should help. I may also do fixed bias on my output tubes (pp).
 
Last edited:
How do plate and cathode resistors (cathode bias) effect the gain of the stage?
Draw two load lines for two different plate loads. Next, see how what the min and max B+ is at the plate for lets say 1 volt of grid variation at each loadline. You do this by drawing vertical lines from the loadlines at the min/max of the given grid voltages. You'll see the bigger plate load will induce a larger voltage swing at the plate, so more amplification.

The cathode resistor (bypassed) sets the operation point. For hifi, it needs to provide enough headroom for the input. If it's too low or high, the input voltage will result in a clipped output, reducing the amplification. The right bias point will give maximum voltage swing and therefor ax amplification.

I was planning on an SMPS with voltage regulation so that should help. I may also do fixed bias on my output tubes (pp).
Have you done SMPS in audio before?? Not easy to get a good clean smps with the proper switching frequencies etc. Linear PSU's are easier in this respect.
 
Last edited:
Draw two load lines for two different plate loads. Next, see how what the min and max B+ is at the plate for lets say 1 volt of grid variation at each loadline. You do this by drawing vertical lines from the loadlines at the min/max of the given grid voltages. You'll see the bigger plate load will induce a larger voltage swing at the plate, so more amplification.

The cathode resistor (bypassed) sets the operation point. For hifi, it needs to provide enough headroom for the input. If it's too low or high, the input voltage will result in a clipped output, reducing the amplification. The right bias point will give maximum voltage swing and therefor ax amplification.

Have you done SMPS in audio before?? Not easy to get a good clean smps with the proper switching frequencies etc. Linear PSU's are easier in this respect.

Ok I'll try a few plots and see what's different. I'm hoping a 12AX7 cathodyne can drive 6L6s in a mobile amplifier. I believe my source signal is 4v peak to peak so I'll have to divide that down to 2v peak to peak for a 12ax7 correct?

I've designed a few switched mode motor controllers (up to 144kw) and modified some SMPS supplies. I have a fairly good understanding of trace layouts so I have a chance. :)

I'm building a 12v to 350v SMPS (to power a stereo 6L6 P-P) but I can't decide on an IC. 494, 3525, 3524.
 
Last edited:
lazzer408 said:
I'm hoping a 12AX7 cathodyne can drive 6L6s in a mobile amplifier. I believe my source signal is 4v peak to peak so I'll have to divide that down to 2v peak to peak for a 12ax7 correct?
No,the amplification grid to cathode of a cathodyne is somewhat less then 1.You have to feed the 12AX7 grid with a least the signal needed to drive the 6L6.
The question of how much can I enter into an amp is simple to answer.As long as Vin x amplification doesn't exceed the max.Vout,go ahead.
Mona
 
Ok I'll try a few plots and see what's different. I'm hoping a 12AX7 cathodyne can drive 6L6s in a mobile amplifier. I believe my source signal is 4v peak to peak so I'll have to divide that down to 2v peak to peak for a 12ax7 correct?
As Ketje pointed out, a cathodyne doesn't work that way. Remember a triode amplifies the 'difference' between grid and cathode as represented by the loadlines. Since the cathode resistor of a cathodyne is large, a voltage will develop at the cathode. In other words, it's swinging up and down together with the grid. This does a couple of things: it reduces the amplification to 1 (or a bit less to be exact), increases input headroom dramaticly and it increases linearity.

I've designed a few switched mode motor controllers (up to 144kw) and modified some SMPS supplies. I have a fairly good understanding of trace layouts so I have a chance. :)
Cool. Keeps us posted on an new developments.
 
As Ketje pointed out, a cathodyne doesn't work that way. Remember a triode amplifies the 'difference' between grid and cathode as represented by the loadlines. Since the cathode resistor of a cathodyne is large, a voltage will develop at the cathode. In other words, it's swinging up and down together with the grid. This does a couple of things: it reduces the amplification to 1 (or a bit less to be exact), increases input headroom dramaticly and it increases linearity.


Cool. Keeps us posted on an new developments.

The first 1/2 of the 12ax7 is being used as the gain stage. My signal hits it first but I'm having very poor results using the tube this way and might go back to a paraphrase. I find I'm having to use huge amounts of NFB to stop this snapping sound in the speakers. It doesn't just distort, it crackles.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.