The signal needs to be a sinewave, and you compare input with output of the amplifier. If you're using oscilloscope probes, it's important that you have calibrated them first. 45 degrees is convenient to view, but not essential. The key point is whether it's a line or an ellipse. Another thing to watch out for is that oscilloscope attenuators may not have quite the same frequency response (which is highlighted by phase response) at different settings.
The trouble with looking at Lissajous figures is that you can't really measure phase accurately. (Good for frequency comparison, though.) You see, there's no problem with phase shift between input and output of an amplifier, what's important is whether it's the right amount. If you plot phase against linear frequency, you ideally want a straight line. The gradient gives the delay. If all frequencies are delayed by the same amount, who cares? The problem is that they often aren't. Achieving constant group delay is very definitely a good thing.
The trouble with looking at Lissajous figures is that you can't really measure phase accurately. (Good for frequency comparison, though.) You see, there's no problem with phase shift between input and output of an amplifier, what's important is whether it's the right amount. If you plot phase against linear frequency, you ideally want a straight line. The gradient gives the delay. If all frequencies are delayed by the same amount, who cares? The problem is that they often aren't. Achieving constant group delay is very definitely a good thing.
Can't you get a rough feel for that though, by seeing if (and how) the shape of the ellipse changes? I paid very little attention in my electronics classis in college, but I seem to remember that the 'width' of the ellipse (or the length of the shorther axis) was proportional to the amount of phase mismatch, or something like that?
And I guess the question "And what could I do to make the group delay more constant" would be the subject of an entire textbook...
And I guess the question "And what could I do to make the group delay more constant" would be the subject of an entire textbook...
It's not very accurate though.
Did you, now? Yes, you could determine phase shift from Lissajous width, but it would be a lot easier to measure it from a standard display.
AC couple, and set both traces to coincide with the centre line. Tweak the variable timebase to make one cycle last 5 major divisions as measured along the centre line (use horizontal shift to help). Count the number of squares between positive-going edges, and divide it by 5 to give a ratio, then multiply by 360 degrees.
Saurav said:
Did you, now? Yes, you could determine phase shift from Lissajous width, but it would be a lot easier to measure it from a standard display.
AC couple, and set both traces to coincide with the centre line. Tweak the variable timebase to make one cycle last 5 major divisions as measured along the centre line (use horizontal shift to help). Count the number of squares between positive-going edges, and divide it by 5 to give a ratio, then multiply by 360 degrees.
Yup. Never thought I'd need to deal with analog electronics again. Software was (and still is) my thing, digital logic circuits I could do OK on, analog electronics was the first thing I threw my hands up at. I even remember the specific topic that did it - feedback. When I look back on it now, I think the professors I had for certain courses played a big role in what I ended up liking.
I use most of my software/firmware training at my job. 10 years ago I didn't think I'd ever need my analog education (or lack thereof) again, so it's kinda ironic that I ended up with a hobby that requires me to re-visit Thevenin's theorem and f = 1 / 2 PI R C.
What were you hinting at with your question, anyway?
Hi,
No, for the simple reason that I'd use an IT xformer here.
I can't quite remember the B+ you have, Joe..something like 250VDC?
In which case you can try to shift the anode resistor in front of the choke, decouple it with say 47µF and put the choke inbetween.
It won't be optimal, I think, but should give you some idea of what the choke does.
Give it a shot,
No, for the simple reason that I'd use an IT xformer here.
I can't quite remember the B+ you have, Joe..something like 250VDC?
In which case you can try to shift the anode resistor in front of the choke, decouple it with say 47µF and put the choke inbetween.
It won't be optimal, I think, but should give you some idea of what the choke does.
Give it a shot,
Sch3mat1c said:
Really depends what you're trying to do doesn't it?
High voltage is hardly required for a choke or IT loaded to drive an SE EL34 with a Va of 300V and a grid of 30V or so.
Running a 6SN7 at say 220V/12mA choke/IT loaded (for arguments sake, a horizontal loadline) gives >110V before it's grid goes positive, and lower and hotter, within spec, gives lower distortion. Still has enough voltage swing to drive a 211 or 300B and though I'd want a bit more headroom in this application it's plenty for all the trioded pents/tets commonly in use.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.