Fet curve tracer

[Gus]

Looking in one of my data books I found a nice little fet curve tracer circuit.

Problem is I can not find it at the NS site to link it here.

FET Curve Tracer
National Semicinductor
John Maxwell
February 1977

Maybe someone can find it at the NS site.

 

[rafafredd]

yes, that would be nice to see...

 

[NewYorkDave]

Does it drive the gate with a stepped waveform and ramp the drain voltage, or is it the simpler one-curve-at-a-time type?

 

[Gus]

very simple one curve on a scope ID vs VGS

But for fet compressors etc this should help picking parts

 

[Gus]

1982 transistor dat book

 

[NewYorkDave]

You don't need a schematic. Just DC-couple your scope and set to X-Y mode, use a small value source resistor, provide a suitable drain supply voltage and drive the gate with a negative ramp (for an n-channel JFET). The gate voltage is the X axis and the voltage at the source (which represents the drain current) is the Y axis.

 

[analag]

Are you going to build it, Gus. I have a curve tracer and (for those who don't know) It's an eye opener. I have not traced one tube that corresponded with the data sheet. The lines are much more curved at the bottom than they would have us believe. Thus I am not impressed by those data sheet mechanics.

analag

 

[clintrubber]

[quote author="analag"]Are you going to build it, Gus. I have a curve tracer and (for those who don't know) It's an eye opener. I have not traced one tube than corresponded with the data sheet. The lines are much more curved at the bottom than they would have us believe. Thus I am not impressed by those data sheet mechanics.

analag[/quote]
Is your curvetracer scheduled for a yearly calibration ? :wink:

Just kidding, I could imagine what you're saying. Don't want to insult your tracer, whatever type it is and however new or old. But is it beyond doubt ?

 

[NewYorkDave]

What analag says about the lowest end of the datasheet curves is true and has been known for a long time. That's why savvy designers generally stay out of that region. Besides, even the approximation given in the datasheet should tell you that it's highly nonlinear down there and you should stay away.

As usual, PRR summed it up pretty well (in a post from two years ago):
[quote author="PRR"]Same on the plate curves. They take a few numbers, mostly at higher currents, and then use a French Curve to draw nice lines. Don't squint too far into the lower left corner-- it may be totally bogus. [/quote]

So, if you wanna play around in the cellar, you'd better bring a flashlight.

 

[analag]

[quote author="clintrubber"]Is your curvetracer scheduled for a yearly calibration ? :wink:

Just kidding, I could imagine what you're saying. Don't want to insult your tracer, whatever type it is and however new or old. But is it beyond doubt ?[/quote]

6021 subminiature tube being traced.

Dave and PRR are indisputable tube gurus....I rest my case.

analag

 

[DerEber]

So after tracing your tubes or FET...... how big is the difference in Audio Quality you realy hear?
Is it very obvious?

 

[kvintus]

It would be really cool to have a tube tracer. I could find a lot of use for one. Is it something I can build? Schematics?

/Anders

 

[analag]

[quote author="DerEber"]So after tracing your tubes or FET...... how big is the difference in Audio Quality you realy hear?
Is it very obvious?[/quote]

This is the kind of question that makes me go smoke a cigar....I'm going now.

analag

 

[clintrubber]

[quote author="analag"]
http://groupdiy.twin-x.com/albums/userpics/10049/normal_Curve.JPG
6021 subminiature tube being traced.

analag[/quote]
I see, it sure can't be the curvetracer (I have the same Tek-tracer)
:thumb:

 

[analag]

I find that Tesla tubes are the closest to the data sheet, even the lines are cleaner.

analag

 

[Gus]

That is the one.

 

[kvintus]

Cool. Thanks!

/A

 

[analag]

You know it can be modded to trace tubes, don't you?

analag

 

[mik]

please, Analag, show us how.
:thumb:

M.

 

[Larrchild]

It's very complicated.
Step 1. Buy a Tektronix 577 curve tracer.

Step 2. Remove transistor test jig and make a plate with tube sockets to go in it's place.

Step 3. Build a filament supply and connect to test jig.

Step 4. Trace Curves!

sorry, triodes only And if you need 100v of step generator bias to show cutoff of varimu tubes, you mite need a dc amplifier on the step-gen.

The way I cheat is that the silver knob on the right is an offset adj for the ramp so I can make the 1st trace start at any voltage. So by adding offset, i can view higher portions of the bias point.

Note: the Tek 577 retains it's functionality as a solid state curve tracer (if you're into that sort of thing) and is not modified .