# New videos: How to test transistors, BJTs and MOSFET

#### tindel

Joined Sep 16, 2012
776
Good videos - a bit of a refresher for me, but I did learn a new trick for testing those mosfet's by charging the gate and discharging through my finger (why didn't I think of that?) Every failed FET I've ever seen has shown a failure mode of a short from drain to source though too.

I do have a question...

To properly match transistors don't you need to see the transistor curves on a curve tracer and try to select transistors with similar curves? Most beta testers only test at a single current (usually about 1mA base current). I've noticed that gain can vary wildly with base current. I assume this is mostly due to the early effects in the transistor. I don't know that measuring a single point beta really matches a pair of transistors. Not in my experience anyway trying to make current mirrors. Care to comment?

#### w2aew

Joined Jan 3, 2012
219
Good videos - a bit of a refresher for me, but I did learn a new trick for testing those mosfet's by charging the gate and discharging through my finger (why didn't I think of that?) Every failed FET I've ever seen has shown a failure mode of a short from drain to source though too.

I do have a question...

To properly match transistors don't you need to see the transistor curves on a curve tracer and try to select transistors with similar curves? Most beta testers only test at a single current (usually about 1mA base current). I've noticed that gain can vary wildly with base current. I assume this is mostly due to the early effects in the transistor. I don't know that measuring a single point beta really matches a pair of transistors. Not in my experience anyway trying to make current mirrors. Care to comment?
Of course, there are much better ways to match transistors as you have correctly pointed out. However, without access to a curve tracer or without building a test circuit for detailed measurements, the simple measurements of beta that these testers can give will at least get you in the ballpark.

The early voltage affect is one variable that will affect collector output impedance as well as make Beta a function of collector voltage.

#### rc3po

Joined Feb 12, 2014
56
I bought a Peak Atlas DCA75 Pro from anatekcorp.com to take care of that.
It works great.

Last edited:

#### studiot

Joined Nov 9, 2007
4,998
You can also compare transistors and other components with a low voltage CT transformer and a scope.

So older scopes has a 'component tester/comparator' based on this.

#### studiot

Joined Nov 9, 2007
4,998
Yes Bertus that is one version and you do not need the centre tap since you can create an artificial centre with resistors.

My version (not mine really) predates octopi and huntrons by several decades from the days of valve equipment, note the selenium rectifier test.

Note the calibration switch.

#### Attachments

• 38.5 KB Views: 45

#### tindel

Joined Sep 16, 2012
776
I've been thinking about designing a device that connects directly to the front end of an oscilloscope that works similar to the curve tracers of old. I'm just not quite sure where to start... I guess studying the old curve tracers to see how they did it would be a good start! I really want to match transistors and tubes... taking a single data point or a single bias point seems inadequate to me. Tubes are a little more difficult due to the the high bias voltages - I'd probably start with just simple solid state devices.

Agilent is the only company that is making curve tracers that I'm aware of these days (and we're talking \$!). I should ask them for a demo at work. The old lab manager - in his infinite wisdom (very tongue in cheek) surplussed our functioning Tek tracer.

#### w2aew

Joined Jan 3, 2012
219

#### studiot

Joined Nov 9, 2007
4,998
Nice videos.

Did you ever do a follow up with more components or component matching?

#### rc3po

Joined Feb 12, 2014
56
Here's the video I did on this a while back: