Is this the proper way to test transistors?

[steele]

Is this the proper way to test transistors?
and to do a comparison?

https://youtu.be/gsMVwq1no_8

[bob91343]

I don't think so.

[Mark19960]

No.

The main issue with this chaps test is that you do not test power transistors without heatsinks. FAIL.
You will get wildly inaccurate and inconsistent results. Some of those devices will go into runaway before others.
That's nothing more than marketing nonsense. The device has been brought to the forum a few times and debunked.

That said... the marketing slogans all over the video about having "consistent results" or "adherence to NIST" are pure comedy when looking at how this so-called test was performed.
NIST and others have publications on how to do this.... and it does not involve floating a power transistor in the air and touching the back with a probe.
So, aherence to NIST went out the window really, really fast.

The reason these do not work is because elements that have an impact on the thermal dissipation of the device are not being changed.
The current flowing through the device is not being changed, no heatsinking is being added/changed, etc.
You might find a few devices out of a tube that are slightly better but whether or not that translates into a meaningful change is doubtful.

There are no test conditions published or specified. Even if those were specified it's still all wrong to start with.
There is no controlled condition whatsoever!
To really test this some jigs would have to be fashioned and a test methodology defined.

I would be curious as to what some of the metrologists on the forum think about it.

When I was researching this I came across several papers on the subject authored or co-authored by David L. Blackburn @ NIST:
https://ieeexplore.ieee.org/author/37283764700
A lot of very good reads. Also see NIST 400-86 that he co-authored.
https://www.nist.gov/publications/semiconductor-measurement-technology-thermal-resistance-measurements

But, marketing BS seems to win over the actual science..... even when the test is an obvious colossal failure.

[bob91343]

The guy who made this video needs to learn a bit more than he knows about electronics.  I agree that this is a ridiculous test.

In fact, changing the output transistor in a CB (ugh) radio probably will change nothing.  The circuit design determines the output, not the device.  Within wide limits.

To go a bit farther, if you change the power output by 10% that translates to 1 dB, 1/6 of an S unit.  And that means that the modulation may drop and make the signal audio weaker.

Don't waste your time with this crap.

[vk6zgo]

His commentary sounds a bit like early onset Alzheimer's.

[WattsThat]

What’s really amazing is that the guy has 15K subscribers and his YouTube screeds read like textbook examples of Dunning–Kruger effect.

An old boss had a favorite phrase for such cases (non-pc comment follows):

“Where the blind live, the one-eyed reign.”

[Mark19960]

He isn't trying to get more power - he is claiming the same as the dodgy marketing material does that these devices will result in less heat.
This component is part of the AM modulator/power regulator circuit. It's basically an adjustable regulator.

The common argument is that it's in a bigger TO3PL and thus will run cooler.
They ignore the maths.. because that means not making the sale. Even simple maths like the ones required to compute dissipation through the device.
The tests I have personally done were not even worth it because the maths just don't work.

All these devices in the larger packages do is reach temperature slightly faster than the smaller form factor TO218 style packaged devices.
And that is obvious... the TO3PL is a larger package and thus has lower thermal resistance. This can be a good thing in the right application, but this isn't it.
it's still dumping the same quantity of heat into the same relatively confined space and heatsink but that does not seem to click.
X watts of heat generated into the exact same heatsink is what it is.
You can either dissipate less heat by reducing the voltage drop thru the device, adding more heatsink,.....but why am I telling you lot this, you know this already

You have a completely ridiculous test with no controlled conditions performed by a chap that says he is a NIST conforming entity, which in of itself is a laugh.
Like I said, it's an advert. A false advert, but one nonetheless.

The tests are ... just wow. wow. batteriser levels of fail.