Electronics > Projects, Designs, and Technical Stuff
Testing 1N4007 with a megger
<< < (5/8) > >>
iMo:
It is a common practice manufacturers produce, for example, 1n4007 chips only, marking them 1n4001-1n4007. The same happens with transistors, memories, mcus, etc., etc.
They also buy chips from other vendors and package them under their brand, when in need.
The most expensive item in semiconductor business (except r&d) is the production of the sets of masks for photo-litography, the production of a silicon die costs almost nothing. You could save say 90% costs making the 1n4007 only.
taydin:

--- Quote from: tooki on November 01, 2019, 01:22:20 pm ---So... they all failed well beyond their maximum ratings. And this is a surprise why?

--- End quote ---

The surprise is that the diode MEASURED to be working fine with a multimeter after it broke down around 1850V. I think I made that clear in my first post.
jaromir:

--- Quote from: tooki on November 01, 2019, 01:22:20 pm ---Well, they can be the same, or not, depending on manufacturing method. E.g. the Vishay ones in the attached datasheet could very well just be binned dies of the same design, since they list identical specs other than breakdown voltage, while another brand's might not be.

--- End quote ---

This is indeed true. Let me share a bit of insider info.

More than decade ago, I worked for major semiconductor manufacturer other than Vishay. For simple rectifying diodes, they manufactured probably three dozens of diode types from one type of silicon. Silicon wafer was doped and gold was sputtered from both sides, making huge "mother diode" (with reverse voltage higher than 1kV) with 30cm diameter PN junction. This basic material was cut into smaller pieces, with variety of perhaps 5 sizes and shapes. For 10A diodes the pieces were larger and hexagonal in shape, 3A diodes were smaller sqares, 1A diodes were the same, with smaller lead cross-section - the limiting factor wasn't current density in silicon or leads, but heat dissipation. When buying larger diode, you are paying for copper too, not only silicon.
All diodes were manufactured as the "highest grade" of the family. For example, all 1N400x were manufactured as 1N4007, the "mother diode" was doped to have reverse breakdown voltage well above 1kV and vast majority of the diodes from manufacturing line had indeed breakdown above 1kV. There were no "mother diodes" for lower breakdown voltage. Lower voltage bins were filled at final diode testing by diodes which failed 1kV test. If those bins weren't full enough (the manufacturing yield was quite good), it was filled from 1N4007 bin. As an example, 1N4006 datasheet states breakdown voltage higher than 800V and 1N4007 (Vbr>1000V) fulfills this criterion. Diodes with higher than 1kV Vbr rating were made as stacks of 1kV chips in one package.

Vast majority of customers were aware of this and ordered 1N4007 exclusively. Reason for having lower grades diodes comes from past, when lower manufacturing yield made binning more reasonable. Nowadays it isn't as attractive and all grades are sold for the same price.
floobydust:
There was a time 1N4005-1N4007 and 1N5405-1N5408 used to be spec'd with higher Vf, lower junction capacitance, longer reverse recovery time, higher thermal resistance - it depended on the manufacturer and their datasheet. The doping or construction was different.

It is in antique databooks the difference but now pretty much the same. It might be that sorting/binning the parts costs more than simply making the 1,000V part, or semi processes have improved.
Almost twice junction capacitance 1N4005-1N4007 Diodes Inc. aww its obsolete now.
Vishay S1A note 800V, 1,000V part differences

So they aren't all the same but who really designs that tight nowadays.
It was discussed here https://electronics.stackexchange.com/questions/84471/what-is-the-difference-between-1n4001-and-1n4007-other-than-their-maximum-revers
tooki:

--- Quote from: taydin on November 01, 2019, 03:42:28 pm ---
--- Quote from: tooki on November 01, 2019, 01:22:20 pm ---So... they all failed well beyond their maximum ratings. And this is a surprise why?

--- End quote ---

The surprise is that the diode MEASURED to be working fine with a multimeter after it broke down around 1850V. I think I made that clear in my first post.

--- End quote ---
I dunno, IMHO if a part has been so grossly abused, it’s suspect, period. And my years of working as a computer tech taught me that parts often semi-fail, leaving them in marginal states where they sorta work.
Navigation
Message Index
Next page
Previous page
There was an error while thanking
Thanking...

Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod