Do IC manufacturers enjoy torturing their customers?

[SharpEars]

So I am looking at linear voltage regulators and I look at the datasheet for the Linear Technology LT3081. This part was just announced on June 21, 2013 (http://cds.linear.com/docs/en/press-release/LT3081.pdf). So, this thing is a little over two years old. I figure it's a novel part with excellent features.

While looking at the datasheet (http://cds.linear.com/docs/en/datasheet/3081fc.pdf), I notice on page 24 an example "Constant-Voltage, Constant-Current 20V/3A Lab Supply" that shows a comprehensive example of how a full 0-20V/3A power supply can be made using this part. Excellent, I think to myself, a complete functional example with voltage and current limiting. Then, I start looking at the parts in this circuit diagram and my eyes are drawn to the enhancement mode N-Channel MOSFET being used, the TPO610T. OK, no problem I figure, let's see where I can acquire this part. After all, the LT3081 was just recently released and so this datasheet is at most two years old, so I should have no trouble finding the part, right?

WRONG! It's like someone at Linear just made this shit up as he went along. What the hell is a TPO610T and where on earth do you find one? Did someone just pull this part number out of their ass? Is this a part made fifty years ago that hasn't been made in the last fourty-five and they just decided to throw it into the circuit to fuck with us? I really would like to know the thought process behind this sort of stupidity in sample circuits found in data sheets. This isn't the first time I've come across this. If they want people to actually be able to build the circuits, why put unobtainium fairy unicorn parts into them?

[Vgkid]

Found it, remove the 'T'.
http://www.mouser.com/Search/m_Search.aspx?Keyword=TP0610

[SharpEars]

Found it, remove the 'T'.
http://www.mouser.com/Search/m_Search.aspx?Keyword=TP0610

I tried that link with and without the T after the 610 (in the link) and it doesn't lead to a part. Could you please post a working URL?

[c4757p]

It's not the T, it's a typo in the datasheet.

It's a TP0610T (TPzero610T), not TPO610T (TPoh610T).

[SharpEars]

It's not the T, it's a typo in the datasheet.

It's a TP0610T (TPzero610T), not TPO610T (TPoh610T).

Haha, OK tell me I'm crazy and that's not the letter O in the datasheet?

Thanks for the correction!!

[c4757p]

It is a letter O in the datasheet. But the real part has a zero.

Nice rant though.

[SharpEars]

It is a letter O in the datasheet. But the real part has a zero.

Nice rant though.

It was worthy of a rant. I am sick and tired of seeing unobtainium parts in sample circuits in datasheets. Until you found the error in their part number , it sure seemed like a typical example of this practice. Somebody really needs to proofread and verify the part numbers on in datasheets. I am sure that I am not the first frustrated person to come across this sort of thing.

[JoeO]

It is time for the US to wake up and solve problems that others have solved years ago.

0(zero) should have a slash through it.
1(one) should have a slash through it.
7(seven) should have a slash through it.

4.7K should be 4K7.

4.7 ohms should be 4R7.

Is there any that I missed?

[c4757p]

1 should have a slash through it?

It's all computerized now, the problem is solved. The Linear engineer who made that document should have copied and pasted the part number from the manufacturer's datasheet to ensure it was exactly, character-for-character, correct.

[JoeO]

I used to work with some German Engineers.  They would send me hand drawn schematics.  The nomenclature I mentioned above is what they used.  Even after the schematic ran through a fax machine, you could still figure out the parts values that were being used.
Here are the numbers that they used.  I seem to remember that they put a slash in their zeros, but I could not find it online.

http://german.stackexchange.com/questions/11027/how-different-is-german-handwriting-from-americans

[georges80]

Having worked at a semiconductor company.... Lasted 1 year before I left in frustration.

You do realise (obviously not) that there is a separate department that takes care of all the formatting/entry/editing/pretty stuff that goes into a published datasheet?? i.e. there is a disconnected between the original datasheet and the published one.

Also, if you really believe that the appnote/schematics are actually proven designs - I'll educate you by stating many have *NEVER* been built/breadboarded etc.

App notes are published to SELL product, that is their #1 justification. The only app notes I would believe are ones that come with an actual real demo board. Linear Tech does have a variety of demo boards/appnotes and those a great and can be tested on the bench. App notes that are only paper I wouldn't trust as far as the flames will reach when set on fire (once printed out on paper)

cheers,
george.

[c4757p]

You do realise (obviously not)...

Obviously? Anyone with half a clue should realize that.

...that there is a separate department that takes care of all the formatting/entry/editing/pretty stuff that goes into a published datasheet?? i.e. there is a disconnected between the original datasheet and the published one.

The departments still pass data between each other somehow. At no point should the data have been handled directly by a human. It should have been copied by computer all the way from Supertex's databases, through the Supertex/Microchip datasheet, through LTC's engineering departments, to the people doing the datasheet formatting, and into the datasheet. Welcome to the 21st century.

[georges80]

Well, the OP seems to think that the schematics are likely good to go...

And the 21st century has nothing to do with documents being re-edited by the publishing department. It happens and that's the reality...

We haven't quite reached perfection/nirvana, that won't happen for a long time...

cheers,
george.

[SharpEars]

Well, the OP seems to think that the schematics are likely good to go...

And the 21st century has nothing to do with documents being re-edited by the publishing department. It happens and that's the reality...

We haven't quite reached perfection/nirvana, that won't happen for a long time...

cheers,
george.

Actually, the OP thinks that after the publishing department has its way, the tech department proof reads for technical accuracy.

[Monkeh]

Well, the OP seems to think that the schematics are likely good to go...

And the 21st century has nothing to do with documents being re-edited by the publishing department. It happens and that's the reality...

We haven't quite reached perfection/nirvana, that won't happen for a long time...

cheers,
george.

Actually, the OP thinks that after the publishing department has its way, the tech department proof reads for technical accuracy.

You may find otherwise.

Copying examples out of datasheets very often does not result in a working, useful circuit. This is not Arduino, you can't just c+p things and have a device.

[georges80]

Well, the OP seems to think that the schematics are likely good to go...

And the 21st century has nothing to do with documents being re-edited by the publishing department. It happens and that's the reality...

We haven't quite reached perfection/nirvana, that won't happen for a long time...

cheers,
george.

Actually, the OP thinks that after the publishing department has its way, the tech department proof reads for technical accuracy.

Yes, but like I said, MANY app notes/design notes are NEVER built...

You have to also understand that the folk writing the datasheets etc are often app engineers and often not senior or that experienced and the chance of picking up/fixing an error is not guaranteed... In some of the companies I've worked at, the app engineering reports to marketing since often that area is sales related.

cheers,
george.

[AF6LJ]

I used to work with some German Engineers.  They would send me hand drawn schematics.  The nomenclature I mentioned above is what they used.  Even after the schematic ran through a fax machine, you could still figure out the parts values that were being used.
Here are the numbers that they used.  I seem to remember that they put a slash in their zeros, but I could not find it online.

http://german.stackexchange.com/questions/11027/how-different-is-german-handwriting-from-americans

I came across this in the seventies and adapted to it just fine.
I am old enough that when I I took drafting we were taught how to draw numbers people could recognize.
This was a good rant, been there....

[T3sl4co1l]

The departments still pass data between each other somehow. At no point should the data have been handled directly by a human. It should have been copied by computer all the way from Supertex's databases, through the Supertex/Microchip datasheet, through LTC's engineering departments, to the people doing the datasheet formatting, and into the datasheet. Welcome to the 21st century.

Ah, but what about character encoding?  Formatting?

When I write documents, I don't expect the PDF output to be 100% verbatim of my input -- for example, LaTeX converts ligatures for me, so copying certain letter pairs ("fl" might be one, I forget?) doesn't quite look right.  (And for that specifically, I think there's a package you can load that makes the selectable/copyable text true to the plain text, I just haven't looked into using it.)

Though I don't think this would happen here, and I would prefer to write a part number in typewriter face or something like that, for clarity.

Anyway, o and 0 are a common confusion.  I like to think I am quite accurate at telling them apart, but the average person, at a casual glance, can make such mistakes easily.

Tim

[c4757p]

Ah, but what about character encoding?

It's not 1998 anymore, do you even utf8 bro?

Formatting?

When I write documents, I don't expect the PDF output to be 100% verbatim of my input -- for example, LaTeX converts ligatures for me, so copying certain letter pairs ("fl" might be one, I forget?) doesn't quite look right.  (And for that specifically, I think there's a package you can load that makes the selectable/copyable text true to the plain text, I just haven't looked into using it.)

Good point, though that doesn't usually apply to part numbers - ligatures in capital letters and intermixed numbers and letters aren't very common.

[T3sl4co1l]

UTF8 is one thing, but UTF16 is another, and ISO-8859-1 is still around, even Ye Olde CP437 from time to time.  And the graphic rendered for a given character varies widely with the font.

The great thing about standards is, there's so many of them  ...and developers just pick whichever they're familiar with, or whatever their tools provide (more or less transparently, at least until something breaks).  Which, in turn, might involve more well-informed, but nonetheless backwards-compatible, decisions.

I don't know that there's even any necessary (required, or even reliable) correspondence between graphic and code with OCR documents.  Considering how ugly some of those algorithms are, it's a wonder they're text-searchable at all.

Tim

[alsetalokin4017]

I ran into a similar "gotcha" yesterday. I was trying to install and run a shell script file that had a very long name and a three letter extension. The "readme" file that gave the instructions for the installation said to run the file by entering ./(longcharacterstringwithdashes_andother-nonalphabetic0-9chars).EXT... and I kept getting "no such file" errors when I knew darn well that the file existed in the directory. Of course I was copy-pasting from the readme file because of all those characters in the filename... and it took me several minutes of careful comparison to finally see that the _actual_ extension of the file in the directory was... not capitalized. Linux is case-sensitive.
 

[Whales]

1 should have a slash through it?

I think they mean the hat/tick on the number 1.  As opposed to drawing it just as a stick.

[Monkeh]

Nobody likes the number I

[dom0]

I used to work with some German Engineers.  They would send me hand drawn schematics.  The nomenclature I mentioned above is what they used.  Even after the schematic ran through a fax machine, you could still figure out the parts values that were being used.
Here are the numbers that they used.  I seem to remember that they put a slash in their zeros, but I could not find it online.

http://german.stackexchange.com/questions/11027/how-different-is-german-handwriting-from-americans

Slashed zeroes are used by some engineering people when hand-writing numbers here, yes.

[Karel]

In the time everybody was using DOS on their pc, the default/builtin font of the graphics cards had always a slash through the zero or a dot in the zero. Engineers/programmers who don't use a similar font should be fired...

By the way, here's your fet at Digikey: www.digikey.com/product-detail/en/TP0610T-G/TP0610T-GCT-ND/5169542