search

KiCad / TO_SOT_Packages_THT.pretty

2 stars 19 forks source link

Verify pinout of TO* footprints #6

Closed odurc closed 7 years ago

odurc commented 8 years ago

According #3 the following footprints may be wrong: TO-12, 18, 43, 72, 3.

evanshultz commented 7 years ago

I was able to find some of these packages at:

TO-18 is correct: image image

TO-72 is correct: image image

TO-3 seems to vary: image image image image image

So I believe the KiCad library has been updated since this issue was posted. I'll keep looking for package drawings of the other 2 to confirm them.

jkriege2 commented 7 years ago

@evanshultz Will you post a PR for TO-3?

evanshultz commented 7 years ago

@jkriege2 Of course! But I'm stuck on the name. Having pins 1 and 2 swapped don't form any kind of natural convention. Pin1_Top and Pin1_Bottom are too lame. Suffixing _Reverse is the only thing I can think of that's universal (and used elsewhere in our library), but that's not very intuitive for the CvPcb/Pcbnew user. Do you have a suggestion?

jkriege2 commented 7 years ago

I had a quick glance: The package drawings above for TO-3 are BOTTOM views, if you rotate them correctly to the TOP view that we see in KiCAD, the numbering in the screenshots matches + all symbols that use the TO-3 should be OK (just checked them again against the current footprint and the respective DSs):

So, in conclusion I would say the TO-3 is correct ... or did I overlook something? Can we close this then?

JAN

evanshultz commented 7 years ago

I (disturbingly) can't find a TO-12 package drawing online, but our footprint has pin 1 as the first pin CCW from the tab, which is consistent with all the other TO footprints in the lib: image

I don't find TO-43 footprints at https://github.com/KiCad/TO_SOT_Packages_THT.pretty, but perhaps that was a typo and should be TO-46? What I do see is that our TO-46 footprint matches the parts I see: http://www.mouser.com/ds/2/187/HFE4070-500_serie-191281.pdf image http://cds.linear.com/docs/en/controlled-drawings/RH129_05-08-5027_SPEC_REV.H.pdf image http://www.ti.com/lit/ds/symlink/lm34.pdf image image

So with regards to everything but TO-3, I think it's fine. I would like to find a TO-12 part datasheet.

@jkriege2 In response to you, Jan, I do have one outstanding question about TO-3. All of your work above looks great and seems to match. But I posted a Microsemi datasheet above which was the only outlier. It's the last package drawing I posted before the KiCad TO-3 footprint. It also appears to be a bottom view but the pin numbers and pinout are reversed.

Let me separate this out a bit:

So if I read this all right, there is a mismatch. Sorry for not going into more details about exactly where I saw the conflict earlier.

jkriege2 commented 7 years ago

@evanshultz yes, now I see ... sorry ;-) ... mmhhh ... looking a round a bit I also found an IRF part that has this pinout: http://www.irf.com/product-info/datasheets/data/irfag50.pdf ... they call this package TO-204AA ... but then that name is also in this 2N3055 datasheet http://www.onsemi.com/pub/Collateral/2N3055-D.PDF ... i.e. a mess ;-)

Currently I'm tempted to say we stick to one symbol and let the user ensure that he get's then pin-mapping right, because creating two that only differ in the pin ordering might raise even more uncertainty? the work to check the pin-ordering against the DS during review is the same with one or two footprints ...

On the other hand, having two footprint options (e.g. TO-3_Style2and TO-3_Style1) might raise the awareness taht there are different interretations in future ...

Don't really know how to proceed on this ...

JAN

evanshultz commented 7 years ago

@jkriege2 Great find! To me, this adds even more reason to make an alternate footprint. Otherwise, since the symbol pinout is fixed, we have a situation where a user could select a part and put a symbol on their schematic, but there's no correct footprint. They might quit KiCad or use the wrong footprint, which would be a bad.

If two footprints show up based on the FPfilter, then at least the user will need to think and find the right one. That seems more safe and just better, to me.

But that doesn't resolve how to name it. I mentioned adding the suffix _Reverse above, but I'm not in love with it. Is there something better you can think of?

jkriege2 commented 7 years ago

What about TO-3_SwappedPins12? Also not really nice, but WYRIWYG (what you read is what you get ;-)

JAN

evanshultz commented 7 years ago

I don't like that footprint name. Since we find the existing footprint more common, but not a standard or obvious (let's say you're most familiar with Microsemi parts), what is it swapped from? And then neither footprint has information to help the user until they look at the footprint. (Neither does adding Reverse, for that matter, but it's more generic.)

Any other footprint name ideas?

bobc commented 7 years ago

However, the Microsemi part calls the top pin pin2 and that's the emitter.

I suspect the Microsemi datasheet is incorrect, the diagram on page 8 does not page 7!

evanshultz commented 7 years ago

@bobc I see what you mean. The table on page 7 does not match the schematic on page 8. They both use T1 and T2, but the pins assignments are swapped between the pages. Probably in error.

With that, I'm closing this case and considering it resolved.