Why would a PCB kit manufacturer provide the wrong footprint for a TO-92 part?

[JoeN]

On a velleman kit, the footprint is "staggered" but the pins on the TO-92 part are normal and in a line.  This causes the part to have to stand up 5mm or so from the board instead of fit flush on the board like every other part (ICs, diodes, resistors, capacitors).  Is there an advantage to doing it that way?  Or was it just a mistake?



P.S.  Velleman PCBs are atrocious.  The pads come up really easily.  Do not use lead-free solder.

Others have found opportunities with this kit where I have only found fault:

 

 

[wraper]

That is completely normal footprint. Sometimes TO-92 come with leads preformed this way too. Second is not perfect at all, useless for wave soldering as part fill fall completely through the holes and won't stay lifted. Also distance between the pads is rather small, especially with old crude PCBs from the past (or just very crappy modern) this might pose a problem. As well will pose a problem with soldering for beginners.

[wraper]

More often leads are formed like this

[JoeN]

That is completely normal footprint. Sometimes TO-92 come with leads preformed this way too. Second is not perfect at all, useless for wave soldering as part fill fall completely through the holes and won't stay lifted. Also distance between the pads is rather small, especially with old crude PCBs from the past (or just very crappy modern) this might pose a problem.

Why would it be preferable for the transistor to not "fall through" the holes and be flush with the PCB where that situation works fine for every other part?

I know about pre-formed legs.  The kit did not come with transistors with pre-formed legs.  Otherwise, I would not have written this thread.

Given the quality of this PCB, I see your point on the pad spacing.  I never have problems with this footprint when I order it from OSHPark, though.  I use the second pattern with TO-92 parts that do not have preformed legs when designing PCBs for myself.

[wraper]

That is completely normal footprint. Sometimes TO-92 come with leads preformed this way too. Second is not perfect at all, useless for wave soldering as part fill fall completely through the holes and won't stay lifted. Also distance between the pads is rather small, especially with old crude PCBs from the past (or just very crappy modern) this might pose a problem.

Why would it be preferable for the transistor to not "fall through" the holes and be flush with the PCB where that situation works fine for every other part?

Read the datasheet of any to-92 part with soldering process description and check how close to the case it is allowed to solder them.

[JoeN]

That is completely normal footprint. Sometimes TO-92 come with leads preformed this way too. Second is not perfect at all, useless for wave soldering as part fill fall completely through the holes and won't stay lifted. Also distance between the pads is rather small, especially with old crude PCBs from the past (or just very crappy modern) this might pose a problem.

Why would it be preferable for the transistor to not "fall through" the holes and be flush with the PCB where that situation works fine for every other part?

Read the datasheet of any to-92 part with soldering process description and check how close to the case it is allowed to solder them.

I'll take your word for it.  I wonder then why this is an issue for transistors and not diodes and ICs which are basically the same sort of silicon junctions.  I guess the wires from the outer edge of IC packages are a little bit further from the solder and when you take into account how the diode's wires usually are given a short run above the pcb to where the wire makes a 90 degree turn and heads down, the diodes are further away too.  So the TO-92 package is prone to soldering damage if it has no vertical offset from wave soldering.  Did not know and now I do.

P.S.  In DipTrace, at least, the second footprint seems to be used in all cases for TO-92 transistor components.  The first one with the offsets is an option (and so is the spaced out one you showed but I didn't because velleman did not use it), but you have to replace the pattern yourself.

[Benta]

If you look at wraper's 1st reply, you'll see that the pins have standoffs. There's even a callout L₁ which tells you how high the device will sit. TO-92 is not designed to sit flush with the PCB.

[Gyro]

I prefer the offset center pin as I think it puts much less stress on the package - the two outer pins (which carry the bond wires) can remain parallel. The stresses involved in splaying the legs to fit an in-line footprint (unless factory preformed) can become very high as the mounting height is reduced, especially as the epoxy package is then heated during soldering.

It also makes reverse insertion much less likely as the pad layout visibly matches the package outline.

[mikeselectricstuff]

I know about pre-formed legs.  The kit did not come with transistors with pre-formed legs.  Otherwise, I would not have written this thread.

May be they have changed supplier since the PCB was designed.

Also for a kit, it would make sense to have the pads spaced out more to reduce the risk of solder shorts

[T3sl4co1l]

I don't like flush-mounting semiconductors.  Transistors, LEDs, diodes (vertical arrangement), etc.  Or any other components really, if I can help it.

There should be lead length to allow for flex, warp and strain in the board, and clearance allows solvent to wash under the component (if the board is being cleaned, especially for a low-leakage application).

Components standing up a bit also allows for hack repairs, like cutting out a failed component and soldering onto the stubs.

Not suitable for everything of course (anything heatsinked through the board, RF applications, etc.), but good as a default method.

Tim

[rdl]

With the pins inline, it is more likely to experience back and forth movement which could lead to failure of the solder joints. The leads in triangle pattern act as a tripod and prevent the part from "flapping around in the breeze".

[free_electron]

old style footprint. philips used to deliver them with staggered pin like this.
once they switched to ammopack style the pins were pent on a 100 mil pitch (2.54mm) so the component shooters had it more easy. in any case the footprint with unbent pins is wrong. as it will be a horrorshow for wave soldering.

[james_s]

I always use the staggered footprint when I design with TO-92 parts. It makes it easy to stick the transistor in the board so it sits level and doesn't easily bend back and forth. I would never place it flush with the board, spacing too tight and no way to hook a test lead to it.