9
\$\begingroup\$

A while ago I read here that it's a good idea to have ceramic capacitors on the input and output pins of 78xx regulators and to put \$10 \mbox{ } \mu F\$ on the input and \$1 \mbox{ } \mu F\$ on the output. I can easily obtain such capacitors only in SMD format and that gave me the idea of soldering them directly to the pins of the regulator. This would give me more or less self-contained units which could be easily used on the dreaded breadboards or anywhere else I may need to use a 78xx or 79xx regulator (Assuming I manage to solder them without losing the capacitors, but that's another topic).

So my question is: Is there any reason not to solder the capacitors directly to the pins of the regulators? I'll most likely be using 1206 components.

\$\endgroup\$
  • 2
    \$\begingroup\$ I believe it is safe. But you can have problems with excessive heat. Capacitors don't like heat. \$\endgroup\$ – Al Kepp Jan 29 '12 at 18:15
  • 1
    \$\begingroup\$ @AlKepp: Ceramic capacitors don't much care about heat. \$\endgroup\$ – Olin Lathrop Jan 30 '12 at 12:53
9
\$\begingroup\$

For one-off bread-boarding this actually sounds like a good idea. Since the center lead is ground, you can put the capacitors between that lead and each outer lead. If you do this high enough up on the TO-220 part, then you still have the lower lengths of the leads free to plug into the breadboard or connect otherwise.

One thing I would do though is still add a capacitor to the output bus on the breadboard separately. It can't hurt to have a little distributed capacitance there. This would help counter some of the inherent resistance of the connection, which can be higher than it should be due to dirt or worn contacts.

\$\endgroup\$
4
\$\begingroup\$

It shouldn't pose a problem for testing, especially not for ceramic capacitors. But it's not so good an idea to do it for a long term with tantalum or electrolitic elements. As Al Kepp allready said - this components prefer room temperature - and those 78xx regulators are linears = they generate a lot of heat.

\$\endgroup\$
3
\$\begingroup\$

I foresee two problems.

First is hand soldering SMT ceramic capacitors. The thermal shock can crack the package. Applying the iron tip to pad (or lead in your situation) is the recommended procedure (see the capacitor datasheet, or soldering application note).

Second is mechanical stress. SMT ceramic capacitors are fragile. Inserting parts into a bread board and bumping the parts while working with the circuit causes significant movement of the leads than can break any capacitors soldered across them.

I have two recommendations. First, find a supplier for leaded components, possibly plan ahead so you can take advantage of worldwide distributors. Second, include capacitor selection as part of the design. It's not a 1 size fits all. Some designs require several capacitors at each IC, while a few require no bypassing (or that provided by a bread board with back plane).

\$\endgroup\$
1
\$\begingroup\$

This is a major issue for manufacturability. Automated pick-and-place machines cannot put two parts on top of one another (manufacturers all have different minimum clearance specs for their equipment).

Basically what you are proposing is point-to-point soldering. For prototypes, one-off designs, and rev 1.0 corrections this is often used, but it is important to understand that it is really bad design.

\$\endgroup\$
  • 1
    \$\begingroup\$ Good hack; Bad Design. \$\endgroup\$ – ObscureRobot Jan 30 '12 at 5:47

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.