I would check with your board house and see if they stock single sided
material. If not (like my board house) they use a double sided material and
strip copper off one side. Might as well make a double sided board. There
are houses that stock the material or if it is high volume, your bare board
supplier should be able to order the material for your job.

I have seen many single sided boards (open up a radio, TV, monitor, UHF
tuner, etc.) used for years with good reliability. If your parts are on the
opposite side as the pad, the parts are properly clinched (or just "spread"
out to the sides) and soldered you should have no problems as long as the
board is not exposed to any significant vibration and there are no heavy
parts.

Good luck!
Tony

-----Original Message-----
From: Abd ul-Rahman Lomax [mailto:[log in to unmask]]
Sent: Wednesday, July 26, 2000 6:42 PM
To: [log in to unmask]
Subject: Re: [DC]


At 05:46 PM 7/26/00 -0500, Mark Mainland wrote:
>A customer of mine wants a design done with a single
>side board (copper on one side). This design has both
>SMT and PTH parts, The SMT parts can't fit on the bottom
>side with the PTH pads, I guess I don't feel right about
>putting PTH parts on the same side as the pads? Has anyone
>ever did that? Thanks for any input.

I wouldn't feel right about it either. That components are on one side and
pads on the other, for through hole components, is the norm because the
components help keep the pads from peeling off the board. The glue is
fairly weak, as many a tech has discovered.

However, if the through hole parts are well-clinched on the other side
(which will prevent the wires from moving), it might not be too bad. The
parts should be clinched and clipped before soldering, and the clinching
should mechanically hold the parts to the board with no play. That might
not be easy. It might also be possible to hold the through-hole parts in
place some other way; good chance someone on this list has more experience
with this.

I did design a 1-sided board once with 1N4148 diodes on the same side as
the pads; as I recall the client abandoned that design because it was not
reliable. If it is going to be done, make the pads as large as possible,
within reason.

The only reason I would know to use a single-sided nonPTH board is to
reduce cost. It might not be worth it, unless the production run is quite
high. And if the production run is going to be high, I would want to see
that PCB thoroughly tested under severe conditions of temperature variation
and vibration, or else there might be a lot of field failures. That's going
to cost money; one might have to produce a lot of units to make the minor
savings on the PCB worth the extra engineering. But it might, I suppose. If
the parts are light there may not be a lot of stress on them.

Usually time to market is a huge factor, more important than shaving off
the last dollar. Balance the second side of the board cost against the cost
of testing, market delay, field failure, and hand soldering.

I do know that the most common cause of failure of consumer electronics
products, in my experience, has been connection failure related to a pad
peeling off the board with, say, a connector held in place only by the pad
glue and solder on a one-sided PCB. Typically the wire and pad and solder
remain as one solid blob, and the track connecting to the blob cracks. It
can be intermittent and nearly invisible; drive you nuts until you
electrically check connectivity.

[log in to unmask]
Abdulrahman Lomax
P.O. Box 690
El Verano, CA 95433