Should I go SMT or not?

[Super Nade]

Guys,

I'm building a servo-loop to control my laser. Should I go SMT or not? How many PCB layers should I use? I have never designed an SMT PCB before, so I hope some of you folks can help me out.

It is a fairly simple ckt which consists of an input buffer, Integrator and a LPF (Butterworth 2 pole) (all have tunable time constants) and an output buffer. So, I' looking at about 4 IC's max, a few capacitors, resistors and 4 multi-turn pots. I'm worried that the advantages of minaturization will be lost because the POT leads would have to be soldered on to the board.

Input signal is low freq < 1kHz.

The o/p will couple to a HV amplifer.

[arneson]

Re: Should I go SMT or not?

Your board layers will lessen the trace difficulty.
I'd be surprised if you needed more than top and bottom but if your going inner then that takes care of ground.
The enclosure and mount options also come to mind.
Don't go by me, it's been years since I laid out and etched small control boards, before surface mount came along.

[Super Nade]

Re: Should I go SMT or not?

The thing is (as always), we are looking to save the few remaining pennies on the NSF grant my Prof has.So cost is a factor.As it is, I'm getting my components by way of samples. The only thing I'm worried about is the mess it would make for me to replace the components, especially if the traces are too close.

[arneson]

Re: Should I go SMT or not?

How miniature do you need to go?
Have you figured a couple of rough layouts?
As we used to say your component placement must be elegant for the traces to work.
Depending on complexity even some of the most beautiful designs have jumper wire
and quite often rev2, at least.