More noise on a perfboard than a breadboard?

Thread Starter

Nate.Kilmer

Joined Apr 4, 2022
12
I'm prototyping a circuit to double the output frequency of a fan tachometer signal, and I've implemented the circuit successfully on a breadboard already. The problem is, when I carefully moved the components over to a perfboard a ton of noise was introduced into the signals. From my experience and from what I've read online, the opposite should be true - there should be more noise on the breadboard implementation than on the perfboard. Has anyone else had this kind of issue? It's a low-frequency circuit (100-300 Hz), but with square waves there may be high-frequency harmonics present.
 

Thread Starter

Nate.Kilmer

Joined Apr 4, 2022
12
Did you layout the circuit the same way on the perfboard as you did on the breadboard?
I compactified things, so instead of long jumper wires connecting terminals, there are shorter wire connections. But the components were transferred directly from the breadboard to the perfboard with the same orientation.
 

Audioguru again

Joined Oct 21, 2019
6,674
Is the noise 50Hz or 60Hz from electricity in walls and floors all around you and is picked up by ordinary wires between the tachometer and the perfboard because you did not use a shielded audio cable?
Or does the fan use very noisy brushes?
 

LowQCab

Joined Nov 6, 2012
4,026
There may be some aspect of the Circuit design that causes it to
be right on the edge of stability.

This may become obvious to someone here
if You would post your Schematic for us to see.
.
.
.
 

Thread Starter

Nate.Kilmer

Joined Apr 4, 2022
12
Make sure you have adequate de-coupling capacitors etc, also ensure that strip traces are as short or shortened as much as possible.
I think MaxHeadRoom might have nailed it. It worked on the breadboard without bypass capacitors, so I moved to the perfboard without them, but the signals got very noisy. Adding back the bypass capacitors helped a lot, but I still need to fine tune the circuit to smooth out a bit more of that noise.

LowQCab asked for the circuit, so here's the schematic. I've got a 24 VDC supply which I tap off with a 5 V zener diode to supply power to an opamp and an XOR configured as described by this article ( https://www.analog.com/en/design-notes/simple-circuit-doubles-input-frequency.html ). The 24 VDC also supplies a 24 VDC brushless fan with an ICL thermistor in series to manage the inrush current. The tachometer output of the fan is an open collector, so I provided a pullup resistor to the 24 VDC line to produce a square wave, but I need the square wave to be 5 V so I made that subcircuit a divider. That tachometer signal feeds back into the frequency doubler circuit as one of the inputs of the XOR gate, and a delayed signal goes into the opamp which is set up as a comparator. When constructed on a breadboard, this gives a nice squarewave output with twice the frequency of the the input tach signal, but transferring the components to a perfboard results in really noisy signals. The bypass capacitors helped, but I'd like to reduce the noise a bit more before implementing this circuit on a final PCB. Any advice to make this circuit more robust is very much appreciated. Thanks again to the folks who have already made suggestions.
tachCircuit.png
 
Top