Hello,
I’m using a binary counter IC to count pulses from an optical encoder before feeding the data into a microcontroller. At low speed it works perfectly, but at higher RPM the count starts missing steps.

The encoder output line is around 20 cm long on the PCB.

Questions:

1. Is this more likely propagation delay, signal ringing, or slow edge transitions?
2. Would a Schmitt trigger buffer help clean the signal?
3. Should encoder traces be routed with ground shielding or series resistors near the source?

 

Propagation delay, ringing, or slow edges?

Most likely signal integrity first, not counter propagation delay.

At higher RPM:

Faster pulse rate makes ringing/reflections more visible
Long 20 cm trace can cause overshoot/undershoot
Slow or noisy edges can violate logic thresholds
Counter propagation delay matters only if pulse frequency approaches the IC timing limits

So check waveform quality before blaming the counter IC. Its a common issue when switching from breadboard to pcb .... REad this guide on Signal Integrity: https://www.aivon.com/blog/pcb-desi...ignal-integrity-and-avoiding-common-pitfalls/

Would a Schmitt trigger buffer help?

Yes .... often a very good fix .... Schmitt trigger:

Cleans noisy or slow transitions
Adds hysteresis, preventing false multiple counts
Restores sharp logic edges before the counter input

It’s commonly used between sensors/encoders and logic counters.

Ground shielding or series resistors?

Use both if needed, but start with series resistors.

Place small series resistor (22–100 Ω) near encoder source/driver to reduce ringing
Route trace over a solid ground reference
Keep away from switching/high-current traces, read this guide: https://www.pcbway.com/blog/PCB_Design_Layout/Considerations_for_High_Power_PCB_Design_25a54c5a.html
Optional grounded guard trace if the area is noisy

For a 20 cm PCB trace, controlled return path with damping resistor is usually more valuable than “shielding” alone.

 

Hello,
I’m using a binary counter IC to count pulses from an optical encoder before feeding the data into a microcontroller. At low speed it works perfectly, but at higher RPM the count starts missing steps.

The encoder output line is around 20 cm long on the PCB.

Questions:

1. Is this more likely propagation delay, signal ringing, or slow edge transitions?
2. Would a Schmitt trigger buffer help clean the signal?
3. Should encoder traces be routed with ground shielding or series resistors near the source?

Are you using both the A and B signals AND their inverses A-bar and B-bar? You get much better signal reliability if you do.

 

Hello,

What are the specifications of the encoder?
It might be the signal drops at high speed.

Bertus