Notifications
Clear all
Topic starter
2021-03-02 1:20 am
I found the linked youtube interesting. It suggests using a counter chip to count a wheels revolutions in place of getting one's board such as an arduino or rpi to do the counting. It shows that getting the microprocessor to do the counting can get inaccurate at the higher revolutions. I though one of your videos that actually explains how to hook a counter chip up would be of interest for all the roboteers.
2021-03-02 7:22 am
I had a similar challenge with an encoder.
I started with a Nano running at 16mhz as the encoder monitor.
I programmed a PIC to output encoder signals starting with 100 slow-going signals in the forward and then in the backwards direction. So the arduino should count up to 100 and then back to zero. After each cycle it would do it again, except a little faster. As the pulses go faster, the pulse-width gets smaller. I wanted to see how fast I could simulate the encoder signals before the Arduino started to lose track. I found that with a pulse width of less than 4 miliseconds it would start to lose track.
I then monitored the encoder in the o-scope and found there were speeds in which it could go beyond that. so the Arduino was not going to provide assurance of 100% tracking.
I got an LS7084 chip on Ebay
This ties directly to the encoder and takes alot of the work the Arduino would have to do.
All the Adruino has to do at that point is count pulses and monitor a direction pin.
But if I recall correctly - the clk output of this chip didn't work
So I added a quad nand gate to monitor the encoder pulses, and used the LS7084 as the direction.
Here is a snapshot of the LS7084
