problems with EPM570

[Altera_Forum]

Hello, 

 I'm an occasional user of Quartus, so please be patient with me.  However, with previous projects based on Max7000 family I have never had this kind of problem.    I'm testing a new project based on a EPM570, containing 3 quadrature inputs and some simple glue logic, and I'm experiencing several problems.   Initially, I was using version 12.1 of Quartus, and the main problem was the instability of the counters 74193. With the external encoder in rotation, I detected abnormal behavior, but only during the positive counting.  After exploding the 74193, and adding some output as test point, I have noted that its internal flip flops were commuting even in the absence of the respective clock.  I also tried to rule out any other signals that could affect the count, but without finding a solution.   After a week of endless changes and tests, I tried to compile the project with another computer where is installed quartus 13 and .. surprise .. this problem disappeared. But I had another problem. As soon as I transfered, everything seemed to work. But if I turned off and on, when I power up instability often reappears.   Curiously, the first quadrature channel counted always well, while the other two behaved differently (with the same encoder connected to the three inputs).    Now, after some minor modifications to manage the marker readout, the original instability is back.   My last attempt was to connect the up/down signals of the first channel even to the counters of the other two channels, and sometimes they don't count the same number of pulses.   I enclose the picture of a single channel. Monitoring test points TP16 and TP17 with a logic analyzer, I'm sure the counter inputs are always correct and without glitches.   It seems the problem is caused by the 74193 counters, but I do not know how to correctly interact with Quartus to try to solve the problem.   I'm pretty desperate and would appreciate any suggestion that makes me able to fix these problems.  Thank you.