It does jump. I'm not 100% satisfied with the replacement.
The problem is, after you turn it quickly, the firmware on the board might get confused. Some of the subsequent turns might be missed.
But if you only ever turn it slowly, it works just fine.
I should've put a scope on the signal traces to debug the issue, but alas. It might be caused by my rework of the PCB pads I tore off. Another possible cause is different signal filtering requirements.
Which one did you use? I am sorry it is not clear to me....the 24/12 or the 30/15? I get the impression it is the 30/15. Did you try the 24/12 and were the results better?
I think I have the SMD version, but have not removed the board yet. I am ordering both the TH and SM versions to cover my bases I would like to get an SM one which works or I have no problem converting a TH to SM.
What I have in my "cart" at Newark...
EC12E2424407 in case board is through hole 30-200 gf-cm torque
EC11J1524413 in case SMD (30:15) 70-170 gf-cm torque
PEC11R-4215F-S0012 in case SMD (24:12) 30-90 gf-cm torque
Folks,
I have completed an encoder upgrade on a 2:1 SMD board. I am super happy with the results.
I used a PEC11R-4215F-S0012 24:12 converted to SMD by gull winging and clipping the leads. To install I cut the leads on the original to remove it. Then it was easy to remove the leftover and not damage the solder pads/lands on the board. This worked out well and I never felt close to lifing the pads. A normal touch with the iron was all that was needed to solder the replacement on the board.
The replacement encoder's shaft sits about 2mm more proud than the original. Ulitmately this is noticable, but not objectionable.
The 24:12 works great. I did not notice any skipping and there is one selection per click. To test it I used a digit selection menu. A fast spin (not roulette wheel fast, but what you might normally do fast) got me from "2" to "7" with what my mind registered as five clicks. I tried that a few times and it all seemed to jive. Selection of an item was flawless. The force to turn the knob is really just about right. It is much easier than the 30:15 in my list above.
Prior to biting the bullet and switching out the encoder I tried adding some friction between the knob and the scope body as suggested earlier in this thread. That did help, but I still needed to be careful and had a good number of missed selections.
While I was in there I added a header to the JTAG port. I have been having a dickens of a time using my SiSpeed debugger (does not dump memory reliablly), so I have an Olimex coming. This will make it that much easier.
73 es God Bless de KK1L, Ron <><