Xinmotek Dual USB Controller Lag

[Alturis]

I recently moved my pi 3 retropie setup from just an xbox 360 wireless on a 22 inch display into an arcade cabinet. Same pi, same setup, same display, just located in a cabinet instead of sitting on a desk. The only thing that is different is that the controller being used is a Xinmotek dual joystick USB as compared to a vanilla 360 wireless.

It was noticed immediately how there appeared to be some input lag especially in registering a change of state to the joystick axes.

I read some previous posts and information here about input lag in general and how it can be improved or is just part of the emulator code and there is nothing that can be done. But in this particular case the only difference is the controller. Everything else is identical in setup.

I was thinking of attempting to update the driver setup as indicated in the docs here: https://retropie.org.uk/docs/Xin-Mo-Controller but that doesnt sound like it would be related as that involves a solve for a lack of left/right input at all.

But I wanted to query if anyone else has had experience with this controller hardware and noticed something similar, done anything about it, and if so what your results were.

[Alturis]

In playing and testing this some more I think what may be happening is a perceived input lag in the stick response simply because it has such a large dead zone before it clicks the switch for that direction.

I am wary of messing with this for fear of making it worse or inconsistent in one direction versus another. But the thought occurs to me that this might be helped if there were a way to bend the resistance panel bits in the stick internals so that it presses the pushbutton sooner as you begin moving the stick in a direction.

Has anyone tried that with any success? Or should I just live with it or consider purchasing better or different joysticks?

[caver01]

@alturis I have HAPP Supers in my cabinet which look similar. The actuator attached to the bottom of the stick can be installed in two directions which amounts to a narrow or wide setting. Obviously, the wider configuration triggers the microswitch levers faster but also enables more reach into the corners. The smaller setting makes it harder to reach the corners and has more dead space. It is dubbed as a 8-way vs. 4-way solution but this is not accurate. My stick does not include a restrictor plate.

However, it does give me some ideas for how you could improve your situation without irreversibly modifying your joystick. You could simply enlarge the actuator by wrapping a bit of tubing around it, or for testing, maybe a number of turns with duct tape just to see how it changes. You could also use an appropriate washer or a 3d printed replacement. I have even seen folks use a piece of wood.

[Alturis]

I am thinking about just going with a completely different stick now.

As I would be replacing my existing Zippyy sticks it would be a requirement that they would fit into the same recesses and ideally have the same screw hole placements. But I could also drill new holes and put a dummy screw in the empty ones for presentation.

Has anyone used these? Thoughts on responsiveness and size of deadzone?

GGG SmoothStick™ Arcade Joystick
http://groovygamegear.com/webstore/index.php?main_page=product_info&cPath=65&products_id=332

[Alturis]

Just to update, I opted to instead go with Sanwa sticks as they are the perfect size with the same screw holes and apparently the top of the line based on the recommendation of those in the know.

https://www.amazon.com/gp/product/B01CRQMWEQ