Membrane switches and overlays. DIY or Proto options?

[fastline]

Guys, I hate to start another thread but this steps away from the Video CRT/LCD area and is a real issue in how we move with that. As in my other thread, I am trying to convert some CRT industrial controls to LCD. Seems like we can get there BUT in all cases, there are membrane switches or tactile switches below the monitor as menu buttons. Those sort of need to line up to make operation easy.

If the screen stays the same size, this is not an issue, but in one machine, we need a bigger display. This would then require the switches to be spaced out more to line up on a bigger monitor. the entire control panel needs a refresh. I always seem to stumble into issues where we need these membrane switches.

Is there a way to make these overlays in a printer? Other methods? easier to 3D print buttons? I could send out for proto PCBs.

On one case, the entire control interface was made with severe overkill HUGE buttons that are cumbersome. I want to convert that entire panel to a membrane overlay and rewire them. All the switches just go to an ADC card wire by wire! lol. It looks like it is from the 80s!

Examples vvvv

https://www.google.com/url?sa=i&sou...aw2-18zo7043Yc1e1FNN8V3h&ust=1545324120578806

https://www.google.com/url?sa=i&sou...aw0cK26aMf_Kc-I3RF-jlSYs&ust=1545324810709027




Where I need to go vvv

https://www.google.com/url?sa=i&sou...aw26FM1-gYhj1mIlX151J86_&ust=1545324882780951

 

[rjenkinsgb]

Making a reliable DIY membrane that also looks good is impossible, as far as I am aware.

You are probably better off with a row of individual low-profile buttons, eg. something like these:
https://uk.rs-online.com/web/p/push-button-switches/6903255/

Note that for anything involving mechanical movements or starting a machine, physical switches are pretty much essential from a safety point of view; they have "feel" and the operator can generally tell if they are about to press the correct thing while looking at the machine, while it's extremely easy to press the wrong membrane button in error.

We use small buttons / keypads / membranes / touchscreens for purely software or configuration stuff, but never for anything involving motion.
They are strictly 22mm buttons & selectors.

(A lot of the stuff we work with is unguarded due to the size, so we are a bit paranoid on the safety and emergency stop side of things - but any consideration that reduces the risk of an accident is worthwhile, regardless of scale).

 

[alec_t]

Other methods?

How about print overlay on good quality paper, laminate, use removable frame to hold laminated print in place? The overlay could then easily be replaced if damaged, worn or updated.

 

[fastline]

Buttons get pushed thousands of times and there is usually coolant and metal chips around so it has to be robust.

RJ, when I refer to membrane buttons, these all certainly have a feel and in actuality, there is some sort of tactile button under the overlay. I know what you are saying about feel and we most certainly will ALWAYS have real switches for Estop, and motion controls. I have seen where hobbiests use a touch panel for all of these and they are proud of themselves but don't realize a real switch is used for a reason and does take more effort to make.

If I have to have an overlay made and printed, anyone know where to go? Seems like PCBs are made everywhere but I don't have too many resources for overlays.

 

[Tony Stewart]

Why not use a wired keyboard that has better tactile and reliable operation than any membrane switches. Judging by the harsh environment, that's what I'd do.

Or better, replace the Fanuc controller with a PC with Windows GRBL Panel2 software and Arduino with CNC Shield and drive Fanuc motor drivers with that. It supports Macro keyboard, mouse buttons more , manual, macro and automation GCode files. For water cooling, spindle and 3 axis controls or more if necessary.

then you can design from Autocad, convert to Gcode automation files and execute on GRBL Panel2 any LCD and keyboard you want with a rubber membrane for water protection.

Thats something like what a pal,of mine did on 2 Bridgeports except the replaced all the electronics.

 

[rjenkinsgb]

There are lots of places that manufacture membrane panels....

https://www.google.com/search?q=mem...rome..69i57.3867j0j1&sourceid=chrome&ie=UTF-8

Tony - there are good reasons a commercial CNC system costs thousands; a DIY setup may be OK for hobby use but it's a completely different ball game in industry.

 

[fastline]

Just swapping for a keyboard sounds easy but the button functions in the control are always unique and never laid out standard in any way. We have a key for "iso assist", "mode select", etc.

But..... I am actually leaving to go look at a machine RIGHT NOW that we will have to decide if we buy and fit a new control on it so keep talking! This is NOT a hobby machine and has to work, not just a theoretical "it should work". I really think I might want to drive Fanuc servos because they are reliable and everywhere. However, I don't know about the amps! Fanuc is VERY proprietary and amps tend to be designed for an exact servo or series. Very high accuracy and speed will be needed and likely 4000+ line encoder feedback.

 

[rjenkinsgb]

If the servo drives are on analog control signals, you can replace the CNC with just about any make. Most still have an analog servo option.
If they are digital, you are stuck with Fanuc unless you replace the whole thing.

It also depends to some extent what controls you have on other machines; if eg. you have mostly Siemens or Heidenhain, it may be an excuse to standardise a bit more so the operators have less types to be concerned with.

 

[fastline]

What control OEMs are doing now is an "emulator". Mitsubishi I believe you can throw it into a "fanuc mode" and you almost don't know the difference. The buttons will be different though. Personally, I have run about everything and none are perfect. I did like Haas but I cannot stand their machines. They are decent commercial grade, but they fall short on stupid things and I am over them. Dig their control though.

 

[Nigel Goodwin]

If I have to have an overlay made and printed, anyone know where to go? Seems like PCBs are made everywhere but I don't have too many resources for overlays.

As others have suggested, you can easily have membrane keyboards made, we use one on a product I designed and developed. There's only three keys on it, and they use 'metal dome' switches within the layers of the self adhesive label.

 

[JonSea]

You can use tactile switches behind a debossed label. When I did this I used a cork or felt pad on top of the switch and a hole of about 15mm through the panel.

This allowed the switches to be mounted on the PCB and the use of an inexpensive label. In my case, the holes through the panel were keyhole-shaped to allow the adjacent LED to shine through a transparent window in the label.

 

[fastline]

JonSea, I think you and Nigel both are onto the same concept and the direction I was probably headed as that seems the most common in my industry. Actual tactile switches that hide behind and overlay. Failure of the switch seems to be pretty quick once the overlay fails but you are supposed to repair it at that point!

On your overlay, how special was this for printing? I remember one company asking about embossing and such that adds quite a bit to the cost.

I guess I envision a new PCB with beefy traces and carefully selected SMD switches, then some sort of plastic grid to shim everything up to about level with the switches, and the overlay to go over it all.

I would be pretty curious to see how Haas is doing it because their buttons are pretty damn slick, literally! They feel like silicone buttons and do not click or detent at all. I think their keypads are actually injection molded. I wish they spent more time on their machine quality than their buttons! lol

 

[JonSea]

I don't recall what the cost for printing the label was - that was 20 years ago. Be clear that you want the switch face to project away from panel. As I recall we had to scrap a batch over confusion between deboss and emboss.

We had them print a "filter layer" over the LEDs and 7-segment displays to enhance constrast.

I did like the nice solid click of these keys when pressed. There was no doubt that the switch had been pressed.

 

[rjenkinsgb]

We once did a system with conventional PCB mount switches on a board set in the back of a machined faceplate, with a separate embossed and windowed panel on the other side.

The membrane could be replaced without touching the electronics if it ever got damaged.
That was specifically made with an embossed surround on the buttons to reduce the chance of cracking over time, with the material flexing to operate the buttons. It was still a lot cheaper than one with internal switches, at that time anyway (20 years ago).

It worked very well, only about two membranes replaced for all the machines built.