EEVblog Electronics Community Forum
Electronics => Projects, Designs, and Technical Stuff => Topic started by: DTJ on April 14, 2016, 12:26:16 pm
-
[I've been having posting problems so hopefully this doesn't double post]
I've been looking at a hand held instrument front panel that needs a simple 5 button keypad and LCD window.
Membrane keypads are expensive, the set up costs are the killer for low volume cottage industry production . $1000 to $2000 for your first 50 or so membrane keypads. After that they are about $15 each.
So I'm thinking about doing a membrane fascia arrangement with tact switches on a pcb behind the front panel. When pressed the front panel label deflects into the hole machined in the front panel activating a tact switch. The label costs come way down to under $5 per piece with no set up costs. There is the added cost of the switches @ ~ $0.20 each
My concerns are:
1) How durable is this set up? Does the front panel deform, split or crack around the key actuator after a few hundred presses? Whats best polyester or polycarbonate? How thick?
2) What sort of minimum size should the actuator end be? Is round better than square?
3) Is 10mm a good diameter for the hole in the front panel for the button to sit in?
4) How critical is maintaining the set back of the PCB from the front panel? Fiddling around for one or two is OK but not for a few dozen.
I'm hoping to get some guidance and ideas from folk who have actually done this. The set up is in the picture below.
Thanks.
-
The use switches in this format on the watercoolers at work.
They seem to last about a month before the membrane breaks through.
That's probably down to the animals I work with but it is a concern.
I'd try to reinforce the membrane contact point or widen the head of the switch to reduce the issue.
Maybe physically limit the travel distance of the switch too.
-
You'll have trouble if the spacing varies by more than a small fraction of the actuator travel. You don't want to have to shim between the PCB and panel to adjust the spacing, and an injection moulded front panel with standoffs and mounting bosses would be $expensive$. If possible, choose switches with an actuator height above the body fractionally less than the panel thickness and locate the panel directly in contact with the switch body.
The fatigue life can be extended by milling a concave depression behind the key face, with gently radiussed edges, with only a small hole in the middle for the actuator. The depth should be such that the actuator can be fully depressed with no over-travel. Because the unsupported area at full travel is minimised, the risk of puncturing the membrane is vastly reduced.
Neither polyester nor polycarbonate are known for their elasticity. You can get away with limited deflection and large key area in relation to the film thickness, but you should set up an automated lifetime tester e.g. a silicone 'finger' on the end of a solenoid plunger operating 10 times a second. Detecting membrane failure automatically may be problematic - maybe simply photograph it every 1000 cycles.
Ideally you'd get the membrane printed on an elastomer with a clear elastomer layer on top to protect the markings.
-
Have you considered capacitive touch sensors? They will work through your label and will be the most physically durable solution. The sensor itself is nothing more than a copper area on the PCB with a trace to the input of your sensor device. Many microcontrollers do this directly, or you can add a dedicated IC that provides either logic level output for each button or some type of serial interface (e.g. I2C).
-
They seem to last about a month before the membrane breaks through.
Yep, that's my worry. I've seen them cracked and split on air conditioner control panels.
You'll have trouble if the spacing varies by more than a small fraction of the actuator travel. You don't want to have to shim between the PCB and panel to adjust the spacing, and an injection moulded front panel with standoffs and mounting bosses would be $expensive$. If possible, choose switches with an actuator height above the body fractionally less than the panel thickness and locate the panel directly in contact with the switch body.
Agreed,I want to avoid fiddling around with assembly. It needs to be plug and go.
Have you considered capacitive touch sensors?
The micro I'm using has capacitive touch inputs. I really like the tactile feedback buttons can provide, either via a mechanical switch or by membrane domes.
-
You can make feedback for capacitive touch using a piezo speaker to make a "tick" sound. Or beeps, but beeps are annoying.
This method of a flexible membrane over tactile buttons often leaves the outline of the switch cap in the membrane.
And you need very low travel switches.
Don't go this way for non-prototypes, it's poor for devices that are operated with high frequency. For set&forget devices, it looks nice and it won't be operated very often.
For often operated equipment, get yourself a stock membrane (or have one made) with internal dome switches.
-
You can use the approach you've suggested, it will work. I have used this approach before, and yes, the chamfered edge is a must, or the label will need to be replaced due to tearing, with monotonous regularity.
Most tact switches activate in typically 0.25mm, and the tolerance on the actuator length is typically +/-0.1mm possibly up to +/-0.2mm depending on source of switches. So the diameter of the opening needs to be widened to allow for the deflection needed to both take up the tolerance and activation.
If you want a membrane-like switch panel - use a proper one. They are not that expensive.
You haven't mentioned the expected activations that this needs to meet, so beyond the above and a "talk to the membrane supplier" comment I don't have much more to add.
Though there was a magnetic-type actuator for use instead of dome switches in membranes, but I'd need to hunt this info down... it was more expensive than standard domes but considerably more reliable.