Flat flex vs Board to Board stackers...

[Smokey]

What is your preferred method for connecting two boards together?

Generally speaking, I'm a Male to Female pin headers board to board stacker kind of guy, so I'm always a little surprised to see flat flex inside products when they are simply rigidly mating two PCBs together inside some assembly. 

I get that flat flex makes it possible to flip the boards apart and probe the bottom side, but that doesn't seem like a good "design for production" kind of decision since its much easier to just plug one board into the other on the assembly line when you start cranking out volume.  Make the RMA department a cable fixture for testing, and that problem mostly goes away.

What mating situations really benefit from flat flex?  Specifically here I'm talking about two boards rigidly mounted together, not like an LCD board that is remoted, or two boards that actually need to be able to move independently like a adjustable keypad or something.

Thoughts?

[poorchava]

Flat flex cables are extremely resistant to vibrations.

Sent from my Samsung i9100 by Tapatalk

[ejeffrey]

Flat flex cables smaller are than most pin header connectors and are cheaper than most other high density options.  It is also automatically strain free.  If you mount a PCB with standoffs and use a connector with no flex, you are guaranteed to put at least some strain on the connector.  This is especially a problem if you want to use surface mount connectors, which you do.  Using only the friction from pin-headers is bad for vibration tolerance, and basically impossible with surface mount connectors.  Flat flex is good for high frequency use.

[Tinkerer]

I work with flat flex regularly. They suck compared to pin headers. One of the main failure points on the devices that use them(the products of my company) is the flat flex cables. I am talking failure percentages of up to 10%. They just launched a new product, guess where a major point of failure is in testing? Flat flex. And these are from different manufacturers too, different shaped cables and all that.

[rob77]

I work with flat flex regularly. They suck compared to pin headers. One of the main failure points on the devices that use them(the products of my company) is the flat flex cables. I am talking failure percentages of up to 10%. They just launched a new product, guess where a major point of failure is in testing? Flat flex. And these are from different manufacturers too, different shaped cables and all that.

10% failure rate at production ? you should really change your supplier/vendor for those flat flex cables

[Tinkerer]

I work with flat flex regularly. They suck compared to pin headers. One of the main failure points on the devices that use them(the products of my company) is the flat flex cables. I am talking failure percentages of up to 10%. They just launched a new product, guess where a major point of failure is in testing? Flat flex. And these are from different manufacturers too, different shaped cables and all that.

10% failure rate at production ? you should really change your supplier/vendor for those flat flex cables

On average, its more about 5%, which is still why reps from one company got dragged in about it.

[ejeffrey]

Still -- given the ubiquity of flat-flex cables in high volume products like hard drives and LCDs I think either you are getting crappy parts from crappy suppliers or you have a problem with your assembly technique or a design flaw.  I don't really think those manufacturers would tolerate a 5% defect rate...

[Gall]

If only low pin count is needed, a good option is to use spring-loaded pins. They require no mating connector, just bare PCB traces. Does not work for more than 8-10 pins.

[Tinkerer]

Still -- given the ubiquity of flat-flex cables in high volume products like hard drives and LCDs I think either you are getting crappy parts from crappy suppliers or you have a problem with your assembly technique or a design flaw.  I don't really think those manufacturers would tolerate a 5% defect rate...

The manufacturer doesnt test them currently. That is going to change real soon though. haha The document is already being produced to send them.

[Richard Head]

I use the 2mm milli-grid connectors from Molex. They are available in SMD and through hole. Available in right-angle also so you can mount the second board perpendicular to the main board, which is what I do quite often.

Dick