Author Topic: Clear flex pcb, someone knows ?  (Read 4185 times)

0 Members and 1 Guest are viewing this topic.

Offline GreggDTopic starter

  • Regular Contributor
  • *
  • Posts: 136
  • Country: us
Clear flex pcb, someone knows ?
« on: November 19, 2017, 09:20:42 pm »
Hi all,
Normally a flex pcb is made of Kapton substrate which is orange-brown in color.
I have seen a photo of a optically clear flex pcb with smd parts soldered on it.
Any one know what this is and who offers it ? Cost is a issue. Quantity would be a few prototypes followed by 5-20 thousand pcs.
 

Offline amyk

  • Super Contributor
  • ***
  • Posts: 8275
Re: Clear flex pcb, someone knows ?
« Reply #1 on: November 19, 2017, 11:46:53 pm »
It could be polyester/mylar (PET). This is cheaper than Kapton but far more temperature-sensitive.
 

Offline daqq

  • Super Contributor
  • ***
  • Posts: 2302
  • Country: sk
    • My site
Re: Clear flex pcb, someone knows ?
« Reply #2 on: November 20, 2017, 12:43:23 pm »
These fellows: https://www.uk-cpi.com/inkjet-flex/

Mind you, there are limitations on the layer count, track width, soldering temperature...
Believe it or not, pointy haired people do exist!
+++Divide By Cucumber Error. Please Reinstall Universe And Reboot +++
 

Online mikeselectricstuff

  • Super Contributor
  • ***
  • Posts: 13748
  • Country: gb
    • Mike's Electric Stuff
Re: Clear flex pcb, someone knows ?
« Reply #3 on: November 20, 2017, 01:04:56 pm »
These fellows: https://www.uk-cpi.com/inkjet-flex/

Mind you, there are limitations on the layer count, track width, soldering temperature...
A while ago they lauched a prototype service, but then got bought and stopped it - I think they're only interested in big customers now.
You're probably looking at needing conductive epoxy rather than solder, though an ultra-low temp Indium based solder might be possible.
Youtube channel:Taking wierd stuff apart. Very apart.
Mike's Electric Stuff: High voltage, vintage electronics etc.
Day Job: Mostly LEDs
 

Offline Gribo

  • Frequent Contributor
  • **
  • Posts: 629
  • Country: ca
Re: Clear flex pcb, someone knows ?
« Reply #4 on: November 20, 2017, 01:42:18 pm »
Some processes also have pitch and trace spacing limitations. When I was prototyping few years ago, silver ink flex had a 1mm pad pitch as the minimum, while Kapton had 0.5mm pad pitch minimum.
I am available for freelance work.
 

Offline HalFET

  • Frequent Contributor
  • **
  • Posts: 512
  • Country: 00
Re: Clear flex pcb, someone knows ?
« Reply #5 on: November 20, 2017, 05:23:17 pm »
Clear polyimide does exist. But probably PET or PEN with SnBi54 solder, I recommend GTS Laminates/Flexible Materials or Induflex/Skultana as supplier.
 

Offline HalFET

  • Frequent Contributor
  • **
  • Posts: 512
  • Country: 00
Re: Clear flex pcb, someone knows ?
« Reply #6 on: November 20, 2017, 08:19:29 pm »
Clear polyimide does exist. But probably PET or PEN with SnBi54 solder, I recommend GTS Laminates/Flexible Materials or Induflex/Skultana as supplier.

Figure I owe you a clearer answer on these things. You have a couple of common flex substrates:
  • Polyimide (PI)
  • Polyester
  • PET/BoPET (or Mylar)
  • PEN

What you'll also run into sometimes in PC (polycarbonate), the main work horses of the industry are PI, Polyester, and PET/BoPET. The way conductors are applied on them depends a bit on what the application is. Screen printing silver or carbon ink on polyester is certainly the most common when you don't need 'intelligence', it's often used for membrane keyboards for example. These inks generally have a flexible binder, meaning they don't crack when you flex them. These are good for high volume low-cost applications, achievable feature size is 120 micron, though you can push it to 100 micron if you use the right inks. You can also print on the other substrates, especially on PET and PEN you'll often see this. They're more high-end than most polyester films, and you can directly screen print interface panels on them as well, this is especially common in combination with capacitive sensing for low cost user interface applications in some markets. Now the minimum feature size the moment you start putting components on becomes tricky, conductive adhesives are troublesome since they don't "self-align" like solder does. So you need to print/dispense a correct quantity at the exact spot and place the component exactly right.  As a result putting QFNs or finer pitched components on these things tends to be challenging, and the conventional flip chip technologies tend to be troublesome on these things unless you resort to ACAs (anisotropic conductive adhesives).

Then you have the other option: etched copper. This is especially the case for polyimide, PET and PEN; never saw it done on polyester myself, but supposedly it exists. The advantage of this is obviously the inherently larger conductivity and the ability to use soldering. Polyimide tends to out-perform FR-4 in material properties so you don't really need to worry too much about it. Track/gap industrially is either 100/100 micron or 120/120 micron in higher volume/lower cost applications, but with thinner copper layers you can go down to 80 micron in high yield and 65 if you're willing to suffer some losses. You can solder on this with regular SAC305 solder without worrying too much about it. Price of a single sided flex in volume production is roughly the same as a single sided FR-4 board, the pricing those shared services offer is quite despicable. (I can list a few board houses that specialise in flex/flex-rigid which offer reasonable pricing if you want.) Now this polyimide tends to be rather yellow/brownish, but transparent polyimide does exist. But there are two factors you should keep in mind when you consider polyimide based circuits: type of copper, and type of bonding of copper to polyimide. Heat bonded polyimide is usually better than adhesive bonded, it's thinner, has far higher peel strength, and is rated up to 250 °C (service temperature) and beyond quite frequently. Adhesive bonded is cheaper, but the adhesive tends to weaken at elevated temperatures, or delaminate when frequently flexed. The copper types are also an important consideration, rolled-annealed copper was squeezed between rollers, causing a sort of interlocking/weaving of the metal's crystaline zones. This tends to improve the mechanical properties, and this is significantly better for dynamic load applications, however it's more expensive than the run-of-the-mill electro-deposited copper. Also the moment you have vias you tend to plate up copper, so you end up with a weaker metal layer usually, but there are ways around these concerns.

Now PET and PEN with copper is where it gets interesting, you can't use SAC305 (250°C+ soldering temperatures) obviously, so you need to use Tin-Bismuth solders. SnBi54 being popular, with a eutectic point around 165°C, is fairly popular for this. Due to the limited adhesion between PET/PEN and copper you always end up with an adhesive layer in between, as a result the minimum track size is often somewhere around 100 µm, and copper thicknesses available are 1/4, 1/2, and 1 oz. You can use just about any component on these, and since the PET/PEN is quite chemically inert you can run it through most regular PCB processes, just don't expect to get reliable via plating on these. An important factor though is that you should ALWAYS use double sided, even if the end result is single sided - just etch away the back side. This is a limitation of the manufacturing process, where the CTE mismatch otherwise causes it to curl up. But overall you can just run this through a regular PCB line, simply avoid usage of extreme cleaning solvents. Main reason to use these is if you want transparency, otherwise it's better to use polyimide substrates, because these are far more common for lower volume applications, and hence a lot cheaper.

As a side note, I should say we sometimes run PET/PEN Cu laminates through our prototyping PCB line at work for external companies, so feel free to give me a prod if you can't get it done elsewhere. (Mind you, we're expensive as hell so it's worth looking around elsewhere.  |O )
 
The following users thanked this post: thm_w, laneboysrc

Offline Kjelt

  • Super Contributor
  • ***
  • Posts: 6460
  • Country: nl
Re: Clear flex pcb, someone knows ?
« Reply #7 on: December 04, 2017, 09:52:45 pm »
If it does not have to be flex AlO (Aluminium Oxyde) looks promising.
Does anyone know a pcb house  that offers this just like the normal pcbs ?

http://www.elite-advanced.com/transparent-pcb/
« Last Edit: December 04, 2017, 10:08:41 pm by Kjelt »
 

Offline HalFET

  • Frequent Contributor
  • **
  • Posts: 512
  • Country: 00
Re: Clear flex pcb, someone knows ?
« Reply #8 on: December 04, 2017, 10:24:47 pm »
If it does not have to be flex AlO (Aluminium Oxyde) looks promising.
Does anyone know a pcb house  that offers this just like the normal pcbs ?

http://www.elite-advanced.com/transparent-pcb/

These guys also go by the name DK Thermal. It's actually a type of LTCC (low temperature co-fired ceramic) substrate, but be warned: they're not 100% transparent. Don't expect to something like a window, it's more like one of those plastic pouches for paper in terms of transparency - quite diffuse until the object is close-by. So if something is stuck against it you'll get a clear image of the thing at the back side, but if it's quite a bit to the back you won't see anything anymore. Setup cost for these things also tends to run quite a bit higher than PCBs (by a factor of ten easily), and maximum size is nowhere near PCB maximums (24x24"). This is something for very special applications, very interesting and cool though! And I should probably add, the sales guy/representative for this company in Europe is quite friendly and very honest, so only contact him if you are genuinely planning to produce something.

But if you only care about transparency you could get away with a glass wafer, sputter or evaporate on some TiW-Cu, etch it, and solder components on that, would probably be cheaper for low volume than the LTCC approach.  ;D
 

Offline Kjelt

  • Super Contributor
  • ***
  • Posts: 6460
  • Country: nl
Re: Clear flex pcb, someone knows ?
« Reply #9 on: December 05, 2017, 01:03:56 pm »
It would be very cool as a large ledmatrix for a window so solar light will still get through and I can make a nice animations/ messages.  8)

 

Offline HalFET

  • Frequent Contributor
  • **
  • Posts: 512
  • Country: 00
Re: Clear flex pcb, someone knows ?
« Reply #10 on: December 05, 2017, 03:22:52 pm »
It would be very cool as a large ledmatrix for a window so solar light will still get through and I can make a nice animations/ messages.  8)

Those also already exist as glass laminates  ;D
http://www.qdel.nl/webwinkel/interlayer-actief/_381_w_15__NL
 

Offline Kjelt

  • Super Contributor
  • ***
  • Posts: 6460
  • Country: nl
Re: Clear flex pcb, someone knows ?
« Reply #11 on: December 05, 2017, 09:53:08 pm »
 

Offline HalFET

  • Frequent Contributor
  • **
  • Posts: 512
  • Country: 00
Re: Clear flex pcb, someone knows ?
« Reply #12 on: December 05, 2017, 10:34:22 pm »
Those also already exist as glass laminates  ;D
http://www.qdel.nl/webwinkel/interlayer-actief/_381_w_15__NL
5€ per led ?  :scared:
Half knowing what they would have had to go through to stick it inside a polymer without causing yellowing, yeah that seems like a fair price...  |O
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf