Author Topic: What bonds the copper to FR4? Or how could I make a custom shaped board?  (Read 12628 times)

0 Members and 1 Guest are viewing this topic.

Offline george graves

  • Super Contributor
  • ***
  • Posts: 1259
  • Country: us
I have a project in mind - it needs to fit into a housing that has a narrow conical shape.  I tried fitting really thin 1/64"(?) FR4 1/2oz copper board - but mechanically it's proving too difficult to bend the board to fit.

But I could easily (well, not that easy) make a mold out of fiberglass - or maybe not....regardless....

My question is - on FR4 - what is the method used to bond the copper layer to the board?  Is the copper sheet added while the fiberglass board is "wet" with epoxy?  Or is the base made, allowed to cure, and then a glue or bonding agent(?) used to glue the copper to the board?

I looked up some resourced of "how PCB's are made" and they all seem to start with the assumption of a copper clad board.

I know this is kinda a crazy idea - but if for nothing else just wanted to explore it.

Thanks for any insight you have.

Offline Psi

  • Super Contributor
  • ***
  • Posts: 6363
  • Country: nz
Over the years lots of research has been done on attachment methods to reduce the problem of tracks and pads lifting off the PCB from the heat of the iron.
So i'm pretty sure it's not easy thing to do.
« Last Edit: May 17, 2012, 05:14:49 pm by Psi »
Greek letter 'Psi' (not Pounds per Square Inch)

Offline JuKu

  • Frequent Contributor
  • **
  • Posts: 566
  • Country: fi
    • LitePlacer - The Low Cost DIY Pick and Place Machine
They also make flexible PCBs. - The Low Cost DIY Pick and Place Machine

Offline mikeselectricstuff

  • Super Contributor
  • ***
  • Posts: 10974
  • Country: gb
    • Mike's Electric Stuff
I've not tried it but FR4 does get quite soft during reflow, and  suspect you could impart some level of curvature to it while hot which would  remain after cooling.
Youtube channel:Taking wierd stuff apart. Very apart.
Mike's Electric Stuff: High voltage, vintage electronics etc.
Day Job: Mostly LEDs


  • Guest
If this is for one-off ? would consider point to point wiring. Perhaps having the layout laser cut from copper foil and then wrapped around conical form and then coated with epoxy?

Offline free_electron

  • Super Contributor
  • ***
  • Posts: 6827
  • Country: us
    • SiliconValleyGarage
first off : there is no such thing as FR4 MATERIAL ... Fr4 is a fire classification. it does not tell you anything about what board material is being used aprt from it being fire retardant category 4.

Laminates are made out of a fiberglass woven contruction and a resin. There are 3 kinds of fiberglass (E , NE and S) , each with their own specific dielectric properties and applications.
It is mainly the choice of fiberglass and the weave ( how the woven construction is made ) that determines the Er and Dk ( dielectric constant and dispersion ).
Materials made by Rogers (some people use the name 'Rogers' to indicate a type of pcb laminate. Rogers Corporation has at least 100 different laminates... This is the equivalent of answering 'Michelin' if someone asks you what car tires you need...)  , Isola (Getek trademark) Nelco and others employ a fiberglass weave. The difference sits in the construction of the weave. ( how many strands per bundle , individual strand thickness , how tightly bundles are woven, how individal mats are stacked and what resin is used)

Prepreg is a laminate that is not fully cured. The fiberglass mat is impregnated with a resin that is half cured so it can be manipulated. if you re-heat it it will melt again. Only if you bring the resin above a certain temperature (the Tg temperature , also called the Glass Transition temperature ) will it solidify completely . ( there is also Td : decomposition. bring it above that and it starts falling apart ... )

For example FR408HR is an FR-4 grade material made by Isola. Tg is about 230 degree C. Td is 350 or so...
Perfect for RoHs compatible processes running at 220 ...260 degrees C.

The copper is bonded by chemically oxidizing the copper foil. They use a process that grows a copper oxide. that stuff is porous enough so that , during lamination , it will bond tightly with the resin.

I have a stack of prepreg and copper foils laying around somewhere at home. i'll take a coupe of pictures and post them here. that way you will see what is going on.

Your best solution would be to use a kapton based laminate. Those are flexible and you can solder them. For example a Dupont Pyralux.  This uses a modified acrylic glue to make the copper stick to the Kapton foil. Any decent PCB fab will have Pyralux in stock as it is one of the common materials used for flex or semi-rigid pcb's.

If you need a sheet of copperfoil that has its back treated with the chemical process let me know.
If you want to do this at home you could go to a local pcb joint , ask to buy a sheet of pre-preg and copper foil and bond it in an autoclave... you will need a mold to apply pressure though.
And then exposure and etching becomes a bit of a problem.

I'd go with Pyralux. Ask the local pcb joint if you can buy a single sided or double sided flex core. get on ebay , buy a couple of sheets of dry-film and send it through a laminator , expose to uv and film and off you go...
« Last Edit: May 18, 2012, 06:25:24 am by free_electron »
Professional Electron Wrangler.
Any comments, or points of view expressed, are my own and not endorsed , induced or compensated by my employer(s).

Offline Jon Chandler

  • Frequent Contributor
  • **
  • Posts: 539
    • Throw Away PIC
In the old days, we made slip rings from PCB material to get data from a slow-speed shaft (say 100 RPM and 2' in diameter).  It was an easy matter to score the copper with a utility knife and peel off the undesired copper between the tracks.  I think the tracks were half-an-inch wide with an equal space between tracks.

Some years later, we had another opportunity to do this.  This time the only place to install the slip ring was on an 8' diameter coupling!  In other accessible locations, the shaft was tapered.  This shaft (a hydroelectric turbine) was only turning 86.5 RPM but with a 24' circumference, the slip ring velocity was scary.

In this application, we needed 4 (maybe 6?) tracks, with about 1/2" width and spacing as I recall.  In the intervening years between the first job and this job, the bonding of the copper to the FR4 substrate had improved tremendously!  It was no longer possible to peel off the copper after scoring.  Excellent for making printed circuit boards but not so good for making slip rings!

After trying several methods, I believe we finally resorted to mechanically removing the undesired copper using an extremely shallow cut on a table saw.

The slip ring was used to transmit 16 temperature signals from the generator rotor to identify hot spots in the windings.  In the days before simple telemetry, we used a 16 channel multiplexer to send one signal at a time across the slip ring.  An analog strip chart recorder displayed each temperature for a few seconds, with an interval of no signal to mark the start of the 16 readings.  It's amazing what you can accomplish with some simple instrumentation  \when you have no alternative.

Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo