Author Topic: surface finish thickness  (Read 7409 times)

0 Members and 1 Guest are viewing this topic.

Offline LaserTazerPhaserTopic starter

  • Regular Contributor
  • *
  • Posts: 203
  • Country: us
surface finish thickness
« on: January 13, 2019, 02:27:32 am »
How thick is this immersion gold thickness for both options?

site is allpcb.com

It is unlikely to be 1 or 2 microns thick, its likely referring to some class or standards.
 

Offline ataradov

  • Super Contributor
  • ***
  • Posts: 11905
  • Country: us
    • Personal site
Re: surface finish thickness
« Reply #1 on: January 13, 2019, 03:28:13 am »
I think they are literally referring to micro-inches. It seems like a reasonable assumption and IPC standards also use "uin" as a unit of measure for this kind of stuff.
Alex
 

Offline wraper

  • Supporter
  • ****
  • Posts: 17952
  • Country: lv
Re: surface finish thickness
« Reply #2 on: January 13, 2019, 03:44:34 am »
I think they are literally referring to micro-inches. It seems like a reasonable assumption and IPC standards also use "uin" as a unit of measure for this kind of stuff.
Unlikely, 1 microinch would mean that gold is so thin there is almost none of it. You shouldn't see even a tad of golden color.
Quote
It seems like a reasonable assumption and IPC standards also use "uin" as a unit of measure for this kind of stuff.
Copper thickness is often measured in um.
« Last Edit: January 13, 2019, 03:50:13 am by wraper »
 

Offline wraper

  • Supporter
  • ****
  • Posts: 17952
  • Country: lv
Re: surface finish thickness
« Reply #3 on: January 13, 2019, 03:50:27 am »
Hmm http://www.unitechindustriesinc.com/services/enig-4552.htm
Seems 1um would be too thick for ENIG. But 1 uin too thin.
 

Offline ataradov

  • Super Contributor
  • ***
  • Posts: 11905
  • Country: us
    • Personal site
Re: surface finish thickness
« Reply #4 on: January 13, 2019, 04:04:37 am »
Quote
ENIG General Applications The minimum immersion gold thickness shall be 0.05 µm [2.0 µin] at -4 sigma (standard deviation)

So the minimum recommended by the standard is 2.0 µin. It seems reasonable that prototyping service will do bare minimum or less. I'm sure AllPcb will do whatever is requested for real orders.
Alex
 

Offline LaserTazerPhaserTopic starter

  • Regular Contributor
  • *
  • Posts: 203
  • Country: us
Re: surface finish thickness
« Reply #5 on: January 13, 2019, 04:29:37 am »
Is 2u" too thin to see?
 

Offline LaserTazerPhaserTopic starter

  • Regular Contributor
  • *
  • Posts: 203
  • Country: us
Re: surface finish thickness
« Reply #6 on: January 13, 2019, 05:18:22 am »
Are there tendencies and even reports of Chinese PCB houses "conserving" on the gold?

I mean big PCB houses like ALLPCB, JLCPCB and elecrow.
 

Offline LaserTazerPhaserTopic starter

  • Regular Contributor
  • *
  • Posts: 203
  • Country: us
Re: surface finish thickness
« Reply #7 on: January 13, 2019, 05:38:32 am »
Are there tendencies and even reports of Chinese PCB houses "conserving" on the gold?

I mean big PCB houses like ALLPCB, JLCPCB and elecrow.

I can see they are consistent from my orders, but I have no way to measure them.
1u" of gold for 10cm*10cm costs $0.2, and most Chinese fabs require additional charge for gold coverage >=30%, so the actual cost of gold for a 10cm*10cm board with maximum free coverage would be $0.06.

JLCPCB quotes $1496 for 2000pcs of 10cm*10cm 2L 1.6mm FR4 w/ ENIG, and $1104 w/o ENIG. That's $0.20 per board difference.
I wouldn't be skimping if my additional cost is $0.06 and I get to make $0.20.

It seems ALLPCB is the most comprehensive and option filled retailer. Unlike other retailers they even permit you to have no silkscreen, determine gold plating thickness and even have no surface finish which is really great for certain boards.
« Last Edit: January 13, 2019, 05:47:07 am by LaserTazerPhaser »
 

Offline LaserTazerPhaserTopic starter

  • Regular Contributor
  • *
  • Posts: 203
  • Country: us
Re: surface finish thickness
« Reply #8 on: January 13, 2019, 08:04:53 am »
I just recently placed an order for 5 single sided ~415mmx423mm boards with ALLPCB they are not circuit boards rather specialty antennas which have much of the copper exposed.

You might want to think twice. The nickel under layer is potentially ferromagnetic. It is supposed to be Ni-P alloy, which will greatly decrease its permeability, but it will still be a bit ferromagnetic.
Unless you have very high frequency, I would recommend covering it up with soldermask or use immersion tin. The best option would be immersion silver, but that would add quite some cost, and is prone to tarnish.

It cannot have any soldermask this is not a typicall antenna. But nice of you to detail magnetic issues the nickel layer can cause. I would have made it 2.4mm thick if that didn't demand 2 or more layers.
 

Offline Zero999

  • Super Contributor
  • ***
  • Posts: 20357
  • Country: gb
  • 0999
Re: surface finish thickness
« Reply #9 on: January 13, 2019, 07:04:55 pm »
The best option would be immersion silver, but that would add quite some cost, and is prone to tarnish.
Would tarnishing be a problem? I would have thought that as long as there's still enough silver, under the layer of oxide it would still be good or does the oxide have high dielectric losses?
 

Offline KL27x

  • Super Contributor
  • ***
  • Posts: 4108
  • Country: us
Re: surface finish thickness
« Reply #10 on: January 15, 2019, 12:57:27 am »
This thread brings back a bad memory on a collaboration. I did the circuit design and firmware, but due to time constraints, physical quirks, and the sheer quantities involved, my client wanted a more experienced PCB guy. The housing was done by another company.

The board was in two parts, with some gold fingers making a contact pressure between the boards. I identified a problem with the prototype because the soldermask on the board encroached on the board with the gold plated contacts. The obvious solution is to enlarge the opening in the soldermask on the bottom board or to reduce the size of the pcb on the top board, so that everything fits without the soldermask getting in between.

I don't know what the big deal was, but he insisted to add carbon to the pads to make them thicker. I knew this would cause other issues and I explained it best I can  that this was unacceptable. The housing was already done, and it didn't allow for this. I wanted to strangle him through the phone. He finally conceded to not add carbon. He claimed he would thicken the pads with gold plating. So I said sure, but only gold (knowing there's no way he is going to add gold of any thickness to the board out of his own contracted fees).

Well, the boards arrived, and they had carbon on the pads. I eventually respun the PCB myself, but not before it cost my client $100,000.
« Last Edit: January 15, 2019, 01:03:36 am by KL27x »
 

Online langwadt

  • Super Contributor
  • ***
  • Posts: 4857
  • Country: dk
Re: surface finish thickness
« Reply #11 on: January 15, 2019, 01:26:28 am »
Quote
ENIG General Applications The minimum immersion gold thickness shall be 0.05 µm [2.0 µin] at -4 sigma (standard deviation)

So the minimum recommended by the standard is 2.0 µin. It seems reasonable that prototyping service will do bare minimum or less. I'm sure AllPcb will do whatever is requested for real orders.

the IPC i looked at said min. 2uin for default ENIG, but min. 1.58uin can be used for soldering only 
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf