TECHNET Archives

December 2013

TechNet@IPC.ORG

Options: Use Monospaced Font
Show Text Part by Default
Show All Mail Headers

Message: [<< First] [< Prev] [Next >] [Last >>]
Topic: [<< First] [< Prev] [Next >] [Last >>]
Author: [<< First] [< Prev] [Next >] [Last >>]

Print Reply
Subject:
From:
Dennis Fritz <[log in to unmask]>
Reply To:
TechNet E-Mail Forum <[log in to unmask]>, Dennis Fritz <[log in to unmask]>
Date:
Thu, 12 Dec 2013 10:17:22 -0500
Content-Type:
text/plain
Parts/Attachments:
text/plain (112 lines)
Good advice., Don, 

The board fabricator can change the "starting copper thickness" any way he wants by either panel plating more copper, or sheet etching the initial copper foil by the process known as 'foil thinning".   Thus, the fabricator can "make" 1/2 ounce, 3/8 ounce, or even 1/4 ounce foil out of one ounce copper.  Of course, he can buy thinner foils, but they may be a premium, and the cost/convenience of etching down will drive his choice. 

Then, the fabricator will lose a bit of starting foil in deburr after drill, microetch before electroless copper, and microetch after image application and before pattern plating.  

The thinner the foil at the time of final etching, the smaller the line width that can be produced.  Remember, the etchant for the surface foil etches into the conductor sidewalls almost as much as it etches downward to the laminate.  AND, conductor etching loses from both sides of the trace.   That is, if you call out and expect 1 ounce starting foil of 1.4 mils or 35 microns, the trace widths may be reduced as much as  2 X 1.4 mils =2.8 mils/2 X 35 = 70 microns.  If you are calling out 5 mil/125 micron conductor widths, the imaged trace on the board may be 5+2.8 = 7.8 mils/125+70 = 195 microns.  If you want equal final lines and spaces, the starting space for etching after removing the pattern photoresist may only be 10-7.8=2.2 mils/250-195=55 microns.   

I hope that you see that fine line circuitry CANNOT be made from thick starting foils.   If at all possible, I recommend NOT giving starting copper foil on the outer layers.  Every board shop knows the characteristics of the plating baths - ratio of copper plated on the surface versus in the holes.  Also, even more importantly, they know their etcher characteristics - how much copper is etched from sidewalls of the traces versus the etching in depth to isolate the copper traces. 

Denny Fritz


-----Original Message-----
From: Don Vischulis <[log in to unmask]>
To: TechNet <[log in to unmask]>
Sent: Thu, Dec 12, 2013 7:14 am
Subject: Re: [TN] IPC-6012C Total Cu thickness


To quote one famous expert, "it depends".  The board designer understands the 
needs of the board.  The thickness specification needs to consider the at least 
the following:  copper surface thickness, through hole copper thickness, plating 
ratio for through hole vs. surface plating (which usually plates more on the 
surface than in the hole), and line width and spacing.  All of these factors 
affect the performance and the producibility of the board.  In some cases the 
cost is also affected.

Just my $0.02.  (Interesting...just noticed that the iPad doesn't have the cents 
symbol...guess that means that the penny is really on the way out)

Don Vischulis

Sent from my iPad

> On Dec 11, 2013, at 7:19 PM, Bernard <[log in to unmask]> wrote:
> 
> Hi Raye----
> While it may be no comfort because I'm out of my area of
> confidence, I believe you are correct.  If the starting thickness
> is to be a requirement, the only way to inspect would be to
> get a coupon from the lot number----- so it seems to me.
> Otherwise, the overall thickness should be specified as the
> requirement.
> Enjoy!!!
> Bernie
> 
>> On 12/11/2013 4:48 PM, Rivera, Raye wrote:
>> Hi Bernard,
>> 
>> As a general rule, copper is copper. On this specific board our engineering 
team confirmed that the ratio of foil to plate does not matter.  While I am no 
expert, the only time I can think of that it might matter is in traces for very 
high frequency signals.  On these even the surface roughness of the trace and 
the cross-sectional profile of the trace can matter.
>> 
>> Here is another consideration.  We determined that the foil layer the PCB 
fabricator used was less than 1/2 oz based on a microsection.  The PCB 
fabricator tell us that they did start with 1/2 oz copper and intentionally 
reduced it to 3/8 oz chemically. If they didn't do this,  by the time they got 
the right amount of plating in the through holes, they would have up to 1.7 oz 
surface copper.  For reasons I don't fully understand they would not be able to 
hold our trace and space requirements when they etched the outer layer traces.
>> 
>> I have always thought a microsection shows the thickness of the starting 
foil. Apparently it only shows the thickness of what remains of the starting 
foil. If we can't measure it, why should we state a minimum starting thickness 
for the foil in the standard? I can see the value of including it in table 3-12 
as a general guideline, but it does read as if it is a requirement.
>> 
>> Best regards,
>> Raye Rivera
>>  
>> 
>> -----Original Message-----
>> From: TechNet [mailto:[log in to unmask]] On Behalf Of Bernard
>> Sent: Wednesday, December 11, 2013 12:38 PM
>> To: [log in to unmask]
>> Subject: Re: [TN] IPC-6012C Total Cu thickness
>> 
>> I ask this question just out of curiosity----- Does the cu/plated metal ratio 
affect in any way the signal transfer rate or electrical characteristics????  If 
no---- all seems academic beyond the total thickness.  If yes, then min cu seems 
appropriate.
>> Bernie Kessler
>> 
>>> On 12/11/2013 12:23 PM, Grunde Gjertsen wrote:
>>> Those tables are very educative when explaining to QA and design people why 
final copper thickness is what it is in relation to base copper and class and 
should be left well alone in my opinion, they are not that difficult to 
understand if you know a bit about PCB manufacture.
>>> At least over here people in the industry is getting more removed from PCB 
manufacture since so many factories has closed down and production gone to far 
east, so to have the standards actually explaining stuff like that and not 
simply stating requirements is not a bad thing.
>>> 
>>> Rgds
>>> Grunde
>>> 

______________________________________________________________________
This email has been scanned by the Symantec Email Security.cloud service.
For more information please contact helpdesk at x2960 or [log in to unmask] 
______________________________________________________________________

 

______________________________________________________________________
This email has been scanned by the Symantec Email Security.cloud service.
For more information please contact helpdesk at x2960 or [log in to unmask] 
______________________________________________________________________

ATOM RSS1 RSS2