Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

PCB copper thickness-How do you measure?

Status
Not open for further replies.

Teebet

Electrical
Apr 27, 2005
5
0
0
CA
I am working on a audio project that I know it works very well. I am building 12 units in which technical all work well but during the audio tests 50% of them fail with a distinctive Buzz being heard on my speakers. By adding a wire I am able to make it dead silent.

The boards that I have ordered are 3 ounce copper but it does not seem to be consistent. Is there a way to measure this amount of copper is deposited onto the board?

Thanks
Robert
 
Replies continue below

Recommended for you

OUCH.. 3oz? You are asking for a headache in my opinion. The board companies find it hard to get that weight so they always use 1 or 2 oz and then try to reflow solder onto them to create 2-3oz stuff. That is what I have found and had explained to me by the board houses.

But buzzing... I would not typically expect buzzing to be caused by too thin copper. More typically buzzing in audio is poor layout/grounding or routing.

Is this wire you add short? Straight? And directly on top of (jumpering and existing trace)?

If it is then you really should redesign the layout to handle 1oz copper and not try to get 2 or 3 oz. This would probably be less expensive to get built too. If this cannot be done then you should place holes and assume you will always add the wire.

How to measure the thickness? You are correct not real easy. Generally you use a depth micrometer. It has a precision rod that protrudes out of the bottom foot.

See this example:
 
There are a couple of alternate approaches to measuring thickness:
> eddy current
> profilometer

In any case, 3 oz copper is quite thick already. If you have problems with that, then you've either got layout problems or possibly circuit topology/design problems. Most boards I've seen, even with very critical analog circuitry worked fine with 1 oz copper PCBs.

Buzzing could be oscillation, which would be quite sensitive to parasitics, etc.

TTFN
 
2 oz. is quite common, particularly with power stuff, but as smoked said, 3 oz. is not a typically stocked item (although obviously available). To slightly change what smoked said, they often copper-plate a board up to thickness (not reflow solder, as he suggested).

Be that as it may, I highly doubt extra thickness is the solution to your problem. If for some odd reason you truly needed the thickness in that area, I would suggest a standard 1 oz. board for the main layout with large contacts between the two areas requiring thicker copper. During assembly, mount a piece of bus bar to those two points... it will be cheaper than buying thicker stock.

You don't often measure the copper thickness of boards, you assume the board house is using what they're saying (unless you suspect otherwise). You may measure it in cases of designs requiring impedance matching, etc., but not for standard designs. In the special design cases, you should try to use a board house that you feel comfortable with, though (i.e., one that tells you the good AND bad).

I'm with IR... I think you have layout issues.


Dan
Owner
 
There are labs that can accurately measure foil thicknesses (I believe these are based on Alpha particle back-scattering).

Your amplifier buzzing sounds like a high frequency oscillation. This is usually avoided by having a filter between the amplifier and the load. Do you have a "Zobel" circuit at the amplifier output?

 
Thanks for the quick feedback.

I did think that it was a board layout or a ground loop but I could not find any. I will continue to do some experiments and try to isolate this problem.

Thanks again!
Teebet
 
Does you board have any inductive components on it? Winding or core vibration is another likely culprit for the sound.



----------------------------------

If we learn from our mistakes,
I'm getting a great education!
 
For PSU boards I have seen some designers place a solid copper wire on the bare tracks and solder it along the whole length, making a very high current parallel path.

If 3 oz copper did solve the buzzing problem, but 2 oz didn’t, the layout design would be very poor.

Are you sure it is not magnetic components (inductors, transformers) buzzing?
 
There are no inductive components nor is the buzz being generated from the transformer. I did check for that.

This audio amplifier has 12 inputs which a source can be shared or isolated to each amp section. If I isolate each input then there is no buzz but once I share the source I then can hear it. It is pushing me towards a ground loop problem at the input instead of the output.

 
Now you are getting there :)

If you put the output into a digital storage scope and lock the trigger to the mains (line) you will see if it is a power frequency loop. You may need to use a differential probe to get the signal on the scope without introducing yet another ground problem.
 
Trace thickness is not generally specified very well. The boards I have seen from three different mfg's are not very consistent with published data on copper plating vs. thickness. If I remember right, this is specifying how much copper is used per unit length or area. I can not recall. I found several board houses who only use 2 oz copper, regardless if you specify 1 oz. If the trace is critical, make sure the board mfg knows this, you will be surprised.
 
For 3 oz. copper boards, I think you can easily recongnize from the edge of the boards. You can see the thickness from the copper cutting surface. We produce charge regulator for solar system. And our regulator can handle 40A work current though 3oz. copper-double side. i think the board material is very good. We bought our boards from China. If you are interesting I can introduce to you. The price for 3oz. copper double side is only 0.03US$.

Best regards,
Feng
 
Teebet

As indicated in some of the above answers,
"how much confusion do you want".
If your noise is not line frequency, or a multiple of it
than it must be a internal oscillation, "most are line fc, or very removed".

EXAMPLE
I have a less than 1 year old Dell, I get 30 db at best, (line fc + many harmonics)
I also have a 10 year old "no name" with a soundblaster card, (also 10 yr old), same programs about 90
to 110 db.

If any one channel workes it should not be in the power
section "good bad news, in the power section one fixes "all".
If it is a noise that cannot be shut off or lowered, with normal volume, or output settings,"it is cap or esr related (loop stability) sometimes frequency or volume will change a little 55%?".
In which case it may be a long fight but only one cure needed.
If the problem is as I think,(input). You have one or more open connections in the pc card (One or more channels with no ground (typical). Try connecting a non working channel
ground to the output or power supply ground.
If it works that is the problem.
If this is true 'YOU NEED TO LET THE PC CARD MAKER KNOW, AND INSEST HE FIX IT'. You can fix it, (add wire in the proper places) " ALL FEED THROUGHTS, both sides, and side to side , NO EXCEPTIONS".
The problem with adding wire is you do not know how many
other places will break, "for twelve it is not much,
only a few tens of hours, but we all benifit when the garbage is taken off of the road, (OR MADE TO FULFILL THEIR CONTRACT WITH YOU).

As to thickness of the plating, "1oz 2oz 3oz," Only above
.1 amps should it make a difference, (very slight) ref mil-275 each oz is equal to .0014 mils thk.
One each .062 card should be (2*.0028 = .062+.0056 = .0676)thicker at the trace area (both sides, for two oz).
Most pc mfg plate up to the reguired thickness either + 1 or 2 oz. this is in mil spec, and is checked, this mesns they start with a 1 oz base card an after etching plate to 2oz,. If they start with two oz and need two oz, they may only plate a .2 oz (save money)
This means there is only .2 oz in through holes, (it breakes with flexing the card).

After these points things get complex.
If you need more info, you must give more detail and
specify the results of these tests.

 
Status
Not open for further replies.
Back
Top