Join us on Wednesday, November 10th at noon Pacific Time, and join Mark Hughes and Greg Ziraldo in the heavy copper PCB hacker chat!

Although printed circuit boards are useful, they are sometimes a bit fragile. Because there are only one or six layers of thin metal on the circuit board, and the ultra-fine traces that must be installed between dense pads and vias, the current-carrying capacity of copper on most PCBs is somewhat limited. This is possible in most cases, especially when it comes to logic levels and small signal currents. But what happens when you really need to increase the juice on the PCB?

Enter the world of heavy copper PCBs, where copper is sometimes as thick as the circuit board substrate itself. Traces as thick as these physically present a variety of challenges, from thermal and electrical considerations to potential manufacturing issues. To help us solve all these problems, Mark and Greg will participate in Hack Chat. They are all working at Advanced Assembly, a fast-forwarding PCB assembly company, with Mark as the research director and Greg as the senior operations director. They understand the ins and outs of heavy copper PCB design, and they will share their wealth with us.

Our hacker chat is a real-time community activity in the Hackaday.io hacker chat group message. This week we will sit down at 12pm Pacific time on Wednesday, November 10th. If the time zone confuses you, we have a convenient time zone converter.

Ha ha. I never realized that these are the same thing. I mean, in hindsight it's obvious that they will become one thing, but it's just something I never thought of.

I can't help but wonder if using more traditional busbars in certain applications is not a better solution.

I have seen some PCBs that use copper strips and occasionally place the legs strategically and solder them like any other through-hole component.

Another question is whether it is possible to add a layer of photoresist to the board and then simply selectively electroplate more copper where needed. However, this may not increase any major numbers and will result in height changes on the circuit board. (It seems that this is not a problem with the circuit board shown in the article above.)

However, I also want to know if it is possible to obtain a better aspect ratio if etching is performed in two or more steps. The photoresist is re-added each time to protect the newly formed wall from further etching. (The circuit board in the picture above seems to be affected by over-etching in some areas.)

In fact, large bus bars are usually just hot-soaked water jet cutting boards, hydraulically crimped cables, and bolted to a ceramic or flame-retardant glass composite material.

Even a 2 ounce PCB with a width of only 12 inches can become a nightmare for hybrid SMD components, because the adhesive layers with different expansion coefficients form the bimorph cantilever. Considering that the pressure caused by temperature changes will usually damage the components, a conservative reflow curve will not help much, because if it is not bolted, the PCB will be severely warped (hence the same fastener fun as the busbar).

Many years ago, I worked at Sylvania Control Devices in Standish, Maine. We have manufactured millions of glass-enclosed circuit breakers embedded in motor windings for over-current and over-temperature protection. To test these and check the distribution of production parameters, we can develop histograms. The circuit breaker is immersed in a hot oil bath, and the lead contacts are connected to the copper pad of the incandescent bulb. When the lamp goes out as the temperature rises, the operator will watch the lamp closely and re-ignite when the temperature drops.

We need more of these 12" x 14" boards, so I decided to try a 10 oz copper-coated G10 1/4" thick board instead of manually fixing the processed copper sheet to the base. It took a while to find the supplier, but I then screen printed the desired pattern, drilled the circuit board directly under the copper layer, and installed the LED instead of the filament bulb. These effects are very good. This is also one of the first LED applications I tried. Compared with current products, they are very sad, but hardly bright enough. The pattern is large enough, and the change of etching is not a problem.

This is a messy operation at best, but the new lighter board makes it easier. It's hard to believe that this was 60 years ago. A fond memory of one of the best jobs I have ever done.

Therefore, it is indeed a very early adopter. The first commercially visible LED I saw appeared in the early 1970s. They are very expensive, but their prices plummeted in the middle of the decade. Soon after, people began to think that digital watches were a very good idea.

We made a board, and we were sure that half of it needed 12 ounces of copper (including the QFN package for switching the 30A load), and the other half only needed 2 ounces. This is a terrible experience. They tried their best, but we had a short circuit between the pads on the QFN, because the aspect ratio was so high, the plating was not uniform enough, and it was impossible to solder all 44 pads on the QFN with lead-free solder ( And it’s troublesome in QFN), and reworking it is also terrible. They did something similar to what you said, they used 2oz foil and then electroplated half of it to 12. Since then, we have used a combination of some hot core sub-boards (for heating) and a large number of wide traces (for pure current) in our subsequent projects, because we are chip designers and are aware of others Will face the same problem, when we do things that require this layout, we make sure that the chip bonding mode allows the required current to be 2 ounces of copper.

Making sure it works in 2 ounces of copper is actually fine.

Yes, adding extra thickness selectively is not a perfect process at all, although it depends on the process.

If the space is limited, I tend to use more than one layer when the current is high enough. Not that this is not its own canned worm...

Sometimes it is quite good to split the board into two completely independent boards. One is used for all high-current materials with large pin pitch, and one is used for high-density logic materials.

I know that this way of thinking has its own flaws... But in this case, I believe that if ultra-thick etching is the best way to do things, then it is already common in the industry.

The bus bar will be cheaper and easier to manufacture. However, those thick marks look cool.

It looks like you can use that board to make a marble game.

This will be a terrible marble game, assuming energy is pumped through it and the marble is made of something like a ball block.

That PCB looks strange. It seems that using half the copper thickness but a wider track can achieve very similar results.

It seems that there is not much consideration of the different currents flowing to different parts on the PCB. For some components, voltage drop is an issue, while for other components, only the current handling capability is important. Take the track leading to the resistor as an example.

A wider trace may not be an option. If your power electronic equipment is operating at the mains voltage, the safety standards stipulate that the minimum spacing between traces is quite large (usually a few millimeters, depending on the voltage and the insulating material present). Doubling the width of the trace may make routing more difficult because, for example, you suddenly cannot get traces between the pads of a standard through-hole resistor and meet the minimum gap. This may just be a bad design, but I will not draw this conclusion based on the picture above.

On normal boards, they will mill grooves to separate tracks that require more clearance. When you will also get the I2R thermal effect, you are not sure what the rules for these types of PCBs are. At the very least, the tracks can be routed on the opposite layer so that wider tracks can be used-less thermal effects, more adhesion.

I recognize that picture! Context: https://www.reddit.com/r/electronics/comments/5i9wxd/i_just_received_my_20_ounce_pcb_soldering_is/ & https://www.reddit.com/r/electronics/comments/5urt3l/remember_that_20oz_700ummade_pcbday_i_i_i_i

thank you very much. PCB summary:

It is https://en.wikipedia.org/wiki/Formula_E. The manufacturing cost of the power bare PCB is obviously US$2250, which is considered a "peanut".

One side shown has a /- 450V input, the other side has 14V @ 150A and 24V @ 80A outputs, and has a 2cm wide track. For a 900V side rail, this is 4kW and approximately 5A.

(Or there are two of these PCBs, one for 14V @ 150A and the other for 24V @ 80A)

The copper is 20 ounces (0.7 mm) thick.

The SMT resistor for reference is 0603.

I glued this type of board around 2012. 6oz copper was then electroplated to produce a thermally conductive double-sided copper clad laminate, processed as pth cappers, then bonded to a thick copper board and processed into pcb, and then finally milled and shaped. From After that, I want to know if it is possible to cnc molding the heavier copper track, then process it, and then use the prepared thermally conductive material to directly bond it to the metal substrate. This will eliminate the step etching problem.

It may be more effective to use this copper thickness to cast it in a mold and then stick some backing plates on it.

There was a project a long time ago. There was a 3D printer here, which was modified to be an inductor with copper wires placed on the glue bed. This is a method that can be used to make custom PCBs using thick copper.

In the past few years, with all electric vehicles (and all need electronic devices to drive their motors), thick copper PCBs are becoming more popular than before.

Products like the PCB at the top of this article may be good for prototypes, but for any product beyond the prototype stage, the price tag is completely absurd. For such a large amount of copper, it may take several hours to add copper only by electrolysis. Even without considering the price, it is not suitable for mass production.

PCB etching and electroplating are very effective for thin copper layers, but thicker copper requires more efficient production methods.

"For such a large amount of copper, electrolysis may take several hours. Even without considering the price, it cannot be mass-produced."

Well. Yes, Faraday is a stern master, but maybe you can double check this sentence. The current price of copper is US$10,000/ton. The current price of aluminum is US$2,600/ton.

How is aluminum made? Yes, electrolysis.

Only use copper heat pipes instead of copper traces; P

Don't forget the effect of skin depth: more than a few hundred kHz, increasing the thickness of the trace will not allow you to gain more conductivity. You need to increase the width (or Litz technology) to get more conductivity.

Please be kind and respectful to help make the comment section great. (Comment Policy)

This website uses Akismet to reduce spam. Learn how to handle your comment data.

By using our website and services, you explicitly agree to the placement of our performance, functionality and advertising cookies. learn more