Tag Archives: Circuit Mechanix

Circuit Mechanix on PUBHTML5

This is just a short blog post to let everyone know that Circuit Mechanix is now available as an online flip magazine from HTML5, I’ve moved this from MagCloud as I felt the magazine was slightly better. MagCloud was also confusing and implied you needed to buy the magazine.

Visit www.PUBHTML5.com and search for Circuit Mechanix to find all the magazines to date available for download.


PCB Mechanic

 

Advertisements

PCB design tools – Selecting the you’re going to curse!

It’s a common question when meeting a group of PCB designers  – “What tool do you use?” There then follows a line of conversation of justifying the choice of tool, maybe highlighting the fact that it wasn’t your choice then after a short discussion on what you’ve managed to achieve. After this is the inevitable list of things that every designer does not like about their design tool, the things that have brought them to such frustration that they’ve been pulling their hair out.

Every PCB designer has something to curse about the PCB design tool that they use. It never, ever runs as smoothly as the designer wants it to all the time. So that begs the question – is the wrong design tool being chosen or does the ‘perfect’ solution not exist?

Let’s get one thing out of the way first – there is no silver bullet in PCB design tools, every single on one of them has their strengths and weaknesses. So the trick (is if the chance arises) to choose the PCB design tool that’s going to suit your needs best.

I’ve been fortunate enough to get some thoughts on this from Jim Patterson, Electronics Group Leader at Evonetix:

So… how to choose an EDA tool? Well, who are you? Start-up with small aspirations where you can rely on contractors? Or do you have big growth plans and a desire to keep control of your design data entry? Are you a medium to large company looking to re-evaluate how effectively you manage your PCB development process? No doubt there are many more situations. I also don’t think there is a one-to-one mapping between your corporate situation and PCB EDA tool. I genuinely believe the most effective approach is to commit to a tool and get on with getting good at it and putting in appropriate procedures to ensure quality design data goes out the door.
Another thing that might influence your choice is the availability of contractors/design-houses who can use your tool of choice. You don’t want to pay to train them as well as do your design.
Enterprise vs Independent: if your business has no plan to make PCB design an important part of the company’s ability to create revenue then there is little point going for the big guns – they require more IT maintenance, more training and more money (eventually). On that last point – it is common to for a start-up to be offered the enterprise tools for free while the get set up. If you go for that be sure you are either financially ready for a big hit around the corner or are well prepared to drop the tool and switch to the competition at short notice.
Thoughts on individual tools…
Altium Designer: most bang-for-buck. Huge number of features backed up by being pretty intuitive to use. I don’t like how “infinitely” configurable it is. Maybe if they focused more on a single usage model it would operate more consistently and lead to fewer utterances of “OH FOR GOD SAKE ALTIUM”. I get the feeling it probably doesn’t scale as well to enterprise level PCB design (i.e. PLM, multi-site design entry, multi-person editing, advanced DFM analysis).
Cadence OrCAD: awesome layout tool, meh schematic entry, relatively open pricing, cheap. I really like the layout tool (PCB Designer) and you can get it very cheap (£549 for a 1y rental, £5500 for top-end options & perpetual license: http://www.parallel-systems.co.uk/pcb/). Should scale up to Allegro enterprise tool relatively easily if you need to.
Mentor PADS: Recently split into 3 types which means it now spreads across the Altium price/features range. Hints at $5k – $18k cost across the option range (http://www.eetimes.com/document.asp?doc_id=1326450). Not used it myself, but the top option now offers a lot of the high-end (xpedition) features. Should scale up to Xpedition enterprise tool relatively easily if you need to.
Honourable mentions…
Pulsonix (http://www.pulsonix.com/), Easy-PC (http://www.numberone.com/) and Design Spark (https://www.rs-online.com/designspark/pcb-software) are all flavours of the same tool. I’ve enjoyed using Design Spark for uni/PhD projects.
I’ve read good things about DipTrace (http://www.diptrace.com/) – very good value for money.

My Own Search

With this in mind a few years ago I was part of a discussion deciding what PCB design tool the company was to use in future. The current tool had failed and it was felt that it was time to move on.
My role in this was to bring the PCB tool options to the group so that the options could be discussed, looking back the summary of what we decided was interesting in many ways. Having had different experiences of the tools and had some demonstrations and quotes this is what was decided:
1) Mentor Pads:
Some of the engineer’s had experience in using Pads, so it was not an unknown. While the tool was felt to be capable it was costly, especially if extra options for 3D visualisation were required. As this was felt to be important and that any ‘extra’s were going to be costly, it was decided that this wasn’t the right way to go.
2) Orcad
There was minimal team experience with Orcad, it was also seen to be a very
Mentor-esque option as only the year before having upgraded to the latest version I was told that a library manager would cost extra. That was as far as Orcad got.
3) Cadstar:
No experience in the team whatsoever with Cadstar, to the point that it wasn’t even discussed as an option.
4) Pulsonix
Pulsonix had been trialed earlier in the year for it’s flex rigid capability, it was the only tool that would import designs and libraries from our current design tool  and came at a good price.
A demo left us feeling like it was a good option, but there were a lot of features that were ‘about to be released in the next version’. It would probably have done the job, but there was the feeling of the customers being guinea pigs for the product.
5) Altium
Altium’s success over recent years, the balance of cost against capability was very attractive. The vault was seen as a very effective tool for documentation control and the options for configuration were obviously huge. 3D integration and some mechanical and simulation capability was built into Altium as well as document release.

Which was best?

Even after deciding that Altium was the tool to use, there was the realisation that it was not going to be perfect. There were bugs, issues and it wasn’t perfect. We hadn’t found a silver bullet – BUT we knew that moving to Altium was going to be able to give us another level in our electronic design capability.
That is why PCB engineer’s tolerate the tools they use, if the tool has been chosen well if it fits into how the company runs it’s going to do the job it needs to do and do the job the electronic and PCB engineer’s demand. What suits one person or company doesn’t always suit another, the only thing that is certain is that at some point the design tool will be, for one reason or another – cursed!

PCB Mechanic
©CircuitMechanix 2018

Circuit Mechanix – Moving Forward

I know that the following on Circuit Mechanix isn’t very big but for those of you that may have been following it probably hasn’t escaped your notice that there hasn’t been much activity this year. Both the magazine and the blog have been a little lacking on fresh material.

So I can try and address this I’m goign to change the way things have been done. Rather than the magazine feeding the blog, the blog will feed the magazine. I’m hoping that this is going to keep fresh content moving into the Circuit Mechanix site and then onto the magazine.

As usual if there are any requests for information or content or any offers of contribution then I would be more than willing to accommodate them.

Thank you for your continued support.

Circuit Mechanix

 

NPL Soldering Defects Database

NPL Defects Database

The Soldering Defect Database is a freely accessible resource from the National Physics Laboratory website:

Defects Database

This little known resource is a fully searchable database of soldering defects with pictures, and information about the causes and solutions of each problem. Once signed up not only is all the information available but the database can be added to with any findings a user wishes to add.

This is a truly a brilliant resource and I can’t think of how many occasions this could have been of use in the past. Have a look, get involved and make sure to pass it onto your colleagues.

Circuit Mechanix Apr-2017

Designing Flexi’s and Flex Rigids – What’s Involved?

Flexible and flex rigid PCB’s are offering solutions to product designers that are not just a luxury in this day and age but a necessity. The technology itself isn’t that new, flex PCB’s have been around for over 30 years. But in today’s world of extremely compact and high speed connections often there is no choice but to use them over conventional electronic assemblies using wired connections to connect between PCB’s and these are to large. This is the main reason for their increased use in the last 10 to 15 years.

But how does the PCB designer go about designing a flexi or flex rigid PCB with components on it and how can the design be made easier to manufacture?
The basic Do’s and don’ts around flexi circuits we can get a picture of what’s involved and why:

a) To avoid the flex material tearing in manufacture or use, put as large a radius as possible on all internal corners.

FCB1

b) Via’s should not be placed in bend area’s as they can crack.

FCB3

c) Tracks running through bend areas should be routed at 90º to the bend.

d) Tracks on flexi’s ideally should be routed with filleted corners to stop them breaking during flexing.

e) Copper pours on flexi’s should ideally be hatched, especially in bend area’s. Because of stresses in the solid copper pour fractures are likely especially during flexing.

FCB2

This simple guide gives the designer a good basis on which to design flexi PCB’S (FPCBs) well. The challenges don’t end here though. Ideally an FPCB will have stiffeners under areas where components are going to be in the circuit. You can bet there will be times where some bright spark will decide for one of many reasons that the FPCB will have components and no stiffeners.

This is not an easy task and the designer is likely to get the task of working out how this can be done. This usually means working out what kind of stiffening frame can be used with the fabricator and assembler and sticking the FPCB down.

The reason for sticking down and FPCB for assembly is clear when it’s understood that the copper on an FPCB is giving at least as much structural rigidity as the substrate itself. Stresses in the copper push and pull the shape into interesting and often unwanted shapes. Assembly would be impossible without sticking the FPCB down to something rigid.

Added to this the copper finish isn’t very durable and can crack if flexed. Keeping these component pads away from bend area’s is necessary to ensure they survive through to assembly, but if it putting the component in a bend area can’t be helped then sticking the FPCB down to secure it is another reason why sticking it down is not be a bad approach to use.

As always it’s never a bad thing to engage manufacturer’s in the design stage and with flex this is even more important, even if only at first until all of the design aspects are better understood.

© Circuit Mechanix 2017

 

Assembling reliability into electronics

Making reliable electronic assemblies is about making a good solder joint isn’t it?

So what’s the problem, plop down some paste in the right area place the component carefully down, apply heat and the solder melts. Hey presto!

Yes, if only it was that easy. The Jedec thermal profile is well known and was mentioned in a previous issue???? The problem is every part of a PCB will have a different thermal profile as will every component and joint. In the ideal world the flux needs to activate at the same time and the solder melt at the same time, but this is never going to happen.

The result of this is an electronic assembly can suffer from dry joints in places and others voids, head in pad or possibly tombstoning. It’s quite a juggling act that needs to be performed. Unfortunately it’s mainly trial an error, but knowing what to consider before going to assembly can help avoid some of these nasty niggly errors.

It’s more difficult to uniformly heat a PCB to the correct Jedec profile. Thermal shadows and hotspots in different parts of the PCB can create problems in achieving a good result. Using a longer thermal ramp up and dwell time can often help but getting heat under leadless components and BGA’s will always create difficulties.

Dry joints and voids under these components will not be immediately obvious with only x-ray inspection able to reveal any issues. Many BGA’s have hundreds of balls on them with every singe one needing to be perfectly soldered in order to operate properly and reliably.

Vapour phase reflow is becoming more popular in reflowing modern high density PCB’s as this gives a far more uniform and reliable thermal profile across the PCB being assembled, even for joints under the component such as BGA’s. It tends to be a slower way of reflowing boards, so it isn’t used as much in higher quantities, but the process gives good results across the board.

 

So now we know how to make a good solder joint that’s it right? The board can be put into service and will work for years – problem solved?

No.

Many PCB’s are put into service in environments that are harsh, with high levels of moisture, dust, or some kind of chemical pollution. Any one of these will affect the long term operation of the product by either forcing the board to overheat or causing problems in the operation of the circuit.

Using underfills for BGA’s and a conformal coating can help protect a circuit from these effects, but a maintenance schedule could still be needed to clean the boards.

An underfill is a liquid that fills in under a BGA or other leadless component to protect the joints and give extra mechanical rigidity.

Conformal coating are sprayed over a PCB to protect the board from moisture and other contaminants. There are many types to use and choosing the right one to protect against the contaminants it’s expected to encounter will need careful consideration.

Both of these processes should only be applied to a tested working board. If these are to be used on a cleaned board, the board must be really clean as any residue flux is likely to corrode any joints or copper over time (such is the nature of no clean fluxes). Once a conformal coating has been applied, there’s little that can be changed on a board, so it’s make or break. If all this works, your PCB will work in some harsh environments for a good time without trouble.

Circuit Mechanix © 2016

The role of fabricators in enhancing the reliability of PCB’s

As is always discussed, how a PCB fabricator makes the PCB’s  that you design can make or break a project. The same is true with reliability, knowing what to ask and what’s needed is most of the battle. If we’re hoping to create a circuit that is going to operate reliably in harsh conditions, what is it that we need to be looking for in our fabricator and asking them to do?

Finding the higher grade PCB materials for use in yours boards is all very well – but can your PCB fabricator get them and use them? Some of the cutting edge materials can be very hard to get hold of, or can have a long lead time attached to them. Similar products are often available from a fabricator if your needs are discussed, who knows, perhaps even something better.

With the materials decided, the circuit has to be manufactured in a way that will enhance it’s reliability. As always, it depends on the application, but if we’re going to consider harsh environments and temperature changes the considering the type of copper will make a difference.

Typically two types are available on PCB laminates, Electro Deposited (ED) and Rolled Annealed copper (RA). Both processes are self explanatory, but rolled annealed will typically be more robust as it’s a sheet of thinly rolled copper presses onto the laminate surface. These sheets will also be used when making inner layers on multilayer assemblies and bonded between layers of prepreg and core materials. It’s not generally known, but a tiny amount of lead is added to increase it’s durability, but not enough to break RoHS directives.

Electro deposited copper will be more fragile and fractures are more likely to be made when subjected to mechanical stresses such as thermal expansion and contraction.

The most likely part of a PCB to fail when subjected to thermal stress are the via’s. The z-axis will expand and could fracture via’s. The via wall thickness will thin in the centre, creating a weakness that can break in thermal cycling.

A fabricator will typically specify the thickness of via wall they’re able to manufacture. Specifying a minimum via wall thickness of 25um will reduce the risk of via fracture considerably. There is no way to eliminate this completely of course especially in cases of extreme temperature change.

The quality and accuracy of via drilling also has a distinct affect on via reliability. Drill bits need to be replaced a regular intervals to ensure the drill is sharp when every via is drilled. A blunt drill will create a rough via wall that could fail, especially when stressed.

Drill accuracy is how on target every via has been drilled. On larger via’s an pads it’s easy to see this, but on smaller.

The condition that needs to be avoided is called the keyhole effect (as shown above) where the drill hole is not on target and can create a potential weakness and failure effect where the track leaves the bad or via.

It needs noting that no via hole will ever be spot on – but it needs to be within the stated manufacturing tolerances of the fabricator. The IPC-A-600 standard will give guidance on the acceptability criteria of this an many other manufacturing defects that may arise from a fabricator. It will also give an idea of the acceptability criteria to specify when asking them to supply boards.

Circuit Mechanix © 2016

How can engineers design in reliability?

Designing electronics that works on the bench is one thing, but lets put these electronics in a harsh environments, which is dirty, where the temperature changes from hot to cold, perhaps rapidly. Maybe the equipment is going to be mounted in a helicopter and experience extreme vibration or perhaps in an environment that subjects the circuits to steam?

Every designer should have a kind of risk assessment in their heads of the factors every design has to face and one of these without a doubt is the expected lifetime of the product or equipment. For example, equipment that only needs to operate only a few hours for a race needs to be thought of differently to the reliability expected for a board operating on an aircraft without fail for 25 years.

The key to knowing how to design the electronics is to have a good knowledge of the environment the equipment is going to operate in. What are the temperatures that it’s expected to operate in? Vibration and mechanical shock are also issues as well as humidity and moisture. The possibilities can be daunting – but if a few simple things are done, the risks around many

Of these issues can be reduced or eliminated.

Temperature

The first thing to remember about temperature changes, is that everything moves. If warming up then all the materials expand, when cooling they contract. If the circuit  undergoes rapid heating or cooling or both the board is going to expand and contact. While this isn’t a big issue in the x and y axis of the board, the z axis is different.

The values of thermal expansion of PCB materials is typically greater in the z axis, but this is widely disregarded as the distance is so short. But if a high degree of change takes place then via’s can fracture, often creating a very irritating failure that will be hard to find and harder to fix.

Reducing the total thickness of the board can help with this if cheaper materials need to be used. Using materials that are designed to operate in higher temperatures can be very effective as these have a reduced coefficient of thermal expansion. The extra cost of these enhanced FR4 materials is often minimal, it only get higher when considering much higher grade materials, the this cost is mostly lost in the processing of the PCB’s.

Another cost effective solution is to ask the fabricator to enhance the via wall thickness. This is covered more in the next article.

Lastly the right components need fitting to the board. If the electronics are going to be operating in an 80C environment, then 50C rated components aren’t going to be good enough. Make sure the components are rated for the task, or it’s likely that something weird will start happening at the operating extremes.

Mechanical Shock – Vibration

This is perhaps the hardest external factor to try and de-risk. Making sure that the pads are large enough, especially for components with a large mass is about all a designer can do. Only when carrying out drop and vibration testing can a designer or engineer get an idea of how well the electronics will hold together and survive in use.

Humidity & Moisture

However undesirable, humidity and moisture can be a big issue for electronics operating outside the home or office. Cars, planes, industrial equipment can all experience this. Where heat can create steam, any electronics subjected to steam are going to have a really tough time, especially if any other chemicals are in the steam as well. External factors like enclosing the electronics or conformally coating them can be used but add extra cost which can be undesirable in some cases. The best alternative is to monitor the moisture the board is subjected to, perhaps allowing extra features to shut down the circuit or warn the end user.

The question is how to do this cost effectively. Adding a conductive test coupon to act as a sensor to any kind of surface pollution is the first step. Adding it to either of the surface copper layers and leaving it exposed and coated in the same finish as the rest of the board will cost very little. Coupling this to a spare op-amp and digital I/O could mean that the only extra expense is the cost of designing it in and testing it.

Designing in reliability is more about thought and understanding than cost. More can be achieved using cheap and simple techniques before spending a lot more money – it just depends where the electronics going and how important it is that it keeps on going.

© Circuit Mechanix 2016

Contact: circuitmechanix@gmail.com

 

Circuit Mechanix December 2016

 

circuit-mechanix-dec-2016

 

Welcome to the fourth and last issue of Circuit Mechanix for 2016.

In this month’s issue the theme of reliability in electronics is looked at and how this can be addressed at each stage of the PCB design and manufacturing process

This issue also has a look at how OrCAD’s Sigrity can assist in design reliability.

Have a look – get in touch and get involved if you like. This is a young project and help with news and features is needed.

 

 

Download the PDF for the magazine here:

circuit-mechanix-dec-2016

 

CircuitMechanix Flipbook:

CircuitMechanix Dec 2016 – Flipbook

 

There is also a LinkedIn Group for the Magazine and discussion around it here:

Circuit Mechanix LinkedIn Group

Circuit Mechanix December 2016

 

circuit-mechanix-dec-2016

 

Welcome to the fourth and last issue of Circuit Mechanix for 2016.

In this month’s issue the theme of reliability in electronics is looked at and how this can be addressed at each stage of the PCB design and manufacturing process

This issue also has a look at how OrCAD’s Sigrity can assist in design reliability.

Have a look – get in touch and get involved if you like. This is a young project and help with news and features is needed.

 

 

Download the PDF for the magazine here:

circuit-mechanix-dec-2016

I’m still not able to provide a flip book – as soon as I get over this technical hurdle I will let you all know!

There is also a LinkedIn Group for the Magazine and discussion around it here:

Circuit Mechanix LinkedIn Group