Solder Jet Printing: Is It the Right Time?
I stopped by the Mycronic booth at Nepcon China 2015 and spoke with Sales Director Nico Coenen. He updated me on the company's new solder jet printing system, a technology that Mycronic has been developing and pioneering for 10 years. The company now believes that now is the time for it to gain real traction. One thing is certain: solder jet printing certainly has advanced and could make a big difference in process and quality.
Barry Matties: Nico, what is Mycronic doing in jet printing?
Nico Coenen: Jet printing in general has been around a while in the industry, but we are unique in that we are pioneering solder-based jet printing. It's a very difficult media to jet.
I've been working for many dispense companies, so I know how challenging it is. Mycronic has developed a unique, patented technology to actually make this process quite mature. We’ve been running in production for about 10 years with this. We are starting to see some competition popping up, which is good, because that means there is more demand for the technology itself.
Matties: It validates what you're doing. What does your typical customer look like?
Coenen: Traditionally, the Mycronic customer would have small- to medium-batch sizes, a lot of changeovers during the day, and they would need a flexible production environment. I would say the first 300–350 systems typically went to these kinds of customers. There were a lot of military, aerospace, and medical customers.
However, we have recently seen a lot of interest from high-volume guys. Not to replace the screen printer, but to have an add-on tool. We have seen that doing fine pitch by itself is not the issue. Having fine pitch and larger components on the same board is where people get into more difficulties.
Typical mobile phone manufacturers have an add-on process where they do screen printing and then a dispense or jet process afterwards.
Matties: Is speed the roadblock for the volume customer?
Coenen: We have a speed issue for volume customers. Let's say a mobile phone today takes about 15 to 20 seconds, max. That's your cycle time. With a screen printer you can do that, no problem. For us, speed is dependent on board size and how many deposits we have to do, so there is quite a bit of variation.
Although we have cranked up the speed 50% from our previous model, we're still not in line with that kind of speed.
Matties: And that is where you need to be to compete with screen printing. Do you see a day when that might happen?
Coenen: Yes, there are people that really would like to have a stencil-free factory for a number of reasons.
Matties: If you go stencil-free you eliminate multiple steps to the process. Preparing the stencil, managing it, painting it—even ordering it.
Coenen: Yes. Basically we are software driven, so something like high-volume mobile phone manufacturing still will have some small variance. Some might have 32 memory, the other 64, and with a stencil, that is a physical change. With jet printing you can do any batch size, and then switch over to the other in seconds.
Matties: I take it your R&D team is focused on increasing speed and breaking the speed barrier? How many jets are in your printer currently?
Coenen: Currently there is one jet. There are options to do four heads in one system, or to go for a modular system where you have up to five. That's all on the table. It looks like we're heading in that direction so that we can offer any capability or any speed, and that would bring us into a whole different league.
Matties: One of the common defects that I hear about is head-in-pillow. Is there an advantage to solving that through a jet system?
Coenen: One of the great benefits of the jet system is that we put down exactly the amount of solder per component because we can program it by piece, by single path, etc. All our initial customers more or less bought this for quality improvements for head-in-pillow and solder balling, which are typical problems you have in the screen printing process.
Matties: What is the most prevalent problem that you see in that process?
Coenen: There are a few. We mentioned head-in-pillow and solder balling; another is handling QFP devices. These are challenging components. Besides that, there are new kinds of problems where you have cavities in boards, flex ridges, and all kinds of different combinations, which make traditional screen printing more complicated. One of the big advantages is that we don't touch. We shoot from a distance, so we're less sensitive to height variations, and so on. Basically, the software can shoot here at this height, or follow by laser, topography, etc.
The system itself has a large opportunity window. We have identified that there is going to be a huge focus on this technology going forward, because there are so many demands on the current technology.
Matties: What do you think is the limit for screen printing? What's the deciding factor where they say screen just isn't what we can use any more?
Coenen: At a certain point, let's say, if your screen is only going to be maybe 50 microns, it would go to 03015. There's only a certain amount of thickness that you can put down, such as 50 microns of solder, and there is a limit on what type of component you can add to that. Again, it's a combination.
There are solutions like step stencils and so on, but again there are some design limitations as to where you can put small components, big components and so on.
Matties: What is the important thing that you think the market should know about this technology? What have we left out?
Coenen: I think the market, or the perception of the market, is maybe that this machine is only for NPIs and very small runs. When we analyze production environments, we more or less see that the utilization of the pick-and-place is very low in high-mix environments, even though you have high-speed machines. If the pick-and-place line is only running 20% of the day then the output is not very high.
With this kind of technology, you can increase the output of the pick-and-place machine as well. Even though it might seem that jet printing is a bottle-neck initially from a speed point of view, if you look at the daily production time you may be able to produce more boards. People have a misperception that it's only for NPI.
Matties: I can only imagine that your R&D department is back there trying to break the speed barrier. To me that is all you should be focused on.
Coenen: We have taken a good step with the new generation where we have increased it by 50%.
Matties: That's a huge breakthrough, but that was the low hanging fruit, right? Now it becomes more difficult.
Coenen: We have done it all in the same platform base. If we want to make the next step, we probably have to think about platform changes.
Matties: What is the price? This must be terribly expensive. What is the expected ROI for an average user?
Coenen: I would say the average price is about $175,000. But the ROI can be as fast as one year, to an average of three to five years.
Matties: So it pays for itself fairly quickly. What sort of maintenance or service requirements are there? Do customers have to buy a contract with you? How does that work?
Coenen: Actually, the system itself is quite maintenance free. The head has no cleaning parts to it. You can just take the head off, put it in a fridge, and in the morning take it out and put it back on the system. There are very few wear parts, so from a maintenance point of view we typically do one check-up on a yearly basis, making sure that all the calibrations are correct. But there is hardly any day-to-day maintenance.
Matties: What about the paste? Can they use anybody's paste? Or are they locked into a certain brand, like yours?
Coenen: We have identified five partners that we have been working with so far, and we are expanding that way.
Matties: Are they reluctant to expand because the market is so small for this application?
Coenen: No, the fluid partners have been demanding parties. We haven't been able to shorten down the qualification process well enough. What we are trying to do now is instead of making it a long, motley process, develop a very good guideline for the paste manufacturers that will increase their chances that the paste will be jettable. Then we can have a much shorter introduction on new paste.
Matties: So you're giving them a recipe?
Coenen: We're giving them a good basis for a recipe. Everybody has their own recipe, but we give them guidelines about viscosity and these kinds of things.
Matties: How many paste guys are excited about this?
Coenen: I have another 5 or 10 waiting for us, because they see that there is a need from their customers that they can't offer. So they have been pushing us, asking us, "Can we qualify?" We're working now to get a much broader spectrum, because that is beneficial on both sides.
Matties: It doesn't sound like that's a roadblock.
Coenen: I don't see it as a roadblock going forward. It has been a little bit of a roadblock over the years where we have had a limited amount of paste.
Matties: I would think that North America would be a prime target for this system.
Coenen: Mycronic has a very good, established customer base in the U.S. It took a little bit longer for the U.S. market to adapt to the technology.
One of the challenging things in the U.S. is that they have a lot of water-soluble paste, which is typically very difficult to jet. We're working on a water-soluble recipe for them. We're pretty close to getting a solution, but that has been a limiting factor for us in the U.S.
Matties: The magic key for this business sounds like broader paste selection and faster cycle times. You mentioned some competitors. How far along are they?
Coenen: What I have seen so far is that they are in the initial stages, but nowhere in comparison to speed levels or board capability levels.
Matties: You have 10 years of experience under your belt already. How many installs do you have at this point?
Coenen: We are close to 400 now.
Matties: That's pretty substantial. Is the concentration in Europe, primarily?
Coenen: I would say yes; the U.S. has caught up some, and actually Asia and China have caught up quite a lot.
Matties: You see more prototypes being built in China, too.
Coenen: Not only that, but here in China we’ve seen success in the high-volume market due to the add-on process, and we’re seeing some customers purchase maybe 10–15 systems, which is very different from our traditional customers.
Matties: Sounds like you're on a good path going forward. I'm glad we had a chance to talk today.
Coenen: It was good talking to you.
Solder paste jet printing demo at the Mycronic booth during Nepcon China 2015.