Better Together: How HDP User Group Showcases the Industry’s Best Side
The High-Density Packaging User Group International Inc. (HDP User Group, or HDPUG) is a non-profit trade organization comprised of members from top companies in the electronics industry, from materials suppliers and manufacturers, to OEMs and end users. Key activities include collaborating on issues facing multiple industries and bringing people together on projects who might not have met otherwise. Barry Matties met HDPUG’s European representative and project facilitator Alun Morgan at the recent EIPC Summer Conference to learn more about the group and current projects.
Barry Matties: Alun, let’s talk about the HDP User Group. Would you mind sharing a quick overview of what the group is all about for our readers?
Alun Morgan: HDP User Group was formed 26 years ago. It's a research and development consortium comprised of members from some of the largest companies in our industry such as Oracle, Cisco, and IBM—many of the big companies who try to share information on common problems. For example, if one industry has an issue with lead-free materials coming around, another one has maybe the same issue. Rather than everybody solve the same issue themselves, we group our resources, and we can save costs. So the whole drive really is to save costs for the electronics industry in order to solve common problems or common issues.
We have nearly 30 projects running now and over 50 member companies. We meet four times a year, twice in North America, once in Europe, once in Asia. We look for new projects, and we carry our projects through to conclusion. Project life cycle varies. Some can be over in a matter of months and others could be a year or even multiple years. Many projects run over multiple phases. I run a few of them now on materials and reliability; we've tested more than 65 materials now in five phases. So with the new materials, we apply common methodology to testing them so we can compare over years how the performance is. Rather than companies having to test new materials, we do it as a group, and we share the data amongst the members so they can do at least a first level screening of new materials and know where they are.
Matties: How do you pick the projects?
Morgan: It's entirely member-driven. We go to the members and say, "What issues are affecting you?" Every meeting we have a call. We go around the room. People say, "Oh yeah, I've got an issue with this, or with that." And we select entirely on the member input.
Matties: Do you see recurring themes from year to year?
Morgan: Yes, we do. We get multiple phases, that's for sure. You can imagine that the lead-free change made a big impact on the industry, so we have a lot of projects on lead-free. We have a range of members. We have members who have to change straight away and members who have exemptions, for example, the military sector. The military and aerospace industries didn't need to go there yet, but of course they are having to go there at some stage, so we had some multiple phase process with them. We tested these things for commercial applications. Now we're looking at much more rigorous testing for military and aerospace applications.
Matties: So when you take a project on in the User Group, is this a project that you take on individually as a company?
Morgan: The HDP User Group appoint a facilitator whose job it is to manage the process of the project. The projects start with an idea. So someone says, "I've got an idea." And then we try to form a group around that. If we can get a group of a few people, half a dozen or so, interested in the project, then we'll have a kickoff call to discuss it. From there we go into define what the project will do—what are the goals of the project? What are the objectives? And how are we going to achieve that? Are we going to build some test vehicles? Are we going to do some initial research or what are we going to do? We move from that to the implementation phase when we actually do the work that we've planned out. And then that's how the projects conclude. The timescales can be varied. Ideas can last for weeks or they can last for months. Definition is usually relatively straightforward. Implementation, again, can be a long time. Some testing takes months.
Matties: It also depends on the participants’ own workload as well? It's still voluntary, so even though they're members, they're volunteers?
Morgan: Every project has a lead for the project. That is from one of the membership. So there's a facilitator who is staff for HDP, and there's a project leader from our membership base. The member basically leads the project. The facilitator does the organizing, the administration, makes sure things arrive on time, and gets resources that are needed as well. Sometimes resources have to be purchased. Sometimes the members offer them, so we generally have most resources given by the membership, but occasionally we have to buy some special components. There might be something we need to do and we take over that side of the work so the project team can focus on the technology and the technical aspects of the project. We take care of the administration.
Matties: You said that you have 30-some projects going. You meet four times a year. I'm assuming that there are a lot of project team meetings that are happening every day.
Morgan: Absolutely. We have them all the time. Every day. A typical cycle would be between two and four weeks for a conference call on a project. It depends on the stage we're at. Sometimes, when we go into testing, there might be a long time where there's nothing really happening. And then when the data starts to arrive we have a lot of calls, because a lot of it is failure analysis and data analysis. So they go that way, but yes, there are calls every single day. We use WebEx for our meetings as well, so you don't have to attend the meeting—you can attend remotely. We had a meeting just last week in Scotland, at Oracle in Linlithgow. We had a number of project leaders call in from Asia and North America. It's just like they're in the room. We just carry on as if they were there.
Matties: That's nice, because you just mentioned the global aspect of this too. You get a nice viewpoint from all the different regions of the world that have their own issues, I guess.
Morgan: They do, but a lot in common. We have membership across the world, of course. I mentioned North American companies, but we have people from companies like Huawei in China, we have Panasonic, we have Hitachi, we have Shikoku, and we have Japanese manufacturers. We have European ones as well, in fact it started in Europe with Ericsson. That was the first member and we try to make sure that we go around the world with this. We try to involve people. Of course many companies, many members have organizations in all the regions of the world anyway, but some are unique to certain regions. Some have specific topics they want to discuss. Some test houses want to be involved in this perhaps. Some material suppliers may be interacting in one region. Solder manufacturers maybe or assemblers, so we have a really broad range of activities.
Matties: So for a material supplier, this is a great opportunity for them to come in and really introduce their materials into a solution and create a market.
Morgan: It's a kind of showcase sometimes for this, and secondly, it’s interesting that people don't think about it so much, but occasionally we get materials that are in beta phase, or late alpha in some cases, to put them through the methodology. Just to get a broad spectrum of testing that we do on a standardized method rather than local methods from the manufacturer. We often get materials that go through the testing phase and actually drop out at some stage. Then the manufacturer actually goes back, he refines the product and then brings it back later on into the process.
Matties: It's a nice proving ground.
Morgan: Exactly, and you're doing it against a standardized methodology that we've been using for probably over 10 years now. We have a really good database of knowledge and we know exactly how materials perform against each other.
Matties: Now the results from your efforts, how are those shared?
Morgan: Well, they're shared to the members primarily. They’re for the members and the members pay for activities, so they're shared with the members.
Matties: So they have the advantage?
Morgan: Yes, they do, but we publish lots of information. We did yesterday in fact. We published results on the high-frequency surface finishes or alternative oxidized finishes on materials. When we publish them, we tend to anonymize the data. We don't tell you which product behaved better than another one. We don't give that information, but we do share the general trends. So I could say, for example on the materials reliability projects, we have shown over time a move towards greater reliability in terms of thermal stress testing. So that's something that we've observed over the years. If we go back to the early days of lead-free, there were some failures going on. Now there are very, very few failures for thermal reasons.
On the other hand, we've seen an increase in failures on reliability testing, for example CAF testing. With features becoming smaller, these tests are becoming more aggressive. We've slightly increased in that area, so we publish that kind of data. In fact, I'm really pleased to say we've won a number of best paper awards. The last two years we've won the best paper award from our projects presented by one of our members at IPC APEX EXPO. So we do well, the papers are of high quality and the information is eventually disseminated.
Matties: The material suppliers are taking it to the open market.
Morgan: They do. Obviously the information is there for people to share. Members can also share information. It's at their discretion. We obviously don't receive any confidential information, by the way. We're very clear on that. The groups that we have are sharing data. If you want to make anything public to the group, then it's public at that point. It's entirely their choice, if they want to share their IP they can do that. Once they've shared it though, it's no longer confidential.
Everyone understands that, and it works well. We focus on things that aren't proprietary or aren't company IP related, but industry common problems that we can all benefit from. Every meeting we have, we start with that discussion just to make it clear to the members so they understand that. It works really well. I have to say, I'm very impressed with the way it runs.
Matties: Over the 26 years that HDP User Group has been around, what's been the greatest success? What's the crown jewel?
Morgan: Well, I'm biased on that. I would say it's probably materials and reliability testing, because that's been my career line.
Matties: How do you measure it though? Because you're talking about lowering costs in your mission statement.
Morgan: Well yes, we do.
Matties: Do you have any metrics that say...?
Morgan: We do. In fact, you'll see in our annual report we actually publish a chart of the leverage. And the leverage is massive, we're talking an average of around 20x on the dollar and some projects up to 40x, so that is really impressive. We take the testing we've done. We value the testing. We then divide that into the cost of the membership and you can come to that figure. We know what each member spends individually. We know what we've committed in terms of resources for the project. So we do measure that. But in the end, I think probably the best thing the group has done is bring the industry together. It's brought people together from all aspects, from the materials supplier to the manufacturers, to the OEMs, to the users, and actually formed a very nice group of very effective working teams whereby we can cover the whole supply chain and actually drive costs down. The whole thing.
I think if we were to quantify that in the way I've just mentioned, you'd probably find that we have really huge savings that have been made over the last 25 or so years.
Matties: That would be interesting to see.
Morgan: It would be a big number, I'm sure.
Matties: Yeah, because you have 51 members currently. It looks like you're maintaining, but are you growing?
Morgan: It goes up and it goes down. We'll probably end this year with a few more than that. There's two or three in the pipeline right now. Some members join the group and then they may come for a single project. We do lose some members by attrition from companies merging, for example.
Matties: When they become a member, what's the expectation of a member? My thought is, can I come in, pay my dues and just collect the reward?
Morgan: You could come and do what you like. As all of the organizations, the more you put in the more you get out. There's no doubt about that.
Matties: But all the results are shared with members.
Morgan: They're all shared, yes, but then the discussions are very important as well. You can get the reports as a member of course, but if you want to really learn what's going on, you need to join the meeting and hear the analysis of those results. Again, the members who we find get the most benefit are the ones who participate the most; they join the most projects.
Matties: I'm thinking there would be companies out there that may want to be a part of this, but they don't have the time or the resource, but they still want to contribute at least financially and share in some of the research.
Morgan: We don't have many like that to be honest. No, very few. Most people who join, join with the intention of being active on at least some projects. We're very broad. We cover a huge range of technologies. For sure, you wouldn't be involved in every single project, nobody would. It wouldn't fit. But mostly you'll find members have two or three projects that they'll really hit. We have a couple of CAF projects that we just started in the last few weeks. We've got a lot of interest in that, of course. So they're two similar projects, but not the same. One is a CAF test vehicle for materials, specifically trying to take away the process design variables. That's had a lot of interest from the material suppliers as you can imagine. And also, OEMs for screening so they can quickly screen materials, and the intention is to reduce the testing cycle.
The other one is actually developing a better equation for the CAF acceleration. In automotive, we have some issues around high voltage testing. We're testing now, instead of 100 volts, we're testing at over 1 kilovolt. That's difficult. It's very difficult to do the testing at those voltages. No one really knows what the relationship is between the high voltages and the low voltages. If we can actually develop a better equation, we could hopefully demonstrate that we could test at low voltages and still simulate behavior or extrapolate behavior at high voltages. That has traction. We have three automotive non-members actually already on that project involved. At the first stages, non-members can join in. At the idea and definition phase. Then when we go to implementation, they can decide whether to join us or not, because at that point the data becomes private. We'll see where that goes, but we're delighted to involve those guys. This is a real issue facing the industry. This testing is expensive. It's difficult and takes a long time. It's slowing up production and development cycles as well. If we can make that quicker in some way, that would be a huge advantage to the industry.
Matties: Absolutely. Do you cover from design to assembly?
Morgan: Everything, even environmental disposal analysis. We cover that as well. We cover the whole thing from cradle to grave, from design right through to disposal. End of life, the whole story.
Matties: Where is the majority of the work done?
Morgan: It's hard to say really. We have a lot on assembly. Projects on the use of new alloys also in harsh environments. A lot on materials; there was a lot on packaging in the early days. There are still a lot of projects running in that area. I wouldn't like to say where it is, and actually, it moves.
Matties: I'm just curious where it is today.
Morgan: Yeah. I am biased again. I've worked in the materials projects. I have five projects there.
Matties: I see so much, as we just saw in the conference today, around direct imaging, solder mask imaging, the inkjets, all these kinds of issues and choices that the industry has to make.
Morgan: Huge choices. For example, chemistry is a big area for us. I should say that. Certainly we've been looking at signal integrity; that's a big topic that covers a lot of things from design. We of course do give inputs to design authorities and standards bodies as well, and many of our projects have a goal to actually publish industry guidelines. I'm doing a project now which is on component rework reliability. So large BGAs, if those fail, they have to be reworked. What's the inference on the rest of the board and the life of that? You take the big BGA off, what does it do to the board? Does it make the board less reliable? How many times can you do that? No one has really quantified it, so we plan to quantify that and we plan to issue design guidelines that say, "OK, this kind of device, this many rework cycles, this temperature are reliable. Beyond that, there's an issue." And try to show that with real data. So that's the kind of thing that we do.
That way we think that we help the whole industry, because we help with design guidance and also with standards. Many of our projects do write straight into standards, at IPC for example.
Matties: So when the fees come in, how is the money managed? Where is that being utilized?
Morgan: Basically we have a small staff. There's a general manager. There's an executive director, administrator, and then project facilitators. It's lean, and the cash goes to running the organization—for the admin basically. And for costs for projects, because we have to sometimes buy materials and components. We also have to organize events, so we have to hire rooms and that kind of thing. It's a pretty low-cost organization—pretty lean. Most of the money that we generate is used for the members’ benefit. Let's put it that way.
Matties: Because a lot of these tests would require some substantial investment, I would think.
Morgan: Many of them are donated by the members, so we're very grateful to the membership for offering their facilities. Occasionally, we have to buy services, but usually members will offer them, because they'll learn from these projects. Sometimes test houses want to show their wares. They want to show what they can do. They want to participate in the project to show the members that they're in the game, that they can do this stuff. Then they will get additional business.
Matties: Well, are there any thoughts that we haven't talked about that you want to share with the industry?
Morgan: Let me say, go to the website HDPUG.org and have a look. See what's on there and look at projects that you're interested in. You'll see which ones are public. You can access them. Get on the call and see what's interesting for you. The more inputs the better for us. You don't have to be a member at this early stage of the projects. Later, who knows, maybe people will join or not join. It's entirely up to them, but I think the more people who share what we're doing the better.
Matties: It seems to me, the more people that talk, the better we are as an industry. To get back to a word you used earlier: showcases. It showcases all the different aspects of these things. Well, Alun, thank you so much.
Morgan: My pleasure Barry.