Printed Circuit Board Assembly & PCB Design SMT Electronics Assembly Manufacturing Forum

Printed Circuit Board Assembly & PCB Design Forum

SMT electronics assembly manufacturing forum.


PCB panellisation

Jacqueline Coia

#16354

PCB panellisation | 3 April, 1998

We are in the process of deciding whether to change a PCB product from a single board to a 3 up panel array. Could you please advise on the most effective removal design, either scoring or snap off, taking into account that material utilisation is important to the decision. Note that these PCBs are to be leaded populated therefore require flow solder. Advice would be appreciated.

reply »

Scott Cook

#16363

Re: PCB panellisation | 3 April, 1998

| We are in the process of deciding whether to change a PCB product from a single board to a 3 up panel array. Could you please advise on the most effective removal design, either scoring or snap off, taking into account that material utilisation is important to the decision. Note that these PCBs are to be leaded populated therefore require flow solder. | Advice would be appreciated. Howdy. Here's my 2 cents, for what it's worth..... Assumption: You will manually depanelize after wave. If this is not true, my scoring issues may go away. 1st, My opinion about scoring on PCA's with SMT is DO NOT DO IT! My reasoning follows: 1) Scoring is a secondary operation, which typically costs extra pennies. I'm a believer in routed panelization, with drilled breakaway tabs. The PCB already gets drilled and routed, so there is effectively no extra charge for this scenario. For a drawing of my recommendation regarding tab configuration e-mail me privately, and give me your fax number. 2) The scoring equipment is basically two adjustable cutting wheels. The kerf alignment of the cuts top to bottom on the PCB are inaccurate to say the least. Just ask your board house how tight they will hold the variance of the cut in a vertical plane, top to bottom....... What happens is that your "1/3" or "1/2" or "2/3" of the thickness of the PCB cut depth becomes a diagonal stress rather than a perfectly vertical stress on the board, due to the tolerance stackup of the variation of cuts. This becomes more and more dangerous for solder joint integrity, particularly with 0603's and 0402's. You'll pop 'em right off the boards! Scott Cook scottc@t-com.com

reply »

Ron Beasley

#16364

Re: PCB panellisation | 3 April, 1998

| | | We are in the process of deciding whether to change a PCB product from a single board to a 3 up panel array. Could you please advise on the most effective removal design, either scoring or snap off, taking into account that material utilisation is important to the decision. Note that these PCBs are to be leaded populated therefore require flow solder. | | Advice would be appreciated. | Howdy. | Here's my 2 cents, for what it's worth..... | Assumption: | You will manually depanelize after wave. If this is not true, my scoring issues may go away. | 1st, My opinion about scoring on PCA's with SMT is DO NOT DO IT! My reasoning follows: | 1) Scoring is a secondary operation, which typically costs extra pennies. I'm a believer in routed panelization, with drilled breakaway tabs. The PCB already gets drilled and routed, so there is effectively no extra charge for this scenario. For a drawing of my recommendation regarding tab configuration e-mail me privately, and give me your fax number. | 2) The scoring equipment is basically two adjustable cutting wheels. The kerf alignment of the cuts top to bottom on the PCB are inaccurate to say the least. Just ask your board house how tight they will hold the variance of the cut in a vertical plane, top to bottom....... | What happens is that your "1/3" or "1/2" or "2/3" of the thickness of the PCB cut depth becomes a diagonal stress rather than a perfectly vertical stress on the board, due to the tolerance stackup of the variation of cuts. This becomes more and more dangerous for solder joint integrity, particularly with 0603's and 0402's. You'll pop 'em right off the boards! | Scott Cook | scottc@t-com.com I agree with Scott. In addition if the boards are large and the are scored or "V groved" there may be some sag in the reflow oven. This impacts bottom side placement, and I have seen sag so bad that the PCB's will not even work in the bottomside machines. In additon if the sag is still present at wave solder it will impact solder quality and you run the risk of flooding the board. I am not so sure about the cost issue. The additional routing and drilling may be more exspensive than "V grove"

reply »

Chrys

#16357

Re: PCB panellisation | 3 April, 1998

I'm also in the "Don't score it" camp. Routed breakaways are much better. If your wave solder process has decent control on the wave height, you shouldn't worry about flooding the boards. The fine pitch consideration for panelizing - Since you normally want to keep anything under 20 mils pitch clustered close together to prevent misregistered screen prints, if you have fine pitch stuff on this board and panelize 3-up, you're spreading it out and risking registration problems.

reply »

Earl Moon

#16365

Re: PCB panellisation | 5 April, 1998

| | | | | We are in the process of deciding whether to change a PCB product from a single board to a 3 up panel array. Could you please advise on the most effective removal design, either scoring or snap off, taking into account that material utilisation is important to the decision. Note that these PCBs are to be leaded populated therefore require flow solder. | | | Advice would be appreciated. | | Howdy. | | Here's my 2 cents, for what it's worth..... | | Assumption: | | You will manually depanelize after wave. If this is not true, my scoring issues may go away. | | 1st, My opinion about scoring on PCA's with SMT is DO NOT DO IT! My reasoning follows: | | 1) Scoring is a secondary operation, which typically costs extra pennies. I'm a believer in routed panelization, with drilled breakaway tabs. The PCB already gets drilled and routed, so there is effectively no extra charge for this scenario. For a drawing of my recommendation regarding tab configuration e-mail me privately, and give me your fax number. | | 2) The scoring equipment is basically two adjustable cutting wheels. The kerf alignment of the cuts top to bottom on the PCB are inaccurate to say the least. Just ask your board house how tight they will hold the variance of the cut in a vertical plane, top to bottom....... | | What happens is that your "1/3" or "1/2" or "2/3" of the thickness of the PCB cut depth becomes a diagonal stress rather than a perfectly vertical stress on the board, due to the tolerance stackup of the variation of cuts. This becomes more and more dangerous for solder joint integrity, particularly with 0603's and 0402's. You'll pop 'em right off the boards! | | Scott Cook | | scottc@t-com.com | I agree with Scott. In addition if the boards are large and the are scored or "V groved" there may be some sag in the reflow oven. This impacts bottom side placement, and I have seen sag so bad that the PCB's will not even work in the bottomside machines. In additon if the sag is still present at wave solder it will impact solder quality and you run the risk of flooding the board. | I am not so sure about the cost issue. The additional routing and drilling may be more exspensive than "V grove" Forget cost. Cost easily is programmed into the panel when a well managed PCB fabrication facility is chosen. One other problem that may be incurred is that of edge deformity or delamination beyond that which is acceptable before or after thermal stress or shock during or after soldering operations.

reply »

Michael Allen

#16358

Re: PCB panellisation | 7 April, 1998

| I'm also in the "Don't score it" camp. Routed breakaways are much better. If your wave solder process has decent control on the wave height, you shouldn't worry about flooding the boards. | The fine pitch consideration for panelizing - Since you normally want to keep anything under 20 mils pitch clustered close together to prevent misregistered screen prints, if you have fine pitch stuff on this board and panelize 3-up, you're spreading it out and risking registration problems. Here's a dissenting opinion: scoring works great in our factory! As you're probably aware, scoring provides a big advantage (compared to "breakaway" tabs) by eliminating the masking process that's usually required to cover the routed areas to prevent flooding during wave soldering. Chrys's comment about fine pitch is a very good one. However, I believe his/her statement about about flooding is only true in special situations where (a) there's no sag/warpage, AND (b) you don't require a turbulent wave. Those conditions don't exist in our application. Our boards are large enough that there's usually some amount of sag (which can be mitigated with stiffeners), and our component mix often requires the turbulent wave. With regard to alignment of score grooves: this has not been a problem for us since we started specifying scored edges 2+ years ago. It might help that we specify a fairly shallow score (0.024" web thickness for an 0.062" board, ie, mat'l remaining between top and bottom grooves) in order to leave the panel as rigid as possible. Using shallow scores means that it's very difficult to snap the board out by hand, but we try to avoid doing that anyway.

reply »

Michael Allen

#16359

Re: PCB panellisation | 7 April, 1998

once more, with my correct e-mail address

reply »

Eldon Sanders

#16360

Re: PCB panellisation | 7 April, 1998

I agree with Michael, scoring works great. We build both and medium size small boards, anywhere from 2 to 100 square inches. Thickness from .030 to .062 Panels contain anywhere from 2 to 16 boards. About half are odd shaped, with cutouts and notches. Almost all have been changed to scored panels or a combination of scoring and routing. The web is spec'ed at 1/3 the thickness of the board. About half are double sided, and fixtures are used to keep the panel flat during first side reflow. The change to scoring has saved countless hours of trimming tabs, and is much easier on the wrist. Most of the panels contain fine pitch, and the rigidity of the scored panel keeps it square, resulting in excellent registration during the printing of the one or both sides. I have never had a problem with the top/bottom score alignment or the web thickness. Nor have I had a problem with stress of snapping the boards apart causing any damage to components or solder joints. I have had damage to a small board using tabs where caps placed near the tab cracked when the board was cut apart. Before scoring, this panel was sent out to be routed apart. Bottom line, I will use tab's only when I can leave the rough edge on the board and assembly it into the product without problems. If the edge required trimming the tab smooth, then the extra cost of the scoring will be saved in not having a trimming operation. And I will always score a large fine pitch panel to gain the registration improvement in printing.

reply »

Steve Gregory

#16362

Re: PCB panellisation | 8 April, 1998

Well...I think we just about beat this one to death, but I'll go ahead and add a couple more points to "pro-score" camp...WITH the caveat that it HAS to be done right! Yes, I will say that I've had to deal with the wrong spec'd or crappy score jobs at one time or another in the past just like a lot of us probably have, and it ain't no fun! You know, a double sided board scored too deep, and the panel falls apart after you got the first side built...fun stuff huh?....or, not scored deep enough so when they're being depanelized, the darn thing delaminates so bad that traces get ripped-up.... that's just peachy ain't it? That's why you need to make sure your fab house can do it correctly. I used to work at a pretty large memory company not too long ago, and we had two, 4-head Kennard routers, and two dual-head machines. When we would get busy (the end of the month or end of the quarter), that was the biggest bottleneck we had in the factory...and we ran 3-shifts too. So we started scoring...and every panel that we converted over to being scored improved the overall cycle time of every other assembly along with shortening it's own...no more bottleneck at the routers because some of the load was being taken off from scoring, which meant boards didn't wait at the routers anymore. Practically every product we built was double sided SMT, we did a lot of PCMCIA flash cards and laptop modems too...all with .020" pitch, and never really had a problem once we got a good fab vendor who had the right scoring machines. There's one company I know of that really are "Guru's" when it comes to scoring, routing, and drilling...that's because that's all they do,...score, route, and drill. The company is called Onanon Inc., there's a gentleman named Dennis Johnson there that really knows his stuff when it comes to scoring. There's more than just a few companies that I know have their fabs drop shipped from their fab vendors to Onanon to get scored. I'll paste a link to their page at the bottom, you should check it out...it's really pretty cool! They've got a animation of one of their actual machines scoring a panel in the page...pretty slick! (but you gotta let it load for a little bit...) -Steve Gregory-

reply »

Scott Cook

#16361

Re: PCB panellisation | 8 April, 1998

| | I agree with Michael, scoring works great. interesting feedback. | The change to scoring has saved countless hours of trimming tabs, and is much easier on the wrist. As mine in scoring, your experience in a routed, drilled web tab breakaway is based upon what sounds like a poor breakaway design. Nor have I had a problem with stress of snapping the boards apart causing any damage to components or solder joints. I do have to take exception, here. What products are you building, and what is the field life expectancy of them? How do you KNOW you didn't harm the integrity of the solder joints? Did you fixture and stabilize the panel while you performed breakaway operations, focusing the stresses to the scored area? I have had damage to a small board using tabs where caps placed near the tab cracked when the board was cut apart. Once again, this sounds simply like a redesign issue for the webbing area location on the breakaway. | Bottom line, I will use tab's only when I can leave the rough edge on the board and assembly it into the product without problems. If the edge required trimming the tab smooth, then the extra cost of the scoring will be saved in not having a trimming operation. Very true; however I use a routed / drilled web design which when broken out of the panel does not require a smoothing or sanding operation. And I will always score a large fine pitch panel to gain the registration improvement in printing. What do you mean here? What registration differences are there between a scored panel and a routed / webbed panel? I'm not being a smarta__ here, I'm just curious where you are coming from in your thoughts. My take on this is that a step and repeat of artwork is a step and repeat; doesn't matter how you panelize. You'll get the same repeatability and accuracy regardless of whether it's scored or routed for the breakaway. This subject has generated great discussion. I've enjoyed it. Scott Cook scottc@t-com.com

reply »

Wayne Bracy

#16356

Re: PCB panellisation | 8 April, 1998

| We are in the process of deciding whether to change a PCB product from a single board to a 3 up panel array. Could you please advise on the most effective removal design, either scoring or snap off, taking into account that material utilisation is important to the decision. Note that these PCBs are to be leaded populated therefore require flow solder. | Advice would be appreciated. Jacquelyne: You might want to contact Cencorp regarding depaneling equipment Wayne

reply »

Wayne

#16355

Re: PCB panellisation | 13 April, 1998

You might want to call CTS and speak with Jim Welch about their depaneling system. His phone number is (630) 293-9470 Wayne

reply »

 Reflow System

soldering station