Printed Circuit Board Assembly & PCB Design Forum

SMT electronics assembly manufacturing forum.

SMTnet
»
Electronics Forum
»
0.031 pcb bowing -- solutions?

0.031 pcb bowing -- solutions?

Subscribe to Thread

We have 1/2 dozens 0.031 pcbs that bow after going through t... - Aug 27, 2012 by Reese

Its no my turn to give some advise. My first question would ... - Aug 27, 2012 by cobhambill

Thank you for your reply. > Its no my turn to give som... - Aug 27, 2012 by Reese

I have run 0.031 boards. One of the things you may want to l... - Aug 27, 2012 by cobhambill

We will look at the oven speed. We considered stiffeners bu... - Aug 27, 2012 by Reese

Ran them with stiffners and I worried about moving parts als... - Aug 27, 2012 by cobhambill

We agree with all of the previous comments, and would add th... - Aug 28, 2012 by robgd3

Thank you for your reply. We will try your suggestions and ... - Aug 28, 2012 by Reese

Reese

#67123

0.031 pcb bowing -- solutions? | 27 August, 2012

We have 1/2 dozens 0.031 pcbs that bow after going through the oven. We are looking for a solution to this problem to prevent bowing of the pcbs. As a result of the bowing, the pcbs, on occasion get caught in the oven, and more often have transfer issues from one conveyor/machine to the next. We use edge belt conveyors throughout.

It has been suggested of using a copper layer,internally on the end substrates. Here is the breakdown of what we have:

Board 1 -- 950 mil end substrate, unscored, no copper layer Board 2 -- 500 mil end substrate, unscored, no copper layer Board 3 -- 500 mil end substrate, unscored, top-side copper layer Board 4 -- 456 mil end substrate, unscored, top-side copper layer Board 5 -- no substrate Board 6 -- 350 mil end substrate, scored, disontinuous top-side copper layer

Board 5 and 6 can be alleviated and some of the changes are obvious; however, the problem persists, even with a copper layer on some of the boards.

Using a center belt conveyor is not feasible since this only treats the symptom and will not effect the transferability into machines that cannot be equipped with a center belt. I would prefer to fix the heart of the problem, the bowing, rather than treat it.

Any suggestions into a possible solution for this problem are welcome. But a solution must be one that is incorporated at the board house, not here are our company. I cannot entertain the idea of adding handling costs to the product.

Has anyone dealt with this before and found a viable solution? Perhaps copper on both sides and an internal layer would do the trick? Thanks in advance.

reply »

cobhambill

#67124

0.031 pcb bowing -- solutions? | 27 August, 2012

Its no my turn to give some advise. My first question would be can you put the boards into a carrier of some sort. There are many wonderful companies that make carriers that are ment for running down a smt line. This can be expensive unless you order more then 5 pallets then the price come down greatly. The next question would be how many zones is your oven and are you running leaded or lead free. If you are running a 5 zone oven with lead free you face some problems. Reason being you are probably hitting the board with to much heat to quickly. You need to do a oven profile and see if you are following your solder paste recommendations. If you are take a look at you peak temps if they are lead you probably dont need to be above 215 if its lead free you need to reveiw the parts on the board and look for the most sensitive part because that will dictate how hot you get. I would recommend not going above 225 because you dont need it for the thickness of PCB you are running. Look at the boards are there weird cut outs that are causing your conveyor sensors to not read if so can you adjust the sensors to a new location or can you panelize the boards to you get you better material area for the sensors to catch and not cause you problems. So can you run in carriers? WHats your profiels indicate? Are you following the solder paste recommendations? Can you panelize the boards?

reply »

Reese

#67125

0.031 pcb bowing -- solutions? | 27 August, 2012

Thank you for your reply.

> Its no my turn to give some advise. My first > question would be can you put the boards into a > carrier of some sort. There are many wonderful > companies that make carriers that are ment for > running down a smt line. This can be expensive > unless you order more then 5 pallets then the > price come down greatly.

We have used carriers in the past and I don't know if this is an option. I'll check with our SMT fellow.

>The next question would > be how many zones is your oven and are you > running leaded or lead free. If you are running a > 5 zone oven with lead free you face some > problems. Reason being you are probably hitting > the board with to much heat to quickly. You need > to do a oven profile and see if you are following > your solder paste recommendations. If you are > take a look at you peak temps if they are lead > you probably dont need to be above 215 if its > lead free you need to reveiw the parts on the > board and look for the most sensitive part > because that will dictate how hot you get. I > would recommend not going above 225 because you > dont need it for the thickness of PCB you are > running.

Yes, we have oven profiles for the boards in question and our temperatures should be correct. Our oven has 14 zones (12 heating and 2 cooling zones).

>Look at the boards are there weird cut > outs that are causing your conveyor sensors to > not read if so can you adjust the sensors to a > new location or can you panelize the boards to > you get you better material area for the sensors > to catch and not cause you problems. So can you > run in carriers? WHats your profiels > indicate? Are you following the solder paste > recommendations? Can you panelize the boards?

Boards are panelized. The profile recommendations are are being followed, as far as I know. we just don't have enough rigidity on the arrays. Have you run any 0.031 pcbs? Have you had similar issues? I will double check the oven profiles to be sure.

reply »

cobhambill

#67126

0.031 pcb bowing -- solutions? | 27 August, 2012

I have run 0.031 boards. One of the things you may want to look at is the speed of the oven. With 14 zones you may be going to slow with the board which could be the reason your boards are starting to bow. You might also think about adding stiffners before you run then throught the reflow process. These are something that are fairly inexpensive and can be added quickly by a operator. Normally stiffners are made of heavy duty fiberglass or metal. Either one works great they allow you to run the current recipe and you strengthen the panel which helps reduce the bowing you are seeing. One other thing you might what to check into could you trun the panel 90degress at the beginning of the process and run the product that way. Its an option if you don't have 0402 or smaller parts if you do I probably would pursue another avenue because you might get bigger problems with tombstoning.

reply »

Reese

#67128

0.031 pcb bowing -- solutions? | 27 August, 2012

We will look at the oven speed. We considered stiffeners but opted against that due to the risk of moving parts pre-reflow due to handling. 90 deg? I don't know. This may or may not help with the bowing. Considering there is little to no substrate on the north/south borders, it will probably be worse, would be my guess.

So you have run 0.031 pcbs. You have not had any issues with them?

reply »

cobhambill

#67132

0.031 pcb bowing -- solutions? | 27 August, 2012

Ran them with stiffners and I worried about moving parts also but had the stiffners cut so that they could only go on so far which made it ok because The distance from the stiffners and parts was 1/16 of an inch no problem with movement.

reply »

robgd3

#67136

0.031 pcb bowing -- solutions? | 28 August, 2012

We agree with all of the previous comments, and would add the following:

Warpage is generally caused by excess moisture in the raw boards. Have you tried baking the boards prior to assembly?

We had this issue on a simple 4-layer board, and found two causes: 1. Poor construction/poor curing of the pre-preg during pcb fabrication; 2. Poor layout of the power/ground planes internally. I'm not a fab expert, but, as I was told in that instance, the nature of the copper pours on the inner layers (and something that wasn't done) caused uneven heating during reflow, which caused the warpage.

I'd recommend asking your fab house to analyze the construction of the board to see if they think there's is a way to address the warpage at that level.

And, turning the boards 90 degrees may well help you. It'll change the order of the preheating and thermal stress on the PCB during reflow.

cheers ..rob

reply »

Reese

#67137

0.031 pcb bowing -- solutions? | 28 August, 2012

Thank you for your reply. We will try your suggestions and keep you posted.

reply »