BGA reflow profile

We are about to run our first BGA devices and would appreciate some confirmation on a few points. First some details: PCB & Device: ============= 1) 16 layer (4 power, 4 gnd, 8 signal) 2) gold on nickel finish 3) ceramic BGA with eutectic balls Reflow equipment: ================= 1) 8 heating, 2 cooling zone DIMA hot air convection reflow oven (400mm wire-mesh belt width) From all I have read, I understand the following: ================================================= 1) I should paste the board - I have a .18mm (.007") stencil from which I get a uniform paste deposit (inspected under microscope) 2) Oven profile?? (help please??) - I understand that the board profile should be a standard one such as prescribed by the solder paste manufacturer. - is this correct. Some pointers would be really appreciated, as we are placing a $3000.00 device. I have some mechanical samples that I am able to experiment with. Thanks again. Robert Steltman

reply »

There is a procedure for profiling on my web site that may be of use to start with www.bobwillis.co.uk | We are about to run our first BGA devices and would appreciate some confirmation on a few points. | First some details: | PCB & Device: | ============= | 1) 16 layer (4 power, 4 gnd, 8 signal) | 2) gold on nickel finish | 3) ceramic BGA with eutectic balls | Reflow equipment: | ================= | 1) 8 heating, 2 cooling zone DIMA hot air convection reflow oven (400mm wire-mesh belt width) | From all I have read, I understand the following: | ================================================= | 1) I should paste the board - I have a .18mm (.007") stencil from which I get a uniform paste deposit (inspected under microscope) | 2) Oven profile?? (help please??) - I understand that the board profile should be a standard one such as prescribed by the solder paste manufacturer. - is this correct. | Some pointers would be really appreciated, as we are placing a $3000.00 device. I have some mechanical samples that I am able to experiment with. | Thanks again. | Robert Steltman

reply »

Here is what I would do faced with your problem. First I would profile a similar board or use an existing profile. From there I would place the dummy device on the surface of a sample board with a thin thermocouple soldered to the board surface under the BGA. The BGA must sit flat on the surface and run a couple of profiles for this test and make ajustments. This is not the normal or the proper way but it will work to get you 90% to your goal. When you are happy with the profile then move forward with the fully pasted board all components and the real BGA. The ceramic BGA will heat sink that area of the board by 15-20C. A normal PBGA will give a difference of 10C from the surface PCB temperature. As I said with the first comments the profile methods on the web site and there is a new CD ROM on profiling and reflow soldering and defects due out soon from ECD in the USA. | We are about to run our first BGA devices and would appreciate some confirmation on a few points. | First some details: | PCB & Device: | ============= | 1) 16 layer (4 power, 4 gnd, 8 signal) | 2) gold on nickel finish | 3) ceramic BGA with eutectic balls | Reflow equipment: | ================= | 1) 8 heating, 2 cooling zone DIMA hot air convection reflow oven (400mm wire-mesh belt width) | From all I have read, I understand the following: | ================================================= | 1) I should paste the board - I have a .18mm (.007") stencil from which I get a uniform paste deposit (inspected under microscope) | 2) Oven profile?? (help please??) - I understand that the board profile should be a standard one such as prescribed by the solder paste manufacturer. - is this correct. | Some pointers would be really appreciated, as we are placing a $3000.00 device. I have some mechanical samples that I am able to experiment with. | Thanks again. | Robert Steltman

reply »

| We are about to run our first BGA devices and would appreciate some confirmation on a few points. | First some details: | PCB & Device: | ============= | 1) 16 layer (4 power, 4 gnd, 8 signal) | 2) gold on nickel finish | 3) ceramic BGA with eutectic balls | Reflow equipment: | ================= | 1) 8 heating, 2 cooling zone DIMA hot air convection reflow oven (400mm wire-mesh belt width) | From all I have read, I understand the following: | ================================================= | 1) I should paste the board - I have a .18mm (.007") stencil from which I get a uniform paste deposit (inspected under microscope) | 2) Oven profile?? (help please??) - I understand that the board profile should be a standard one such as prescribed by the solder paste manufacturer. - is this correct. | Some pointers would be really appreciated, as we are placing a $3000.00 device. I have some mechanical samples that I am able to experiment with. | Thanks again. | Robert Steltman Robert, You've got a tough one. I need a little more info first. Are you running a no-clean or a cleaning process?(NC vs. RMA or H2O Soluble) Is the BGA near other components or is it fairly isolated? What type of thermal profiler do you have? First of all, you have a very heavy / thick assembly so you are going to need increased temps or a slower belt speed (the paste is going to dictate belt speed eg. how long in soak and when do the nasties go volatile?) The next thing is the surface finish. You've got gold over nickel which is going to further jump up the temperatures. Bottom line, don't be suprised by a really "hot looking" profile. Bob's got the method on profiling that device. You need to get under it with a thermocouple. You can either go right under the part, if you have the appropriate wire or you can drill through the bottom of the board. Personally, when I have destructive samples, I hog out a channel for the wire to sit in. Unless you own an "x-ray drill" (who's got that one) you can't really see where you're going to come through. Well, you can use the vias as a guide but I'm still partial to the channel approach. Just do it with a dremmel or something similar so the channel is roughly the same diameter as the TC wire. Another thing you're going to have to watch for is frying the other components. Some components may only be able to see 220C. These guys are tough to deal with and nearly impossible on your assembly without turning the BGA insertion into a secondary process. Regards, Justin Medernach Flextronics International

reply »

| | We are about to run our first BGA devices and would appreciate some confirmation on a few points. | | First some details: | | PCB & Device: | | ============= | | 1) 16 layer (4 power, 4 gnd, 8 signal) | | 2) gold on nickel finish | | 3) ceramic BGA with eutectic balls | | Reflow equipment: | | ================= | | 1) 8 heating, 2 cooling zone DIMA hot air convection reflow oven (400mm wire-mesh belt width) | | From all I have read, I understand the following: | | ================================================= | | 1) I should paste the board - I have a .18mm (.007") stencil from which I get a uniform paste deposit (inspected under microscope) | | 2) Oven profile?? (help please??) - I understand that the board profile should be a standard one such as prescribed by the solder paste manufacturer. - is this correct. | | Some pointers would be really appreciated, as we are placing a $3000.00 device. I have some mechanical samples that I am able to experiment with. | | Thanks again. | | Robert Steltman | Robert, | You've got a tough one. I need a little more info first. Are you running a no-clean or a cleaning process?(NC vs. RMA or H2O Soluble) Is the BGA near other components or is it fairly isolated? What type of thermal profiler do you have? First of all, you have a very heavy / thick assembly so you are going to need increased temps or a slower belt speed (the paste is going to dictate belt speed eg. how long in soak and when do the nasties go volatile?) The next thing is the surface finish. You've got gold over nickel which is going to further jump up the temperatures. Bottom line, don't be suprised by a really "hot looking" profile. Bob's got the method on profiling that device. You need to get under it with a thermocouple. You can either go right under the part, if you have the appropriate wire or you can drill through the bottom of the board. Personally, when I have destructive samples, I hog out a channel for the wire to sit in. Unless you own an "x-ray drill" (who's got that one) you can't | really see where you're going to come through. Well, you can use the vias as a guide but I'm still partial to the channel approach. Just do it with a dremmel or something similar so the channel is roughly the same diameter as the TC wire. Another thing you're going to have to watch for is frying the other components. Some components may only be able to see 220C. These guys are tough to deal with and nearly impossible on your assembly without turning the BGA insertion into a secondary process. | Regards, | Justin Medernach | Flextronics International Hi Justin Thanks for your reply. OK, I have made some discoveries since I posted my first query. I made a mistake, the device is plastic with a metal top and eutectic balls, not ceramic as I first thought. The PCB has been manufactured using an electroless gold process. I will be using a water soluble paste from Amtech - WS-488. I have the recommended profile for this paste and will attempt to mimic it. I have a "dead" board that was destroyed by another manufacturer and have used this to run a few tests. The oven I have allows for 2 TC's to be coupled to the PCB and then profiled while capturing the profile on the PC. I have watched Bob's video on BGA and followed his suggestion on placing the 1 TC under the device and the other next to it. I am quite surprised to see how low the temps really are with this PCB. Your predictions are spot on. I will try bump up the settings tomorrow. There is a mix of PLCC, QFP, SOIC and PLCC sockets around the BGA. As you say I will have to be very careful with the temperatures. Luckily this is a very low volume product and we can afford to put quite a bit of manual work into it. Again thanks for all the help - I look forward to your reply. Robert Steltman elprom@stp.co.za robls@mweb.co.za

reply »

| Here is what I would do faced with your problem. First I would profile a similar board or use an existing profile. From there I would place the dummy device on the surface of a sample board with a thin thermocouple soldered to the board surface under the BGA. The BGA must sit flat on the surface and run a couple of profiles for this test and make ajustments. This is not the normal or the proper way but it will work to get you 90% to your goal. | When you are happy with the profile then move forward with the fully pasted board all components and the real BGA. | The ceramic BGA will heat sink that area of the board by 15-20C. A normal PBGA will give a difference of 10C from the surface PCB temperature. As I said with the first comments the profile methods on the web site and there is a new CD ROM on profiling and reflow soldering and defects due out soon from ECD in the USA.

Hi Bob Thanks for the reply. I have watched your video and found it most informative. I will do some more tests and let you know the results. I will check out your site as well. Thanks again for all the help. Robert Steltman elprom@stp.co.za robls@mweb.co.za

reply »