BGA Voiding Problem

We are just starting to manufacture a PCB with a .039" fine-line PBGA. We are currently seeing voids, mainly in the center, that are as much as 45%. We also intermittently see the corner balls of the PBGA smaller than the center balls. We are using a 6-mil stencil with a 1:1 aperture size (.019" mil). We are using a convection oven and reflowing in air. The solder paste is an OA 62SN36PB02AG and the reflow profile is: Preheat - 0-150 degrees C in 90 sec Soak - 150-189 degrees C for 120 sec Reflow - 189-210 degrees C for 60 sec Thermocouples were attached to both the corner and center of the PBGA and they both had the exact same profile with a max temp of 210 Degrees C. Does anyone have any recommendations for minimizing the amount of voiding? Is 45% voiding acceptable? What would explain the corner balls being smaller than the center balls?

Thank you for your input.

reply »

| We are just starting to manufacture a PCB with a .039" fine-line PBGA. We are currently seeing voids, mainly in the center, that are as much as 45%. We also intermittently see the corner balls of the PBGA smaller than the center balls. We are using a 6-mil stencil with a 1:1 aperture size (.019" mil). We are using a convection oven and reflowing in air. The solder paste is an OA 62SN36PB02AG and the reflow profile is: | Preheat - 0-150 degrees C in 90 sec | Soak - 150-189 degrees C for 120 sec | Reflow - 189-210 degrees C for 60 sec | Thermocouples were attached to both the corner and center of the PBGA and they both had the exact same profile with a max temp of 210 Degrees C. | | Does anyone have any recommendations for minimizing the amount of voiding? Is 45% voiding acceptable? What would explain the corner balls being smaller than the center balls? | | Thank you for your input. | Seems a bit too much voiding. We run traditional profiles for our paste types (3) in both atmosperic conditions and with nitrogen only. We don't exceed the 183 degree C. window more than 100 seconds, nor the 205 degree C. window for more than 45 seconds (our paste types). Maximum liquidous, of course, does not exceed 219.

I would only do what the paste guys tell you unless, of course, it doesn't work. Then you run the experiment. Of course, it doesn't hurt to have the "right" reflow equipment. I know you do and proof it performs as secified.

Earl Moon

reply »

| | We are just starting to manufacture a PCB with a .039" fine-line PBGA. We are currently seeing voids, mainly in the center, that are as much as 45%. We also intermittently see the corner balls of the PBGA smaller than the center balls. We are using a 6-mil stencil with a 1:1 aperture size (.019" mil). We are using a convection oven and reflowing in air. The solder paste is an OA 62SN36PB02AG and the reflow profile is: | | Preheat - 0-150 degrees C in 90 sec | | Soak - 150-189 degrees C for 120 sec | | Reflow - 189-210 degrees C for 60 sec | | Thermocouples were attached to both the corner and center of the PBGA and they both had the exact same profile with a max temp of 210 Degrees C. | | | | Does anyone have any recommendations for minimizing the amount of voiding? Is 45% voiding acceptable? What would explain the corner balls being smaller than the center balls? | | | | Thank you for your input. | | | Seems a bit too much voiding. We run traditional profiles for our paste types (3) in both atmosperic conditions and with nitrogen only. We don't exceed the 183 degree C. window more than 100 seconds, nor the 205 degree C. window for more than 45 seconds (our paste types). Maximum liquidous, of course, does not exceed 219. | | I would only do what the paste guys tell you unless, of course, it doesn't work. Then you run the experiment. Of course, it doesn't hurt to have the "right" reflow equipment. I know you do and proof it performs as secified. | | Earl Moon | Let me clarify - though I'm not very good at this. With our system, we do not exceed 100 seconds in the 183 window. We do not exceed 45 seconds in the 205 window. Also, we don't use 1:1 on anything. For 20 mil diameter pads, we go to about 17 mils though the only difference this makes is to minimize bridging - not voiding.

Earl Moon

reply »

| | | We are just starting to manufacture a PCB with a .039" fine-line PBGA. We are currently seeing voids, mainly in the center, that are as much as 45%. We also intermittently see the corner balls of the PBGA smaller than the center balls. We are using a 6-mil stencil with a 1:1 aperture size (.019" mil). We are using a convection oven and reflowing in air. The solder paste is an OA 62SN36PB02AG and the reflow profile is: | | | Preheat - 0-150 degrees C in 90 sec | | | Soak - 150-189 degrees C for 120 sec | | | Reflow - 189-210 degrees C for 60 sec | | | Thermocouples were attached to both the corner and center of the PBGA and they both had the exact same profile with a max temp of 210 Degrees C. | | | | | | Does anyone have any recommendations for minimizing the amount of voiding? Is 45% voiding acceptable? What would explain the corner balls being smaller than the center balls? | | | | | | Thank you for your input. | | | | | Seems a bit too much voiding. We run traditional profiles for our paste types (3) in both atmosperic conditions and with nitrogen only. We don't exceed the 183 degree C. window more than 100 seconds, nor the 205 degree C. window for more than 45 seconds (our paste types). Maximum liquidous, of course, does not exceed 219. | | | | I would only do what the paste guys tell you unless, of course, it doesn't work. Then you run the experiment. Of course, it doesn't hurt to have the "right" reflow equipment. I know you do and proof it performs as secified. | | | | Earl Moon | | | Let me clarify - though I'm not very good at this. With our system, we do not exceed 100 seconds in the 183 window. We do not exceed 45 seconds in the 205 window. Also, we don't use 1:1 on anything. For 20 mil diameter pads, we go to about 17 mils though the only difference this makes is to minimize bridging - not voiding. | | Earl Moon | Why am I doing this on a Sunday? Why am I doing this at all? I don't go to church and there's no senorita besid me in bed, so what's going on? Ah, too much cerveza last night!

Looking again at your profile it seems to be close to what I said, or not. You decide. However, upon further thought processing (I flatter myself), consider oxidation or other solder termination contamination. Damn, I almost forgot about that stuff though went through a recent bout of HASL crap and people touching my private areas (board surfaces) with their dirtly little hands.

Just some more stuff, but remember I ain't all here or there,

Moonman

reply »

| | | | We are just starting to manufacture a PCB with a .039" fine-line PBGA. We are currently seeing voids, mainly in the center, that are as much as 45%. We also intermittently see the corner balls of the PBGA smaller than the center balls. We are using a 6-mil stencil with a 1:1 aperture size (.019" mil). We are using a convection oven and reflowing in air. The solder paste is an OA 62SN36PB02AG and the reflow profile is: | | | | Preheat - 0-150 degrees C in 90 sec | | | | Soak - 150-189 degrees C for 120 sec | | | | Reflow - 189-210 degrees C for 60 sec | | | | Thermocouples were attached to both the corner and center of the PBGA and they both had the exact same profile with a max temp of 210 Degrees C. | | | | | | | | Does anyone have any recommendations for minimizing the amount of voiding? Is 45% voiding acceptable? What would explain the corner balls being smaller than the center balls? | | | | | | | | Thank you for your input. | | | | | | | Seems a bit too much voiding. We run traditional profiles for our paste types (3) in both atmosperic conditions and with nitrogen only. We don't exceed the 183 degree C. window more than 100 seconds, nor the 205 degree C. window for more than 45 seconds (our paste types). Maximum liquidous, of course, does not exceed 219. | | | | | | I would only do what the paste guys tell you unless, of course, it doesn't work. Then you run the experiment. Of course, it doesn't hurt to have the "right" reflow equipment. I know you do and proof it performs as secified. | | | | | | Earl Moon | | | | | Let me clarify - though I'm not very good at this. With our system, we do not exceed 100 seconds in the 183 window. We do not exceed 45 seconds in the 205 window. Also, we don't use 1:1 on anything. For 20 mil diameter pads, we go to about 17 mils though the only difference this makes is to minimize bridging - not voiding. | | | | Earl Moon | | | Why am I doing this on a Sunday? Why am I doing this at all? I don't go to church and there's no senorita besid me in bed, so what's going on? Ah, too much cerveza last night! | | Looking again at your profile it seems to be close to what I said, or not. You decide. However, upon further thought processing (I flatter myself), consider oxidation or other solder termination contamination. Damn, I almost forgot about that stuff though went through a recent bout of HASL crap and people touching my private areas (board surfaces) with their dirtly little hands. | | Just some more stuff, but remember I ain't all here or there, | | Moonman |

Hey,

Looking at the problem i guess the solder paste and the composition of the paste might be worth investigating. From the little studies we have done we have kind of analysed the reflow profile, the solder paste composition( flux, flux content, solvent) and the solderability of the surface to be critical. The reflow profile u have mentioned seems to be ok in a generic sense but i guess the reflow profile specified by the paste manufacturer might be important.

In my studies and experiments i have seen huge voids ( 50%, 60%)and the reliability testing for these are in progress and it would be intresting to see the results.

I hope the information is usefull to me.

Thanx

yunus

reply »