Solder FINES vs. Solder Balls

Can someone explain the difference between solder "fines" and solder balls? Is there an official specification where "fines" are defined? What is the spec for "fines"?

I consider "fines" individual unmelted metal spheres that are found in solder paste. It seems likely that when millions of 35micron balls are deposited (as in paste) some may not melt and/or not become a part of the solder joint. We use Kester R596, a water soluable flux-based paste and wash the boards immediately following reflow but we are seeing "fines". These "fines" are visible ONLY under a microscope and vary in quantity from 5 per 9"x14" assembly to 25 over the same area after washing. We have tried scrubbing to no avail.

I feel someone is making a issue of a non-issue. Ramp-up is about 2.5 degrees C per second in a BTU VIP 70 forced convection oven. All other solder issues are great and we have re-verified the profile and it is still within the paste specs.

Any help (official specs are important) we can get on this matter would be greatly appreciated.

Mark

reply »

| Can someone explain the difference between solder "fines" and solder balls? Is there an official specification where "fines" are defined? What is the spec for "fines"? | | I consider "fines" individual unmelted metal spheres that are found in solder paste. It seems likely that when millions of 35micron balls are deposited (as in paste) some may not melt and/or not become a part of the solder joint. We use Kester R596, a water soluable flux-based paste and wash the boards immediately following reflow but we are seeing "fines". These "fines" are visible ONLY under a microscope and vary in quantity from 5 per 9"x14" assembly to 25 over the same area after washing. We have tried scrubbing to no avail. | | I feel someone is making a issue of a non-issue. Ramp-up is about 2.5 degrees C per second in a BTU VIP 70 forced convection oven. All other solder issues are great and we have re-verified the profile and it is still within the paste specs. | | Any help (official specs are important) we can get on this matter would be greatly appreciated. | | Mark | Mark: Several points:

Solder Paste Specification. ANSI/J- STD-005, Requirements For Solder Pastes

Fines. In my parlance, fines are the powder of a powder manufacturing process, if that�s not too convoluted. But continuing with your definition of fines, our venerable ( or is it venerated??) friend Bob Willis (www.bobwillis.co.uk/) differentiates between reflowed solder ball sizes and has words to describe the types.

Raw Solder Powder Sizing. A common powder size is -200/+325 mesh. In this powder, 99% by weight will pass through a 200 (holes per square inch) mesh screen and less than 20% of the powder by weight will pass through a 325 mesh screen. From this, you can figure the upside potential for fines. Your supplier can give you greater detail on your paste and on raw solder fines. (At least what I call fines.)

Small Solder Balls Remaining After Washing. Mark, something is wrong here!!! You should be able to wash the solder balls from the boards. With limited information, it sounds like your solder mask is improperly cured, partially curing during reflow, creating solder balls, and bonding the balls to the board during this partial curing. Our recently departed (from the forum) friend, Earl Moon, made major contributions to a thread on mask curing recently. Check the archives.

Have your fab house check you boards for proper curing, if you don�t have MEK or some other method for checking mask cure. MEK is a hazardous material. I do not condone using it for testing boards. (Pfeu!!! I got that out my system.)

Good luck

Dave F

reply »

| | Can someone explain the difference between solder "fines" and solder balls? Is there an official specification where "fines" are defined? What is the spec for "fines"? | | | | I consider "fines" individual unmelted metal spheres that are found in solder paste. It seems likely that when millions of 35micron balls are deposited (as in paste) some may not melt and/or not become a part of the solder joint. We use Kester R596, a water soluable flux-based paste and wash the boards immediately following reflow but we are seeing "fines". These "fines" are visible ONLY under a microscope and vary in quantity from 5 per 9"x14" assembly to 25 over the same area after washing. We have tried scrubbing to no avail. | | | | I feel someone is making a issue of a non-issue. Ramp-up is about 2.5 degrees C per second in a BTU VIP 70 forced convection oven. All other solder issues are great and we have re-verified the profile and it is still within the paste specs. | | | | Any help (official specs are important) we can get on this matter would be greatly appreciated. | | | | Mark | | | Mark: Several points: | | Solder Paste Specification. ANSI/J- STD-005, Requirements For Solder Pastes | | Fines. In my parlance, fines are the powder of a powder manufacturing process, if that�s not too convoluted. But continuing with your definition of fines, our venerable ( or is it venerated??) friend Bob Willis (www.bobwillis.co.uk/) differentiates between reflowed solder ball sizes and has words to describe the types. | | Raw Solder Powder Sizing. A common powder size is -200/+325 mesh. In this powder, 99% by weight will pass through a 200 (holes per square inch) mesh screen and less than 20% of the powder by weight will pass through a 325 mesh screen. From this, you can figure the upside potential for fines. Your supplier can give you greater detail on your paste and on raw solder fines. (At least what I call fines.) | | Small Solder Balls Remaining After Washing. Mark, something is wrong here!!! You should be able to wash the solder balls from the boards. With limited information, it sounds like your solder mask is improperly cured, partially curing during reflow, creating solder balls, and bonding the balls to the board during this partial curing. Our recently departed (from the forum) friend, Earl Moon, made major contributions to a thread on mask curing recently. Check the archives. | | Have your fab house check you boards for proper curing, if you don�t have MEK or some other method for checking mask cure. MEK is a hazardous material. I do not condone using it for testing boards. (Pfeu!!! I got that out my system.) | | Good luck | | Dave F | | | My definition of "fines":

IPC Sizing classifications go down in size

Type 2 45 to 75 um dia -200/+325 mesh Type 3 25 to 45 um dia -325/+500 mesh Type 4 20 to 38 um dia -400/+635 mesh

"Fines" are smaller than 20 um. They actually sneak into in all types of solder paste, but you tend to get more in the type 4, because of the automatic sorting machines and the smaller ball sizes involved.

The problem with the fines is that solder balls have surface oxides that hamper wetting and coalescense. Larger balls have more inside material than surface area, so the ratio of oxide to solder is lower. Fines have a much higher oxide/solder ratio and too much of them can cause solderability problems. Hence the desire to use type 3 pastes on everything but extremely fine pitch.

As for checking to see if mask is cured, a little ethyl acetate on a Q-tip will tell you. Drag it across the mask and see if it comes up green. If it does, you're not fully cured. I think ethyl acetate is a little more user-friendly than MEK.

My 2 cents,

Chrys

reply »

| | | Can someone explain the difference between solder "fines" and solder balls? Is there an official specification where "fines" are defined? What is the spec for "fines"? | | | | | | I consider "fines" individual unmelted metal spheres that are found in solder paste. It seems likely that when millions of 35micron balls are deposited (as in paste) some may not melt and/or not become a part of the solder joint. We use Kester R596, a water soluable flux-based paste and wash the boards immediately following reflow but we are seeing "fines". These "fines" are visible ONLY under a microscope and vary in quantity from 5 per 9"x14" assembly to 25 over the same area after washing. We have tried scrubbing to no avail. | | | | | | I feel someone is making a issue of a non-issue. Ramp-up is about 2.5 degrees C per second in a BTU VIP 70 forced convection oven. All other solder issues are great and we have re-verified the profile and it is still within the paste specs. | | | | | | Any help (official specs are important) we can get on this matter would be greatly appreciated. | | | | | | Mark | | | | | Mark: Several points: | | | | Solder Paste Specification. ANSI/J- STD-005, Requirements For Solder Pastes | | | | Fines. In my parlance, fines are the powder of a powder manufacturing process, if that�s not too convoluted. But continuing with your definition of fines, our venerable ( or is it venerated??) friend Bob Willis (www.bobwillis.co.uk/) differentiates between reflowed solder ball sizes and has words to describe the types. | | | | Raw Solder Powder Sizing. A common powder size is -200/+325 mesh. In this powder, 99% by weight will pass through a 200 (holes per square inch) mesh screen and less than 20% of the powder by weight will pass through a 325 mesh screen. From this, you can figure the upside potential for fines. Your supplier can give you greater detail on your paste and on raw solder fines. (At least what I call fines.) | | | | Small Solder Balls Remaining After Washing. Mark, something is wrong here!!! You should be able to wash the solder balls from the boards. With limited information, it sounds like your solder mask is improperly cured, partially curing during reflow, creating solder balls, and bonding the balls to the board during this partial curing. Our recently departed (from the forum) friend, Earl Moon, made major contributions to a thread on mask curing recently. Check the archives. | | | | Have your fab house check you boards for proper curing, if you don�t have MEK or some other method for checking mask cure. MEK is a hazardous material. I do not condone using it for testing boards. (Pfeu!!! I got that out my system.) | | | | Good luck | | | | Dave F | | | | | | | My definition of "fines": | | IPC Sizing classifications go down in size | | Type 2 45 to 75 um dia -200/+325 mesh | Type 3 25 to 45 um dia -325/+500 mesh | Type 4 20 to 38 um dia -400/+635 mesh | | "Fines" are smaller than 20 um. They actually sneak into in all types of solder paste, but you tend to get more in the type 4, because of the automatic sorting machines and the smaller ball sizes involved. | | The problem with the fines is that solder balls have surface oxides that hamper wetting and coalescense. Larger balls have more inside material than surface area, so the ratio of oxide to solder is lower. Fines have a much higher oxide/solder ratio and too much of them can cause solderability problems. Hence the desire to use type 3 pastes on everything but extremely fine pitch. | | As for checking to see if mask is cured, a little ethyl acetate on a Q-tip will tell you. Drag it across the mask and see if it comes up green. If it does, you're not fully cured. I think ethyl acetate is a little more user-friendly than MEK. | | My 2 cents, | | Chrys | Thanks for the advice Dave and Chrys. Perhaps some expansion of this issue can help us pinpoint what's happening.

1. The fines we see are NEVER on the soldermask, they do not appear to be solderballs in the classic sense we remember from wave soldering (this assembly never sees a wave solder). 2. They are always next to the lead (physically contacting) of a device and situated in the area of no mask between the pad and soldermask. It would require at least 10 of these files to create a short.

3. These fines are present whether a device has no soldermask between leads or there is masking between the leads. They are generally on 31mil QFP's. They are easily removed using a "VERY" fine pick and will not come off by shaking or tapping the board in that area.

4. They occur on only this one assembly.

5. They occur generally (95%) in one section of the PCB.

6. Profile verifications in this section of the PCB show no more than 1 to 2 degree variations compared to other locations on PCB.

In the meantime, we are reviewing and auditing our solderpaste handling procedures, screen printing process, stencil underside wiping, and solder print accuracy and repeatability.

Thanks again for your help and the directions to Bob Willis. There was some great information there.

Mark

reply »

| | | | Can someone explain the difference between solder "fines" and solder balls? Is there an official specification where "fines" are defined? What is the spec for "fines"? | | | | | | | | I consider "fines" individual unmelted metal spheres that are found in solder paste. It seems likely that when millions of 35micron balls are deposited (as in paste) some may not melt and/or not become a part of the solder joint. We use Kester R596, a water soluable flux-based paste and wash the boards immediately following reflow but we are seeing "fines". These "fines" are visible ONLY under a microscope and vary in quantity from 5 per 9"x14" assembly to 25 over the same area after washing. We have tried scrubbing to no avail. | | | | | | | | I feel someone is making a issue of a non-issue. Ramp-up is about 2.5 degrees C per second in a BTU VIP 70 forced convection oven. All other solder issues are great and we have re-verified the profile and it is still within the paste specs. | | | | | | | | Any help (official specs are important) we can get on this matter would be greatly appreciated. | | | | | | | | Mark | | | | | | | Mark: Several points: | | | | | | Solder Paste Specification. ANSI/J- STD-005, Requirements For Solder Pastes | | | | | | Fines. In my parlance, fines are the powder of a powder manufacturing process, if that�s not too convoluted. But continuing with your definition of fines, our venerable ( or is it venerated??) friend Bob Willis (www.bobwillis.co.uk/) differentiates between reflowed solder ball sizes and has words to describe the types. | | | | | | Raw Solder Powder Sizing. A common powder size is -200/+325 mesh. In this powder, 99% by weight will pass through a 200 (holes per square inch) mesh screen and less than 20% of the powder by weight will pass through a 325 mesh screen. From this, you can figure the upside potential for fines. Your supplier can give you greater detail on your paste and on raw solder fines. (At least what I call fines.) | | | | | | Small Solder Balls Remaining After Washing. Mark, something is wrong here!!! You should be able to wash the solder balls from the boards. With limited information, it sounds like your solder mask is improperly cured, partially curing during reflow, creating solder balls, and bonding the balls to the board during this partial curing. Our recently departed (from the forum) friend, Earl Moon, made major contributions to a thread on mask curing recently. Check the archives. | | | | | | Have your fab house check you boards for proper curing, if you don�t have MEK or some other method for checking mask cure. MEK is a hazardous material. I do not condone using it for testing boards. (Pfeu!!! I got that out my system.) | | | | | | Good luck | | | | | | Dave F | | | | | | | | | | | My definition of "fines": | | | | IPC Sizing classifications go down in size | | | | Type 2 45 to 75 um dia -200/+325 mesh | | Type 3 25 to 45 um dia -325/+500 mesh | | Type 4 20 to 38 um dia -400/+635 mesh | | | | "Fines" are smaller than 20 um. They actually sneak into in all types of solder paste, but you tend to get more in the type 4, because of the automatic sorting machines and the smaller ball sizes involved. | | | | The problem with the fines is that solder balls have surface oxides that hamper wetting and coalescense. Larger balls have more inside material than surface area, so the ratio of oxide to solder is lower. Fines have a much higher oxide/solder ratio and too much of them can cause solderability problems. Hence the desire to use type 3 pastes on everything but extremely fine pitch. | | | | As for checking to see if mask is cured, a little ethyl acetate on a Q-tip will tell you. Drag it across the mask and see if it comes up green. If it does, you're not fully cured. I think ethyl acetate is a little more user-friendly than MEK. | | | | My 2 cents, | | | | Chrys | | | Thanks for the advice Dave and Chrys. | | Perhaps some expansion of this issue can help us pinpoint what's happening. | | 1. The fines we see are NEVER on the soldermask, they do not appear to be solderballs in the classic sense we remember from wave soldering (this assembly never sees a wave solder). | | 2. They are always next to the lead (physically contacting) of a device and situated in the area of no mask between the pad and soldermask. It would require at least 10 of these files to create a short. | | 3. These fines are present whether a device has no soldermask between leads or there is masking between the leads. They are generally on 31mil QFP's. They are easily removed using a "VERY" fine pick and will not come off by shaking or tapping the board in that area. | | 4. They occur on only this one assembly. | | 5. They occur generally (95%) in one section of the PCB. | | 6. Profile verifications in this section of the PCB show no more than 1 to 2 degree variations compared to other locations on PCB. | | In the meantime, we are reviewing and auditing our solderpaste handling procedures, screen printing process, stencil underside wiping, and solder print accuracy and repeatability. | | Thanks again for your help and the directions to Bob Willis. There was some great information there. | | Mark | Mark,

If i'm picking this up right your getting the "fines" on your 31 thou flatpacks and it's not getting removed by washing...and yoiu are picking them off...

couple of thought's here, your obviously using a water wash paste wich is far more forgiving to stencil life and so forth so that's probably not a problem. But what about slumping ???? have you checked you appeture to pad ratio ? you may need to undersize it to avoid the paste wondering off the pad. In term's of removing it with a pick tool, why remove it at all ?/ you say that it aint creating a short, you cnat wash it off which suggets's that's it's trapped in flux and going nowhere anyhow. something else crosses my mind, if it's mainly concentrated in one area of the PCB is what about any of the following:- Screen printing - how aare the support's ?? Placement - again suppoort's but also the placement pressures oven - pad sizes and large ground planes can cause mysterious things to occur in reflow ovens...and are u using N2 if so is the PPM enough or are the N2 curtains doing thier job ???

good luck

reply »

| | | | | Can someone explain the difference between solder "fines" and solder balls? Is there an official specification where "fines" are defined? What is the spec for "fines"? | | | | | | | | | | I consider "fines" individual unmelted metal spheres that are found in solder paste. It seems likely that when millions of 35micron balls are deposited (as in paste) some may not melt and/or not become a part of the solder joint. We use Kester R596, a water soluable flux-based paste and wash the boards immediately following reflow but we are seeing "fines". These "fines" are visible ONLY under a microscope and vary in quantity from 5 per 9"x14" assembly to 25 over the same area after washing. We have tried scrubbing to no avail. | | | | | | | | | | I feel someone is making a issue of a non-issue. Ramp-up is about 2.5 degrees C per second in a BTU VIP 70 forced convection oven. All other solder issues are great and we have re-verified the profile and it is still within the paste specs. | | | | | | | | | | Any help (official specs are important) we can get on this matter would be greatly appreciated. | | | | | | | | | | Mark | | | | | | | | | Mark: Several points: | | | | | | | | Solder Paste Specification. ANSI/J- STD-005, Requirements For Solder Pastes | | | | | | | | Fines. In my parlance, fines are the powder of a powder manufacturing process, if that�s not too convoluted. But continuing with your definition of fines, our venerable ( or is it venerated??) friend Bob Willis (www.bobwillis.co.uk/) differentiates between reflowed solder ball sizes and has words to describe the types. | | | | | | | | Raw Solder Powder Sizing. A common powder size is -200/+325 mesh. In this powder, 99% by weight will pass through a 200 (holes per square inch) mesh screen and less than 20% of the powder by weight will pass through a 325 mesh screen. From this, you can figure the upside potential for fines. Your supplier can give you greater detail on your paste and on raw solder fines. (At least what I call fines.) | | | | | | | | Small Solder Balls Remaining After Washing. Mark, something is wrong here!!! You should be able to wash the solder balls from the boards. With limited information, it sounds like your solder mask is improperly cured, partially curing during reflow, creating solder balls, and bonding the balls to the board during this partial curing. Our recently departed (from the forum) friend, Earl Moon, made major contributions to a thread on mask curing recently. Check the archives. | | | | | | | | Have your fab house check you boards for proper curing, if you don�t have MEK or some other method for checking mask cure. MEK is a hazardous material. I do not condone using it for testing boards. (Pfeu!!! I got that out my system.) | | | | | | | | Good luck | | | | | | | | Dave F | | | | | | | | | | | | | | | My definition of "fines": | | | | | | IPC Sizing classifications go down in size | | | | | | Type 2 45 to 75 um dia -200/+325 mesh | | | Type 3 25 to 45 um dia -325/+500 mesh | | | Type 4 20 to 38 um dia -400/+635 mesh | | | | | | "Fines" are smaller than 20 um. They actually sneak into in all types of solder paste, but you tend to get more in the type 4, because of the automatic sorting machines and the smaller ball sizes involved. | | | | | | The problem with the fines is that solder balls have surface oxides that hamper wetting and coalescense. Larger balls have more inside material than surface area, so the ratio of oxide to solder is lower. Fines have a much higher oxide/solder ratio and too much of them can cause solderability problems. Hence the desire to use type 3 pastes on everything but extremely fine pitch. | | | | | | As for checking to see if mask is cured, a little ethyl acetate on a Q-tip will tell you. Drag it across the mask and see if it comes up green. If it does, you're not fully cured. I think ethyl acetate is a little more user-friendly than MEK. | | | | | | My 2 cents, | | | | | | Chrys | | | | | Thanks for the advice Dave and Chrys. | | | | Perhaps some expansion of this issue can help us pinpoint what's happening. | | | | 1. The fines we see are NEVER on the soldermask, they do not appear to be solderballs in the classic sense we remember from wave soldering (this assembly never sees a wave solder). | | | | 2. They are always next to the lead (physically contacting) of a device and situated in the area of no mask between the pad and soldermask. It would require at least 10 of these files to create a short. | | | | 3. These fines are present whether a device has no soldermask between leads or there is masking between the leads. They are generally on 31mil QFP's. They are easily removed using a "VERY" fine pick and will not come off by shaking or tapping the board in that area. | | | | 4. They occur on only this one assembly. | | | | 5. They occur generally (95%) in one section of the PCB. | | | | 6. Profile verifications in this section of the PCB show no more than 1 to 2 degree variations compared to other locations on PCB. | | | | In the meantime, we are reviewing and auditing our solderpaste handling procedures, screen printing process, stencil underside wiping, and solder print accuracy and repeatability. | | | | Thanks again for your help and the directions to Bob Willis. There was some great information there. | | | | Mark | | | Mark, | | If i'm picking this up right your getting the "fines" on your 31 thou flatpacks and it's not getting removed by washing...and yoiu are picking them off... | | couple of thought's here, your obviously using a water wash paste wich is far more forgiving to stencil life and so forth so that's probably not a problem. But what about slumping ???? have you checked you appeture to pad ratio ? you may need to undersize it to avoid the paste wondering off the pad. | In term's of removing it with a pick tool, why remove it at all ?/ you say that it aint creating a short, you cnat wash it off which suggets's that's it's trapped in flux and going nowhere anyhow. | something else crosses my mind, if it's mainly concentrated in one area of the PCB is what about any of the following:- | Screen printing - how aare the support's ?? | Placement - again suppoort's but also the placement pressures | oven - pad sizes and large ground planes can cause mysterious things to occur in reflow ovens...and are u using N2 if so is the PPM enough or are the N2 curtains doing thier job ??? | | good luck | Dear Mark,

I just wonder, is the paste you are now using is active enough to remove any oxide on solder. I encounter this problem previously, however, once I try on some stronger paste, (of better wetting and spreading ability), this problem seems to reduce.

Vic

reply »