Tin-Lead thickness on PWB's

I'm reviewing my board fab spec. It calls for a minimum SnPb thickness of 50 microinches on HASL PWB's. I've looked at other specs that call out anything from 30 to 80 microinches, and others that just say the copper pad must be covered and solderable. What's reasonable? How thick are the intermetallics (typically) on HASL boards? What about the surface layer of oxide? Thanks for your input.

reply »

| I'm reviewing my board fab spec. It calls for a minimum SnPb thickness of 50 microinches on HASL PWB's. I've looked at other specs that call out anything from 30 to 80 microinches, and others that just say the copper pad must be covered and solderable. | What's reasonable? How thick are the intermetallics (typically) on HASL boards? What about the surface layer of oxide? | Thanks for your input. Chrys, This is old Earl. Don't mean to be a bore. I won't cover it all again but there is no min/max for HASL. As far as intermetallics go - they don't form for some time as solder wetting (when controlled and this process is not controllable) is defined, once again, as the joining of two metallic surfaces, with a specified solder medium, without difussion or intermetallic formation. HASL is uncontrollable because it relies on something that cannot be accomplished effectively. Anyway, any coating thickness will do as long as it covers the copper surface to keep it from oxidizing or other contamination. I know you're trying to put together a spec. It's just not in the cards for HASL. It is for any other plating or coating process wherein either electroplating or immersion is done. Hope I don't continue to be a pain, but industry has gone all through this problem and concluded there are better alternatives to HASL. Sincerely and best wishes, Earl Moon

reply »

Isn't there an IPC spec on this? I was looking for it a couple of days ago and couldn't find it. We have an internal spec of no more than .002" but I don't know how realistic this is. If you can get your PWB vendor to sign up to this I think you should go for it. We are soldering mostly >.025" pitch so the only problem we run into is too much HASL on the adhesive side causing improper glue dispense. As Earl keeps saying HASL thickness is hard to impossible to control. I think we need to look hard at OSP's again but then you get into a whole set of new requirements of storage and questionable looking solder joints. Nothing solders like solder and it is hard to break this mentality especially when you are building >.025" pitch.

reply »

| Isn't there an IPC spec on this? I was looking for it a couple of days ago and couldn't find it. We have an internal spec of no more than .002" but I don't know how realistic this is. If you can get your PWB vendor to sign up to this I think you should go for it. We are soldering mostly >.025" pitch so the only problem we run into is too much HASL on the adhesive side causing improper glue dispense. | As Earl keeps saying HASL thickness is hard to impossible to control. I think we need to look hard at OSP's again but then you get into a whole set of new requirements of storage and questionable looking solder joints. Nothing solders like solder and it is hard to break this mentality especially when you are building >.025" pitch. I agree nothing solders like solder except contamination, oxide free surfaces. Unfortunately, HASL is contaminated by default. There is no way to keep copper surfaces clean before introduction to the HASL process. The copper is oxided before introduction and causes solderability problems - no matter the thickness. OSP's are a limited answer as their shelf life is short. Again, immersion gold (7 millionths thick max over nickel) is an answer as is silver now that migration problems are resolved with the very thin coating via immersion. Palladium also works - with some work. I recognize HASL isn't going away asap, but it should be replaced asap. Also, again, there is no possible way to control HASL thickness as I have described in my article in the December (I think) issue of Printed Circuit Fabrication Magazine. Thanks again, Earl Moon

reply »

| | Isn't there an IPC spec on this? I was looking for it a couple of days ago and couldn't find it. We have an internal spec of no more than .002" but I don't know how realistic this is. If you can get your PWB vendor to sign up to this I think you should go for it. We are soldering mostly >.025" pitch so the only problem we run into is too much HASL on the adhesive side causing improper glue dispense. | | As Earl keeps saying HASL thickness is hard to impossible to control. I think we need to look hard at OSP's again but then you get into a whole set of new requirements of storage and questionable looking solder joints. Nothing solders like solder and it is hard to break this mentality especially when you are building >.025" pitch. | I agree nothing solders like solder except contamination, oxide free surfaces. Unfortunately, HASL is contaminated by default. There is no way to keep copper surfaces clean before introduction to the HASL process. The copper is oxided before introduction and causes solderability problems - no matter the thickness. OSP's are a limited answer as their shelf life is short. Again, immersion gold (7 millionths thick max over nickel) | is an answer as is silver now that migration problems are resolved with the very thin coating via immersion. Palladium also works - with some work. I recognize HASL isn't going away asap, but it should be replaced asap. Also, again, there is no possible way to control HASL thickness as I have described in my article in the December (I think) issue of Printed Circuit Fabrication Magazine. | Thanks again, | Earl Moon Wouldn't you know it?? December 1997 is one of the few issues that Printed Circuit Fabrication Magazine. doesn't have on-line. Dave F

reply »

Earl, I admit that I missed your article since I seemed to have let my subscription to Circuit Fabrication Magazine expire. I am interested in your experience in this area. Is there a way to get a copy of your paper? Could you maybe post it to a web site? Thanks, Jerry

reply »

| Earl, | I admit that I missed your article since I seemed to have let my subscription to Circuit Fabrication Magazine expire. I am interested in your experience in this area. Is there a way to get a copy of your paper? Could you maybe post it to a web site? | Thanks, | Jerry Oh damn! You didn't let that happen. Hey, I really am on your side. You've said it all right. It is me that made the mistake. My wonderful work appeared in December 1882 or was it 1992. Yeh, It was that last one. I would be very happy to post the article, but I don't have it anymore either. However, if you wish, I will recreate vital parts. How 'bout that guys and gals. Not matter, this is such an important issue. It was when HASL began hassling us, and it will continue until we all get away from its use - and we are doing just that. I know we all want a spec for everything, and well we should. However, due to the way boards come out of the copper plating process (microetched and rinsed wet) they oxidize immediately. If kept on the shelf before introduction into the tin/lead coating process, they become more contaminated. When they are introduced to the flux mechanism, it often cannot remove the oxidation. Even if it does, after traveling up out of the molten solder pot, air knives provide a very uneven, "puddled" glop on the copper surface. I'll bet you all have noticed this condition. Could this be why it is impossible to spec the condition ranging from nearly no solder to over .003" of the stuff. We all are faced with a problem (we in my limited place in space call it an issue as we refuse to deal with problems) that has no good answer but replacement. I really do appreciate what you are going through and each of you is right searching for the answer, but a spec is not possible except to say each and every pad must be totally wetted. Total wetting is not possible, either, with HASL. We've all seen de or nonwetted areas on SMT solder termination areas. This gets back to the panel's removal from copper plating and its introduction into a process that cannot wet anything so oxidized. What more can I say? Besides, It's all been said before - oh so many times. Best Wishes for a bright soldery future and may all your joints be acceptable, Earl Moon

reply »

| Earl, | I admit that I missed your article since I seemed to have let my subscription to Circuit Fabrication Magazine expire. I am interested in your experience in this area. Is there a way to get a copy of your paper? Could you maybe post it to a web site? | Thanks, | Jerry Jerry, I did find my article concerning HASL process control, or the lack thereof. If you really want it, I will scan it and send it to you free of charge and without any obligation to buy or rent. Let me know how to get hold of you, or if you wish - what forum would be best to post the article without peeing off a bunch of anti spamers? Also, because so many bare fab questions are being asked, repeatedly, I have other writings concerning everything from design through test. Some are quite informative and answer constantly recurring questions. Any ideas would be appreciated. Sincerely, Earl Moon

reply »

| | Earl, | | I admit that I missed your article since I seemed to have let my subscription to Circuit Fabrication Magazine expire. I am interested in your experience in this area. Is there a way to get a copy of your paper? Could you maybe post it to a web site? | | Thanks, | | Jerry | Jerry, | I did find my article concerning HASL process control, or the lack thereof. | If you really want it, I will scan it and send it to you free of charge | and without any obligation to buy or rent. Let me know how to get hold of | you, or if you wish - what forum would be best to post the article without | peeing off a bunch of anti spamers? Also, because so many bare fab questions | are being asked, repeatedly, I have other writings concerning everything from | design through test. Some are quite informative and answer constantly recurring | questions. Any ideas would be appreciated. | Sincerely, | Earl Moon Earl, Please send it to me. I am embroiled in a "discussion" on this topic. I would be eternally grateful. I'm even willing to barter. How would you like my last paper on controlling the wave solder process? How about the one on air entrainment from hot N2 knives? My first born? (can't have him - I'd never trade my yellow lab) Thanks, Earl

reply »

| | | Earl, | | | I admit that I missed your article since I seemed to have let my subscription to Circuit Fabrication Magazine expire. I am interested in your experience in this area. Is there a way to get a copy of your paper? Could you maybe post it to a web site? | | | Thanks, | | | Jerry | | Jerry, | | I did find my article concerning HASL process control, or the lack thereof. | | If you really want it, I will scan it and send it to you free of charge | | and without any obligation to buy or rent. Let me know how to get hold of | | you, or if you wish - what forum would be best to post the article without | | peeing off a bunch of anti spamers? Also, because so many bare fab questions | | are being asked, repeatedly, I have other writings concerning everything from | | design through test. Some are quite informative and answer constantly recurring | | questions. Any ideas would be appreciated. | | Sincerely, | | Earl Moon | Earl, | Please send it to me. I am embroiled in a "discussion" on this topic. I would be eternally grateful. I'm even willing to barter. How would you like my last paper on controlling the wave solder process? How about the one on air entrainment from hot N2 knives? My first born? (can't have him - I'd never trade my yellow lab) | Thanks, Earl

Boy, You guys are something else. You won't even trade your good stuff for my good stuff. What the hell, it's all for the common good. Just one or two things. I'm probably missing something but how do I get this thing to you? Again, is there a place on the net I could just dump it for all to see? Also, it will take me a couple days to scan and send. Talk to you soon, Earl Moon

reply »

| | Earl, | | I admit that I missed your article since I seemed to have let my subscription to Circuit Fabrication Magazine expire. I am interested in your experience in this area. Is there a way to get a copy of your paper? Could you maybe post it to a web site? | | Thanks, | | Jerry | Jerry, | I did find my article concerning HASL process control, or the lack thereof. | If you really want it, I will scan it and send it to you free of charge | and without any obligation to buy or rent. Let me know how to get hold of | you, or if you wish - what forum would be best to post the article without | peeing off a bunch of anti spamers? Also, because so many bare fab questions | are being asked, repeatedly, I have other writings concerning everything from | design through test. Some are quite informative and answer constantly recurring | questions. Any ideas would be appreciated. | Sincerely, | Earl Moon Earl: Can you post it in the SMTNet Library? Dave F

reply »

| I'm reviewing my board fab spec. It calls for a minimum SnPb thickness of 50 microinches on HASL PWB's. I've looked at other specs that call out anything from 30 to 80 microinches, and others that just say the copper pad must be covered and solderable. | What's reasonable? How thick are the intermetallics (typically) on HASL boards? What about the surface layer of oxide? | Thanks for your input. Chrys, Everybody is being a little rough on HASL. There are a host of other problems with alternative surface finishes. Earl knows a ton about this stuff. I'm pretty well versed on the alternative surface finishes from a process standpoint. Seen 'em all. Worked with most. OSP's are tough. You need to approve a supplier with tight process controls and definitely a conveyorized process. I have seen boards where the OSP was too thick and Alpha 609 couldn't even break it down. That OA is ACTIVE. I used it to clean rust off my golf clubs!! Nuff said. We got around it by sending the fabs through the reflow oven 3 times before processing in order to break down the OSP (Entek Plus)coating. Imidazole is a pretty cool OSP. It's a proprietary OSP at Via Systems. Dr. Lee Parker introduced it. It's pretty neat stuff. It really doesn't break down in thermal cycles. Good for double sided reflow, something that is tough for Entek. Enthone has a pretty adept chemistry staff. It's a nice coating too but you have to apply it properly. A ton of board shops throw this stuff on their boards but they don't know what the heck they're doing. Some don't even test for thickness of coating application. Imidazole is done primarily at Via Systems. They invented it. They apply it. Everything is internal. Thus, better control. As for the noble metal crap. That's what it is. I use it. I have to. If you get Seven microns of gold over nickel, you are all done. 3 to 5 microns is necessary for a decent solder joint. Any more, you'll see embrittlement, proved it. Any less, you'll see voids and oxidized nickel, proved it. By the way, good luck soldering to oxidized nickel. You also need to increase your reflow temperatures to get a shiny solder joint. Great. Now I'm thermal cycling between 25 and 255 in my reflow oven. Some components can't take that heat. If the assy. is double sided, you have to go through that thermal cycle at least twice. Silver is in it's "toddler" phase. People are still learning how to use it. You need to change your paste to allow for this in process, which will change all of your profiles. Bottom line, flat pads look like a great concept. The flatness doesn't come for free!!! I do .016" pitch all the time with HASL boards. I don't have a problem. Choosing a good vendor, having proper pad sizes and aperture geometries can cover up all the geometric anomalies of a HASL pad. As for oxidation and solderability, everything rusts or has a cost. With HASL, it can oxidize. Get an active paste. With OSP, it breaks down and you have exposed copper. Use 'em fast. With gold over nickel, you need the right thickness and the proper paste and parts. Like I said, it's all got a cost. We have to deal with these issues soon anyway. Lead won't be around forever.... I'm oudahere! That was long winded. Sorry. Best Regards, Justin

reply »

| | I'm reviewing my board fab spec. It calls for a minimum SnPb thickness of 50 microinches on HASL PWB's. I've looked at other specs that call out anything from 30 to 80 microinches, and others that just say the copper pad must be covered and solderable. | | What's reasonable? How thick are the intermetallics (typically) on HASL boards? What about the surface layer of oxide? | | Thanks for your input. | Chrys, | Everybody is being a little rough on HASL. There are a host of other problems with alternative surface finishes. Earl knows a ton about this stuff. I'm pretty well versed on the alternative surface finishes from a process standpoint. Seen 'em all. Worked with most. OSP's are tough. You need to approve a supplier with tight process controls and definitely a conveyorized process. I have seen boards where the OSP was too thick and Alpha 609 couldn't even break it down. That OA is ACTIVE. I used it to clean rust off my golf clubs!! Nuff said. We got around it by sending the fabs through the reflow oven 3 times before processing in order to break down the OSP (Entek Plus)coating. Imidazole is a pretty cool OSP. It's a proprietary OSP at Via Systems. Dr. Lee Parker introduced it. It's pretty neat stuff. It really doesn't break down in thermal cycles. Good for double sided reflow, something that is tough for Entek. Enthone has a pretty adept chemistry staff. It's a nice coating too but you have | to apply it properly. A ton of board shops throw this stuff on their boards but they don't know what the heck they're doing. Some don't even test for thickness of coating application. Imidazole is done primarily at Via Systems. They invented it. They apply it. Everything is internal. Thus, better control. As for the noble metal crap. That's what it is. I use it. I have to. If you get Seven microns of gold over nickel, you are all done. 3 to 5 microns is necessary for a decent solder joint. Any more, you'll see embrittlement, proved it. Any less, you'll see voids and oxidized nickel, proved it. By the way, good luck soldering to oxidized nickel. You also need to increase your reflow temperatures to get a shiny solder joint. Great. Now I'm thermal cycling between 25 and 255 in my reflow oven. Some components can't take that heat. If the assy. is double sided, you have to go through that thermal cycle at least twice. Silver is in it's "toddler" phase. People are still learning how to use it. You need to change your paste to allow for this in process, which will change all of your profiles. Bottom line, flat pads look like a great concept. The flatness doesn't come for free!!! I do .016" pitch all the time with HASL boards. I don't have a problem. Choosing a good vendor, having proper pad sizes and aperture geometries can cover up all the geometric anomalies of a HASL pad. As for oxidation and solderability, everything rusts or has a cost. With HASL, it can oxidize. Get an active paste. With OSP, it breaks down and you have exposed copper. Use 'em fast. With gold over nickel, you need the right thickness and the proper paste and parts. Like I said, it's all got a cost. We have to deal with these issues soon anyway. Lead won't be around forever.... | I'm oudahere! That was long winded. Sorry. | Best Regards, | Justin Justin: Thanks for your interesting comments. Questions they raise are: 1 When you talk about OSP suppliers, what do you mean by "tight process controls?" 2 When you talk about OSP suppliers, why "definitely conveyorized process?" 3 On Entek: When you say that "you have to apply it properly," how should it be done? 4 On Entek: When you say that "some don�t even check for thickness of coating application," what should they be doing? 5 When you speak of not having problems doing 16 pitch HASL boards: What do you mean by: � Proper pad sizes? � Proper aperature geometries? 6 Continuing with this last question: When laying-out proper pad sizes for 16 pitch components on boards for fab, how do you spec the web between pads? Thanks Dave F

reply »

| | | I'm reviewing my board fab spec. It calls for a minimum SnPb thickness of 50 microinches on HASL PWB's. I've looked at other specs that call out anything from 30 to 80 microinches, and others that just say the copper pad must be covered and solderable. | | | What's reasonable? How thick are the intermetallics (typically) on HASL boards? What about the surface layer of oxide? | | | Thanks for your input. | | Chrys, | | Everybody is being a little rough on HASL. There are a host of other problems with alternative surface finishes. Earl knows a ton about this stuff. I'm pretty well versed on the alternative surface finishes from a process standpoint. Seen 'em all. Worked with most. OSP's are tough. You need to approve a supplier with tight process controls and definitely a conveyorized process. I have seen boards where the OSP was too thick and Alpha 609 couldn't even break it down. That OA is ACTIVE. I used it to clean rust off my golf clubs!! Nuff said. We got around it by sending the fabs through the reflow oven 3 times before processing in order to break down the OSP (Entek Plus)coating. Imidazole is a pretty cool OSP. It's a proprietary OSP at Via Systems. Dr. Lee Parker introduced it. It's pretty neat stuff. It really doesn't break down in thermal cycles. Good for double sided reflow, something that is tough for Entek. Enthone has a pretty adept chemistry staff. It's a nice coating too but you have | | to apply it properly. A ton of board shops throw this stuff on their boards but they don't know what the heck they're doing. Some don't even test for thickness of coating application. Imidazole is done primarily at Via Systems. They invented it. They apply it. Everything is internal. Thus, better control. As for the noble metal crap. That's what it is. I use it. I have to. If you get Seven microns of gold over nickel, you are all done. 3 to 5 microns is necessary for a decent solder joint. Any more, you'll see embrittlement, proved it. Any less, you'll see voids and oxidized nickel, proved it. By the way, good luck soldering to oxidized nickel. You also need to increase your reflow temperatures to get a shiny solder joint. Great. Now I'm thermal cycling between 25 and 255 in my reflow oven. Some components can't take that heat. If the assy. is double sided, you have to go through that thermal cycle at least twice. Silver is in it's "toddler" phase. People are still learning how to use it. You need to change your paste to allow for this in process, which will change all of your profiles. Bottom line, flat pads look like a great concept. The flatness doesn't come for free!!! I do .016" pitch all the time with HASL boards. I don't have a problem. Choosing a good vendor, having proper pad sizes and aperture geometries can cover up all the geometric anomalies of a HASL pad. As for oxidation and solderability, everything rusts or has a cost. With HASL, it can oxidize. Get an active paste. With OSP, it breaks down and you have exposed copper. Use 'em fast. With gold over nickel, you need the right thickness and the proper paste and parts. Like I said, it's all got a cost. We have to deal with these issues soon anyway. Lead won't be around forever.... | | I'm oudahere! That was long winded. Sorry. | | Best Regards, | | Justin | Justin: Thanks for your interesting comments. Questions they raise are: | 1 When you talk about OSP suppliers, what do you mean by "tight process controls?" | 2 When you talk about OSP suppliers, why "definitely conveyorized process?" | 3 On Entek: When you say that "you have to apply it properly," how should it be done? | 4 On Entek: When you say that "some don�t even check for thickness of coating application," what should they be doing? | 5 When you speak of not having problems doing 16 pitch HASL boards: What do you mean by: | � Proper pad sizes? | � Proper aperature geometries? | 6 Continuing with this last question: When laying-out proper pad sizes for 16 pitch components on boards for fab, how do you spec the web between pads? | Thanks | Dave F Dave, I can go through everything except the pad / aperture sizes. These are proprietary and the reason why Flex can build better than most. Sounds conceited but that's the contract game. You need a new fab supplier if they can't get a web in a .016" pitch device. As far as the process control question is concerned. You want to know that they are controlling their chemical changeover times. Are they using control charts on their etching process? Are they using them on their coating process? As for question 2: Conveyorization means control. These coatings are sprayed or dipped. Manual processes such as these yeild deviation. Mechanized processes aid in repeatability. That's the name of the game with this stuff. For question 3 regarding application, I only have one thing to say. We should never say how to do something, only to do it right. I don't know if dipping or spraying is better. I do know that thickness is the critical attribute. These leads to the answer to question 4. A spectrophotometer should be used to check coating thickness. If they don't have one of these devices or they say they can eyeball it, don't use them. Like I said, coating thickness is everything but I think it hangs somewhere around 2500 to 3500 Angstroms (don't take that as gospel). Best Regards and keep 'em coming, Justin Medernach

reply »

| | | | I'm reviewing my board fab spec. It calls for a minimum SnPb thickness of 50 microinches on HASL PWB's. I've looked at other specs that call out anything from 30 to 80 microinches, and others that just say the copper pad must be covered and solderable. | | | | What's reasonable? How thick are the intermetallics (typically) on HASL boards? What about the surface layer of oxide? | | | | Thanks for your input. | | | Chrys, | | | Everybody is being a little rough on HASL. There are a host of other problems with alternative surface finishes. Earl knows a ton about this stuff. I'm pretty well versed on the alternative surface finishes from a process standpoint. Seen 'em all. Worked with most. OSP's are tough. You need to approve a supplier with tight process controls and definitely a conveyorized process. I have seen boards where the OSP was too thick and Alpha 609 couldn't even break it down. That OA is ACTIVE. I used it to clean rust off my golf clubs!! Nuff said. We got around it by sending the fabs through the reflow oven 3 times before processing in order to break down the OSP (Entek Plus)coating. Imidazole is a pretty cool OSP. It's a proprietary OSP at Via Systems. Dr. Lee Parker introduced it. It's pretty neat stuff. It really doesn't break down in thermal cycles. Good for double sided reflow, something that is tough for Entek. Enthone has a pretty adept chemistry staff. It's a nice coating too but you have | | | to apply it properly. A ton of board shops throw this stuff on their boards but they don't know what the heck they're doing. Some don't even test for thickness of coating application. Imidazole is done primarily at Via Systems. They invented it. They apply it. Everything is internal. Thus, better control. As for the noble metal crap. That's what it is. I use it. I have to. If you get Seven microns of gold over nickel, you are all done. 3 to 5 microns is necessary for a decent solder joint. Any more, you'll see embrittlement, proved it. Any less, you'll see voids and oxidized nickel, proved it. By the way, good luck soldering to oxidized nickel. You also need to increase your reflow temperatures to get a shiny solder joint. Great. Now I'm thermal cycling between 25 and 255 in my reflow oven. Some components can't take that heat. If the assy. is double sided, you have to go through that thermal cycle at least twice. Silver is in it's "toddler" phase. People are still learning how to use it. You need to change your paste to allow for this in process, which will change all of your profiles. Bottom line, flat pads look like a great concept. The flatness doesn't come for free!!! I do .016" pitch all the time with HASL boards. I don't have a problem. Choosing a good vendor, having proper pad sizes and aperture geometries can cover up all the geometric anomalies of a HASL pad. As for oxidation and solderability, everything rusts or has a cost. With HASL, it can oxidize. Get an active paste. With OSP, it breaks down and you have exposed copper. Use 'em fast. With gold over nickel, you need the right thickness and the proper paste and parts. Like I said, it's all got a cost. We have to deal with these issues soon anyway. Lead won't be around forever.... | | | I'm oudahere! That was long winded. Sorry. | | | Best Regards, | | | Justin | | Justin: Thanks for your interesting comments. Questions they raise are: | | 1 When you talk about OSP suppliers, what do you mean by "tight process controls?" | | 2 When you talk about OSP suppliers, why "definitely conveyorized process?" | | 3 On Entek: When you say that "you have to apply it properly," how should it be done? | | 4 On Entek: When you say that "some don�t even check for thickness of coating application," what should they be doing? | | 5 When you speak of not having problems doing 16 pitch HASL boards: What do you mean by: | | � Proper pad sizes? | | � Proper aperature geometries? | | 6 Continuing with this last question: When laying-out proper pad sizes for 16 pitch components on boards for fab, how do you spec the web between pads? | | Thanks | | Dave F | Dave, | I can go through everything except the pad / aperture sizes. These are proprietary and the reason why Flex can build better than most. Sounds conceited but that's the contract game. You need a new fab supplier if they can't get a web in a .016" pitch device. As far as the process control question is concerned. You want to know that they are controlling their chemical changeover times. Are they using control charts on their etching process? Are they using them on their coating process? As for question 2: Conveyorization means control. These coatings are sprayed or dipped. Manual processes such as these yeild deviation. Mechanized processes aid in repeatability. That's the name of the game with this stuff. For question 3 regarding application, I only have one thing to say. We should never say how to do something, only to do it right. I don't know if dipping or spraying is better. I do know that thickness is the critical attribute. These leads to the answer to question 4. A spectrophotometer | should be used to check coating thickness. If they don't have one of these devices or they say they can eyeball it, don't use them. Like I said, coating thickness is everything but I think it hangs somewhere around 2500 to 3500 Angstroms (don't take that as gospel). | Best Regards and keep 'em coming, | Justin Medernach

Man, you guys just don't give up. I don't blame you. Never give up until it is right. I would simply add again that the only reason we apply any type copper surface coating is to prevent oxidation so solder wetting may be achieved. The problem with this and other coating processes (as with HASL) first is panels come out of some kind of wet chemical process (a microetch and rinse), are dried (causing immediate oxidation), then coated whether with organics or metals - electroplated, electrolessly deposited, sprayed, dipped, or? One thing organics have going for them is they are essentially fluxes. In the good old days, instead of Entek, we used what was known as "Seal Brite." Somewhat the same difference. The situation then, as now, at first the stuff actually fought or inhibited oxidation. Over time, the battle was lost depending mostly on storage and useage conditions. Therefore, a supplier must assure process control first to minimize oxidation. Then, the supplier must demonstrate coating thickness control. With the new gold, silver, and palladium metal coatings, the thickness must not exceed about 7 millionths of an inch or the obvious will be effected. For organics, because of their short shelf lives, and some suppliers inability to effectively manage processes, more surface contamination will be effected quicker with thinner coating thicknesses. As has been so well stated by you all - when it's too thick other problems are manifested. I guess, with all this, I vote for electrolessly deposited metals with carefully specified and audited process control. This means keeping oxidation to an absolute minimum (function of time copper exposed to air) and applying specified thicknesses so solder joint contamination will not be effected. I won't say again how much I am against HASL, or will I. And on it goes, Earl Moon

reply »

| | | | | I'm reviewing my board fab spec. It calls for a minimum SnPb thickness of 50 microinches on HASL PWB's. I've looked at other specs that call out anything from 30 to 80 microinches, and others that just say the copper pad must be covered and solderable. | | | | | What's reasonable? How thick are the intermetallics (typically) on HASL boards? What about the surface layer of oxide? | | | | | Thanks for your input. | | | | Chrys, | | | | Everybody is being a little rough on HASL. There are a host of other problems with alternative surface finishes. Earl knows a ton about this stuff. I'm pretty well versed on the alternative surface finishes from a process standpoint. Seen 'em all. Worked with most. OSP's are tough. You need to approve a supplier with tight process controls and definitely a conveyorized process. I have seen boards where the OSP was too thick and Alpha 609 couldn't even break it down. That OA is ACTIVE. I used it to clean rust off my golf clubs!! Nuff said. We got around it by sending the fabs through the reflow oven 3 times before processing in order to break down the OSP (Entek Plus)coating. Imidazole is a pretty cool OSP. It's a proprietary OSP at Via Systems. Dr. Lee Parker introduced it. It's pretty neat stuff. It really doesn't break down in thermal cycles. Good for double sided reflow, something that is tough for Entek. Enthone has a pretty adept chemistry staff. It's a nice coating too but you have | | | | to apply it properly. A ton of board shops throw this stuff on their boards but they don't know what the heck they're doing. Some don't even test for thickness of coating application. Imidazole is done primarily at Via Systems. They invented it. They apply it. Everything is internal. Thus, better control. As for the noble metal crap. That's what it is. I use it. I have to. If you get Seven microns of gold over nickel, you are all done. 3 to 5 microns is necessary for a decent solder joint. Any more, you'll see embrittlement, proved it. Any less, you'll see voids and oxidized nickel, proved it. By the way, good luck soldering to oxidized nickel. You also need to increase your reflow temperatures to get a shiny solder joint. Great. Now I'm thermal cycling between 25 and 255 in my reflow oven. Some components can't take that heat. If the assy. is double sided, you have to go through that thermal cycle at least twice. Silver is in it's "toddler" phase. People are still learning how to use it. You need to change your paste to allow for this in process, which will change all of your profiles. Bottom line, flat pads look like a great concept. The flatness doesn't come for free!!! I do .016" pitch all the time with HASL boards. I don't have a problem. Choosing a good vendor, having proper pad sizes and aperture geometries can cover up all the geometric anomalies of a HASL pad. As for oxidation and solderability, everything rusts or has a cost. With HASL, it can oxidize. Get an active paste. With OSP, it breaks down and you have exposed copper. Use 'em fast. With gold over nickel, you need the right thickness and the proper paste and parts. Like I said, it's all got a cost. We have to deal with these issues soon anyway. Lead won't be around forever.... | | | | I'm oudahere! That was long winded. Sorry. | | | | Best Regards, | | | | Justin | | | Justin: Thanks for your interesting comments. Questions they raise are: | | | 1 When you talk about OSP suppliers, what do you mean by "tight process controls?" | | | 2 When you talk about OSP suppliers, why "definitely conveyorized process?" | | | 3 On Entek: When you say that "you have to apply it properly," how should it be done? | | | 4 On Entek: When you say that "some don�t even check for thickness of coating application," what should they be doing? | | | 5 When you speak of not having problems doing 16 pitch HASL boards: What do you mean by: | | | � Proper pad sizes? | | | � Proper aperature geometries? | | | 6 Continuing with this last question: When laying-out proper pad sizes for 16 pitch components on boards for fab, how do you spec the web between pads? | | | Thanks | | | Dave F | | Dave, | | I can go through everything except the pad / aperture sizes. These are proprietary and the reason why Flex can build better than most. Sounds conceited but that's the contract game. You need a new fab supplier if they can't get a web in a .016" pitch device. As far as the process control question is concerned. You want to know that they are controlling their chemical changeover times. Are they using control charts on their etching process? Are they using them on their coating process? As for question 2: Conveyorization means control. These coatings are sprayed or dipped. Manual processes such as these yeild deviation. Mechanized processes aid in repeatability. That's the name of the game with this stuff. For question 3 regarding application, I only have one thing to say. We should never say how to do something, only to do it right. I don't know if dipping or spraying is better. I do know that thickness is the critical attribute. These leads to the answer to question 4. A spectrophotometer | | should be used to check coating thickness. If they don't have one of these devices or they say they can eyeball it, don't use them. Like I said, coating thickness is everything but I think it hangs somewhere around 2500 to 3500 Angstroms (don't take that as gospel). | | Best Regards and keep 'em coming, | | Justin Medernach | | Man, you guys just don't give up. I don't blame you. Never give up until it is right. | I would simply add again that the only reason we apply any type copper surface coating | is to prevent oxidation so solder wetting may be achieved. The problem with this and | other coating processes (as with HASL) first is panels come out of some kind of wet chemical | process (a microetch and rinse), are dried (causing immediate oxidation), then coated whether | with organics or metals - electroplated, electrolessly deposited, sprayed, dipped, or? | One thing organics have going for them is they are essentially fluxes. In the good old days, | instead of Entek, we used what was known as "Seal Brite." Somewhat the same difference. The | situation then, as now, at first the stuff actually fought or inhibited oxidation. Over time, | the battle was lost depending mostly on storage and useage conditions. | Therefore, a supplier must assure process control first to minimize oxidation. Then, the supplier | must demonstrate coating thickness control. With the new gold, silver, and palladium metal | coatings, the thickness must not exceed about 7 millionths of an inch or the obvious will be | effected. For organics, because of their short shelf lives, and some suppliers inability to | effectively manage processes, more surface contamination will be effected quicker with thinner | coating thicknesses. As has been so well stated by you all - when it's too thick other problems | are manifested. | I guess, with all this, I vote for electrolessly deposited metals with carefully specified and | audited process control. This means keeping oxidation to an absolute minimum (function of time | copper exposed to air) and applying specified thicknesses so solder joint contamination will not | be effected. I won't say again how much I am against HASL, or will I. | And on it goes, | Earl Moon Oh, I forgot to add the most important part (besides organic thickness, composition, and application). Immersion applied (electroless) metals are coated immediately after the previous bath (rinse) thereby avoiding any possible contamination (when process control effectively administered). This means, obviously, that these metals are the purest, flatest, most solderable coating available (again - with effective process management). Question - as surface is defined as having no depth (thickness) - how many angstroms does it take to get depth? I just forgot along with all the other stuff I wish I could remember. Enough from me forever, Earl Moon

reply »

| | | | | | I'm reviewing my board fab spec. It calls for a minimum SnPb thickness of 50 microinches on HASL PWB's. I've looked at other specs that call out anything from 30 to 80 microinches, and others that just say the copper pad must be covered and solderable. | | | | | | What's reasonable? How thick are the intermetallics (typically) on HASL boards? What about the surface layer of oxide? | | | | | | Thanks for your input. | | | | | Chrys, | | | | | Everybody is being a little rough on HASL. There are a host of other problems with alternative surface finishes. Earl knows a ton about this stuff. I'm pretty well versed on the alternative surface finishes from a process standpoint. Seen 'em all. Worked with most. OSP's are tough. You need to approve a supplier with tight process controls and definitely a conveyorized process. I have seen boards where the OSP was too thick and Alpha 609 couldn't even break it down. That OA is ACTIVE. I used it to clean rust off my golf clubs!! Nuff said. We got around it by sending the fabs through the reflow oven 3 times before processing in order to break down the OSP (Entek Plus)coating. Imidazole is a pretty cool OSP. It's a proprietary OSP at Via Systems. Dr. Lee Parker introduced it. It's pretty neat stuff. It really doesn't break down in thermal cycles. Good for double sided reflow, something that is tough for Entek. Enthone has a pretty adept chemistry staff. It's a nice coating too but you have | | | | | to apply it properly. A ton of board shops throw this stuff on their boards but they don't know what the heck they're doing. Some don't even test for thickness of coating application. Imidazole is done primarily at Via Systems. They invented it. They apply it. Everything is internal. Thus, better control. As for the noble metal crap. That's what it is. I use it. I have to. If you get Seven microns of gold over nickel, you are all done. 3 to 5 microns is necessary for a decent solder joint. Any more, you'll see embrittlement, proved it. Any less, you'll see voids and oxidized nickel, proved it. By the way, good luck soldering to oxidized nickel. You also need to increase your reflow temperatures to get a shiny solder joint. Great. Now I'm thermal cycling between 25 and 255 in my reflow oven. Some components can't take that heat. If the assy. is double sided, you have to go through that thermal cycle at least twice. Silver is in it's "toddler" phase. People are still learning how to use it. You need to change your paste to allow for this in process, which will change all of your profiles. Bottom line, flat pads look like a great concept. The flatness doesn't come for free!!! I do .016" pitch all the time with HASL boards. I don't have a problem. Choosing a good vendor, having proper pad sizes and aperture geometries can cover up all the geometric anomalies of a HASL pad. As for oxidation and solderability, everything rusts or has a cost. With HASL, it can oxidize. Get an active paste. With OSP, it breaks down and you have exposed copper. Use 'em fast. With gold over nickel, you need the right thickness and the proper paste and parts. Like I said, it's all got a cost. We have to deal with these issues soon anyway. Lead won't be around forever.... | | | | | I'm oudahere! That was long winded. Sorry. | | | | | Best Regards, | | | | | Justin | | | | Justin: Thanks for your interesting comments. Questions they raise are: | | | | 1 When you talk about OSP suppliers, what do you mean by "tight process controls?" | | | | 2 When you talk about OSP suppliers, why "definitely conveyorized process?" | | | | 3 On Entek: When you say that "you have to apply it properly," how should it be done? | | | | 4 On Entek: When you say that "some don�t even check for thickness of coating application," what should they be doing? | | | | 5 When you speak of not having problems doing 16 pitch HASL boards: What do you mean by: | | | | � Proper pad sizes? | | | | � Proper aperature geometries? | | | | 6 Continuing with this last question: When laying-out proper pad sizes for 16 pitch components on boards for fab, how do you spec the web between pads? | | | | Thanks | | | | Dave F | | | Dave, | | | I can go through everything except the pad / aperture sizes. These are proprietary and the reason why Flex can build better than most. Sounds conceited but that's the contract game. You need a new fab supplier if they can't get a web in a .016" pitch device. As far as the process control question is concerned. You want to know that they are controlling their chemical changeover times. Are they using control charts on their etching process? Are they using them on their coating process? As for question 2: Conveyorization means control. These coatings are sprayed or dipped. Manual processes such as these yeild deviation. Mechanized processes aid in repeatability. That's the name of the game with this stuff. For question 3 regarding application, I only have one thing to say. We should never say how to do something, only to do it right. I don't know if dipping or spraying is better. I do know that thickness is the critical attribute. These leads to the answer to question 4. A spectrophotometer | | | should be used to check coating thickness. If they don't have one of these devices or they say they can eyeball it, don't use them. Like I said, coating thickness is everything but I think it hangs somewhere around 2500 to 3500 Angstroms (don't take that as gospel). | | | Best Regards and keep 'em coming, | | | Justin Medernach | | | | Man, you guys just don't give up. I don't blame you. Never give up until it is right. | | I would simply add again that the only reason we apply any type copper surface coating | | is to prevent oxidation so solder wetting may be achieved. The problem with this and | | other coating processes (as with HASL) first is panels come out of some kind of wet chemical | | process (a microetch and rinse), are dried (causing immediate oxidation), then coated whether | | with organics or metals - electroplated, electrolessly deposited, sprayed, dipped, or? | | One thing organics have going for them is they are essentially fluxes. In the good old days, | | instead of Entek, we used what was known as "Seal Brite." Somewhat the same difference. The | | situation then, as now, at first the stuff actually fought or inhibited oxidation. Over time, | | the battle was lost depending mostly on storage and useage conditions. | | Therefore, a supplier must assure process control first to minimize oxidation. Then, the supplier | | must demonstrate coating thickness control. With the new gold, silver, and palladium metal | | coatings, the thickness must not exceed about 7 millionths of an inch or the obvious will be | | effected. For organics, because of their short shelf lives, and some suppliers inability to | | effectively manage processes, more surface contamination will be effected quicker with thinner | | coating thicknesses. As has been so well stated by you all - when it's too thick other problems | | are manifested. | | I guess, with all this, I vote for electrolessly deposited metals with carefully specified and | | audited process control. This means keeping oxidation to an absolute minimum (function of time | | copper exposed to air) and applying specified thicknesses so solder joint contamination will not | | be effected. I won't say again how much I am against HASL, or will I. | | And on it goes, | | Earl Moon | Oh, I forgot to add the most important part (besides organic thickness, composition, and application). | Immersion applied (electroless) metals are coated immediately after the previous bath (rinse) thereby | avoiding any possible contamination (when process control effectively administered). This means, obviously, | that these metals are the purest, flatest, most solderable coating available (again - with effective | process management). | Question - as surface is defined as having no depth (thickness) - how many angstroms does it take | to get depth? I just forgot along with all the other stuff I wish I could remember. | Enough from me forever, | Earl Moon I wish I knew all the stuff you forgot, too....

reply »

Hey yooze guys! Enough of the jokes between you and Earl (GRIN), let's get back to the topic...well, okay... you can tell a few more...(but good ones, 'kay?) Anyhoo...I'm not sure if ya'll are on the IPC Technet list or not, I've not seen any posts from ya'll, but there was a post that I just read that I thought pretty interesting. One; because I've never heard of this before, and two; it relates to this discussion quite well. I'll paste it here below: Subj: Re: [TN] HASL alternatives Date: 5/27/98 8:02:50 AM Pacific Daylight Time From: ed@GEICIRCUITS.COM (Ed Cosper) Sender: TechNet@IPC.ORG (TechNet) Reply-to: TechNet@IPC.ORG (TechNet E-Mail Forum), ed@GEICIRCUITS.COM (Ed Cosper) To: TechNet@IPC.ORG

I have seen a lot of traffic on different alternatives to HASL. Most seem to refer to various OSP's or some form of precious metal. However, I would like to offer a different approach. Several years ago I worked with a PCB supplier who was requested to supply parts with a bare copper finish. No protective coating at all. The last operation in house was to clean the surface, rinse with DI, dry and then parts were packaged in groups of 10 and vacuum sealed. We placed a strip of gray anti-oxidation paper from 3M in the package before sealing. According to our customer, it was easier for them to open a package of 10, and clean before fluxing with a sulfuric than it was to fight coplanarity issues of the solder. They had tried an OSP but found it to be inconsistent. The practice spread to two other customers before I had left that supplier. I haven't found anyone else even remotely interested in that practice since, but apparently it worked for at least 3 of our customers for over a year without any reported issues. I'm not sure if it is still being used, but I thought I'd share that experience with the rest of the net. Thanks, Ed Cosper Graphic Electronics Tulsa, OK So whatcha' think? Is this something that sounds like it would work? -Steve Gregory-

reply »

| Hey yooze guys! | Enough of the jokes between you and Earl (GRIN), let's get back to the topic...well, okay... you can tell a few more...(but good ones, 'kay?) | Anyhoo...I'm not sure if ya'll are on the IPC Technet list or not, I've not seen any posts from ya'll, but there was a post that I just read that I thought pretty interesting. One; because I've never heard of this before, and two; it relates to this discussion quite well. I'll paste it here below: | Subj: Re: [TN] HASL alternatives | Date: 5/27/98 8:02:50 AM Pacific Daylight Time | From: ed@GEICIRCUITS.COM (Ed Cosper) | Sender: TechNet@IPC.ORG (TechNet) | Reply-to: TechNet@IPC.ORG (TechNet E-Mail Forum), ed@GEICIRCUITS.COM (Ed Cosper) | To: TechNet@IPC.ORG | | I have seen a lot of traffic on different alternatives to HASL. Most seem to refer to various OSP's or some form of precious metal. However, I would like to offer a different approach. | Several years ago I worked with a PCB supplier who was requested to supply parts with a bare copper finish. No protective coating at all. | The last operation in house was to clean the surface, rinse with DI, dry and then parts were packaged in groups of 10 and vacuum sealed. We placed a strip of gray anti-oxidation paper from 3M in the package before sealing. | According to our customer, it was easier for them to open a package of 10, and clean before fluxing with a sulfuric than it was to fight coplanarity issues of the solder. | They had tried an OSP but found it to be inconsistent. The practice spread to two other customers before I had left that supplier. I haven't found anyone else even remotely interested in that practice since, but apparently it worked for at least 3 of our customers for over a year without any reported issues. I'm not sure if it is still being used, but I thought I'd share that experience with the rest of the net. | Thanks, | Ed Cosper | Graphic Electronics | Tulsa, OK | So whatcha' think? Is this something that sounds like it would work? | | -Steve Gregory- Boy, New startups have dampened your acute sense of humor, or mine. All we've ever wanted, in the history of "boardom," was a bare metal surface (copper obviously being first choice) that would need no coating. The reason we had to find coatings was to protect from oxidation. The idea Ed discusses is not new. It has been tried, and failed (in most common applications) first because of stringent controls needed to clean and protect the metal surface before packaging (scratch a penny and see how fast it looses its shine in the scratched area). Second, who wants to install an acidic rins in their assembly operation. Third, now you have to rinse the panels after cleaning and dry them (causing oxidation that may not be removable) before assembling and soldering operations. Under ideal conditions, as described, with everything described above, the system would work fine. Remember, IBM was the first real champion of OSP's in the mid to late 1980's. They were heavy into SMT before most of us (not me of course) had an oxide free gleam in our eyes. They did it because it meant freedom from HASL. Again, now we want it all. A smooth, flat, shiny, solderable surface. Well, you can have it with rapidly becoming acceptable, more available (as demand increases - and it will) precious metal finishes. After all, these are relatively new processes (over and over again throughout PCB fab history), and now we are talking about 7 millionths of an inch of electrolessly (dipped) deposited metal versus 7-50 or more millionths of an inch of gold that was electroplated by, as an example, HP back in the $36/oz. pre gold rush days. When demand increases, available will increase and prices will drop. Other adversities will be overcome, as they always are, concerning solder mask compatibility during processing. I just don't see any other answers unless an additive precious metal (no oxidizing, solderable, conductive surface) process is devised for all us soldering junkies. This is a world where there are few things new when it comes to PCB's. We just keep doing it over until it gets better. It's called evolution and re-evolution. The breakthroughs, therefore demands, are coming from all you supersonic SMT war lords taking no prisoners. Earl Moon

reply »

| Hey yooze guys! | Enough of the jokes between you and Earl (GRIN), let's get back to the topic...well, okay... you can tell a few more...(but good ones, 'kay?) | Anyhoo...I'm not sure if ya'll are on the IPC Technet list or not, I've not seen any posts from ya'll, but there was a post that I just read that I thought pretty interesting. One; because I've never heard of this before, and two; it relates to this discussion quite well. I'll paste it here below: | Subj: Re: [TN] HASL alternatives | Date: 5/27/98 8:02:50 AM Pacific Daylight Time | From: ed@GEICIRCUITS.COM (Ed Cosper) | Sender: TechNet@IPC.ORG (TechNet) | Reply-to: TechNet@IPC.ORG (TechNet E-Mail Forum), ed@GEICIRCUITS.COM (Ed Cosper) | To: TechNet@IPC.ORG | | I have seen a lot of traffic on different alternatives to HASL. Most seem to refer to various OSP's or some form of precious metal. However, I would like to offer a different approach. | Several years ago I worked with a PCB supplier who was requested to supply parts with a bare copper finish. No protective coating at all. | The last operation in house was to clean the surface, rinse with DI, dry and then parts were packaged in groups of 10 and vacuum sealed. We placed a strip of gray anti-oxidation paper from 3M in the package before sealing. | According to our customer, it was easier for them to open a package of 10, and clean before fluxing with a sulfuric than it was to fight coplanarity issues of the solder. | They had tried an OSP but found it to be inconsistent. The practice spread to two other customers before I had left that supplier. I haven't found anyone else even remotely interested in that practice since, but apparently it worked for at least 3 of our customers for over a year without any reported issues. I'm not sure if it is still being used, but I thought I'd share that experience with the rest of the net. | Thanks, | Ed Cosper | Graphic Electronics | Tulsa, OK | So whatcha' think? Is this something that sounds like it would work? | | -Steve Gregory- Boy, New startups have dampened your acute sense of humor, or mine. All we've ever wanted, in the history of "boardom," was a bare metal surface (copper obviously being first choice) that would need no coating. The reason we had to find coatings was to protect from oxidation. The idea Ed discusses is not new. It has been tried, and failed (in most common applications) first because of stringent controls needed to clean and protect the metal surface before packaging (scratch a penny and see how fast it looses its shine in the scratched area). Second, who wants to install an acidic rins in their assembly operation. Third, now you have to rinse the panels after cleaning and dry them (causing oxidation that may not be removable) before assembling and soldering operations. Under ideal conditions, as described, with everything described above, the system would work fine. Remember, IBM was the first real champion of OSP's in the mid to late 1980's. They were heavy into SMT before most of us (not me of course) had an oxide free gleam in our eyes. They did it because it meant freedom from HASL. Again, now we want it all. A smooth, flat, shiny, solderable surface. Well, you can have it with rapidly becoming acceptable, more available (as demand increases - and it will) precious metal finishes. After all, these are relatively new processes (over and over again throughout PCB fab history), and now we are talking about 7 millionths of an inch of electrolessly (dipped) deposited metal versus 7-50 or more millionths of an inch of gold that was electroplated by, as an example, HP back in the $36/oz. pre gold rush days. When demand increases, available will increase and prices will drop. Other adversities will be overcome, as they always are, concerning solder mask compatibility during processing. I just don't see any other answers unless an additive precious metal (no oxidizing, solderable, conductive surface) process is devised for all us soldering junkies. This is a world where there are few things new when it comes to PCB's. We just keep doing it over until it gets better. It's called evolution and re-evolution. The breakthroughs, therefore demands, are coming from all you supersonic SMT war lords taking no prisoners. Earl Moon

reply »