Suspected Pink Ring

While inspecting incoming PCBs we detected around all plated-through-holes (only there and all in the same place) some halfmoonshaped light rings that match obviously only with the pink ring symptom shown in the IPC-A-600 chapter 2.5.2.. That would mean the PCBs are acceptable but we are not quite sure about our judgement. Can anybody tell me what a Pink Ring actually is and what the cause of it is or if we face maybe another more serious effect we have to worry about ? The PCB-vendor isn�t much help for he always says it�s OK and I�m rather more in DFM (CE is what I dream of) and SMT-processes than in the material stuff.

Thanks in advance

Wolfgang

PS: Could have attached a picture if I already knew how to put it somewhere in the net.

reply »

| While inspecting incoming PCBs we detected around all plated-through-holes (only there and all in the same place) some halfmoonshaped light rings that match obviously only with the pink ring symptom shown in the IPC-A-600 chapter 2.5.2.. That would mean the PCBs are acceptable but we are not quite sure about our judgement. | Can anybody tell me what a Pink Ring actually is and what the cause of it is or if we face maybe another more serious effect we have to worry about ? | The PCB-vendor isn�t much help for he always says it�s OK and I�m rather more in DFM (CE is what I dream of) and SMT-processes than in the material stuff. | | Thanks in advance | | Wolfgang | | PS: Could have attached a picture if I already knew how to put it somewhere in the net. | | | Wolfgang,

I suspect you already know most major pink ring causes. One is an acid solution (H2SO4 usually) in a wet chemical process line (plating) reacting with an oxide treatment used for lamination effectiveness. This especially is possible with black (cupric) oxide copper/innerlayer treatments. It usually is not present with red/brown (cuprous) oxide treatments associated mostly with non epoxy resin systems - polyimide as an example.

If pink ring conditions exceed specified requirements (as in your well stated IPC rendition), possible failure may result under certain conditions (usually thermal). These failures manifest themselves as delamination and propagation into hole wall plating as cracking thus rendering unacceptable defects. This especially is true concerning unsupported holes used in SMT applications/designs.

Ask your supplier about which oxide treatment is used and how it is controlled to prevent pink ring. Ask other suppliers, if your present supplier remains "ignorant or stupid" about other options and what process management requirements are in place to prevent such possible anomalies.

Regards,

Earl Moon

reply »

| | While inspecting incoming PCBs we detected around all plated-through-holes (only there and all in the same place) some halfmoonshaped light rings that match obviously only with the pink ring symptom shown in the IPC-A-600 chapter 2.5.2.. That would mean the PCBs are acceptable but we are not quite sure about our judgement. | | Can anybody tell me what a Pink Ring actually is and what the cause of it is or if we face maybe another more serious effect we have to worry about ? | | The PCB-vendor isn�t much help for he always says it�s OK and I�m rather more in DFM (CE is what I dream of) and SMT-processes than in the material stuff. | | | | Thanks in advance | | | | Wolfgang | | | | PS: Could have attached a picture if I already knew how to put it somewhere in the net. | | | | | | | Wolfgang, | | I suspect you already know most major pink ring causes. One is an acid solution (H2SO4 usually) in a wet chemical process line | (plating) reacting with an oxide treatment used for lamination effectiveness. This especially is possible with black (cupric) oxide copper/innerlayer treatments. It usually is not present with red/brown (cuprous) oxide treatments associated mostly with non epoxy resin systems - polyimide as an example. | | If pink ring conditions exceed specified requirements (as in your well stated IPC rendition), possible failure may result under certain conditions (usually thermal). These failures manifest themselves as delamination and propagation into hole wall plating as cracking thus rendering unacceptable defects. This especially is true concerning unsupported holes used in SMT applications/designs. | | Ask your supplier about which oxide treatment is used and how it is controlled to prevent pink ring. Ask other suppliers, if your present supplier remains "ignorant or stupid" about other options and what process management requirements are in place to prevent such possible anomalies. | | Regards, | | Earl Moon | hi Earl, it�s the first time I noticed something like this. Strange thing for me is that it occurs on all boards round all through-holes (about 2500 to 3000 per board) always in the same sector of the hole surounding it like a halo for about 120� and a width of about .5mm to .8mm. It�s a 6-layer PCB in FR4 and I suppose they used this black oxide treatment you mentioned. For we had enough trouble of all kind with this particular supplier we already changed our source for PCBs this is just his last delivery and due to no-time-to-get-new-ones we have to decide if we can use them (IPC rates Pink Ring acceptable Class 1,2,3) or if we have to bite in that lemon called get them new. Being not so familiar with all english technical expressions I suppose Pink Ring is just used to describe anomalies with a special similar look but different causes as you mentioned. Different causes I can think of are delamination or mechanical influence while drilling. Prior to soldering and after one test in the vapourphase they are electrically ok and no further increase of the defect is noticed. So maybe we use this lot at our own risk but keep an eye on it and maybe learn something for the future.

Thanks Wolfgang

reply »

| | | While inspecting incoming PCBs we detected around all plated-through-holes (only there and all in the same place) some halfmoonshaped light rings that match obviously only with the pink ring symptom shown in the IPC-A-600 chapter 2.5.2.. That would mean the PCBs are acceptable but we are not quite sure about our judgement. | | | Can anybody tell me what a Pink Ring actually is and what the cause of it is or if we face maybe another more serious effect we have to worry about ? | | | The PCB-vendor isn�t much help for he always says it�s OK and I�m rather more in DFM (CE is what I dream of) and SMT-processes than in the material stuff. | | | | | | Thanks in advance | | | | | | Wolfgang | | | | | | PS: Could have attached a picture if I already knew how to put it somewhere in the net. | | | | | | | | | | | Wolfgang, | | | | I suspect you already know most major pink ring causes. One is an acid solution (H2SO4 usually) in a wet chemical process line | | (plating) reacting with an oxide treatment used for lamination effectiveness. This especially is possible with black (cupric) oxide copper/innerlayer treatments. It usually is not present with red/brown (cuprous) oxide treatments associated mostly with non epoxy resin systems - polyimide as an example. | | | | If pink ring conditions exceed specified requirements (as in your well stated IPC rendition), possible failure may result under certain conditions (usually thermal). These failures manifest themselves as delamination and propagation into hole wall plating as cracking thus rendering unacceptable defects. This especially is true concerning unsupported holes used in SMT applications/designs. | | | | Ask your supplier about which oxide treatment is used and how it is controlled to prevent pink ring. Ask other suppliers, if your present supplier remains "ignorant or stupid" about other options and what process management requirements are in place to prevent such possible anomalies. | | | | Regards, | | | | Earl Moon | | | hi Earl, | it�s the first time I noticed something like this. Strange thing for me is that it occurs on all boards round all through-holes (about 2500 to 3000 per board) always in the same sector of the hole surounding it like a halo for about 120� and a width of about .5mm to .8mm. | It�s a 6-layer PCB in FR4 and I suppose they used this black oxide treatment you mentioned. | For we had enough trouble of all kind with this particular supplier we already changed our source for PCBs this is just his last delivery and due to no-time-to-get-new-ones we have to decide if we can use them (IPC rates Pink Ring acceptable Class 1,2,3) or if we have to bite in that lemon called get them new. | Being not so familiar with all english technical expressions I suppose Pink Ring is just used to describe anomalies with a special similar look but different causes as you mentioned. | Different causes I can think of are delamination or mechanical influence while drilling. | Prior to soldering and after one test in the vapourphase they are electrically ok and no further increase of the defect is noticed. So maybe we use this lot at our own risk but keep an eye on it and maybe learn something for the future. | | Thanks Wolfgang | Wolfgang,

It would be nice if you could qualify suppliers in the too brief manner I described in my last article for the SMTnet Express. Failing that, here's the news.

First, pink ring is a form of delamination on a scale from acceptable (never to me) and just beyond. I mean the acid attacks the oxide treatment, now part of the MLB, and renders it mush (not adhesion or laminate bond strength in the construction area affected around the plated hole barrels and pads). X-Sectional analysis, using quality conformance test circuitry and attendant test coupons, provides all the evidence you need.

Second, if you've no test circuitry you'll have to cut up a board at defect locations. Then, qualify the pink ring/delamination as acceptable or not, or is it really what we're talking about.

Third, use the thermal stress test requirements in IPC somewhere, now lost to me after the demise of 55110 days. It may be in 6012 or 650. Someone keeping track of all this stuff knows. Anyway, perform the test using a specimen prepared as required and perform the solder float test at 550 degrees F. for 10 seconds. Then, pot the specimen and examine it in accordance with whatever requirement.

If the defect/anomaly has propagated into the hole walls, serious defect has been detected and the board lots are subject to rejection. If you proceed with using the boards as received, and you find opens (as an example), trace it back to the pink ring and the delamination into the hole wall.

There's much more and on it goes.

Best wishes for a happy holiday season, whateve that may be,

Earl Moon

reply »

| While inspecting incoming PCBs we detected around all plated-through-holes (only there and all in the same place) some halfmoonshaped light rings that match obviously only with the pink ring symptom shown in the IPC-A-600 chapter 2.5.2.. That would mean the PCBs are acceptable but we are not quite sure about our judgement. | Can anybody tell me what a Pink Ring actually is and what the cause of it is or if we face maybe another more serious effect we have to worry about ? | The PCB-vendor isn�t much help for he always says it�s OK and I�m rather more in DFM (CE is what I dream of) and SMT-processes than in the material stuff. | | Thanks in advance | | Wolfgang | | PS: Could have attached a picture if I already knew how to put it somewhere in the net. | | | Wolfgang: Earl is spot on, but let me spin this a little differently.

Epoxy does not adhere well to untreated copper surfaces. So, many fabricators use some color of copper-oxide treatment to rough-up the copper surface. The color varies according to the density of the copper-oxide crystals and the ratio of cupric-oxide to cuprous-oxide. If drilling damages the epoxy-copper interface, the dark copper-oxide will dissolve during the plating process, leaving a distinct pink ring around holes.

Many consider this a cosmetic defect. The effect on long term reliability is in controversy. You must decide for your self, as Earl has indicated.

Baking after drilling is a fix for boards damaged by aggressive drilling. It is a repair operation that increases the fabricator's COGS. It�s not the prevention you desire. I�d be interested in your fabricator�s approach to preventing the occurrence of pink ring. You should hear thing like:

� Replacing the black oxide surface treatment with a reduced oxide treatment, as Earl mentioned, is a good approach. � A better approach would be less aggressive drilling with (1) fewer hits on a drill before changing it out and (2) drilling with fewer panels in a book. � A better approach yet would be combining (1) replacing the black oxide surface treatment with a reduced oxide treatment and (2) less aggressive drilling.

Good luck

Dave F

reply »

| | While inspecting incoming PCBs we detected around all plated-through-holes (only there and all in the same place) some halfmoonshaped light rings that match obviously only with the pink ring symptom shown in the IPC-A-600 chapter 2.5.2.. That would mean the PCBs are acceptable but we are not quite sure about our judgement. | | Can anybody tell me what a Pink Ring actually is and what the cause of it is or if we face maybe another more serious effect we have to worry about ? | | The PCB-vendor isn�t much help for he always says it�s OK and I�m rather more in DFM (CE is what I dream of) and SMT-processes than in the material stuff. | | | | Thanks in advance | | | | Wolfgang | | | | PS: Could have attached a picture if I already knew how to put it somewhere in the net. | | | | | | | Wolfgang: Earl is spot on, but let me spin this a little differently. | | Epoxy does not adhere well to untreated copper surfaces. So, many fabricators use some color of copper-oxide treatment to rough-up the copper surface. The color varies according to the density of the copper-oxide crystals and the ratio of cupric-oxide to cuprous-oxide. If drilling damages the epoxy-copper interface, the dark copper-oxide will dissolve during the plating process, leaving a distinct pink ring around holes. | | Many consider this a cosmetic defect. The effect on long term reliability is in controversy. You must decide for your self, as Earl has indicated. | | Baking after drilling is a fix for boards damaged by aggressive drilling. It is a repair operation that increases the fabricator's COGS. It�s not the prevention you desire. I�d be interested in your fabricator�s approach to preventing the occurrence of pink ring. You should hear thing like: | | � Replacing the black oxide surface treatment with a reduced oxide treatment, as Earl mentioned, is a good approach. | � A better approach would be less aggressive drilling with (1) fewer hits on a drill before changing it out and (2) drilling with fewer panels in a book. | � A better approach yet would be combining (1) replacing the black oxide surface treatment with a reduced oxide treatment and (2) less aggressive drilling. | | Good luck | | Dave F | Dave, Nice going.

Now that we've drive Wolfgang away to other issues more controllable, I'll add that your gentle description of aggressive drilling is very nice. Actually, when drill rates, as feeds and speeds primarily, are out of control, holes get punched and gouged. Likewise, drill spindle runout causes problems.

My issue with all this is simple. I haven't seen or heard of pink ring in at least 15 years. Too much good living I guess. Supplier re-evaluation is in order, but I think you know that Wolfgang though it was a nice stroll down unmemorable lane.

Earl Moon

reply »

| | | While inspecting incoming PCBs we detected around all plated-through-holes (only there and all in the same place) some halfmoonshaped light rings that match obviously only with the pink ring symptom shown in the IPC-A-600 chapter 2.5.2.. That would mean the PCBs are acceptable but we are not quite sure about our judgement. | | | Can anybody tell me what a Pink Ring actually is and what the cause of it is or if we face maybe another more serious effect we have to worry about ? | | | The PCB-vendor isn�t much help for he always says it�s OK and I�m rather more in DFM (CE is what I dream of) and SMT-processes than in the material stuff. | | | | | | Thanks in advance | | | | | | Wolfgang | | | | | | PS: Could have attached a picture if I already knew how to put it somewhere in the net. | | | | | | | | | | | Wolfgang: Earl is spot on, but let me spin this a little differently. | | | | Epoxy does not adhere well to untreated copper surfaces. So, many fabricators use some color of copper-oxide treatment to rough-up the copper surface. The color varies according to the density of the copper-oxide crystals and the ratio of cupric-oxide to cuprous-oxide. If drilling damages the epoxy-copper interface, the dark copper-oxide will dissolve during the plating process, leaving a distinct pink ring around holes. | | | | Many consider this a cosmetic defect. The effect on long term reliability is in controversy. You must decide for your self, as Earl has indicated. | | | | Baking after drilling is a fix for boards damaged by aggressive drilling. It is a repair operation that increases the fabricator's COGS. It�s not the prevention you desire. I�d be interested in your fabricator�s approach to preventing the occurrence of pink ring. You should hear thing like: | | | | � Replacing the black oxide surface treatment with a reduced oxide treatment, as Earl mentioned, is a good approach. | | � A better approach would be less aggressive drilling with (1) fewer hits on a drill before changing it out and (2) drilling with fewer panels in a book. | | � A better approach yet would be combining (1) replacing the black oxide surface treatment with a reduced oxide treatment and (2) less aggressive drilling. | | | | Good luck | | | | Dave F | | | Dave, Nice going. | | Now that we've drive Wolfgang away to other issues more controllable, I'll add that your gentle description of aggressive drilling is very nice. Actually, when drill rates, as feeds and speeds primarily, are out of control, holes get punched and gouged. Likewise, drill spindle runout causes problems. | | My issue with all this is simple. I haven't seen or heard of pink ring in at least 15 years. Too much good living I guess. Supplier re-evaluation is in order, but I think you know that Wolfgang though it was a nice stroll down unmemorable lane. | | Earl Moon | | I�m in the SMT business now for about 10 years, starting with two solderirons, a lot off empty space, a bunch of catalogues and an almost completely in SMT CPU-board for a VME-based-highspeed-computersystem to build. It ended up in two inlneplacers, IR-reflow and wavesoldering and a couple of 1000 units throughput. The last years there was a new start with rising a prototyp and teaching place and doing some consulting for the local industry around Hamburg. Now things changed and I�m at the front again trying to improve the design of my present company to meet the needs of the assembly. As you haven�t seen pink ring for years, I have seen it the first time and the answer of the supplier was that we shouldn�t look with such a magnification at the boards. What you both said was a lot of help for me and I think I know what to do.

Didn�t get the last sentence Wolfgang

reply »

| | | | While inspecting incoming PCBs we detected around all plated-through-holes (only there and all in the same place) some halfmoonshaped light rings that match obviously only with the pink ring symptom shown in the IPC-A-600 chapter 2.5.2.. That would mean the PCBs are acceptable but we are not quite sure about our judgement. | | | | Can anybody tell me what a Pink Ring actually is and what the cause of it is or if we face maybe another more serious effect we have to worry about ? | | | | The PCB-vendor isn�t much help for he always says it�s OK and I�m rather more in DFM (CE is what I dream of) and SMT-processes than in the material stuff. | | | | | | | | Thanks in advance | | | | | | | | Wolfgang | | | | | | | | PS: Could have attached a picture if I already knew how to put it somewhere in the net. | | | | | | | | | | | | | | | Wolfgang: Earl is spot on, but let me spin this a little differently. | | | | | | Epoxy does not adhere well to untreated copper surfaces. So, many fabricators use some color of copper-oxide treatment to rough-up the copper surface. The color varies according to the density of the copper-oxide crystals and the ratio of cupric-oxide to cuprous-oxide. If drilling damages the epoxy-copper interface, the dark copper-oxide will dissolve during the plating process, leaving a distinct pink ring around holes. | | | | | | Many consider this a cosmetic defect. The effect on long term reliability is in controversy. You must decide for your self, as Earl has indicated. | | | | | | Baking after drilling is a fix for boards damaged by aggressive drilling. It is a repair operation that increases the fabricator's COGS. It�s not the prevention you desire. I�d be interested in your fabricator�s approach to preventing the occurrence of pink ring. You should hear thing like: | | | | | | � Replacing the black oxide surface treatment with a reduced oxide treatment, as Earl mentioned, is a good approach. | | | � A better approach would be less aggressive drilling with (1) fewer hits on a drill before changing it out and (2) drilling with fewer panels in a book. | | | � A better approach yet would be combining (1) replacing the black oxide surface treatment with a reduced oxide treatment and (2) less aggressive drilling. | | | | | | Good luck | | | | | | Dave F | | | | | Dave, Nice going. | | | | Now that we've drive Wolfgang away to other issues more controllable, I'll add that your gentle description of aggressive drilling is very nice. Actually, when drill rates, as feeds and speeds primarily, are out of control, holes get punched and gouged. Likewise, drill spindle runout causes problems. | | | | My issue with all this is simple. I haven't seen or heard of pink ring in at least 15 years. Too much good living I guess. Supplier re-evaluation is in order, but I think you know that Wolfgang though it was a nice stroll down unmemorable lane. | | | | Earl Moon | | | | | I�m in the SMT business now for about 10 years, starting with two solderirons, a lot off empty space, a bunch of catalogues and an almost completely in SMT CPU-board for a VME-based-highspeed-computersystem to build. It ended up in two inlneplacers, IR-reflow and wavesoldering and a couple of 1000 units throughput. The last years there was a new start with rising a prototyp and teaching place and doing some consulting for the local industry around Hamburg. Now things changed and I�m at the front again trying to improve the design of my present company to meet the needs of the assembly. | As you haven�t seen pink ring for years, I have seen it the first time and the answer of the supplier was that we shouldn�t look with such a magnification at the boards. | What you both said was a lot of help for me and I think I know what to do. | | Didn�t get the last sentence | Wolfgang | | Wolfgang,

Thanks for the insight. I was sure you were very successful and would make the right decision.

The last sentence referred to my inability to remember all I forgot. Also, it refers to an issue best left in the past - until it rears its ugly head, as it does always.

Respectfully,

Earl Moon

reply »