Metal based PCB reflow.

I am starting a new project which involve 3 oz copper layer with 20 mil components on the top of the board and 0.040" thick copper with D2Pak on the bottom of the board. There is anybody has experience on metal-based PCB reflow process? I guess preheat will be a challenge.

Thanks

Gang

reply »

| I am starting a new project which involve 3 oz copper layer with 20 mil components on the top of the board and 0.040" thick copper with D2Pak on the bottom of the board. | There is anybody has experience on metal-based PCB reflow process? I guess preheat will be a challenge. | | | Thanks | | Gang | | Looks like it has Military/Aerospace written all over it. Your challenge will be more than one. I can only conclude from your posting you're using 3 oz. copper foil (4.5 mils) to cover the entire surface as a heat sink/electrical plane. You're using 40 mil thick copper on the bottom as a heat sink.

If this is true, you will have a potato chip after soldering. I'm sure you've considered this and have a fix available.

I've done this type reflow before. However it was accomplished using vapor-phase soldering techniques. I have no experience doing it with convection type reflow, though it probably is possible depending on how robust the reflow system and its ramp/soak capability.

I will say our big bad BTU even stalls with boards having too much thermal mass. Even so, its profile can be adjusted to accomplish difficult tasks as it has about 15' or so to convey product through many zones and thermal settings.

Earl Moon

reply »

| | I am starting a new project which involve 3 oz copper layer with 20 mil components on the top of the board and 0.040" thick copper with D2Pak on the bottom of the board. | | There is anybody has experience on metal-based PCB reflow process? I guess preheat will be a challenge. | | | | | | Thanks | | | | Gang | | | | | Looks like it has Military/Aerospace written all over it. Your challenge will be more than one. I can only conclude from your posting you're using 3 oz. copper foil (4.5 mils) to cover the entire surface as a heat sink/electrical plane. You're using 40 mil thick copper on the bottom as a heat sink. | | If this is true, you will have a potato chip after soldering. I'm sure you've considered this and have a fix available. | | I've done this type reflow before. However it was accomplished using vapor-phase soldering techniques. I have no experience doing it with convection type reflow, though it probably is possible depending on how robust the reflow system and its ramp/soak capability. | | I will say our big bad BTU even stalls with boards having too much thermal mass. Even so, its profile can be adjusted to accomplish difficult tasks as it has about 15' or so to convey product through many zones and thermal settings. | | Earl Moon |

Thanks for your response. I am going to use top layer foil for most of circuit traces and bottom layer to act like buss bars to handle high current.So it won't be a one big flat copper. Do you think I am still going to get "potato chips" ?

Gang

reply »

| | | I am starting a new project which involve 3 oz copper layer with 20 mil components on the top of the board and 0.040" thick copper with D2Pak on the bottom of the board. | | | There is anybody has experience on metal-based PCB reflow process? I guess preheat will be a challenge. | | | | | | | | | Thanks | | | | | | Gang | | | | | | | | Looks like it has Military/Aerospace written all over it. Your challenge will be more than one. I can only conclude from your posting you're using 3 oz. copper foil (4.5 mils) to cover the entire surface as a heat sink/electrical plane. You're using 40 mil thick copper on the bottom as a heat sink. | | | | If this is true, you will have a potato chip after soldering. I'm sure you've considered this and have a fix available. | | | | I've done this type reflow before. However it was accomplished using vapor-phase soldering techniques. I have no experience doing it with convection type reflow, though it probably is possible depending on how robust the reflow system and its ramp/soak capability. | | | | I will say our big bad BTU even stalls with boards having too much thermal mass. Even so, its profile can be adjusted to accomplish difficult tasks as it has about 15' or so to convey product through many zones and thermal settings. | | | | Earl Moon | | | | Thanks for your response. I am going to use top layer foil for most of circuit traces and bottom layer to act like buss bars to handle high current.So it won't be a one big flat copper. Do you think I am still going to get "potato chips" ? | | Gang | I still don't know the MLB construction, assuming it is a multilayer type. Yes, you are going to have some warp and twist problems from where I sit at this time.

Many years ago, I made the mistake of accepting a huge order for 5 layer boards consisting of three ounce copper top side and eight ounce on the other much as you describe. We were able to etch the top outers but had much trouble etching and assembling the bottom. Even when we used "noflow" prepreg to attach the heavy copper, we corrupted the whole process during assembly even with stiffners and other support mechanisms.

Of course, the best way to be sure is to experiment. You might use an unbalanced MLB construction with a heavy copper plane next to the surface of the light weight copper side. There are options, but balanced constructions are best.

I'm sure you cannot, but can you place the heavy copper after assembly? Just a thought.

Earl Moon

reply »