Frank here from PulsarProFX.com. We have found a much better way to reflow your solder paste. We've used the "hot plate" (closed coil type) from ToastMaster "Eclipse" model #6431. The idea here is to radiate heat directly up the board and you get very fast reflows in about 30 seconds when dialed up to "2" on the thermostat. The trick is to also make a simple "holder" out of thin aluminum sheet ... about 3" wide and form a "Z" bend. Top is the handle and bottom is where you rest your board. The objective is to be able to lift the board off the hot plate after the solder paste has all reflowed without shaking any of the parts out of position while the solder is still molten. Just carefully pick it up and lay it down over a wood surface or something that can stand the heat. The paste sets up in about 5 seconds and voila!
I've always used SRA-solder.com (Stan Rubenstein's business of making solder paste dispensers. He was a real nice guy but is now pushing up the daisies.) I think they still sell leaded paste. Anyway, they might give you a hard time about "Oh, it must ship 2nd Day Air" due to stability", etc. For hobby use, no sweat as it can be a room temp or even hot for quite a period of time. I've gone over a year on one tube and it still pushes out just fine for proto's and 1-ups.
This method will also work with our new "FlexPCB" but may take a few more seconds since it's not "heavy" to lay flat all the time because it's so thin.
I'm a fan of Zephryrtronics paste -- they make it fresh, ship it right away, and it has a 1-year shelf life if refrigerated. Truth is, I've use 18-month-old paste with no problems (I seal it in a bag and second container, then store in a refrigerator).
The other day I needed to knock-up a couple quick PCBs for a miniature IronMan display for a Hollywood FX shop. I used the DipTrace to layout the boards, the Pulsar Pro system to etch, and ZephPaste to solder in a toaster over.
One tube of ZephPaste will last a long time. Be sure to order plungers as they are not included with the tubes.
Amen on the plungers. That stuff is d@54d hard to put through a needle applicator without them. Also, zeph sells rework tool kits including everything; plungers, tube of paste, tweezers, solder sucker etc. Also includes precut up bits of LowMelt (their brand). Both Pb and Pb free kits. Some tools are cheezy, rather buy the tweezers separately from cooper tools or similar. But they are convenient.
Glad to see that others are having good luck with toaster oven reflow. I still use an old heating element style toaster oven by Magic Chef. Here's how I build my PowerTwigs. I did a full blown toaster oven PCB build demonstration at UPEW 2010 with my wife (my pick-n-place person)
Apply paste to stencil (using a laser cut kapton stencil)
lay PCB on ESD mat
lay stencil over PCB and align (visually)
using putty knife, spread the bead of paste across the PCB apertures using enough pressure so that almost no paste is left across the surface. (ensures that paste is forced into the apertures, but also that the putty knife does not scoop into the aperture)
lift stencil up slowly and check paste print
use tweezers to place parts onto paste; give each a slight push to "stick" into the paste.
place boards on metal tray into "warmed" temp toaster oven (run at 400 for about 30 seconds just so it is above room temp)
set toaster oven to "350+" setting
monitor temp of candy thermometer in corner of oven and Fluke 377 thermocouple in center of tray while watching for paste to reflow.
once paste reflows (visually see it liquify and get really shiny), leave oven on for about 20 seconds
turn off oven, wait 15 seconds
open oven door, wait 1 minute
pull tray from oven and set away from oven (away from heat)
once at room temp remove from tray
In regards to a profile for the reflow process, the target times and temps should be modeled after the profile recommendations from the solderpaste you are using. The solderpaste's datasheet will have the parameters for the optimum performance of the flux that is within the paste. In conjunction with the paste parameters, some component datasheets will contain maximum temperatures or max rising/cooling slopes that need to be taken into account. There are cases where these two conflict enough to make profiling difficult.
For example, I recently encountered a 16 layer lead free PCBA with a large FPGA that needed a lot of heat to get it to the liquidous temperature of the paste, but also on the board was a unique sensor with a maximum allowable temp of 240C. Since SAC305 lead free solder is liquidous at 217C, we only had 23 degrees to work with to penetrate the body of the FPGA to ensure it reflowed properly. Running a sample board through 3 profiles with a KIC2000 profiler gave us an oven recipe for our 10 zone BTUs that enable us to accommodate both limitations.
As for the comment about how much flux is in solderpaste....typically solderpaste is 85-90% metal, meaning 10-15% flux. I have some really good microscope shots of solder paste prints on a microBGA that are from 14 mil wide octagon shaped apertures. You can really see the solder balls and how they are suspended in the flux. I will crop a few to post here tomorrow.
You can forget "revitalizing" solder paste. When it's gone, it's gone. At least that is my experience of trying to keep it alive. You can have a new batch at your door in 24 hours versus wasting a lot of time and energy trying to save $50.
I've learned I still have a bit to learn about using solder paste.
As I've mentioned in the NonaProp thread, I didn't apply the paste as carefully as I should have in some areas of the board.
The footprint on the right is the same area with the solder bridges seen in this next photo.
I followed TChap's instructions in post #9. According to my DMM, the oven reached 459 degrees. I reduced the 400 degrees time to one and a half minutes since it looked like the solder was flowing pretty well. I think I may try to reduce the time at the peak temperature a bit more in future runs.
Duane, the 400 runs a bit long but the thing is that my timer only allows 1 minute increments, I always way for the timer to beep to change. So if you wanted to watch a real stop watch you could get it better. I don't like 400 at 1 minute using the oven timersince it never gets hot enough in that time.
You have a lot of solder on there. 5 mils high is fine for those parts. There are numerous places that will laser a mylar( or other plastic) and ship to you within a day or so. I always order stainless framed stencils. For a QFN 5m is nice, 6m is too high. If the board is something you will make a lot of, just go for the stainless with frame for about 200 average price.
A few of these hinges do the job. There is a bit of play in these hinges, so often I will put a support on the left and right side of the frame to avoid left to right play. But these hinges allow fast change out to other frames. Certainly there are better choices for hinges out there. If you don't use a hinge, a percentage of the screenings will have the solder slid over, on your boards that may not be such a big deal. On fine pitch parts, it can make for inconsistent results.
I find syringes to be a big hassle when you can easily order a mylar for dirt cheap these days.
Fixing the bridges is fast and easy using a glob of flux spread over the row of pins, just drag the excess solder over all the pins to disburse it. Let the flux be the heat transfer agent to the pins and solder, not the tip of the pencil. That way the pins are not feeling the direct temp of the pencil.
Question? Would it be possible to reintroduce flux to "dried out paste"? Dig it out of the syringe, warm the mass, introduce flux and stir?
Looks like the only thing that would degrade in the paste would be the flux...
Yes, flux can be added to freshen paste but it is no longer standard practice. Back when I was a n equipment operator at HP in ~1994, I used flux often to freshen paste because we ran only prototypes, so usually only 10 boards each run. Just when the paste was really worked in and running smooth, we were done. We would return paste to the jars for the next run (we averaged two builds a day for the Roseville Networks Division which meant hubs, routers, switches, and NIC cards for LaserJets) I hate to admit it, but I could tell if paste would work well simply by how it smelled.
Yes, the alloy/metal in paste is not affected by time. Only the flux performance is affected. It varies across pastes due to the different chemical make ups. At my last company, we took expired paste and dumped it into the solder pot in the wave. Smoked like crazy till all the flux burned off, but after that, 63/37 SnPb is just 63/37 SnPb....
T Chap is right about the height, if I was to build that board in production with tin/lead, I would be using a 5 mil stencil with apertures reduced 10% in width and length from the pad size. Target paste height would be 5.5 mils. As lead free, I may go with a 4 mil stencil because apertures would be 1:1 with the pads. Target paste height would be 4.6 mils.
Comments
Frank here from PulsarProFX.com. We have found a much better way to reflow your solder paste. We've used the "hot plate" (closed coil type) from ToastMaster "Eclipse" model #6431. The idea here is to radiate heat directly up the board and you get very fast reflows in about 30 seconds when dialed up to "2" on the thermostat. The trick is to also make a simple "holder" out of thin aluminum sheet ... about 3" wide and form a "Z" bend. Top is the handle and bottom is where you rest your board. The objective is to be able to lift the board off the hot plate after the solder paste has all reflowed without shaking any of the parts out of position while the solder is still molten. Just carefully pick it up and lay it down over a wood surface or something that can stand the heat. The paste sets up in about 5 seconds and voila!
I've always used SRA-solder.com (Stan Rubenstein's business of making solder paste dispensers. He was a real nice guy but is now pushing up the daisies.) I think they still sell leaded paste. Anyway, they might give you a hard time about "Oh, it must ship 2nd Day Air" due to stability", etc. For hobby use, no sweat as it can be a room temp or even hot for quite a period of time. I've gone over a year on one tube and it still pushes out just fine for proto's and 1-ups.
This method will also work with our new "FlexPCB" but may take a few more seconds since it's not "heavy" to lay flat all the time because it's so thin.
Frank
Amen on the plungers. That stuff is d@54d hard to put through a needle applicator without them. Also, zeph sells rework tool kits including everything; plungers, tube of paste, tweezers, solder sucker etc. Also includes precut up bits of LowMelt (their brand). Both Pb and Pb free kits. Some tools are cheezy, rather buy the tweezers separately from cooper tools or similar. But they are convenient.
In regards to a profile for the reflow process, the target times and temps should be modeled after the profile recommendations from the solderpaste you are using. The solderpaste's datasheet will have the parameters for the optimum performance of the flux that is within the paste. In conjunction with the paste parameters, some component datasheets will contain maximum temperatures or max rising/cooling slopes that need to be taken into account. There are cases where these two conflict enough to make profiling difficult.
For example, I recently encountered a 16 layer lead free PCBA with a large FPGA that needed a lot of heat to get it to the liquidous temperature of the paste, but also on the board was a unique sensor with a maximum allowable temp of 240C. Since SAC305 lead free solder is liquidous at 217C, we only had 23 degrees to work with to penetrate the body of the FPGA to ensure it reflowed properly. Running a sample board through 3 profiles with a KIC2000 profiler gave us an oven recipe for our 10 zone BTUs that enable us to accommodate both limitations.
As for the comment about how much flux is in solderpaste....typically solderpaste is 85-90% metal, meaning 10-15% flux. I have some really good microscope shots of solder paste prints on a microBGA that are from 14 mil wide octagon shaped apertures. You can really see the solder balls and how they are suspended in the flux. I will crop a few to post here tomorrow.
Question? Would it be possible to reintroduce flux to "dried out paste"? Dig it out of the syringe, warm the mass, introduce flux and stir?
Looks like the only thing that would degrade in the paste would be the flux...
As I typed this I realized that this is a MacGyver question.
Possible in a "survival situation" vs. practical....
I will order paste.
P.S. I made a few more tests with photos. I have to resize my photos for the current version of this forum, so to view them go here. http://www.savagecircuits.com/showthread.php?461-Toaster-Oven-Reflow-Oven&p=4593#post4593
Thank you to all that have replied! I am learning quite a bit!
traVis.
As I've mentioned in the NonaProp thread, I didn't apply the paste as carefully as I should have in some areas of the board.
The footprint on the right is the same area with the solder bridges seen in this next photo.
I followed TChap's instructions in post #9. According to my DMM, the oven reached 459 degrees. I reduced the 400 degrees time to one and a half minutes since it looked like the solder was flowing pretty well. I think I may try to reduce the time at the peak temperature a bit more in future runs.
The board changed color while cooking.
You have a lot of solder on there. 5 mils high is fine for those parts. There are numerous places that will laser a mylar( or other plastic) and ship to you within a day or so. I always order stainless framed stencils. For a QFN 5m is nice, 6m is too high. If the board is something you will make a lot of, just go for the stainless with frame for about 200 average price.
http://www.ebay.com/itm/2-pc-Silk-Screen-Printing-Butterfly-Frame-Hinge-Clamps-DIY-Hobby-Equipment-Press-/170937426783
A few of these hinges do the job. There is a bit of play in these hinges, so often I will put a support on the left and right side of the frame to avoid left to right play. But these hinges allow fast change out to other frames. Certainly there are better choices for hinges out there. If you don't use a hinge, a percentage of the screenings will have the solder slid over, on your boards that may not be such a big deal. On fine pitch parts, it can make for inconsistent results.
I find syringes to be a big hassle when you can easily order a mylar for dirt cheap these days.
Fixing the bridges is fast and easy using a glob of flux spread over the row of pins, just drag the excess solder over all the pins to disburse it. Let the flux be the heat transfer agent to the pins and solder, not the tip of the pencil. That way the pins are not feeling the direct temp of the pencil.
Yes, flux can be added to freshen paste but it is no longer standard practice. Back when I was a n equipment operator at HP in ~1994, I used flux often to freshen paste because we ran only prototypes, so usually only 10 boards each run. Just when the paste was really worked in and running smooth, we were done. We would return paste to the jars for the next run (we averaged two builds a day for the Roseville Networks Division which meant hubs, routers, switches, and NIC cards for LaserJets) I hate to admit it, but I could tell if paste would work well simply by how it smelled.
Yes, the alloy/metal in paste is not affected by time. Only the flux performance is affected. It varies across pastes due to the different chemical make ups. At my last company, we took expired paste and dumped it into the solder pot in the wave. Smoked like crazy till all the flux burned off, but after that, 63/37 SnPb is just 63/37 SnPb....
T Chap is right about the height, if I was to build that board in production with tin/lead, I would be using a 5 mil stencil with apertures reduced 10% in width and length from the pad size. Target paste height would be 5.5 mils. As lead free, I may go with a 4 mil stencil because apertures would be 1:1 with the pads. Target paste height would be 4.6 mils.
http://tymkrs.tumblr.com/post/79464970494/so-you-wanna-solder-smt - has a lot of links to the different items we use for our SMT stuffs!
It looks like you used ChipQuik, which is what we use as well. Our one 10ml syringe of it is still working strong!