Shop OBEX P1 Docs P2 Docs Learn Events
SMT oven attempt — Parallax Forums

SMT oven attempt

T ChapT Chap Posts: 4,223
edited 2006-11-21 09:13 in General Discussion
I am discovering that with a slight learning curve, smt is actually easier to do. I am working on a method to laser some material for the stencil overlay. The stencil gets put over the board, squeegied with alloyed paste, then place the parts and bake.

The oven needs to follow some general profiles to get things consistant. Here is an app I am working up that can be OS X or PC, it sends out the temp values as a string via USB2ser. Your favorite microm converts the string to a pwm rate( or just on off) that turns on the heating element on the convection oven. THere are two metheds to use:
616 x 380 - 45K

Comments

  • Paul BakerPaul Baker Posts: 6,351
    edited 2006-11-12 20:50
    The thermocouple kit works well: http://www.parallax.com/detail.asp?product_id=28022
    however you'll have to program a one-wire interface. There are othe chips out there that can read the millivolt range output by the thermocouple. I believe Circuit Celler as an article on it.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Paul Baker
    Propeller Applications Engineer

    Parallax, Inc.
  • T ChapT Chap Posts: 4,223
    edited 2006-11-12 21:38
    Thanks Paul. It looks like it would be much easier just to use a BS2p. For a one-off there is no sense going to too much trouble when the Stamp is already set up with code ready to go. I assume the temperature values are somehow obtained, whether through a request by the c, or automatically obtained through a loop. I'd like to have the temp value sent out via serial to usb2serial adapter for monitoring in the software app. It looks llike this may be a simple solution. If there is an SXB or Spin example of serial to 1wire interaface, please let me know.
  • bubbleheadbubblehead Posts: 36
    edited 2006-11-12 23:52
    Here's a link to a Seattle Robotics club article on reflow soldering with a toaster oven:http://www.seattlerobotics.org/encoder/200006/oven_art.htm

    I think there was a recent article in Circuit Cellar about using a toaster oven too.

    Here's a spark fun artcle on using an electric frying pan!
    http://www.sparkfun.com/tutorial/ReflowToaster/reflow-hotplate.htm
  • mediaprefectmediaprefect Posts: 31
    edited 2006-11-13 00:22
  • James LongJames Long Posts: 1,181
    edited 2006-11-13 03:27
    I'm actually going to try the skillet method found on the Spark Fun website.....sounds easier that the oven.

    But note I said try....I haven't yet.



    James L
  • James LongJames Long Posts: 1,181
    edited 2006-11-13 03:31
    Paul,

    I thought the DS2760 would only go to 127 C........



    But it may be my misunderstanding.

    James L
  • T ChapT Chap Posts: 4,223
    edited 2006-11-13 04:18
    Not sure what the dif between cold junction is but:

    K-type (Chromel / Alumel) Range: 32 F to 1873 F (0 to 1023 C)

    J-type (Iron / Constantan) Range: 32 F to 1873 F (0 to 1023 C)

    T-type (Copper / Constantan) Range: 32 F to 752 F (0 to 400 C)
  • Paul BakerPaul Baker Posts: 6,351
    edited 2006-11-13 06:17
    James, you are correct, there are devices which provide the range of temperatures required.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Paul Baker
    Propeller Applications Engineer

    Parallax, Inc.
  • T ChapT Chap Posts: 4,223
    edited 2006-11-13 08:07
    Guys, the Parallax blurb says the following.
    Somebody said...

    "Thermocouples provide a low-cost, reliable means of measuring a wide temperature range (32 F to 1873 F or 0 to 1023 C). This cost-effective kit includes three thermocouples, documentation and the DS2760 module"

    "Cold Junction measurement: 0°C to +127°C (0.125°C resolution)"

    Using the DS2760 we can measure the Seebeck voltage from the thermocouple with a resolution of
    15.625 microvolts, then measure the cold junction temperature with a resolution of 0.125 degrees
    Celsius. A simple table look-up using the cold junction temperature will give us the cold junction
    compensation voltage. This is combined with the Seebeck voltage and, using a modified binary search
    algorithm, we can determine the compensated temperature from the thermocouple data table.

    I think this actually does read the higher temps, called Seebeck voltages.

    Post Edited (originator99) : 11/13/2006 8:21:23 AM GMT
  • T ChapT Chap Posts: 4,223
    edited 2006-11-16 06:38
    Here is the latest rendition of the Temp Profiler for SMT soldering. The application sends out a pseudo pwm on the RTS line (pin 7 I think on the rs232 port), the frequency range for ovens is adjustable from 500ms to 5000ms cycles. There are 6 phases that are adjusted for their Goal Temp, Time at Goal, and ramp rates(Temp Change Rates). The settings can be stored as a profile in the preset menu. If you want to check out the early stages app(PC or OSX), the link has the zips for the folders. Right now the rates and temps are not working, but the PWM is fixed at 2000ms on and 2000ms offtimes just for testing. You can set some values, enter them in presets and recall them. Locate ports first, then Select port prior to Start. Stop does not work. You can set some values and step through the phases. The WaitTemp button is for servo mode, it waits to see the Goad Temp set by the temp slider input, then runs for the time specified, then goes to the next phase. If not WaitTemp, the profile just runs the times specified.

    I am controlling an oven via USB2serial adapter straight to the SSR, no interface, just the + input on the SSR from the RTS line.

    Next is to tackle the DS2760 I just got in today(Parallax kit). There is a chip that interfaces 1wire to serial that may work to minimize hardware and do all the work in the computer. The 1wire to serial chip is linked also for those interested. If any one has done this type interface before I'd love to find out how difficult it could be to access the DS2760 from the computer to get the temp.

    pdfserv.maxim-ic.com/en/ds/DS2480B.pdf

    studiotoyz.com/TempProfile.zip

    studiotoyz.com/TempProfileOSX.zip

    Post Edited (originator99) : 11/16/2006 9:36:25 AM GMT
    1112 x 523 - 104K
  • James LongJames Long Posts: 1,181
    edited 2006-11-16 15:23
    Keep us updated....I'm interested in how it works out.

    I like your interface so far.....looks great.. Did you write that yourself?

    James L
  • T ChapT Chap Posts: 4,223
    edited 2006-11-16 18:30
    Yes I am doing it myself in Realbasic. Thanks, I'll send you the app when done.

    Post Edited (originator99) : 11/16/2006 7:40:29 PM GMT
  • James LongJames Long Posts: 1,181
    edited 2006-11-17 02:17
    Realbasic......hmm never heard of it.

    Keep up the good work.

    Also ....can I ask what "oven" you are using. There are so many out there.



    James L
  • T ChapT Chap Posts: 4,223
    edited 2006-11-17 02:32
    Just some basic convection oven, Quisinart or something like that. Thanks to Bean's thermocouple SX code, I almost have the temp into the computer, just have to hook up the MAX3232 to get it back in as a string.
  • James LongJames Long Posts: 1,181
    edited 2006-11-17 02:44
    If I build a "oven" it will be self contained with a LCD display. Probably a graphical display. I will use a Stamp or something like that. The Prop is just too much for such a simple project. I just wish the stamps were cheaper.....but hey......I can't have everything.

    But I do have a problem. I have one board that can not be put into an oven. It has a package on it that will melt. I think I will have to "skillet" that board.

    I would like to look at some of your electrical parts(photos) to get an idea on how you did it.



    James L
  • T ChapT Chap Posts: 4,223
    edited 2006-11-17 03:17
    You could use a heat gun on the parts that wont fit in the oven. I am using the SX28 not Propeller. It doesn't get simpler than an SX, and they are 4$. The code is basically done I just have to mod it to send out the temp serially. The code for the LCD version is already done, posted in the SX forum by Bean. Just plug and go. You can't just add the meltalbe part last, or remove it temporarily?
  • James LongJames Long Posts: 1,181
    edited 2006-11-17 03:30
    No....the part will melt in an oven....it is small enough to go in....but the plastic case can't be oven reflowed. That is why the "skillet" on that board. The board is the only thing heated....not the components. Well the components don't get as hot.


    Hmmm....souds interesting.....but I've never used the SX chip.

    James L
  • T ChapT Chap Posts: 4,223
    edited 2006-11-17 03:52
    That is a tough call, maybe submerge the board in liquid but if the plastic part sits above the board. Otherwise I would use direct heat with a heat guy for instance.
  • James LongJames Long Posts: 1,181
    edited 2006-11-17 04:12
    The component can't be submerged....it is a new Barometric sensor from invensense. It has a thin plastic case that will melt. I'm not sure who designed it...but they should be shot. It works great....but the fact that the case melts make mounting it a real pain.

    My other boards that I'm designing can go into an oven. I just have that one design...that will have to have "special" care.

    I just ordered two boards...and going to order components for them tomorrow. I haven't even started on an oven yet. I do have a skillet. I have read a lot about people doing their smt boards this way. These are prototypes....so if I mess them up....it's no· big deal.

    I'm going to start designing my oven this weekend. Should be fun. I'll have to look at Beans info in the forum.

    James L
  • T ChapT Chap Posts: 4,223
    edited 2006-11-21 09:13
    James. here are all the files needed to get the oven set up. You can run it straight off the rs232 to an SSR, or use the circuit shown. I will make some boards but not this week. There are notes in the folder. Keep everything inside the folder or it wont find the Profiles file.

    www.studiotoyz.com/tempprofilepc.zip

    www.studiotoyz.com/tempprofileosx.zip

    Post Edited (originator99) : 11/21/2006 9:34:35 AM GMT
    1081 x 526 - 109K
    1085 x 678 - 80K
    1004 x 748 - 88K
Sign In or Register to comment.