Conversion of a toaster oven into a SMT reflow oven

liyf · 发表于 2017-2-19 16:15:19

Conversion of a toaster oven into a SMT reflow oven.

Having got fed up of hand soldering SMT components (and being unable toafford assembly at a C.E.M.) I thought I would have a go at a toasteroven conversion.
There are many such conversions on the web and most have their own design of external temperature controller.
Since you can get a brand new 1/16 DIN PID controller on eBay for under£30 GBP shipped or a second hand one for less than £20 Ithought there was little point buying one specifically aimed at a SMToven conversion.  Also if the 1/16 DIN PID controller could bebuilt in to the toaster oven it would save on an external enclosure andinterconnects.

The sum total of parts I used was as follows.

1)  Toaster Oven 800W (2x250W at the top, 2x150W at the bottom) £20 GBP
2) PID controller (Second hand £19 GBP or new dual setpoint for £25 to £30 GBP)
3)  Solid state relay (Second hand from ebay £4.50 GBP)
4) Thermocouple (1mm Stainless Steel K-type) £10 GBP
5) Soldering mat (Intended for plumbing, good for 600 deg C £4 GBP)
6) High Temperature Silicone RTV (good for 350 deg C) £5 GBP
7) Assorted crimp terminals.
8) Solder Paste (Time Expired 500g Lead Free - ebay £10 GBP)
9) Window Paint Scraper (in lieu of solder squeegee £1 GBP)
10) Aluminium sheet (ebay £5 GBP)
11) Solder stencil £14 GBP

Total project cost £92.50 GBP

The toaster oven was purchased for £20 from T J Hughes.

Pulling the lid revealed the clockwork timer with bell and above it theselector switch for top elements, bottom elements or both.
Probing this area with a thermocouple whilst the lid was still onrevealed an air temp of over 100 degrees C. (not good for a PIDcontroller)

The knob for the timer was forcibly removed revealing the screws forthe time switch. This was taken out and a suitable hole cut forthe PID controller.
A Dremel (or dremel clone) with a diamond slot cutting disk (£3GBP for 3 on ebay) makes short work of the pressed steel front panel.

At this point before fitting the PID controller I decided to sort outthat 100 deg C ambient temperature in this part of the oven.
What you need is some high performance thermal insulation which is goodfor 300 deg C plus. I found something suitable in a localhardware store for under £4.
This mat is intended for plumbers to solder on and is good for 600degrees C. One slight drawback is that the weave contains a smallquantity of fine metal threads to help hold it together.
The mat should therefore be considered conductive and suitableclearance or insulation be applied between it and anything at mainspotential.

The wires to the heater elements were removed to allow fitting of themat. Clearance holes were made in the mat by cutting a smallcross aligned with the weave.
Ideally 2 layers of the thermal insulation should be used. Thepack I bought gave 1 full layer and one partial layer on top.
As I mentioned earlier the mat should be considered electricallyconductive so trim back the loose weave surrounding the ceramic elementmounts.
The mat may be held in place by very small dabs of thermal RTV adhesive (make sure it is the 350 deg C type)
An additional 2.5mm hole was drilled to the right of the PID controller to pass the thermocouple wire through.

Viewed from the front now the PID controller is in place

Now the solid state relay was fitted on a piece of bent aluminiumsheet. Also the vents in the case were bent further in to allowbetter airflow.

Lid back on and the conversion is complete.  The element selector switch and Power ON neon have been retained.
Note the excess thermocouple cable curled up under the wire rack. The thermocouple is a 1000 degrees C rated 1mm stainlesssteel  K type.
This was another ebay special.  You can buy them new on ebay with100mm long probes ends rather than the 1m one I got.  Price isapprox $15.
I painted the end of the thermocouple probe with matt black Plasti-kote Hot Paint.  See testing below.

Testing time....

I screened a test board with solder paste and added a couple of representative components.
I put the PID controller in manual mode and slowly ramped up the temperature.
The lead free solder paste (Melting Point 217 Deg C) melted at an indicated 180 degrees C.
I put this down to the board temperature being higher than the air temperature measured by the thermocouple.
Ideally the thermocouple would be in contact with the PCB but thiscould be awkward to arrange and ensure consistency of contact from board to board.
The board is heated by infrared energy from the top and bottomelements.  Spraying the thermocouple black makes it absorb some ofthe IR however it is still cooled by the ambient air.
With the black paint in place the indicated temperature at which the Lead Free solder melted was 190 C.
I think the easiest way to compensate for this is to set a fixed offset of approx. 30C on the reflow profile.
So an initial ramp to preheat at an indicated 150C then a heat to reflow with a maximum indicated peak of 220C.
I would suggest a calibration based indicated solder melt temperatureon a representative test board whenever the PCB dimensions or layers /copper weight change significantly.
On my panel the temperature seemed fairly even across the board withthe solder under all components reflowing within 5 seconds of eachother..
Parts were from 0805 resistors to D-Pack MOSFETS.

I had to enter the temperatures manually with the PID controllerup/down buttons.  Some of the new PID controllers on ebay have 2setpoints.
This would allow the dwell and reflow temperatures to be enteredseparately and swapped between using a switch or front panel control.

The measured temperature ramp rate was approx. 1 degree C per second.Whilst this is a little on the low side it can still achieve thedesired result.
For most lead free processes you are allowed 60 to 150 seconds abovethe solder melt temperature.  This would easily let you hit a peakof 245 C within 60 seconds.
*Remember to tune (or run auto-tune on) the PID controller else the ramp rate may be considerably lower.*

Ok - now on to putting down the solder paste.
The cheapest stencils are laser cut from polyester sheet.  Theseare somewhat fragile so are not suitable for large production runshowever their price makes them very attractive for prototype and smallvolume runs. At the time I bought mine a 21 x 22 cm piece of polyester(with 2 stencil designs laser cut on it) cost me £14 includingshipping.
Quick plug here - only used the chap once but the price and service were great.  http://www.smtstencil.co.uk/

To use the stencil you need some sort of jig to hold the stencil inplace over the PCB.  Avoid buying a jig of the shelf as most areintended for use with framed stencils.
A simple aluminium plate with scrap PCB glued around the outside can make a serviceable jig.
The purpose of the scrap PCB is to locate the PCB to be printed and toprovide a level edge at the PCB height all the way around it.
This prevents the stencil from bending at the edges as you put the paste down.

Now you need a squeegee for applying the solder paste. Some ofthe manual pasting jigs come with a squeegee however the price forthe cheapest on ebay is over $100 by the time you add shipping.
In the quest for a cheap solution I found that a £1 window paintscraper with retractable 90mm blade provided reasonable results.

A little paste goes a long way. Using the scraper at a veryshallow angle move it backwards to push the paste through the stencil.

Once the panel is pasted you can carefully lift the stencil.

Now the components need to be put down. I find the best method isto drop the component lightly on to the paste then gently press down onit until it sinks a little.
Take care not to smudge the paste with your hands as you are placing components.

Now carefully transfer the board to the reflow oven. Knocking the board at this stage may dislodge several components.

Reflow the board and allow to cool. This particular oven has 3latching positions for the glass door - open, closed and vented.
Once the oven hits peak reflow temperature I move the element selectorswitch to all elements off and also move the glass door to the ventedposition.
Once the temperature is below 140C I open the door fully. Thismethod does not exceed the maximum recommended cool-down of 6 deg. Cper second.
Once the oven is below 50C you can remove the board and admire your handiwork.

Usual disclaimers apply. The author of this page takes noresponsibility for any damage to you or your property resulting frommodifying a toaster oven.

Converting a toaster oven:
1) Voids the warranty.
2) Provides a possible electrocution hazard.
3) Provides a possible fire hazard.

Always check earth continuity to case after modification.
(Preferably PAT test the oven after modification.)
Never operate unattended