Improving a T-962A Reflow Oven Part2

[@seabreezecoffeeroasters7994]

Good sensible solutions. Lots of parallels with Coffee Roasting and hot spot issues to mix convective and conductive heating and avoiding hot spots.

[@UnexpectedMaker]

Great job Jerry! I look forward to seeing how this pans out.. I had thought about modding mine by adding reflective tape above the elements to get more bounce off the top, but I'd not thought about restricting the intensity of the hot spot like you are doing.

[@nheng6913]

Hi Jerry, my T-962 arrives tomorrow and I can't wait to start playing with it, although I have some small boards to assemble first. While the gears are cranking rapidly on ways to improve this design, here's a thought I had, drawing upon your baffle findings. With an IR camera, take a wide angle view of the entire heating pattern produced on a piece of laminate. No circuit pattern because for this you'd want a uniform "imaging" plane. Or perhaps the laminate does have just a simple pour with some "average" density of copper. Anyway, this is your imaging surface for what follows. Take the thermal image and use it to generate, from the inverse of the temperature, an aperture mask with holes of varying diameter and/or spacing. The mask could be fabricated by readily available photo etching processes, which are not terribly expensive (well, maybe too much for a single unit). I've done many electronic shields in nickel and brass by this method. I don't recall what a full sheet costs but think it was 9x18" or so. In the case of this mask, the holes would be etched fully through, not just bending lines. This aperture mask sits some distance under the heaters and now "perfectly" attenuates the IR energy so that the received IR "field" is completely uniform. Unfortunately, I have no time to play to this degree but perhaps someone else can have fun with it. You could also just walk to the nearest drill press and drill such a mask into thin aluminum flashing, using a glued on CAD print as the template. added: water-jet is another option for fabrication of an aperture mask. It's cutting "kerf" is much coarser than photo etching but probably adequate.

[@GodzillaGoesGaga]

Nice. Curious if a bit of coat hanger wire with hook bends in at each end would work ? Basically a hanging baffle that is relatively thin. Maybe double up the coat hangar wire ?

[@alfonsopelizzo6166]

The parabolic reflector with the lamp in the focus axis is what these machines need.

[@jamesdarrah1696]

Please give information on the motor you used & how did you power it? Thanks

[@hangebicom]

The amount of work you've done is really amazing, but why just summarize in 3 lines of text what was the issue and how it was fixed? watching 22 hour minute video just to discover that some kapton tape was glued over is overkill imo.

[@jontscott]

Adding metal strips does seem like a good solution. Where does the reflow oven measure the board temp to govern the heater power? Or does it depend on the whole chamber being about the same temp? Also when running boards should they be sitting on the metal tray or slightly raised?

[@JerryWalker001]

It seems to be working so far. It is just a matter of figuring out the width and length of the strips. The chamber temp is measured by 2 probes hanging down into the chamber. They measure air temperature rather than board temp but it does not matter if the chamber temp is even across a board because the profile can be adjusted or the probes 'weighted' to correct the response. The main goal is to get an even temperature otherwise a controller will never be able to give good results. Fitting a baffle to the fan has also improved things (I show this in the next video). Sitting boards on the metal will cause problems because the tray is not heated so will draw energy out of the board. I sit the boards on inverted ic's (faulty ones) so that they are sitting on the tips of the pins. Also bear in mind that many boards warp when they are procesed so only part of it may be in contact with the metal.

[@gyorka]

Hi Jerry, I've watched your series of t-962A improovements serveral times, and would like to implement what you did on my T-962C - first issue I have with it is that i've bought it used, and the previous owner wanted to modify it and stop hafl a way. Heating elements (6 of them) are forming a bow, and I don't like the idea there - if this should be uniform, how this bow would help to do that - i'ts not a pizza or cake to expect that edges would be lower and center higher, so here is my question: should I modify their height so they would form straight line as in T-962A, or there is some science in this, and I just don't know enough about it to make a judgment? Thanks!

[@gyorka]

did some calculations (very simplified and naive ones) and it seems that for any given IR element, except for outmost two, power factor will be way above 2.5 (close to 2.9 probably), and for remaining two it wil be something around 2.3 - this is directly under element, in the space between two of them power factor will drop to 2.4 (2.8) - that is if i put these elements evenly spaced on the same height. Values were calculated up to 60 degrees, these in brackets up to 80 degrees, but i don't know if i should also include some sort of correction factor for the distance IR has to travel to reach specific point. Thinking of it, that bow could help with outmost elements IR light to reach far ends, beyond certain degree that would be blocked if these lamps would be on the same plane - if that would not look as if it was chiseled or biten out of that metal sheet i would have some more confidence of initial design. Unfortunatly, I don't have an IR camera, and buying one just for this hardly can be justified

[@runebverrud6596]

Great mod - I'm going to try this out! What length and width of the strips did you end up with, and how did you fasten them?

[@MrOrangeman18]

hi. any follow up vids on this? cheers

[@hugos31]

Acerque los sensores a la placa pcb

[@alfonsopelizzo6166]

Instead of inserting baffles above the heaters to reduce the hot spot (so the hot spot now will be created on the surface of the heater and damaging the heater) redirect that energy evenly with a proper parabolic reflector. Check the front lights of a 80' Mini Cooper and see if the reflector is flat..

[@JerryWalker001]

The baffles are not touching the heaters so do not create hot spots on the heaters. They also do not return the energy to the heater, they absorb it. A parabolic reflector would not work in this application as the heater radiation overlaps from each heater and the distance from the heater to board is very short so a reflector would create uneven temperatures across the boards. My oven now produces very good results and has been used for a long time without issues. I would however encourage everyone to experiment with different solutions. As an experiment try driving a mini cooper close to a wall with the lights on and you will see the issue in using reflectors at short distances (emitter separation > distance to target).

[@alfonsopelizzo6166]

@@JerryWalker001 I meant one parabolic reflector for each heater like one of those in bathroom heaters. The reflector (in the end 4 of them) can be designed with the beam divergence needed to make a uniform coverage of the absorbing surface. The system will also be more efficient as in the current design some energy is lost in the chamber above the heaters and: less energy storage and therefore faster modulation. Also note the design of one of those lights with linear lamp used in night road works for example. The surface is not smooth and that scrambles the reflections of the light in more directions and gives a more uniform coverage. With the mini cooper I was talking about the reflector but you are making an example of the whole system which is a projector. Note that the light has lenses in front of the reflector. I got your point though about the use of the baffles. Hi beam and low beam of the mini use a similar technique :)

[@alfonsopelizzo6166]

The heaters generate infrared waves in the electromagnetic spectrum and can be reflected. The drawing in your video shows the distances between the heater and the flat PCB tray and explains why the hot and cold spots. That is also the case for the above panel of the chamber. There is a loss of energy above the heater and therefore the heater must compensate for the loss by becoming hotter and therefore creating the hot spot that you try to deplete with the baffle. If you redirect the energy otherwise lost above the heater downwards, the heater itself would run less hot and therefore it reduces the direct energy to the PCB board. It will be a more balanced and efficient system.

[@JerryWalker001]

@@alfonsopelizzo6166 Yes I understood what you meant and as I said it would not really work very well due to the heaters being so close to the boards. The reflectors would have a 'focus' and the amount of 'overlap' would be dependant on the distance of the board and components from the heaters. You would end up with hot spots all over the board and how would you deal with the 'direct' heating? To see what I mean you could try it with some strip lights and reflectors and try varying the distance of the board and components. What you need in an oven like this is non focused and diffuse light. But like I said you could try your idea to see if you can make it work. Note that the inside of the chamber is highly reflective (especially at IR) so almost all the energy already ends up in the chamber after multiple (diffuse) reflections.

[@JerryWalker001]

@@alfonsopelizzo6166 I spent 20 years working as an optical specialist so I am familiar with how the EM spectrum works. The temperature of the heater is what generates the IR so it is best operated as hot as possible as this reduces heating time and makes the cycle more efficient. At IR frequencies almost all the IR energy is already reflected due to the nature of the chamber material. I suggest you try it as I spent several months trying various ideas and the simplest and most effective is what I showed but if you can show something that works better then that would be great but you should demonstrate your ideas as I did.

[@jero1836]

.

[@Townie_au]

So what widths did you find for the aluminium bars over the elements?