Finally starting on this convection oven -> reflow oven conversion project

[mindcrime]

Project Update 06-09-2021

Got off work today and was feeling pretty lazy, so I didn't make much progress here. What I did do was finally take some baby steps towards getting this "control unit" thing packaged up properly. Specifically I worked out a way to mount the LCD to the box, using some small "L"-brackets.

It took a little  massaging with the dremel tool and a grinding wheel, but it all fits together reasonably nicely now.









I'm torn on whether or not I feel like putting in the effort to cut a rectangle in the top lid of this thing, that exactly fits around that inner frame of the keypad, and mount if from inside. It's definitely cleaner, but I've also considered just bolting it down from the top like you see it pictured here, and cut a slot right below that row of pads where the wires come up and connect, to route the wiring. Flip a coin I guess. Aesthetics aren't that critical for this thing, since it's a one-off and besides myself and you lot, pretty much nobody else will ever see it.

[mindcrime]

Project Update 06-12-2021

OK, made some progress on mounting that keypad. In the end, yes, I went the lazy way and just mounted it from the top, and dremel'd out a slot underneath for the wires. But c'mon, who are you kidding, we all knew it was going to go that way, now didn't we? 



The keypad bolted securely in place.



Doesn't look too bad - from the side you can't see! 



Not the prettiest job in the world machining that slot in the plastic. But oh well. Visual aesthetics was never a super high priority for this project.



Wires go through here...



This actually wouldn't be too bad, if I had not goofed and machined the bottom edge of the slot down too low, making it visible from the top. Oh well.. dem's da breaks.



A more or less final view of what it'll look like from the top, when finished. There will be some wires coming out the back somewhere, but otherwise most of the parts that involve drilling holes and cutting material are done.

Now the big push will be to finish getting all the electronic bits consolidated onto one (or maybe two) PCB's, and then mount the boards, including the Arduino and the thermocouple amplifier, inside somehow. I'm almost tempted to whack an edge card connector in there, and do the final assembly on a board that fits one of those (ISA style) so I can slide it in and out easily. 

[mindcrime]

Project Update 06-15-2021

I'm starting to get an idea of how this control box is going to come together, I think.



I added female header connectors to the wires coming off the keypad, which will terminate at a row of male header pins that will live on the board that hosts the keypad decoder IC.



The keypad decoder stuff is still on the breadboard for now, but will soon move to a dedicated board of its own. Which will couple with:



This power supply board, which I've decided to leave unmodified (ie, I'm not moving that one lonely transistor to this board after all). Because...



I realized that the board with the one lonely transistor is a perfect spot to cram the arduino and the thermocouple amp.

This gives me 3 boards, which I'll stack vertically using threaded hex standoffs. I'll then mount the board assembly by securing some threaded standoffs to the bottom of the enclosure with JB Weld or resin, or whatever.

I also need to add a way to get the 12VAC feed into the enclosure, and an interface for the signal to the SSR. I'm leaning towards mounting banana plug connectors on the back of the oven and on this enclosure, and making the connection with a set of what would basically be banana plug jumper cables.
 

[mindcrime]

Project Update v2, 06-15-2021

Got started making that board that will hold the keypad decoder stuff. Not a lot of progress, but enough to count for something I guess.



Soldered some headers on the board, and starting to put in a socket for the 74C922 decoder IC.



Keypad wires just plug into the headers. Simple.



Get that socket soldered in place and the battle is half won!



I screwed up on one pin and lifted the little copper pad off the board. No biggie though. When I bring in a wire to solder to that pin, the pin and the wire will be locked together and the wire can bond to the pad around the hole it comes through. And it's not like the socket is going to go anywhere with the rest of the pins soldered in place.



Now to put the socket in place for the 74HC04 inverter. Those two chips and a couple of capacitors are all that go on this board, besides header pins and wire.



Both sockets in place and the IC's (temporarily) seated. Looking snug!

Now to add the headers for output and power connections, add the capacitors, and wire everything up. At that point, this thing will be very close to finished!

[fourfathom]

Way to go!  FYI, I prefer to use double-sided perfboard with plated-through holes for this kind of breadboarding.  It's a bit more difficult to modify (than the single-sided phenolic board) once the components are soldered, but the connections are way more secure and the boards themselves are much stronger.

I can't wait to see how runs when you've got it all put together!

[coppercone2]

I can only stand veroboard and all the other ones are torture to use

just leave it on a breadboard and melt the first pcb with it

[fourfathom]

I can only stand veroboard and all the other ones are torture to use

just leave it on a breadboard and melt the first pcb with it

Veroboard with the copper strips, right?  That's never seemed handy to me.  I always needed so many cuts and jumpers that I would rather just strip and solder wire-wrap wire between isolated pads, and lay down / tack-solder a few bare wires as ground and power busses.

I'm really dating myself here, but I used wire-wrap to build a very early 6800 processor board for an industrial control system.  Later I used that thin wire with the solder-through insulation.  These days I either build simple stuff on a perfboard, or mill a simple circuit board, or for more complicated designs send the gerbers off for fab.  I just had a bunch of simple 1"x1" boards made for SMA attenuators.  Ridiculously cheap at about $0.20 each. and about the same for shipping (qty 50).

[beanflying]

The few boards like Mindcrime used here I had got trashed for assorted Spacers in woodworking jigs and non Electronics use. I got tired of the pads departing from the substrate.  Anything that advertises itself as 'Double Sided Prototype Board' is a much better thing eBay auction: #233880123981 I keep a few sizes and also some alternate pitch versions for odd ball jobs.

Back on topic looking good 

[coppercone2]

if you use the stripping tool and its sharp for veroboard (or a drill if you don't mind it being weaker), it is efficient

if you use an exacto instead of a rotary cutter then thats your problem.

[mindcrime]

if you use the stripping tool and its sharp for veroboard (or a drill if you don't mind it being weaker), it is efficient

if you use an exacto instead of a rotary cutter then thats your problem.

I bought a new drill bit set the other week, just to keep at my bench so I can use one of the bits for cutting Veroboard strips. Haven't even used it yet. I keep Veroboard, single sided protoboard, double sided protoboard and all sorts of miscellaneous stuff here, but those cheap little single sided boards just happened to be close at hand when I started on this. It would probably have been easier to use the Veroboard... oh well. Next time.

The few boards like Mindcrime used here I had got trashed for assorted Spacers in woodworking jigs and non Electronics use.

Heh. I do the same thing, FWIW. These things are so handy for all sorts of miscellaneous uses. Sometimes I grab one and use the bare side as a sacrificial surface to solder on (when joining wire to wire, or wire to connector, etc.) instead of leaving the thing laying directly on the blue ESD matting. They also make handy shims / spacers for all sorts of applications. 

[mindcrime]

Project Update 06-19-2021

In the previous episode we saw the bare outlines of a circuit board for the keypad decoder logic starting to take shape. Today we see how things have progressed from there.





And it's done. Or is it "done"? Hmm.. time will tell. But all the parts are on the board and everything is wired up. Nothing to do now but remove left over flux and test it. *gasp* "test"? That can't be a good thing...



Keypad connected, Arduino connected, and no more breadboard crap in the way. Time to see what this baby can do!





It's this point in the story where our erstwhile hero realizes that critical mistakes have been made...   





Fortunately the error was no more than mounting the display upside down. 



After a bit of fiddling with the correct ordering of the jumpers coming from the keypad to the board, we're back to being able to correctly decode keypresses and respond to them.

Not much left on the hardware side now, aside from mounting these three boards in the box, wiring everything up, and finalizing the arrangements for how the control box is going to connect to the oven. I continue to lean towards the idea of using banana plug jacks on both ends (oven and control box) with the connections made with cables with male banana plug connectors on each end.

After that, I can continue to tune and iterate on the software part, but I think the software already provides something usable.

I also need to integrate the piezo buzzer that is going to signal when a profile run is finished.

Oh, and then there's the whole door opener thing. I'm still punting on that, largely driven by the fact that the linear actuator I bought, while it would work fine, is honestly too big for this application. I am going to evaluate buying one that's better suited,or perhaps even consider switching to using a servo instead. No hurry on that bit though. I can open the door by hand for now.

My SOIC-14 stencil and some more NE555 IC's in SOIC-14 just came in, so soon I'll resume doing some test solder jobs and see how much nicer things turn out when the solder is properly applied. 

[mindcrime]

It's always the dumbest things that trip you up. So I'm sitting here working on assembling / packaging all of this stuff, and I get the thermocouple amplifier board bolted to the piece of perfboard that is going to hold it. Then I start to bolt the Arduino Nano 33 IoT down and ... promptly come-a-gutsa. The stupid mounting holes on the Nano 33 board are like 0.07" in diameter... too small for the smallest machine screws I have here (#2). And there's really no room to drill the holes out any larger, without risk of damaging something.

Sssssooooooo... off to McMaster Carr to order a bag of #1 machine screws and matching nuts. Those supposedly work with 0.073" holes, which should be close enough. Worst case I take a tapered file and ream the hole out by 3 thousandths of an inch. Maybe I should have bought some #0 screws as well, just to be on the safe side...

*sigh*
 
Maybe I should just hot-melt glue the bloody thing into place.

[fourfathom]

Could those be "M2" screws (2 mm dia, slightly smaller than "#2")?  These days a lot of small boards use the metric M2 hardware.  Me, I grew up using #6 and #4 screws for this stuff, but recently I was assimilated and bought some M2 assortments: various screw, nut, washer, standoff, etc kits in stainless, brass, and plastic.  Cheap, from China, via Amazon.

[mindcrime]

Could those be "M2" screws (2 mm dia, slightly smaller than "#2")?  These days a lot of small boards use the metric M2 hardware.  Me, I grew up using #6 and #4 screws for this stuff, but recently I was assimilated and bought some M2 assortments: various screw, nut, washer, standoff, etc kits in stainless, brass, and plastic.  Cheap, from China, via Amazon.

Probably. And yeah, I've thought about ordering an assortment of the metric ones. I'm not, even as a red-blooded 'Murican, opposed to metric stuff at all. I keep using SAE / Imperial stuff out of habit more than anything. I guess it's time to catch up with the modern world and switch over wholesale.

[beanflying]

Arduino sizes and screw placements generally SUCK. I keep some 2 and 2.5mm plastic hardware principally for them. In the case of the UNO you risk shorting bits out with metal fastners and in particular if you enlarge the holes.

As an R/C flier and being of a certain Age and an Aussie I still talk 2-56 and 4-40 and have even been known to have conversations with farmers in Rods and Chains 

[mindcrime]

Arduino sizes and screw placements generally SUCK. I keep some 2 and 2.5mm plastic hardware principally for them. In the case of the UNO you risk shorting bits out with metal fastners and in particular if you enlarge the holes.

Good point. Maybe I should order up some nylon screws just to be on the safe side.

As an R/C flier and being of a certain Age and an Aussie I still talk 2-56 and 4-40 and have even been known to have conversations with farmers in Rods and Chains 

Do you measure weights in "Stone" as well? 

[mindcrime]

More baby steps towards making this a finished project.




Instead of just having wires flapping around in the breeze, I added header pins to connect the output from this board. One set for ground, one set for 5V and one set for 12V.  No, the placement isn't optimal, but honestly, these were an afterthought. Not to suggest that much "thought" went into the layout of the board in the first place. 







A handy little adapter to power the Arduino from the header pins on the power supply board.



A pair of banana plug connectors for the SSR control.



And a pair to receive the 12VAC feed from the transformer mounted in the oven case.



Now, I just need all of this crap, er stuff, erm, I mean,  "electronics" securely and neatly mounted in this box somehow, and all the wiring connected. 

Idea #1 - wire everything up, cram the boards in there more or less as  you see, and then fill the whole dang thing with resin! 

Idea #2 - run back and forth to McMaster Carr, the local Ace hardware, my parts bin, etc., frantically scrambling to cobble together the right combination of standoffs, washers, machine screws, nuts, epoxy, etc. necessary to allpw me to spend hours painstakingly drilling holes, bolting things together, and tediously assembling this into something that I could be proud of (if I were on a really bad acid trip, and had no standards, and knew no one else would ever see it).

*sigh* as much as I hate both ideas, I suppose #2 will carry the day. Maybe I should make a version doing the resin thing though, and see if I can get it displayed at the Museum of Modern Art or something. 

[mindcrime]

OK, it's not quite Idea #2, but it's not quite as bad as "pot the entire assembly in resin" either. In the end, I went with the one mounting technique I most specifically did not want to use. And that's "drill holes all the way through the sides of the box and use machine screws and nuts to bolt the boards to the sidewalls".











And yet, that's exactly what I did in the end.

"So why", you may be asking, "did you do it the one way you didn't want to do it? And why were you so opposed to doing it that way in the first place?"

Glad you asked. To answer those questions in reverse order:

1. I didn't want to do that because it's ugly. And while visual aesthetics aren't goal #1 of this project, I also wasn't exactly looking for the "Mad Max aesthetic" look and feel. Or would this be the "Dr. Frankenstein's Lab aesthetic" look and feel?

2. I did it because it was the one technique that let me make progress with the hardware I had handy, without waiting for a bunch of bolts and standoffs and other hardware to be delivered. Also, my original plan was to stack the boards vertically with standoffs separating the boards, but the more I look at it the more I think that would make it really hard to deal with all those jumper wires. And for now, I really don't feel like getting rid of those. Having the boards mounted vertically on the sidewalls leaves all that room in the middle for wires to bend and flow and makes repair access for the future relatively easy.

Also, and maybe the biggest deciding factor, was an epiphany of sorts that I had. I think there will be a clean way to come back and get rid of the gnarly looking bolts and nuts on the sides. Since I already drilled holes all the way through, I'm thinking I can come back later and add those heat-press-fit threaded inserts where the holes are now, and press them to where they're flush with the edge of the case (or even slightly countersunk). Then I can just paint over the outside and/or putty over the blemishes with a little JB Weld or something. Then the mounting from the inside will be exactly the same, there just won't be a nut on the other side of the wall. And if that doesn't work, well, I can always keep the Dr. Frankenstein's Lab look, or buy a new box and do something different altogether.

Also, yes, I know that one board is bowed pretty badly. I need to dig my bag of nylon spacers out of the big-heap-o-parts-boxes later. I'll find some and redo all of those with spacers to keep the boards from bowing like that.

[mindcrime]

Project Update 06-29-2021

So, in the end, I did have to order more hardware and tools and wait for that stuff to be delivered. The existing mounting holes on the Arduino Nano board turned out to be sized for a metric M1.6 screw, and I had no M1.6 screws, nuts, etc. nor a 3.5mm nut-driver, here. But, I got that stuff ordered, after some fits and starts, and now all the board mounting is worked out. I just need to un-mount all of these, add spacers as appropriate to keep from flexing the boards, and then re-assemble it, and the physical fabrication on this box is down to about the next to last punch-list item (cutting a slot in the top for the wires that go to the LCD display to pass through).






I did goof on one aspect of this though. That board that carries the Arduino Nano got mounted too high up and didn't leave enough room for the (previously made) straight USB cable power adapter. Luckily I keep a bunch of the right-angle USB cables around here as well, so I cut one of those up and am re-making the power adapter with that.

[mindcrime]

Project Update 06-29-2021 episode II

Managed to make some more progress this evening after work. And I think finally, mercifully, I can now say I'm done with everything that involves opening up the actual oven. Even better, there are now no more unsightly (and unsafe) wires snaking through razor-edged ventilation slots.



Banana plugs. Ugly, yes. But effective.



It is physically impossible for me to drill holes for mounting banana plugs and get them to line up properly. Of course, if I would measure and mark this stuff instead of trying to "eyeball" it all the time....



Looks even worse from the inside. But nobody will ever know but us, right?






Connectors affixed, and now I can close this baby up!



These all run to the matching connectors on the control box.



Green is my 12VAC output which provides power to the control box. Why green? I dunno. It was handy. And I tend to associate the color green with AC wiring for some weird reason.



AC output still works. Yay!



The other end of the connection. You can start to kinda see how it all comes together...







Lid re-installed, and the control box attached to the oven. This is very close to finished now. Just a few more tweaks on the control unit and we're ready to solder.

[beanflying]

Strange but some of your images are blocked for me 



And some of them really are too. Whoever or wherever you are hosting your pictures seems to be bandwidth limiting 

[mindcrime]

Strange but some of your images are blocked for me 



And some of them really are too. Whoever or wherever you are hosting your pictures seems to be bandwidth limiting 

Aarggh... Google Photos is stupid. I need to switch to a different image hosting setup, but it's just so convenient to rely on GPhotos since my Android phone automatically uploads pictures for me, etc. **sigh**

Aaah well. Not sure what to say. I did try loading this thread from an incognito window in a different browser that had not visited this thread before, to eliminate both login cookies and cached images, and yet all the recent pictures showed up OK for me. Maybe it's just some transient network error or something? Who knows...

[fourfathom]

I can't see the images in this post: https://www.eevblog.com/forum/projects/finally-starting-on-this-convection-oven-gt-reflow-oven-conversion-project/msg3592043/#msg3592043
All the others, including your latest, look fine.

And congrats on your good progress.  A couple of days ago I used my oven to reflow a batch of RF attenuators.  I needed a better assortment of SMA attenuators and didn't feel like buying them.  So I did the complicated and time-consuming thing of laying out a small PCB, using custom footprints for the resistors that would accept anything from 0402 to 1206 surface-mount sizes.  It's just a simple Pi network, with parallel resistor locations that let me use my assortment of 1% resistors (I stock the 5% series) to get the desired resistance.  So far I've made 10dB and 20dB pads, and they are flat within 1/10 dB from DC to 1 GHz, and within 1/2 dB to 2 GHz -- way better than I had anticipated.  The return loss is decent for my needs, better than -30dB to 1 GHz.  I have through-holes around the board perimeter to let me add a shield cover if I want to try that. I used JLCPCB, 50 boards for 20 cents each, plus $13 for shipping.  Crazy!

I finally got a small-diameter syringe tip  for the solder paste, and while it takes a whole lot of force to work the syringe I can put small dots of paste on the pads.  I hand-soldered the SMA connectors, but I'm going to try a few in the oven to see how they hold up.

[mindcrime]

I can't see the images in this post: https://www.eevblog.com/forum/projects/finally-starting-on-this-convection-oven-gt-reflow-oven-conversion-project/msg3592043/#msg3592043
All the others, including your latest, look fine.

Weird. I don't always understand how GPhotos decides what to allow and not allow. Once I added the Chrome "get GPhotos link" extension to my browser and started using those links, it seems to mostly work, but even then I guess it's hit or miss. **sigh**

I finally got a small-diameter syringe tip  for the solder paste, and while it takes a whole lot of force to work the syringe I can put small dots of paste on the pads.  I hand-soldered the SMA connectors, but I'm going to try a few in the oven to see how they hold up.

Nice. What size tip did you wind up using? I might look into getting some of those and see how it works.

I did get a bunch of additional "stock" stencils from ChipQuik the other day. Now I have stencils for single chips in most of the common SOIC layouts and a couple of other common ones.

[fourfathom]

I used this syringe kit from Amazon: https://www.amazon.com/gp/product/B08G8XNFKK
The short #18 needle gave me a reasonable solder dot for my resistor pads, but it would have been much too big for my 0.5mm-pitch IC pads.  I suspect that if I had used a smaller needle I would have cracked the plastic syringe from the necessary pressure.  It took some practice to get the solder to stick to the pads rather than stay adhered to the tip, and I sometimes used a small wire to push the solder around (and then the solder would want to stick to the wire).  But by the third board I was getting the hang of it.  The solder was at around 70 deg F).  This needle was 1" long, probably a shorter needle would require less pressure to get the paste to flow.  It's a work in progress and stencils are cheap, so "good enough" is OK with me.

I had originally got the syringe needles to use with a vacuum-pickup for grabbing and placing components, but so far I'm still using tweezers.