Imdes Jumbo vapour phase oven

[48X24X48X]

The machine construction reminds me of the ultrasonic cleaner that you can find on Aliexpress and Alibaba but at least that one doesn't require that much of tight control and accuracy.

[trevwhite]

At present I think the only big issue appears to be the cooling stage 

[trevwhite]

I did some runs with the unit. It seems so close to being really good if honest. I am not talking about the 'unique' construction of the electronics, etc. I am also not here talking about the cost. This is about does it produce good results.


I tested the profile on a large 200mm x 200mm panel and on a 50mm x 70mm board and the profile was pretty consistent for both. This allowed me to just put a scrap piece of PCB in with the thermocouple attached as the main sensor for the oven.


I tweaked some settings a little and the best I could get so far is the attached temperature profile. It is the cooling phase that is really letting it down. I read that the wetting phase where its around 185C going up to peak and then back down to 185 should be a max of 150seconds. I am seeing more like 180-200.

[trevwhite]

The soldering, to me looks really clean. I am just uncomfortable with the cooling phase. I really want to find a solution to that as the oven appears to work otherwise.

[trevwhite]

A note on the anti tombstone setting. It is currently set to 200C and I think 7 seconds. What this does is at the threshold temperature ( 200C ) it turns on the fans whilst keeping the heaters on( I think ). This adjusts the slope of the profile and you can see this on the image previously attached. It is quite a neat feature to be able to adjust the slope like that.

[bson]

Is it me, or do those solder joints look a little pitted, lumpy, or gritty?  Is this to be expected?

[kylehunter]

Is it me, or do those solder joints look a little pitted, lumpy, or gritty?  Is this to be expected?

Agreed. The joints don't look great to my eyes.

[48X24X48X]

It does look like the paste didn't fully entirely reflow completely. You can still see the fine particles with your naked eyes.

Looking back at your graph, it might look like your soaking period is happening too fast and too short. There is when the process drags a bit and curve should be flatter.

[trevwhite]

I think the temperature may have been a little low for the paste I was using. I do though think the cooling needs work.


Anyone got any ideas on how to cool down better without major work to this unit?


Trev

[cgroen]

I think the temperature may have been a little low for the paste I was using. I do though think the cooling needs work.


Anyone got any ideas on how to cool down better without major work to this unit?


Trev

Have you tried lifting it up from the table, perhaps 5 to 8 cm ?

[trevwhite]

Yeah, did not make a great deal of difference. I even put another fan underneath and that did not seem to affect the cool down rate.


Maybe its the air flow path. The air comes out of the holes at the side so maybe helping suck the air up the side and out might help

[trevwhite]

Okay, I actually think I might have misread things. The wetting stage for unleaded is from 217C. Lead based profiles are 187C. I was thinking it needed to go up from 187C, peak and then back down again for unleaded. So frustrating!


This means that the oven does appear to be actually working okay. The one thing I change was to place the thermocouple on a spare piece of PCB and sat that on the grill rather than use the attached thermocouple on the side. I am sure this side one can be calibrated to work but I am okay using a scrap piece of board.


So for me the oven does work. I think the electronics is rough considering the list price but I think I can live with it.

[Marsupilami]

I have the mini and I have mixed feelings about it. What helped me is to realize how dumb the controller was. I got a cheapo thermocouple datalogger from amazon and probed a bunch of locations, at the board, at the built in thermocouple etc. Things that matter: make sure that the built in thermocouple is at the exact same height as the board. Mine was off. Also the amount of liquid inside makes a huuuge difference. Basically you have to do thermal profiling with a known liquid level tweaking the preheat and main heating times and then stick to that exact level.



The liquid quantity has the biggest effect on the main heat ramp rate, if there's too little it heats up way too quickly, if there's too much it will overshoot due to the excess thermal mass.

I have a controlleo toaster oven controller that I plan to retrofit into that, that thing has an adaptive PID controller built into it, maybe with that I could make the thing a bit smarter.

So overall meh, but once figured out it gives nice, consistent results.

[trevwhite]

I also have one of these that I was thinking might do a better job whilst also making it look a bit more professional but at the moment I have run iut of time.   

https://www.whizoo.com/

[trevwhite]

Marsupilami, did you analyse the control process? I wondered if it pulses the heater on and off to start with and then does full power once it reaches a certain tempersture?.
With the soak profile I thought it would have put the fans on a bit as well to level off the temperature but it just turns the heater off and the temperature still rises. 

[Marsupilami]

I was hoping for such a feature but it doesn't do any of that. The heating element is connected via an electromechanical relay so you can hear it whenever it turns on or off.

What it does:

• Full blast until thermocouple registers set preheat temp, e.g. 150°C
• Heater off for set profile time, e.g. 60s
• If anti tombstoning is on it also runs the fan for set time
• Full blast for main heating phase until thermocouple registers boiling temperature, e.g. 230°C
• Full blast for set time above liquidus, e.g. 30s
• Cooldown. Heater off, fan on.

There's not a whole lot of sophisticated signal processing going on to adjust ramp rates and so on.
I have the whizoo controller too, that has proper PID control and could do PWM on the heater with the SSR thus controlling climb rates.

[trevwhite]

Be great to get a profile working for the whizoo. Sounds like it might not be too much work. Also sounds like it doesn't need to be too complicated either.


 

[sandalcandal]

By default you are almost guaranteed to have a lot of tombstoning due to lack of proper temperature profile or even adequate temperature sensing. The way temperature sensor is located very close to the wall basically eliminates sensing temperature which would be similar to PCB temperature. Also you will get galden drops on the board since it will condensate on the lid as this garbage do not have a cooling coil which prevents galden reaching to the top and escaping. Also you cannot really take it apart, it stitched with pop rivets everywhere. Not to say someone who designed this has barely any knowledge about electronics. Nor CE certification bears any compliance with safety standards, since this thing is electrically unsafe  from what I've seen. It probably poses a fire hazard too as thermal switch is mounted in a way that it does not trip in absence of galden. Got bitten by it myself when I forgot that I removed it. PCB turned into a black smoking cracker, yet temperature sensor did not even reach the temperature to switch off the heater.

I noticed in Marco Reps' DIY vapour phase reflow setup he seems to also have issues with tombstoning https://youtu.be/D28uSzCs7-k?t=619

[trevwhite]

I have not experienced any serious issues yet with tombstones. I did ten boards with 50 components on each and got just one tombstone in total. 

[sandalcandal]

I have not experienced any serious issues yet with tombstones. I did ten boards with 50 components on each and got just one tombstone in total.

IIRC tombstoning is also heavily related to pad design and solder/flux used so it might be a YMMV situation not entirely dependent on the reflow oven.

Edit: BTW looking to buy or get recommendations for reflow equipment in another thread here

[Marsupilami]

I bought a small enclosure. I plan to put the Controlleo and the SSR, a small DC psu and stuff in it with a thermocouple input, a general IO terminal and a C13 power output connected from the SSR.
Then I'll gut the oven, connect the C14 power inlet directly to the heating element and put a connector on the back for the built in thermocouple and one for the fans/lid switch. This will minimize the stuff I have to build/wire within the crowded imdes box.



The only thing I need now is time. *le sigh*

[bson]

So the verdict is this product isn't usable out of the box and requires significant modifications and tweaks for decent results?

[trevwhite]

I think you need to double check it's doing what you expect. Don't just hope it works out of the box. I put a thermocouple on a scrap piece of PCB and placed it on the tray with other boards to get the best temperature measurement. I also played around a bit with profiles to get used to it. The great thing is you can't burn boards as the galden limits the peak temperature.
I do think it is over priced for what it is but it all comes down to how much time you have, how much money you have and your end goal..I wanted something that I could get working quickly and then build on later. I did not pay list price though. 1800 euros for the jumbo would have left me a bit upset really. It also appears that given a hot plate, appropriate container and say a controlleo 3 board you probably could get something working relatively quickly I think. But for me the basis of my learning was to see what the Imdes unit did so how long would it have taken starting from scratch?
 



 


 

[Marsupilami]

So the verdict is this product isn't usable out of the box and requires significant modifications and tweaks for decent results?

The problem is that there aren't really any affordable professional options. Cheapest real alternative costs like 3x as much.
I'm thinking about it and I can't tell whether I would've bought it if I had known what I know now. Maybe. It works. The operating principle makes it a lot more robust then cheap IR and convection ovens.
Could be better? Yeah.
Is there 1600EUR craftsmanship in it? Nope.
On the upside both the reseller and the guy making them were helpful.

[wraper]

So the verdict is this product isn't usable out of the box and requires significant modifications and tweaks for decent results?

The problem is that there aren't really any affordable professional options. Cheapest real alternative costs like 3x as much.
I'm thinking about it and I can't tell whether I would've bought it if I had known what I know now. Maybe. It works. The operating principle makes it a lot more robust then cheap IR and convection ovens.
Could be better? Yeah.
Is there 1600EUR craftsmanship in it? Nope.
On the upside both the reseller and the guy making them were helpful.

Frankly the same cheap horeca container placed on top of mini electric stove should work about as well. In addition you could remove it from the stove once it reflowed and place onto some metal or stone plate to improve cooling speed.