Options for reflowing 12x14 pcb

Erik Friesen

My modified toaster oven is too small for the pcb for my next project.

Has anyone tried -

A: A regular oven

B: A hotplate

C: A hotplate + hot air to finish

Brian Carpenter

a hotplate works fine. Do not attempt to do it in your oven, although it may work, lead in your food is no good.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔


It's Only A Stupid Question If You Have Not Googled It First!!

Erik Friesen

I have access to an old oven. I am a little worried about using it because of slow heating, and I am afraid that the infrared will cook spots on the board with no convection fan.

Is there any specific type of hotplate that works better than any other?

Phil Pilgrim (PhiPi)

With that large a board, I think a convection fan would be a must.

-Phil

Erik Friesen

Problem is, that old oven has cooked chicken litter (literally, for a gasifier project) and I don't feel like using it in my house or garage.

Phil Pilgrim (PhiPi)

Heck, with that kind of stuff wafting around inside, you may not even need to use flux!

-Phil

mctrivia

my toaster only does 12" stove would probably work fine if you have a fan inside. you could wire in a cheap metal fan. would wear out quicker then normal but would last for a while.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
My new unsecure propmod both 1x1 and full size arriving soon.

kwinn

If you can get a motor with a long shaft you can mount it outside the oven with the shaft going through a hole and the fan mounted inside. Works well on an oven with a bottom drawer.

parts-man73

I would think the hotplate would have hot and cold zones. You'll have spots that aren't soldered, or possibly burnt spots.

How about a hot air reflow station with a wand. It would take awhile to do the whole board, but I'd think you'd have better control and it'd get done right.

Is this a one-off design, or making many off them?

Hot plate + hot air might help...

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Brian

uController.com - home of SpinStudio - the modular Development system for the Propeller

PropNIC - Add ethernet ability to your Propeller! PropJoy - Plug in a joystick and play some games!

SD card Adapter - mass storage for the masses Audio/Video adapter add composite video and sound to your Proto Board

mctrivia

a good hot plate is actually very evenly distributed in heat at least the ones i played around with in chemistry class. never tried soldering with one though.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
My new unsecure propmod both 1x1 and full size arriving soon.

James Long

Erik Friesen said...
My modified toaster oven is too small for the pcb for my next project.

Has anyone tried -

A: A regular oven

B: A hotplate

C: A hotplate + hot air to finish

There are quite a few drawbacks to all the above.

• First getting the entire board up to temperature. This is the biggest challenge. With a PCB that size, convection circulated heat or vapor phase reflowing is going to be the most consistent. This is important due to the board trying to warp on heating and cooling (mostly on cooling).
  
• Heating the board fast enough to prevent the flux from flashing off. If you are doing lead, the issue is not as great.

If the board is one single piece, using a hotplate with hot air can cause it to warp when cooled. Even with a fancy reflow oven, boards will occasionally still warp.

12 x 14 is a pretty large board to do in one shot. I will say just about any of the above can be done, if you are extremely careful. Your mileage may vary on your success rate.

We did do boards up to 10 x 12 before our oven was found to be too small to do the job (it has an 11 inch wide belt, but it's heating abilities only allow about 40 sq. in). We are working on a vapor phase to replace it. It will be able to do just about any reasonable size board, although our Pick & Place will still limit the board size.

I personally vote for the hot plate (alone). It is not the way I would do it (not a big fan of hot plates), but I think it would cause the least amount of stress on the board. If you have metal USB connectors, or SD card holder, you may need to touch those with some hot air. I would use the hot air with the hotplate in a limited amount.

The oven is a throw of the dice. You may not heat fast enough, or you may over temp the board (or spots), it is an unknown.

A hot plate and hot air, may cause stress points on the solder joints (uneven shrinkage).

Whatever you choose, I wish you luck. Keep us informed of the outcome.

James L

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
James L
Partner/Designer

Lil Brother SMT Assembly Services

Erik Friesen

My thought with using the hot plate was to preheat the board to about 125-150c and finish with my hot air pencil.

The board is a 2mm thick 4 layer.

I will be using lead paste on this one. I have had success with lead free em907 but that peak 240-250c is something you have to watch carefully as it is fairly close to plastic, fr4 fry points.

I am not a fan of the hot plates idea, but I don't have any solid evidence against it. I wonder if I need to look for a piece of aluminum to buffer the heat a little.

mctrivia

copper would work best but any flat metal plate would work.

Curious would coating the board in thermal grease be a good idea? never tried using a hot plate to solder just etch boards back when i made my own boards.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
My new unsecure propmod both 1x1 and full size arriving soon.

James Long

mctrivia said...
copper would work best but any flat metal plate would work.

Curious would coating the board in thermal grease be a good idea? never tried using a hot plate to solder just etch boards back when i made my own boards.

I personally do not think I would even try that. Thermal grease has all kinds of things in it that may be hard to get off the board. If doing the reverse side, I definitely wouldn't use thermal grease. It may leave a film which could affect the solder paste.

Just putting a board on a electric skillet is enough to let the board absorb the thermal energy. I have talked to a lot of people who have used the skillet method, and most have great success. Only the metal USB connectors and SD card holders usually take a little hot air to finish.

I even know of one person who used a clothes iron and an aluminum plate to reflow a board. Not a recommended method, but it can be done.

James L

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
James L
Partner/Designer

Lil Brother SMT Assembly Services

Erik Friesen

I am thinking about this some more, and waiting on the arrival of my boards.

Supposing I use the above mentioned oven, I am thinking about preheating the oven to around 225c, and using an airbake pan, or other method to slow the direct thermal effect on the board. Sort of like when you put a slice of pizza in the oven in a piece of tinfoil, it can take 3-4 minutes to get the outside warmed up. An airbake pan has a layer of air sandwiched between two pieces of aluminum. The reason for my thinking on this would be that once the preheat is finished on an oven, the bake coil on the bottom would be the only source of heat, and the heat should be fairly even.

Would this method realistically be way outside of the recommended ramp - soak - and cooldown recommendations for lead based solder paste? I am just afraid that otherwise the oven will heat too slow and unevenly.

My instinct keeps rebelling at the thought of using the hotplate method, as the hotplates I have looked at have a piece of aluminum about 1/8" or less with one heating coil ring going around the outside of the plate. If you read the sparkfun tutorial on their hotplate experiment, you will see that they had to move the board around to find the hot spots.

I have perused the web for a small convection toaster oven, and there seem to be none large enough.

I may adapt an old fan to this oven if necessary.

T Chap

I would adapt a new door to a convection oven, one that allowed half the PCB to stick out. Most regular convections will handle 12 by 8. Make a door that allows the board to stick in half way (12 x 7), mount parts on one side only and cook that side. After cooling, mount the other side and cook it. The trick will be applying solder in two stages, but should be doable, just find a nice line to divide the parts up a long. Even a tin foil makeshift door will be fine to contain the hot air, or piece of glass that fits the opening.

Post Edited (TChapman) : 5/10/2009 10:26:55 PM GMT

James Long

Erik Friesen said...
I am thinking about this some more, and waiting on the arrival of my boards.

Supposing I use the above mentioned oven, I am thinking about preheating the oven to around 225c, and using an airbake pan, or other method to slow the direct thermal effect on the board. Sort of like when you put a slice of pizza in the oven in a piece of tinfoil, it can take 3-4 minutes to get the outside warmed up. An airbake pan has a layer of air sandwiched between two pieces of aluminum. The reason for my thinking on this would be that once the preheat is finished on an oven, the bake coil on the bottom would be the only source of heat, and the heat should be fairly even.

Would this method realistically be way outside of the recommended ramp - soak - and cooldown recommendations for lead based solder paste? I am just afraid that otherwise the oven will heat too slow and unevenly.

My instinct keeps rebelling at the thought of using the hotplate method, as the hotplates I have looked at have a piece of aluminum about 1/8" or less with one heating coil ring going around the outside of the plate. If you read the sparkfun tutorial on their hotplate experiment, you will see that they had to move the board around to find the hot spots.

I have perused the web for a small convection toaster oven, and there seem to be none large enough.

I may adapt an old fan to this oven if necessary.

Erik,

There can be problems with your method. First most parts are a maximum ramp of 3 Deg C per second. If you exceed that ramp you may have part failures.

An airbake pan may prevent the solder from reflowing fully. If the board is kept below the reflow temperature (or is slow to come to temperature) you can have cold solder joints, or the parts will tombstone like crazy (especially if you use copper pours/planes).

There are going to be hot spots and cold spots in a regular oven without some kind of air circulation.

All of these are factors which govern reflow oven design. I wish I knew of a method which would work, but none come to mind.

I really believe a vapor phase would be the best method to use, but the design is tricky, and expensive. If we were further along with our design, I would share some of the information, but we are not to full scale yet. When finished it would easily do a 16 x 16 board (that would be about the maximum though).

If nothing else, just go with a hot air station. PCB's are usually pretty forgiving.

James L

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
James L
Partner/Designer

Lil Brother SMT Assembly Services

mctrivia

i recently tried soldering with a torch and it worked very well on 10 gauge wire. would be stupid on ic's but got me thinking that if you do not care about the shape of your oven you could speed up the heat up phase by cutting a small hole in for a torch and spark ignitor. as long as it was pointing away from the board with a shield to stop direct flow this method would give you faster heat up times then using an electric coil. It also would have the added benefit of causing convection without a fan because of the force caused by the exiting gas.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
My new unsecure propmod both 1x1 and full size arriving soon.

Erik Friesen

I have a hot air gun, but it seems to me that a hot air gun surely effectively ramps at a greater rate than 3 deg C per second.

I have done a little checking on some IR ovens like this - www.ntscope.com/SMD-2004A.html but I have a feeling that using direct IR has some serious drawbacks.

Seriously though, is the ramp up rate on a typical range slow enough to cause an issue?

mctrivia

i don't actually know. Why not plug it in crank the temperature and time how long it takes to get to say 300 C. Simple math to calculate if slower then 3deg/sec. My guess is 5min or 1 deg/sec

As for infared they work extremely well on QFP and SOIC packages. they are less eficient on QFN or BGA packs because the pins are under the ic body. I have no numbers though on how good or bad they work. Every place I have worked has used hot air for reflow and that is the method I use here at home.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
My new unsecure propmod both 1x1 and full size arriving soon.

Post Edited (mctrivia) : 5/11/2009 1:46:03 AM GMT

Erik Friesen

The ramp rate is around 0.4c second from 40c to 185c for my cooking range. I have a direct thermocouple test from my hot air gun that is around 80F/second up to 500F.

WBA Consulting

@James and Vapor Phase:

Wow, I was surprised to hear someone considering vapor phase as a solution, even though I just read an interesting article about it's possible comeback. I know there has been some in the industry revisiting vapor phase to take advantage of the precise temperature control but reliable, cost effective solutions are still limited. There's a lot of work going on right now to improve vapor phase equipment when used with lead free alloys like SAC305, but I would be hesitant to go that route yet. Back in the day, so to speak, I remember all the horror stories about the number of variables involved with the overall vapor phase process (preheat, slope, etc) and the "Apollo 13" like efforts that sometimes would take place to dial in a board. Once you had it dialed in, the boards were near perfect everytime since the temperature characteristics of the vapor fluid are consistent, but it was so much quicker to get an "above acceptable" profile with an IR or convection oven.

Anyhow, there's a really good article in the April 2009 edition of Circuits Assembly about "Revisiting Vapor Phase". www.circuitsassembly.com/cms/component/content/article/201/8104-vapor-phase-revisited

I'll be interested in hearing more about your progress in this endeavor. Most of my colleagues don't even know what vapor phase is!

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Andrew Williams
WBA Consulting
IT / Web / PCB / Audio

James Long

WBA Consulting said...
@James and Vapor Phase:

Wow, I was surprised to hear someone considering vapor phase as a solution, even though I just read an interesting article about it's possible comeback. I know there has been some in the industry revisiting vapor phase to take advantage of the precise temperature control but reliable, cost effective solutions are still limited. There's a lot of work going on right now to improve vapor phase equipment when used with lead free alloys like SAC305, but I would be hesitant to go that route yet. Back in the day, so to speak, I remember all the horror stories about the number of variables involved with the overall vapor phase process (preheat, slope, etc) and the "Apollo 13" like efforts that sometimes would take place to dial in a board. Once you had it dialed in, the boards were near perfect everytime since the temperature characteristics of the vapor fluid are consistent, but it was so much quicker to get an "above acceptable" profile with an IR or convection oven.

Anyhow, there's a really good article in the April 2009 edition of Circuits Assembly about "Revisiting Vapor Phase". www.circuitsassembly.com/cms/component/content/article/201/8104-vapor-phase-revisited

I'll be interested in hearing more about your progress in this endeavor. Most of my colleagues don't even know what vapor phase is!

Andrew,

Well to be perfectly blunt, we are copying Asscons method of the vapor phase reflow system. If you look around and read what they are doing, they are using the fluid to preheat the boards.

This is the draw back to vapor phase. Ever system previously (except for sealed chamber vapor phase) used heating elements to preheat the boards.

I have figured a way to reliably preheat/solder all in the same sealed container. I am doing this differently than Asscon. I know their method, which I felt could be improved upon.

There are two things that really hep with our current design. First the reflow chamber is always sealed when at temperature, so no reflow fluid can escape in vapor form. Second, because the board is not being moved while in process, there is little chance of parts being displaced. Our system based on a batch type reflow operation.

Also, because of our design, very little vapor phase fluid is needed for the system. We have calculated about 2 litres total.

There are ways to prevent the "dial in" to a board, to some degree. I can't explain it all, because we are planning to sell the systems once they are completed, but suffice to say, it deals directly with wattage of the heating elements, and thermal mass within the reflow chamber (static thermal mass).

Just to insure there is not a misunderstanding. We are not dipping boards into the vapor, but building a vapor blanket very slowly (controlled) around the PCB. It will rise from the bottom of the PCB up to the highest component.

We do have a small prototype working really well. It was a math game to get the specific parts right. Cooling (vapor blanket collapse) has been our biggest hurdle, but we now have devised a system to counter that.

We have found reasonably priced ovens which can provide a consistent reflow curve to be scarce. We have a conveyor oven at the moment which is to be sold once the system is operational. We do have a batch oven which is in operation, and will be our system for components, which can not be vapor reflowed.

But as always the biggest expense of the system is the fluid itself. With a gallon being right at $1000.00 it hurts with the initial investment.

James L

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
James L
Partner/Designer

Lil Brother SMT Assembly Services

WBA Consulting

James,
Well, it sounds like you have done a lot more research than average. I applaud you for that. Most people go off of sales pitches to make decisions. I'll be looking forward to seeing your final design. I don't recall any personal experience with machines that used the vapor blanket around the board to create a more controlled preheat process. If you get that down, I believe you will have a marketable product. If you can keep the volume of liquid down to 2 litres, that alone will be a selling point. I remember the stress that went into the handling procedures of the liquids. It was treated like gold.

I hope others who are unfamiliar with the vapor phase process will read the article. It does have a good theory/history essence to it.

good luck!

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Andrew Williams
WBA Consulting
IT / Web / PCB / Audio

James Long

WBA Consulting said...
James,
Well, it sounds like you have done a lot more research than average. I applaud you for that. Most people go off of sales pitches to make decisions. I'll be looking forward to seeing your final design. I don't recall any personal experience with machines that used the vapor blanket around the board to create a more controlled preheat process. If you get that down, I believe you will have a marketable product. If you can keep the volume of liquid down to 2 litres, that alone will be a selling point. I remember the stress that went into the handling procedures of the liquids. It was treated like gold.

I hope others who are unfamiliar with the vapor phase process will read the article. It does have a good theory/history essence to it.

good luck!

Andrew,

Well with anything, there are drawbacks. If you do get a chance, read Asscon's website (others should too).www.asscon.de/e/pages/products/quicky300.html
With the fluid being $1000.00/gal I don't doubt the critical aspect of working with the liquid.

I just hope we can make a product small companies can afford. I would like to beat out the American made 3 zone ovens price wise. I know beating the Chinese made batch oven is impossible, or the home built toaster reflow.

We have set a target of $5000.00, but hope to come in under that (Retail price).

James L

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
James L
Partner/Designer

Lil Brother SMT Assembly Services

Erik Friesen

I have a curve from the modified toastmaster that I have been using. Each line represents one minute. I used this to do lead free without too much difficulty.


Speaking of chinese competition, can someone explain to me how ourpcb can do 5 of these 4-layer boards for around $650 + shipping, while other USA sources quote 5 at $500 each?

I think pkfg?? has a reflow oven that retails for around $6000.

James Long

Erik Friesen said...
I have a curve from the modified toastmaster that I have been using. Each line represents one minute. I used this to do lead free without too much difficulty.


Speaking of chinese competition, can someone explain to me how ourpcb can do 5 of these 4-layer boards for around $650 + shipping, while other USA sources quote 5 at $500 each?

I think pkfg?? has a reflow oven that retails for around $6000.

Your curve is pretty nice, although a little on the slow ramp side.

As for the PKFG reflow oven, how big of a board can it handle, and how fast do you think it could be fed? That is the biggest problem with small ovens. The boards just remove all the heat, and the chambers must re stabilize. Money wise a small conveyor oven is a waste. it is running continuously, but can only be fed a board every 6- 20 minutes depending on the model. A toaster or batch oven is much better for small producers, home or otherwise.

If their oven doesn't suffer from temp oscillation, I would think that is a great price.

As for the Chinese being able to make boards for so cheap. First their labor costs are almost nothing, and they do not have to deal with our legal system nor our environment regulations. That is most of the cost involved with PCB manufacturer. Plus their quality can not compare. USA boards are most definitely better quality. But if you are not worried about how the board looks, as long as it works, you can get by with a Chinese board. We see many of both kinds.

James L

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
James L
Partner/Designer

Lil Brother SMT Assembly Services

Erik Friesen

I have done some heavy mods to this oven, but it has a better profile than my toaster. I put the elements on switches, and put the toastmaster fan inside the oven.

kwinn

I got a real deal ( $5.00 ) on a Black & Decker 12" Pizza sized toaster oven last week. It looked like it had been dropped, but the only damage seemed to be the bent housing and the door not closing properly. Worst case I thought I could remove and use the heating elements, but as luck would have it I straightened the sheet metal and realigned the door and everything worked. Unfortunately my wife saw it and liked it better than our current toaster oven. Oh well.

Erik, this was the biggest toaster oven I have seen, and it was 14" x 11"+a bit. Still not quite up to your requirement.

James Long

kwinn said...
I got a real deal ( $5.00 ) on a Black & Decker 12" Pizza sized toaster oven last week. It looked like it had been dropped, but the only damage seemed to be the bent housing and the door not closing properly. Worst case I thought I could remove and use the heating elements, but as luck would have it I straightened the sheet metal and realigned the door and everything worked. Unfortunately my wife saw it and liked it better than our current toaster oven. Oh well.

Erik, this was the biggest toaster oven I have seen, and it was 14" x 11"+a bit. Still not quite up to your requirement.

Do you have a model number. I would like to have a smaller batch oven, which is a little larger than the one I own.

James L

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
James L
Partner/Designer

Lil Brother SMT Assembly Services

kwinn

James, here is all the information I could find on it.

Black & Decker counter top convection oven
Catalog number CTO6301 Type 1
120V 60 Hz 1500W
Made in China
APPLICA CONSUMER PRODUCTS INC.
Miramar FL 33027