Electronics > Manufacturing & Assembly

Pick&Place assembly line

<< < (3/4) > >>

EEMarc:
In that case, I can see why you are working hard while starving. The thing that a startup would answer that I still haven't figured out is why would a company that has been in business for over a decade not have experience in making boards, a key component. I can come up with two guesses outside of a startup. One is a switch from through hole to surface mount. The other is a new set of products that use electronics to the extent of requiring PCBs where the old products did not.

For very low volume, it is often better to do it by hand. Like you mentioned, the setup time both initial to teach the machine how to operate, and to switch to that job in the future are quite time consuming. Add in the stencil cost, the time it takes to setup the reflow oven and stencils, and the fact that the machine will make most of it's mistakes on the first few boards/panels made results in additional cost in labor, parts, and expense. In this case, focus on getting good equipment for hand soldering SMD. This equipment will be good for low volumes stuff and also the inevitable rework for higher volume stuff. The keys to good hand soldering is very good lighting. Good magnification for fine pitch stuff. Good soldering equipment. High quality component handling equipment such as tweezers, vacuum pick up tools. Foot pedal operated paste dispenser. In many cases, I can place just as fast as a low end machine with the right setup. I just don't like to do this for extended periods of time which is why I use a PnP machine. When you are done, stick it in the reflow oven and let it work it's magic. You can get away with a less than good setup for a few boards, but when it comes to working a full shift soldering, you really need a good setup. Especially very bright lighting and good optics to relieve eye strain.

If a low volume product picks up in sales, upgrade it to PnP status. Just make sure that all of the designs are PnP friendly for best results. Some PnP machines have the ability to attach a syringe with attached pneumatic solenoid to apply it to the board. You can put paste or glue down where you want it. For the low volume example, you can get away with not getting a stencil for a given board by having the PnP machine paste the pads before placing the parts albeit, you are taking away machine time so a stencil is often a good investment. Better use is for the ability to paste is a new revision with the addition of an extra cap  or two or something else simple where other parts aren't moved. The stencil doesn't place paste where is doesn't belong, so fill in the new parts and save the cost for a new stencil. Alternatively, you can use glue to hold parts down. This is the secret for making 2 side component boards where you place parts on one side and run it through the oven. The second run through, the upside down components will be firmly attached except during the reflow portion. This also plays a roll in the choice of reflow oven. If you have a type of oven that will touch the bottom of the board, you will need to make a frame block for the boards/panels to lift them up. Small parts such as ICs and resistors, small caps, will be held in place by the solder's surface tension. The large components will fall off so glue them in place. If a part falls off, use your rework station to fix it and add glue next time. It will take some effort to get 2 side component boards to run well, but it is well worth the effort.

When it comes to pick and place, feeders are very important. Both in how many you get and how many places the machine can hold. You can have 10 different products or more that can run with 1 setup because they use a large number of the same components. Simply formulate a setup plan with all of the products and feeder locations. Simply load the the feeder plan, load product A's PnP file, put the stencil in and set the paster up, and set the reflow oven up. Then run the product A board through. When done, change the paster for product B, Change the PnP program, Change the reflow oven if needed, and run product B. Then C. And so on. If you have boards with similar reflow oven profiles, same it's setup time by running them back to back. This is the kind of thing you can do if you have enough feeders and enough feeder slots. If you run out of feeder slots but have enough feeders, set the later feeders up and simply swap a few feeders out when you need to.

If machine time is a limiting factor you can optimize the placement rate by selecting prime feeder locations for the most popular components such as 0.1uF capacitors and certain resistor values. Usually the closes feeder slots to the actual board are prime feeder slots. The time it takes to move to a feeder location, pick the part up, move to the board, and place can be very different for each feeder slot. Far away feeders take much longer than feeders right next to the board. For long runs especially, it can save a lot of time to move popular parts to prime locations. Notice in the uCurrent video that the locations selected are very close to the board allowing placement rates of roughly 6 parts per second. The optimization is also important for where to place the feeders relative to each other. Notice that the simultaneous pickup picks from specific reels. If not optimized, the head will have to move between picking up the parts slowing down the action. Low end machine often will not have simultaneous pickups but there is still gains to be made from relative feeder locations. There are also machine specific optimizations to be made.

For larger components that need to use the bottom vision as opposed to on the fly vision, remember that the next place after going to the feeder is to the bottom vision camera. This means that the prime location for bottom vision components is different than vision on the fly components. Also notice that the uCurrent video, the machine only uses 2 heads and the placement rate is very slow compared to on the fly vision placements. A tiny fraction in fact.

As for hookups, you need the right hookups. If you have single phase service only, try your best to get single phase machines. It is expensive to get a single to 3 phase converter or have 3 phase service put in. You may need to have dedicated circuits and special high current outlets installed depending on the reflow oven in particular. Nothing will ruin your day more than having the circuit breakers pop. 

If your machine needs air, either tap into shop air or you will need a good heavy duty oil filled compressor sized for your needs. If you have shop air service, you need to make sure that there is enough excess capacity to handle your machine. Venturi effect created vacuum uses a large volume of air. Keep that in mind. I prefer to use a good quality vacuum pump as it is more efficient than using shop air. Even if the machine doesn't use it, shop air is still very useful. Also, once shop air is available, others will probably want to put there stuff on it.  Make sure that you have some capacity for them if this will happen and also make sure that they don't cause problems by using too much air. This is especially a problem with some blower attachments that use a ton of air. If you do install some shop air, make sure to control condensed water. Usually the level plumbing will be at a slight slope to allow the water to flow. The best options is to have it flow back towards the compressor to a drop down line with the incoming air in a T. This is where the excess water is allowed to drain down to a stub pipe that collects water. Put a valve there to drain the water out of the line. For drop downs to machines and other workstations, the best practice is to use and upside down U pipe going up then back down. This will prevent the T from dumping water into the machine. It is good practice to also have a drain at the bottom of the drop down as well to collect and release any water that does condense there. With shop air, I'm considering adding a refrigeration unit to the shop air outlet. This is called a drier and it will eliminate all water from condensing in the system beyond the drier. The drier will need to be drained of condensed water.

Machine placement is also important. You don't necessarily need to have an assembly line setup. With low end machines, it is a good idea to inspect the boards after placement to fix the mistakes before putting the board through the oven. This can save time over reworking a board later. As for ideas for setting up a line, there are many good youtube videos that show some interesting setups. You can pick up some very useful hints and techniques from many unlikely sources. My favorite on there that I have seen is the Apple Macintosh Fremont California factory. I know a couple people that worked on the original Macintosh. It's funny to see Macintosh computers make more Macintosh computers.

Rundstedt:

--- Quote from: Psi on July 13, 2012, 01:12:04 am ---haha yeah, he raises a good point.

Also, you can add   "Setup a full pick & place / reflow assembly line"   to your CV, which sounds awesome :D

--- End quote ---
Well that would be awesome indeed.

--- Quote from: EEMarc on July 13, 2012, 03:33:59 pm ---In that case, I can see why you are working hard while starving. The thing that a startup would answer that I still haven't figured out is why would a company that has been in business for over a decade not have experience in making boards, a key component. I can come up with two guesses outside of a startup. One is a switch from through hole to surface mount. The other is a new set of products that use electronics to the extent of requiring PCBs where the old products did not.

--- End quote ---
Well while they were on market, their electronic stuff was (and still is) produced by outside companies, they look into integrating everything into company that's why I'm in charge of developing electronic products, and finding machines to buy.

--- Quote from: EEMarc on July 13, 2012, 03:33:59 pm ---For very low volume, it is often better to do it by hand.

--- End quote ---
Well it would be far too much to produce by hand, even now. At least 3000 MAP sensors a year. And a few other products.

--- Quote from: EEMarc on July 13, 2012, 03:33:59 pm ---When it comes to pick and place, feeders are very important. Both in how many you get and how many places the machine can hold. You can have 10 different products or more that can run with 1 setup because they use a large number of the same components. Simply formulate a setup plan with all of the products and feeder locations. Simply load the the feeder plan, load product A's PnP file, put the stencil in and set the paster up, and set the reflow oven up. Then run the product A board through. When done, change the paster for product B, Change the PnP program, Change the reflow oven if needed, and run product B. Then C. And so on. If you have boards with similar reflow oven profiles, same it's setup time by running them back to back. This is the kind of thing you can do if you have enough feeders and enough feeder slots. If you run out of feeder slots but have enough feeders, set the later feeders up and simply swap a few feeders out when you need to.
--- End quote ---
Good tip, but probably most of them will share few resistors and capacitors

--- Quote from: EEMarc on July 13, 2012, 03:33:59 pm ---If machine time is a limiting factor you can optimize the placement rate by selecting prime feeder locations for the most popular components such as 0.1uF capacitors and certain resistor values. Usually the closes feeder slots to the actual board are prime feeder slots.
--- End quote ---
I do not think that time will be a factor now, but tis is another great tip to consider.

--- Quote from: EEMarc on July 13, 2012, 03:33:59 pm ---As for hookups, you need the right hookups. If you have single phase service only, try your best to get single phase machines. It is expensive to get a single to 3 phase converter or have 3 phase service put in. You may need to have dedicated circuits and special high current outlets installed depending on the reflow oven in particular. Nothing will ruin your day more than having the circuit breakers pop. 
--- End quote ---
No, that's not a problem. We have our own transformer, compressed air or fume extraction also isn't a problem.

--- Quote from: mikeselectricstuff on July 12, 2012, 04:50:54 pm ---
--- Quote ---after that period I will renegotiate my wage.
--- End quote ---
Have they given you a target budget?

--- End quote ---
No they haven't cause they don't know how much they have to spend (yeah, I see the irony). They now want to buy used line to see in production what we need, and then buy new one. But I'm trying to force buying new one, with EU help we would probably pay half or even less but to get EU help we need to write proposal for specific machines. And usually when someone sells that kind of line there is a reason to sell it (I figured), either mechanics are getting screwed up or it can't pick up some elements.
Now I look though some responses from dealers to get orientation in prices.

Thank you all for the tips, if you have others, please share.

poorchava:
Heh, what company do you work for? I'm just curious.

Btw. If you feel like you're nor earning much, then maybe you can look for some company with foreign captital. Dunno how in your location, but here in Wroclaw (and nearby) we have quite many companies from electronics & IT market: Nokia Siemens, IBM, Espotel, Diehl Electronics, WABCO, TechniSat and others. And usually starters net salary for person with masters degree in whatever engineering type is around $800-900 (base is around 1100 but taxes and worthless social security take like 1/4...1/3).

On the other hand it's quite nice ocasion to put yourself in a position, where your employer can't afford to fire you :)

IanB:

--- Quote from: Rundstedt on July 13, 2012, 09:05:25 pm ---Well while they were on market, their electronic stuff was (and still is) produced by outside companies, they look into integrating everything into company that's why I'm in charge of developing electronic products, and finding machines to buy.
--- End quote ---

Bringing manufacturing in house seems like an odd step to be taking. Usually the companies you would contract out to would have at their disposal much more sophisticated machines with greater capabilities due to the volume of work they undertake. Since they would be specializing in manufacture you would expect them to be more efficient and more consistent too.

What is driving the in-sourcing of the work? Is it to have greater flexibility and control with faster turnaround times perhaps? Or are there quality problems with the outside contractors?

I have no great knowledge of this, but I'm curious. One normally would expect to concentrate on the part where you have most value to add (product design and customer satisfaction) and to delegate the routine aspects like circuit board assembly to other companies that specialize in that area.

Rundstedt:

--- Quote from: poorchava on July 14, 2012, 09:06:22 pm ---Heh, what company do you work for? I'm just curious.

Btw. If you feel like you're nor earning much, then maybe you can look for some company with foreign captital. Dunno how in your location, but here in Wroclaw (and nearby) we have quite many companies from electronics & IT market: Nokia Siemens, IBM, Espotel, Diehl Electronics, WABCO, TechniSat and others. And usually starters net salary for person with masters degree in whatever engineering type is around $800-900 (base is around 1100 but taxes and worthless social security take like 1/4...1/3).

On the other hand it's quite nice ocasion to put yourself in a position, where your employer can't afford to fire you :)

--- End quote ---
I don't even have engineering degree yet
--- Quote from: IanB on July 14, 2012, 10:56:26 pm ---Bringing manufacturing in house seems like an odd step to be taking. Usually the companies you would contract out to would have at their disposal much more sophisticated machines with greater capabilities due to the volume of work they undertake. Since they would be specializing in manufacture you would expect them to be more efficient and more consistent too.

What is driving the in-sourcing of the work? Is it to have greater flexibility and control with faster turnaround times perhaps? Or are there quality problems with the outside contractors?

I have no great knowledge of this, but I'm curious. One normally would expect to concentrate on the part where you have most value to add (product design and customer satisfaction) and to delegate the routine aspects like circuit board assembly to other companies that specialize in that area.

--- End quote ---
Well there are many factors:
- there are little to no after-warranty services. Since we didn't design product there isn't a lot what we can do to repair it.
- there are some little difficulties with our contractors (some minor errors in product that they refuse to acknowledge)
- we export a lot to China/India/Thailand and other cost sensitive markets, and producing in-house will cut expenses (you say that outside companies have better machines, but we are looking at similar line)
- we have no control over producing capabilities of companies we hire, sometimes when they have too much other orders they will put ours at back burner.

There are a few other reasons to.



Navigation

[0] Message Index

[#] Next page

[*] Previous page

There was an error while thanking
Thanking...
Go to full version