SMT Quick-Tips: Selecting a Pick and Place Machine
Robert Voigt, DDM Novastar
KF-SMT contains other products and information you need, so please check it out.
How To Select A Pick and Place Machine
A Pick and Place machine is the second step in a paste, place, reflow assembly operation. The “Place” function follows the “solder paste” function (stencil printer). The ‘place’ operation selects and delivers a component over the board and drops it into position. The simplest form of pick and place operation is by hand, that is, manually picking a component from a bin and, with the aid of a pair of tweezers and a magnifying glass, positioning it on the board and completing the operation with a hand-held soldering iron.
This method works well if you are only doing occasional boards. Other things to consider – the size of the components (big or small) – affect the time required to hand place and solder. Fine pitch components are another issue, where more precision and accuracy are required, and the human factor comes into play. The work then become more tedious and time consuming.
First, we will focus on machine-assisted manual systems for users interested in going from a couple boards a day to much higher production volumes. Fully automatic systems are complex enough that we’ll cover them separately.
PRODUCTION VOLUME
Let’s start by addressing production ranges for various types of machine-assisted manual systems. For purposes of comparison, since all circuit boards vary in size and complexity, we’ll talk about volumes in terms of components per hour, or CPH. This will help you to decide what level of automation you’ll need.
On the very low side of the scale – using a manual hand system – the only expense is the appropriate hand tools for non-machine assisted manual placement. At the high end of the spectrum, these machines are often modular or customized for high speed unattended operation. Buyers in this market are likely looking more at ROI than initial cost.
Manual and semi-automatic systems
A manual pick and place system is desirable for small, growing operations that need to increase their hand-held production volumes incrementally while also improving quality, thus reducing rework or rejects; however, the accuracy of placement is still limited by the capability of the operator. Benefits of a machine-assisted manual system include:
- Less operator fatigue
- Fewer placement errors
- Better control
- Improved yield, less rework
A machine-assisted manual system can be equipped with features such as an X-Y indexing table with vacuum pickup head or pen; ergonomic fixturing to help relieve operator fatigue; and additional fixturing for θ (rotation) and Z (height) positioning in addition to X and Y.
Fig 2: Component trays and feeders
Some machines offer an optional liquid solder paste dispenser, which is applied just before placing the component on the board if a stencil printer has not been used. Additional options include:
- Component Handling Trays
- Liquid Dispenser
- Tape Feeders
- Feeder Racks
- Vision Assist option
- Optional Stands
In most cases, machine assisted manual systems can be purchased with just the bare necessities, and desirable options can be added later as they’re needed.
Fig 3: A vision-assisted manual machine
Semi-automatic systems
Today there are very few semiautomatic machines still being made because of the increasing affordability of some of the more automated systems on the market. They were originally introduced at a time when the leap from manual to fully automatic systems was just too cost prohibitive and were made available with some features to assist the manual operation.
More correctly referred to as “enhanced manual” systems, semi-automatic pick and place machines typically include a computer interface with a vision system that shows where the components go, but the placement itself is still done manually. This type of machine helps the operator position ultra-fine pitch components more accurately for low-volume applications, an operation that is very difficult to accomplish using a simple machine-assisted manual machine.
Ease of use
Most pick and place machines will handle a fairly wide variety of board sizes, with a work table designed to accommodate boards up to 16” x 24”. There is also ease of control over the components, which aids in accuracy, along with a simple learning curve. In most cases, no training is required.
Don’t overlook electrical requirements. Make sure the machine you buy will plug-and-play in your environment without pulling in new wiring or else plan on an adapter/transformer.
AUTOMATIC PICK AND PLACE MACHINES
We will start by talking about two aspects of machine capability – accuracy and repeatability, and Pick and Place centering methods.
ACCURACY AND REPEATABILITY
For production machines, we typically recommend looking for a machine with accuracy of +/- .001” and down to fine pitch capability of 12 mil on a repeated basis. Less expensive machines often don’t meet this spec, so that’s something to be aware of.
Most low-cost machines will also not come standard with a computer or software which could help with the repeatability aspects if not the accuracy. While some may offer enhanced technology – most do not.
Check out DDM Novastar's wide range of Pick and Place Machines.
Have questions or need help specifying a machine? Use our contact form for assistance.
PICK AND PLACE CENTERING METHODS
There are four (4) methods for pickup and placement:
- No centering mechanism
- Laser centering
- Mechanical (jaws)
- Vision centering
1. Method 1: No centering mechanism other than relying on the component’s pick-up point for placement. In other words, the part is not physically centered after being picked up by the tool head, and if it’s picked off-center on the tool, it will be off-center when placed on the board. This is obviously not a very accurate placement method because there is no definable tolerance. You can expect to find this method used by hobbyists or instructors, but certainly not in any type of precision production environment. There are not many options available either, and long-term reliability is questionable.
- a. Pros: Low cost.
- b. Cons: Low accuracy, repeatability and long-term reliability, no options, or spare parts.
- c. Size range: No definable tolerances
Fig 4: Mechanical centering
2. Method 2: Mechanical centering jaws or fingers In this method, the component is picked up and moved into its center position in the X and Y axes on the pick-up head. Typically, this method is easy to set up and repeatable within +/-.001” accuracy. This centering method is generally found in low- to mid-range machines.
- a. Pros: Easy to learn and set up; repeatable; one of the fastest method currently available; a true “on-the-fly” system; low cost.
- b. Cons: Physically touches the component which may not be appropriate for certain types of parts, especially those with delicate leads.
- c. Size range: packages up to 35 mm square.
3. Method 3: Laser centering In this method, the component is picked up inline with a laser beam which detects the component’s center position on the tool head and recalculates the zero point of the part according to its position in the X, Y axes and rotational position relative to the head for an accurate placement on the board.
Fig 5: Mechanical squaring (older version)
- a. Pros: Touchless; on-the-fly (similar to mechanical method).
- b. Cons: It is less reliable. There are limitations on the types of parts it can handle, such as very thin components (If .050 thin, they may need to be reset because of part variations, even from the same vendor); requires longer setup time, since the Z axis (part thickness) must be defined; more costly than Mechanical Centering, but about the same as Vision.
- c. Size range: cannot center parts below packages or larger than 35 mm square.
4. Method 4: Vision centering Here, there are two types, Look-Down and Look-Up. Look-down vision will view the top of the component prior to picking it up for its pick-up location. It then calculate its center, compares it to its image file from the stored database, then picks up the component and transports it to its position on the board.
- a. Pros: True touchless centering; can handle odd-shaped and delicate components; Look-Down Vision Centering placement is accurate to +/-.004”.
- b. Cons: Typically longer setup times due to the need to teach the vision system how to identify part images which are stored in the machine’s database; a slower method of centering due to time slice required for processing; Vision is more costly than the mechanical method; for Look-Down vision, the part may move from its pickup point to its placement on the board.
- c. Size range: – 15 mm
Fig 6: Look-up and look-down vision centering
The Look-Up Vision method is the most accurate centering method available. The component is first picked up from the pickup area, moved to a camera station that looks at the bottom of the component, and calculates its center position.
- a. Pros: True touchless centering, handles delicate components; accurate down to +/- .001” positioning capability
- b. Cons: Typically, a longer setup time due to the need to teach the vision system how to identify the image, stored in the machine’s database; a slower method of centering due to processing time; Vision is more costly than the mechanical method.
- c. Size range: – 50 mm (can see smaller and more detail)
The Pick-Up and Centering Method you choose will have a great deal of influence on the quality and speed of your production needs, along with how to relate this accuracy back to the machine. But, that’s just the beginning.
As with any complex machine, there will be tradeoffs between cost and capabilities, some of which specifically relate to production accuracy and yield. We will address next:
- Mechanical positioning methods
- Machine construction
- Solder paste fluid dispensing
- Component feeders
To review, when starting the evaluation process, there are two defining factors to keep in mind which determine what category fits your machine needs. The first principal factor is CPH (components per hour), and the secondary factor is machine capability. While it’s constructive to start by understanding how production rates affect the type and performance of a pick-and-place machine, please refer to the prior two chapters for those ranges.
Machine capability is the second defining factor in helping choose the correct auto pick and place machine for your needs. In this chapter, we will address three aspects of machine capability that have a direct impact on final board quality and production yield.
COMPONENT POSITIONING SYSTEMS
For more information, please visit Smt Feeder Parts.
Fig 7: Component Pickup Strip
After each component is picked up and centered in the tool by one of the methods described in the previous chapter, it must then be positioned accurately on the board, in an X-Y position. There are three methods commonly used for positioning:
- Positioning with no feedback system (open loop system)
- Positioning with rotary encoders (closed loop system)
- Positioning with linear encoders (closed loop system)
Method 1: No positioning feedback loop In this system, the motor drives the part to a location on the board defined in the program by the number of steps in each X-Y axis, but there’s no way to tell if it actually ends up in the right place. These systems use stepper motors for positioning.
- a. Pros: Low cost
- b. Cons: Unreliable accuracy; not recommended for high quality production
Method 2: Positioning with rotary encoder In this method, an encoder is mounted directly on the motor shaft and delivers position feedback to the control system; however, it only reports the motor position, and not the actual position of the x-y axis. This is dependent upon the remainder of the mechanical components that make up the machine. These machines can use stepper or servo motors. (and usually associated with cost)
- c. Pros: Low cost; this system is widely used on entry-level machines
- d. Cons: Typical positioning accuracy of +/- .005”
Method 3: Positioning with linear encoder In this method, linear scales are mounted to the machine’s X-Y axes table and an encoder is mounted on the traveling beam that will be carrying the components. This method will report its actual position back to the control system and make corrections to the position programmed, if needed, to within a few microns of actual X & Y location for the component placement (which is typically 12,800 increments – or steps – for each inch of travel). The best machines in the this category uses servo motors.
- e. Pros: Very high accuracy, to within +/- .”; very repeatable
- f. Cons: More costly, but necessary for high value production
NOTE: The quality of the encoder (the position feedback sensor), is an important element in the whole system and does affect accuracy.
Fig 8: All-welded construction
MACHINE CONSTRUCTION
When selecting a pick-and-place machine, you should be aware that its construction will dictate its effective CPH range and footprint, including considerations for the number of component feeders it can accommodate.
1. All-welded steel: The most accurate machine will have a frame that is constructed of solid welded structural steel tube. This provides significant stability necessary for accurate positioning and high-speed movement of X & Y axes. This construction method is recommended for ANY production environments, and it will remain stable without requiring ongoing calibration.
2. Bolt-together frame: Extruded aluminum or formed sheet metal frame will come with a lower initial accuracy than a welded frame and will need to run more slowly because it can’t handle the rapid inertia shifts of X – Y axis movement. Further, it will likely go out of calibration frequently, which will adversely impact labor time, downtime and yield. (Lower cost usually reflects a weaker construction.)
SOLDER PASTE / FLUID DISPENSING
Any pick and place machine should be capable of offering fluid dispensing systems. Most common liquids include solder pastes, adhesives, lubricants, epoxies, fluxes, glue, sealants, and more. This is a valuable option when building prototypes or one-off PCB assemblies that do not warrant the cost of a dedicated printer stencil or foil.
Fig 9: Work deck with component feeders
COMPONENT FEEDERS
If the machine’s production will be dedicated to a small number of components and type of job, it’s very easy to identify the number and type of feeders. However, that is not usually the case with contract assembly shops, since they don’t know what type of board and how many different components the next job will require. Some OEMs also need flexibility for a wide range of board configurations, especially if they intend to use the same machine for prototypes and several different production boards. So, it’s useful in those cases to consider a machine with the greatest number of feeder position and options that can accommodate the footprint your space can handle.
Types of feeders include:
- Cut strip holders are usually associated with the low volume world.
- Matrix tray holders are used for components that are not available on tape.
- Tube feeders dispense components supplied in tubes.
- Electric tape (and reel) feeders are usually more costly initially, but offer the best long-term investment. Electric tape feeders are available as single units in a variety of sizes, and cover the range of components up to 56 mm large components. Many manufacturers now offer a multiple feeder (known as bank feeder). These are available for 8 mm tape, and can come with up to twelve 8mm feeder lanes per unit.
Fig 10: Tape feeder
Since components are packaged in many forms, e.g., discrete components on tape, quad packs, matrix trays, tubes, cut strips, etc., your choice of feeders would depend on your production but also on any size restrictions you may have. A good starting point is to purchase the most feeders you can get in the footprint you have available.
SOFTWARE
When considering the purchase of a Pick and Place machine, one of the most important considerations is the software interface. There are three primary goals of a good operating system for users in the low to mid-volume range, defined as up to 8,000 CPH:
- Maximizing ease of use
- Providing broad flexibility
- Optimizing performance
Fig 11: Machine ground holding plate
Ease of Use
Because small to mid-volume assembly operations must switch projects often, ease of setup and use is a much more important factor than it is for large volume operation where a single setup may handle a run of hundreds of thousands of components. The job-shop assembler must be agile enough to switch among a wide range of board sizes and component selections quickly to meet a variety of production demands. The machine must also be able to handle a broad range of component sizes, from very small to very large, without a strain on setup and testing.
In contrast, large production machines are often made up of several modules of pick and place systems, positioned in-line where they’re needed most for fine pitch parts, chip shooters, or optional tasks. This allows the high volume producer to customize a line to optimize production speed, efficiency and quality. In these environments, a longer setup can be tolerated, because it will be made up in production efficiency.
First, some basic questions:
- Does the model you’re looking at come with a computer, or just the software? This is neither good nor bad, since some users prefer to install the software on their own PCs; however, a fully integrated system assures there will be no software compatibility issues, and this can streamline the installation and setup.
- Does the machine run on a familiar graphical user interface (GUI) such as Windows™ or a proprietary system? Most every operator will be immediately familiar with the intuitive interface of Windows conventions, a key factor in accelerating usability, particularly for a new machine. A proprietary GUI may require a longer learning curve.
Skill sets of the operator should be augmented by the machine provider with:
- Good documentation
- Hands-on training or videos
- A utility to teach the machine common components and repetitive routines
For machines handling over 8,000 CPH, expect a higher learning curve because complexity increases significantly.
Flexibility
An important utility to look for that gives the custom assembler great flexibility is a Universal CAD Translator function (UCT). The UCT allows users to import the pick-and-place data to the machine’s database to help create the program and scale it. When a project is begun, the user selects the program to run from an archived set of files. This permits quick change over from one board to another since all the programming is memorized.
Fig 12: Screen of Universal CAD Translator (UCT) Software Program
01. Text window displays the file to be imported and its location /path.
02. Pressing button opens window to allow user to change the CAD file to be imported.
03. Displays the folder to which the converted file in the .prg format will be saved.
04. Press Browse opens window to allow user to change location in which the folder for the .prg file will be saved.
05. Default name of exported file. Can be changed by editing in this text box.
06. Import Sides.
07. Top- if selected/checked, the top side component placement will be imported.
08. Bottom- if selected/checked, the bottom side component placement will be imported.
09. Displays filename and path/location of the file after importing is completed.
10. Input File Preview Filtered (up to 50 lines).
11. Character entered to ignore unnecessary information not required for the .prg file.
12. Enter the number of lines to be ignored from pick & place CAD during importing process.
13. Using either cComment Char or Ignor Lines displays file in preview window without comment lines.
14. Displays file in preview window in its original format (including comments).
15. Information Box.
16. Displays program lines that will be imported from the user defined selections.
17. Displays lines actually imported.
18. Displays lines not imported in .prg format.
19. Displays number of lines the UCT processed for importing the file into the .prg format.
20. Adds angle to each placement.
Another feature to look for is a master feeder and component database. Once the operator stores component data, it’s there forever and can be accessed and imported into any new board configuration setup. This database grows as you add components, so over time, you’ll spend less time programming and more time producing. Often, the database will remember inventory, so as you use components, the remaining stock will always be available to check. This is a great feature for scheduling and inventory planning.
Be sure to see if the system you’re considering only stores data for particular boards rather than an entire component database. If so, it will only remember specific board component data, and it won’t show all available inventory.
Fig 13: Screen of offline simulation for optimization
Optimization
Certain utilities are often provided with a well-designed machine to aid in setting up and programming the system. One of the most important utilities affecting optimized performance is offline software.
Offline software allows the user to simulate the pick and place machine’s routine in a remote environment for programming. It can be installed on any computer and looks just like the machine’s GUI. It allows the user to manipulate the program to sort functions and modify program lines for most efficient use and speed, for instance, by aggregating like-type components in the same sequence, minimizing tool changes and the time it takes to perform those functions. It can also create board references for multi-boards prior to running on the machine.
To speed job change-over, software interface should include sub-routines for common operations such as setting up matrix trays, identifying strip feeders, and teaching for vision centering. Expanding on the last point, image interpretation should be clear and straightforward; if it isn’t – and the machine has a hard time recognizing a component – the result could be a misplaced component, resulting in a lot of unnecessary rework. A well designed software interface will capture a range of image qualities for each component type that are all representative of an acceptable part, and store it as an approved file. This improves speed, repeatability and efficiency, and final board quality.
Other considerations
Just as important as the physical features of a quality pick and place machine are the “soft” features. Make sure to check on:
- Availability of on-site or factory training?
- Remote diagnostics – can your vendor provide this via online support?
- Critical software updates – Do they come free or with a cost?
- Is the software interface available for pre-sale review?
VENDOR SUPPORT
When evaluating any type of SMT machine, consider factory support as one of the most important assets of your purchase. The best way to learn how a company treats its customers is by word of mouth. Talk to several customers to find out how happy they are with the machine, the seller, and the support they provide. Where is the manufacturing plant? Can they help troubleshoot alignment issues over the ? Do they offer field service? Do they have spare parts in stock for immediate shipment? While there isn’t much of a used market for manual, machine-assisted or enhanced manual pick and place machines, it’s still a good idea to ask your supplier about their older machines in the field, and if down the road, spare parts are available, and about their capability to customize a spare part if the machine becomes obsolescent. Ask what the expected life-cycle of the product is. The industry standard is seven years. Remember, there is a difference between a true manufacturer and an equipment supplier or distributor.
0 Members and 1 Guest are viewing this topic.
Want more information on Juki Replacement Parts? Feel free to contact us.
ajb
Setting up a 'real' SMT line - advice?
«
on: June 26, , 05:01:36 pm »
We're at a point where we're seriously planning out a real in-house SMT production line. This decision is not based on volume/cost, but on the need to support high mix/low volume production with often very tight timelines, and support shorter R&D cycles on boards we don't want to be hand assembling. We've considered going with something like the XY or Neoden equipment to start with, but it doesn't seem like that equipment will last us long enough to be worth trouble of setting up and swapping out when we inevitably want to upgrade. So I'm currently looking at options for 'real' equipment, with a new placement machine, probably an improved manual stencil printer, and possibly a used reflow oven if we run into a good opportunity. (Unfortunately most of the used inventory for this stuff is on the wrong side of the country from us.) At some point in the future we would upgrade to an automatic stencil printer, and otherwise expand in feeders and the overall line from there. We have a good amount of floor space picked out, with plenty of power and air capacity. We also have experience with bringing up in-house CNC manufacturing capacity, so have some idea of the amount of workflow and tooling support that these sorts of things inevitably entail.
I've got pricing for Hanwha, Yamaha, and Panasonic equipment so far (and yes, the pricing is eye-watering), and I can look at brochures and promo videos all day, but that all only tells me so much. For those of you who have gone down this road:
- what's not covered in the sales pitch that you found was important to know when getting started?
- what's been your experience with getting started in any particular brand/class of equipment?
- per the above, any particular machines or brands you found to be easier or harder to get started on?
- what other equipment or accessories did you find were essential to support the SMT line? (aside from 'more feeders', that's already on the list!)
loki42
Re: Setting up a 'real' SMT line - advice?
«
Reply #1 on: June 27, , 01:38:34 am »
I've had minimal problems with the pick and place section (with my Universal Genesis machines) apart from trim pots which have been very picky about nozzle choice. Check you're getting the nozzles you need, the UIC stuff has different heads with different nozzles and it's pretty confusing. There's a big spreadsheet for what picks what on what head. Oh trays have been a bit of trouble too, I think because my belt alignment isn't prefect and I don't know how to set it up.
Printing has been a lot of trouble for me, mostly around board support and getting the parameters right. I've just bought spi to try to catch errors. The on machine 2d inspection is handy but it can't capture volume errors (coverage is good but too thick/thin).
Is the Panasonic similar price to the hanwha and Yamaha? I thought they were much higher end, similar to Fuji.
The following users thanked this post: ajb
SMTech
Re: Setting up a 'real' SMT line - advice?
«
Reply #2 on: June 27, , 07:35:51 am »
Understand your requirement first and foremost. Work out your inventory, your changeover, what parts might need to be loaded into feeders ready to switch between products. What feeders those parts need. Genuine low volume high mix is directly targeted by two manufacturers not on your list; Europlacer and MyData. Both of these have solutions that make owning a lot of "feeders" cheaper in a space efficient manner. And save yourself the pain and buy an automatic printer.
Get an idea of pricing on consumable parts like nozzles. Get hands on time with feeders and software see how intuitive or fiddly things are. Establish if you think you would be best served with "smart" feeders. Understand the software options included and your own requirements for things like traceability, these things cost extra and are sometimes part of expensive suites.
@loki42 have you looked at GridLok/QuikTool/Matrix https://quik-tool.com/ https://psasystems.com/smt ?
The following users thanked this post: ajb
loki42
Re: Setting up a 'real' SMT line - advice?
«
Reply #3 on: June 27, , 01:39:46 pm »
I've looked into grid loc etc for my printer and given I'm only doing a very low number of different boards (very low mix) dedicated tooling makes more sense for us i think. I attempted to get some 3d printed by a professional 3d printer but the flatness was terrible so I just made some on our press brake. It's working well. I'm planning to machine some vacuum fixtures at some point too. I've got some of the official UIC board supports too and the similar ones from count on tools.
Other thing I forgot pnp wise that was wisely said by smtech, make sure you can put down the parts you want with the machines you are thinking about. Some are very limited with part height or can't feed from trays. If you need to place odd form you might either need to place it by hand or limit your machine options.
The following users thanked this post: ajb
ajb
Re: Setting up a 'real' SMT line - advice?
«
Reply #4 on: June 27, , 02:56:47 pm »
Thanks folks, this is all helpful. I am planning to arrange some hands-on time with a couple of machines before we make a final decision to suss out some of the software/workflow considerations.
Is the Panasonic similar price to the hanwha and Yamaha? I thought they were much higher end, similar to Fuji.
They're all in the same general ballpark, but the Panasonic is a bit more expensive--or the same price for a few less feeders. Seems like everyone gives meaningless list prices with a big discount on everything, which makes comparisons annoying depending on how they break out that discount and option charges. For reference, this is for the Hanwha SM485, Panasonic AM100, and Yamaha YMR20-1. Several other manufacturers to look into of course, there are more options than I would have guessed before I started counting.
Genuine low volume high mix is directly targeted by two manufacturers not on your list; Europlacer and MyData. Both of these have solutions that make owning a lot of "feeders" cheaper in a space efficient manner.
Thanks, will bump those up the list to look at.
Some are very limited with part height or can't feed from trays. If you need to place odd form you might either need to place it by hand or limit your machine options.
Lack of internal trays is a downside for the Hanwha, it's either give up 23 feeder lanes for two trays, or pay $50k for the side tray magazine, neither is attractive for our current part mix.
jmelson
Re: Setting up a 'real' SMT line - advice?
«
Reply #5 on: June 27, , 03:10:16 pm »
Big items to consider :
1. height of tallest component - especially electrolytic capacitors
2. width of widest component - like CPUs and FPGAs
3. What nozzles are available for the machine
4. what's the smallest part you will be placing
5. finest lead pitch of ICs
If you only will have one P&P, you want a machien that can handle the FULL range of parts you will use. On a multi-nozzle machine, if the nozzles are too close together, you have issues with the parts bumping into each other and messing up the alignment.
Does the machine have up-looking cameras or flying vision? The flying vision is a lot faster.
And, of course, feeders!!! The quality of the feeders is really key! My first P&P had mechanical feeders that were always jamming and not pulling cover tape reliably. It was a HUGE pain. My newer machine has electronic feeders that do quite a bit better.
How will you set up a job on the P&P? For both of my machines, I wrote programs to convert the placement file of my CAD system to what the machine needed for import, then I had to assign feeders and get the machine to accept the setup. Complexity and ease of use of the software during setup can make a HUGE difference. Also, self-diagnostic features are pretty important. A machine that tells you exactly what went wrong "Component from feeder # 43 X dimension too small 0.3mm" is much better than one that is vague "error ocurred".
Jon
The following users thanked this post: ajb
SMTech
Re: Setting up a 'real' SMT line - advice?
«
Reply #6 on: June 27, , 03:15:23 pm »
Thanks folks, this is all helpful. I am planning to arrange some hands-on time with a couple of machines before we make a final decision to suss out some of the software/workflow considerations.
Is the Panasonic similar price to the hanwha and Yamaha? I thought they were much higher end, similar to Fuji.
They're all in the same general ballpark, but the Panasonic is a bit more expensive--or the same price for a few less feeders. Seems like everyone gives meaningless list prices with a big discount on everything, which makes comparisons annoying depending on how they break out that discount and option charges. For reference, this is for the Hanwha SM485, Panasonic AM100, and Yamaha YMR20-1. Several other manufacturers to look into of course, there are more options than I would have guessed before I started counting.
Genuine low volume high mix is directly targeted by two manufacturers not on your list; Europlacer and MyData. Both of these have solutions that make owning a lot of "feeders" cheaper in a space efficient manner.
Thanks, will bump those up the list to look at.
Some are very limited with part height or can't feed from trays. If you need to place odd form you might either need to place it by hand or limit your machine options.
Lack of internal trays is a downside for the Hanwha, it's either give up 23 feeder lanes for two trays, or pay $50k for the side tray magazine, neither is attractive for our current part mix.
Internal tray king is a Eurplacer iineo, it can be fitted with a tray area in the middle of the machine over the conveyor which can hold 10 trays, or a tray of lose parts or strips on sticky tape, your choice. Grippers, tall parts it does it all if you pay enough.
The following users thanked this post: ajb
loki42
Re: Setting up a 'real' SMT line - advice?
«
Reply #7 on: June 28, , 12:57:25 am »
Internal maybe... but a universal instruments machine with PTFs you can have 100s of stacks of trays. The trays are automatically removed when empty and parts can be gang picked from the PTF conveyors. The PTFs are also pretty cheap second hand. Inspection during new part introduction is more annoying than with direct pick though as there's no camera on the ptf side so you have to pick the part with the ptf so see if the orientation is correct. I assume the Fuji, Panasonic etc tray feeding options would be similar.
« Last Edit: June 28, , 12:21:18 pm by loki42 »
The following users thanked this post: ajb
SMTech
Re: Setting up a 'real' SMT line - advice?
«
Reply #8 on: June 28, , 07:50:49 am »
Internal maybe... but a universal instruments machine with PTFs you can have 100s of stacks of trays. The trays are automatically removed when empty and parts can be gang picked from the PTF conveyors. The PTFs are also pretty cheap second hand. Inspection during new part introduction is more annoying than with direct pick though as there's no camera on the ptf side so you have to pick the part with the ptf so see if the orientation is correct. I assume the Fuji, Panasonic etc tray feeding options would be similar. Europlacer is probably better for very short run prototype stuff but I'd say Panasonic, Fuji and ASM are the higher end stuff.
Every machine has its gimmick and its ideal target audience, every machine also has a PTF option, they mostly take away feeder slots and cost $$ whereas internal space is free(ish) (the Europlacer table does cost money). You just have to work out which niche you are in, "higher end" is a difficult thing to quantify, you'd never try and built millions of iPhones or Pi's with a Europlacer line, they're not designed to compete with an NXT. NXT lines have their own limitations.
We had similar deliberations when we last looked @ p&p we knew we needed to fit a certain number of feeders and trays and that basically was not possible with a single "conventional" placer, the double sized machines back then were not flex placers. That left Europlacers, long MyDatas, maybe the iPulse M20 or standalone machines that take feeders on all sides and have a placement area you can use as you like (Essemtec/Dima/Fritsch/Autotronik). In theory it could also have included Universal (ouch); the Panasonic AM100 (poorly sold then); Contact C5 (looks sorta cool, nobody buys it); or Mimot (because we all know them ). And there we were boxed in our niche, luckily the choice is better today, more machines like the RS1 are truly flexible but we would still need 2 of most and at that point you can trigger the gripe mentioned elsewhere, load balancing software suites that manage your multi machine line.
The following users thanked this post: ajb
loki42
Re: Setting up a 'real' SMT line - advice?
«
Reply #9 on: June 28, , 12:20:32 pm »
Yeah, you loose a nozzle changer for each PTF with UIC but who needs 4 x 77 nozzles... you don't loose any feeder slots. There are for sure better machines for specific jobs. Feeders for UIC are giant compared to Mycronic and the handles are designed by people who haven't seen a human. I think used UIC are a decent deal used but I wouldn't touch them new. The company is all running on Lotus domino and I don't think they've updated any of their software in the last 20 years... I think finding a company that is actually friendly and easy to deal with would be much better.
I think some of the machines like the Europlace, Mycronic and maybe Essemtec can feed from cut tape easily which would be great for prototyping. You mentioned sticking them down or is there specific cut tape feeders? Like these things that I'm not sure how they work... https://www.nozzles4smt.com/STRIPFEEDERS-mod-Ver-2-Series_c_135.html
The following users thanked this post: ajb
SMTech
Re: Setting up a 'real' SMT line - advice?
«
Reply #10 on: June 28, , 01:01:36 pm »
Yeah, you loose a nozzle changer for each PTF with UIC but who needs 4 x 77 nozzles... you don't loose any feeder slots. There are for sure better machines for specific jobs. Feeders for UIC are giant compared to Mycronic and the handles are designed by people who haven't seen a human. I think used UIC are a decent deal used but I wouldn't touch them new. The company is all running on Lotus domino and I don't think they've updated any of their software in the last 20 years... I think finding a company that is actually friendly and easy to deal with would be much better.
I think some of the machines like the Europlace, Mycronic and maybe Essemtec can feed from cut tape easily which would be great for prototyping. You mentioned sticking them down or is there specific cut tape feeders? Like these things that I'm not sure how they work... https://www.nozzles4smt.com/STRIPFEEDERS-mod-Ver-2-Series_c_135.html
Essemtec has an equivalent of that, it just works like custom tray, its a bit of pain to use to be honest, I tend to try and use a real feeder. The count on tools solution probably fits in a PTF and is more limited in component height. I can't recall what Mycronics solution was, Europlacer you can literally use sticky tape, and they will custom machine you things too. Europlacer also used to have a feeder element specifically for cut tape, I have no idea how it was different. The loose part thing however is very new. Yamaha offered the CoT tray, I think Hanwha has a fancy feeder that can handle tape strips but I'm not sure how short they can be or what the wastage is like. There are also "FeederFingers" for many brands, I think they try and copy the Agilis method onto a standard feeder.
I agree finding a company you can work with is also a good idea, if getting information and quotes out of them is difficult, imagine the support. We are doing some investment right now, I have one failure to reply, I'm not chasing. There are also some very different approaches to servicing...
The following users thanked this post: ajb
ajb
Re: Setting up a 'real' SMT line - advice?
«
Reply #11 on: June 28, , 04:16:00 pm »
Any thoughts on backup pins, height mapping, placement force monitoring features? I don't think we need to worry about it, but those are features that show up in some brochures but not others, so I'm curious how widely applicable they are. We're not doing max width boards/panels, but some of our current tab routed panels have a fair bit of flex. Of course those could be redesigned into smaller panels and/or wider frames.
Reflow ovens: Sales reps are telling me 7-8 zones minimum for lead free, but I've also heard from a couple of people who are getting good results with 3-zone systems. Given that throughput isn't a big concern, I'm interested in what anyone else has to say on that.
Re: Trays, not sure about other brands, but Hanwha offer a PTF that installs on the conveyer side. So you don't lose any feeder slots, but do lose the dual conveyer/shuttle option, for whatever that's worth. Still more expensive than I can justify without more tray parts, but would be a decent future upgrade option.
Re: Cut tape, Panasonic also have one of those cut-tape-on-tray thing, which seems like any machine that supports trays ought to be able to offer. A fixed piece of cut tape is just a 1xN matrix tray after all. I guess the difference would be in how easy they make it to program that sort of thing. Those Feeder Fingers look interesting, will add those to the list. I would imagine the fancy autoloading feeders from any brand would be able to handle cut tape, but I'm sure a standard electronic feeder plus a Feeder Finger is a cheaper way to go.
For our purposes, being able to put two trays inside the machine, one for tray parts and one for cut tape, should work pretty well. By the time we need more tray capacity I will be much less concerned about cut tape. (Not that I'm even that worried about cut tape now, but it would be nice to use up what we have on hand.)
I agree finding a company you can work with is also a good idea, if getting information and quotes out of them is difficult, imagine the support. We are doing some investment right now, I have one failure to reply, I'm not chasing. There are also some very different approaches to servicing...
For sure. Of course that super helpful and responsive rep who holds your hand through the whole purchase process might vanish the moment their commission comes in, so hard to gauge that up front sometimes....
Thanks all!
SparkyStan
Re: Setting up a 'real' SMT line - advice?
«
Reply #12 on: June 28, , 08:56:17 pm »
Any thoughts on backup pins, height mapping, placement force monitoring features? I don't think we need to worry about it, but those are features that show up in some brochures but not others, so I'm curious how widely applicable they are. We're not doing max width boards/panels, but some of our current tab routed panels have a fair bit of flex. Of course those could be redesigned into smaller panels and/or wider frames.
One thing you can consider for board flex is to use carriers, basically a rigid fixture for thin and flex PCBs usually. I haven't used one yet as we alleviated our flex by decreasing panel size but I have spoken to a company that manufacturers small IoT devices and they use carriers for their panels to support the, through pick and place and reflow with good success.
Reflow ovens: Sales reps are telling me 7-8 zones minimum for lead free, but I've also heard from a couple of people who are getting good results with 3-zone systems. Given that throughput isn't a big concern, I'm interested in what anyone else has to say on that.
In my experience, our 3 zone oven has worked well for panels less than 12x12" in size (1-6 layers so far, 2oz outer layers, 0.5oz inner layers). One thing to consider is if you plan on larger boards with heavy copper and large planes, the 3 zone might show it's limitations and require more tuning of your profile to ensure good soldering (i.e slower/longer profile and/or higher temperatures).
Re: Cut tape, Panasonic also have one of those cut-tape-on-tray thing, which seems like any machine that supports trays ought to be able to offer. A fixed piece of cut tape is just a 1xN matrix tray after all. I guess the difference would be in how easy they make it to program that sort of thing. Those Feeder Fingers look interesting, will add those to the list. I would imagine the fancy autoloading feeders from any brand would be able to handle cut tape, but I'm sure a standard electronic feeder plus a Feeder Finger is a cheaper way to go.
I'm only familiar with the ASM E/X feeders but I can just load cut tapes into feeders as normal, they just need a cover tape extension to go back to the peeling mechanism. This is nice as I just treat cut tapes as you would a reel. If you're not maxing out your feeders slots, consider seeing what the feeders of your options are capable of cut tape wise. I've definitely used some 10 part cut tapes from a sample kit in a pinch a few times but we have set minimums for parts depending on pitch as they are cumbersome to handle below a certain length. IIRC the Essmetec sales guy mentioned they could feed cut tapes down to 1 inch as well.
The following users thanked this post: ajb
jayx
Re: Setting up a 'real' SMT line - advice?
«
Reply #13 on: June 28, , 10:43:51 pm »
Would be usefull if you let us know what you exactly mean by "high mix/low volume". You need to take into consideration how many feeders you need in the machine to assembly your most complex PCB, how often you're going to swap the components in the feeders (sales pitch/brochures rarely mention how long it takes to swap a tape in the feeder and set it up for use in the machine), is setup changeover time priority over placement speed or the other way round.
It will also have some impact on the other machines selection. For paste printing, it may be better to look at jet printing. For reflow oven, perhaps vapour phase.
Other equipment to consider is dry storage for moisture sensitive components, wahing machine for the PCBs, x-ray inspection, rework station for bottom terminated components (BGAs, QFNs, etc.).
The following users thanked this post: ajb
SMTech
Re: Setting up a 'real' SMT line - advice?
«
Reply #14 on: June 29, , 12:13:04 pm »
A larger longer oven will increase throughput. A short oven with few zones tries to match the profile your paste would like to function against what is possible with the size of your board and the mass of components placed upon it. If you read a solder paste datasheet you will note there are 3-4 key periods. Fewer zones gives you less opportunity to bring every area on the board up to the windows requested by your paste profile and leaves you more reliant on conveyor speed, which in turn makes it more likely you will break the time rules instead. 6-8 vertical zones is a decent compromise, fits in an oven that is not too ridiculous (4m or less) and this is also where many major brands start with their ranges. My experience when we purchased ours was that of the major brands Heller was easily the value choice, it was cheaper and had more features than its rivals, BTU wasn't far behind. Similar spec'd well known Chinese brands were cheaper (about 25%). Its possible once its 20 years old you'd choose a more expensive brand 2nd user, but Hellers are everywhere. That said I visited a firm that primarily made boards for the European Space Agency and they ran a 3 zone oven, but it was a good one from ERSA.
Pipped and routed panels with a lot of flex will need support, you can do that with pins or systems like grid-lok. I wouldn't underestimate the importance of board support, on our Essemtec, boards are supported by manually adjusted pins, these are hard to place accurately under a double side board, hard to pack densely for a complex flexible panel and hard to set at just the right height. There was grid-lok style option, I wish I had chosen it but at the time it wasn't needed. This can have a radical affect on placement accuracy as the board surface isn't exactly where the machine expects it to be. Some but not all placement machines measure the placement force which would help in these cases or with warpage.
Other toys like X-ray, washing and even AOI are for later when volume, customer or complexity demands them
The following users thanked this post: ajb
loki42
Re: Setting up a 'real' SMT line - advice?
«
Reply #15 on: June 29, , 12:29:38 pm »
I totally agree again with board support. I'm running 0.6 - 1.2mm boards and a lot of the problems have been board support related with printing and placement.
The following users thanked this post: ajb
SMTech
Re: Setting up a 'real' SMT line - advice?
«
Reply #16 on: June 29, , 02:53:37 pm »
Its been mentioned in here, but I would suggest that while Jet printing sounds cool, no stencils to store, no waiting for stencils etc etc. You cannot buy those platforms without spending well into 6 figures, so you need to be not buying a lot of stencils before they pay for themselves, older Mycronics are out of support, they work best with special pastes and I bet servicing/care is a delight. There are no good manual printers, Reprint and PBT make decent(ish) Semi-auto's - you can buy old EKRA, MPM or DEKs everywhere and you should be able to find spares & support too.
The following users thanked this post: ajb
mikeselectricstuff
Re: Setting up a 'real' SMT line - advice?
«
Reply #17 on: June 29, , 03:40:46 pm »
Its been mentioned in here, but I would suggest that while Jet printing sounds cool, no stencils to store, no waiting for stencils etc etc. You cannot buy those platforms without spending well into 6 figures, so you need to be not buying a lot of stencils before they pay for themselves, older Mycronics are out of support, they work best with special pastes and I bet servicing/care is a delight. There are no good manual printers, Reprint and PBT make decent(ish) Semi-auto's - you can buy old EKRA, MPM or DEKs everywhere and you should be able to find spares & support too.
Yeah, stencil jetting sounds good at first sight, but seems an expensive option and only really useful for niche cases. Leadtime isn't a good argument as you always have the PCB leadtime, and stencil leadtime will rarely be longer.
If storage is an issue, use unframed stenciles - there is at least one system, don't recall the name, that uses holes around the edge of an unframed stencil to mount it into a special frame that tensions it to make it like a framed one.
Maybe for places running very high mixes perhaps ?
Youtube channel:Taking wierd stuff apart. Very apart.
Mike's Electric Stuff: High voltage, vintage electronics etc.
Day Job: Mostly LEDs
The following users thanked this post: ajb
ajb
Re: Setting up a 'real' SMT line - advice?
«
Reply #18 on: June 29, , 08:27:43 pm »
Would be usefull if you let us know what you exactly mean by "high mix/low volume". You need to take into consideration how many feeders you need in the machine to assembly your most complex PCB, how often you're going to swap the components in the feeders (sales pitch/brochures rarely mention how long it takes to swap a tape in the feeder and set it up for use in the machine), is setup changeover time priority over placement speed or the other way round.
Sure. For reference, our most complex board currently in production is ~90 BOM lines and ~300 total placements between top and bottom, roughly 130 8mm equivalent feeder lanes. BUT, we do NOT need to be doing that particular one in-house right off the bat or possibly ever, and there's room to knock a few BOM lines off the build anyway. Our initial board builds will be less than half that (both BOM lines and total placements), so I'm thinking we might start with about 50 feeders with room to grow from there. No BGAs, a couple QFNs, everything else is (imperial) up to ~25mm QFP. A day to set up and a day to run 50-100 boards would be great to start, with the idea that we can add feeders/carts later on to speed up changeover.
It will also have some impact on the other machines selection. For paste printing, it may be better to look at jet printing. For reflow oven, perhaps vapour phase.
From conversations I've had so far, it seems like a vapor phase batch oven wouldn't save much in terms of purchase cost, especially compared to used conventional inline ovens, plus we'd eventually want to replace it with an inline setup anyway. So I don't think that makes a lot of sense for us. On printing, coincidentally both jet printing and used MPM stencil printers came up in a conversation I had earlier today, and the latter seems a much better way to go as SMTech notes.
jayx
Re: Setting up a 'real' SMT line - advice?
«
Reply #19 on: June 29, , 09:46:33 pm »
I've seen a demo of Essemtec (Fox if I remember correctly) with paste jetting option which cost (again if remember correctly) ~£20k, and was quite impressed with the speed and accuracy. So may be worth to consider if you go for this machine, as a stand alone probably not worth the price. If you decide to go for stencil printer, you should consider DEK VectorGuard system.
As for vapour phase, there is also potential running cost saving (less energy usage, no need for nitrogen), and I think much better solder joints quality, possibly less solder bridges and tombstones, less problems if you have large and heavy components, finally potential to significantly reduce voiding (if you need low voiding) if you go for vacuum option. Obviously there is the liquid (Galden) cost to consider, and not many second hand machines available.
For changeover consider how many tapes you'll need to swap if you change between a different PCBs. If you have 50 BOM lines on one board and 50 on the other, but only 10 are common, there is a lot to swap if you only have 50 feeders.
The following users thanked this post: ajb
Mangozac
Re: Setting up a 'real' SMT line - advice?
«
Reply #20 on: June 29, , 11:01:09 pm »
On printing, coincidentally both jet printing and used MPM stencil printers came up in a conversation I had earlier today, and the latter seems a much better way to go as SMTech notes.
We run a 25 year old MPM UP printer and it still runs like a champion (albeit with the odd repair but nothing that has been catastrophic). Not that I'm suggesting you should go for something that old but based on our experience I would certainly consider a newer MPM printer when it comes time to upgrade (which I will likely do in another year or two before this one gets that bit too old).
I was told that jet printers require special paste and there is a lot of waste/hassle of cleaning the jetting head but that may have just been marketing talk.
The following users thanked this post: ajb
Reckless
Re: Setting up a 'real' SMT line - advice?
«
Reply #21 on: June 30, , 05:39:53 am »
From conversations I've had so far, it seems like a vapor phase batch oven wouldn't save much in terms of purchase cost, especially compared to used conventional inline ovens, plus we'd eventually want to replace it with an inline setup anyway. So I don't think that makes a lot of sense for us. On printing, coincidentally both jet printing and used MPM stencil printers came up in a conversation I had earlier today, and the latter seems a much better way to go as SMTech notes.
I have 3 used MPM UP's for sale if interested in Chicago. Highly recommended, last forever and if ever a part is needed generally very cheap to find as it was the most popular machine sold. MPM still supports them in terms of technical support for FREE. The only SMT manufacturer that has friendly support.
The following users thanked this post: ajb
Reckless
Re: Setting up a 'real' SMT line - advice?
«
Reply #22 on: June 30, , 06:06:02 am »
I've seen a demo of Essemtec (Fox if I remember correctly) with paste jetting option which cost (again if remember correctly) ~£20k, and was quite impressed with the speed and accuracy. So may be worth to consider if you go for this machine, as a stand alone probably not worth the price. If you decide to go for stencil printer, you should consider DEK VectorGuard system.
I have seen the Essemtec Fox live in person, it's definitely my favorite PNP in terms of form factor but I have reservations on it for actual production use (40+ hours/week). I saw it spill parts at 100% speed back in . My feeling is that it is the most expensive prototyping machine and is slow when compared to other brands. The jetting feature only started working recently in last year or so. They have sold ~250 units in 8 years since launch in . I think its because they are pricing it too high. It does have nifty features like doing parts (Smaller than , , ) using 4mm feeders:
The following users thanked this post: ajb
Reckless
Re: Setting up a 'real' SMT line - advice?
«
Reply #23 on: June 30, , 06:36:48 am »
A larger longer oven will increase throughput. A short oven with few zones tries to match the profile your paste would like to function against what is possible with the size of your board and the mass of components placed upon it. If you read a solder paste datasheet you will note there are 3-4 key periods. Fewer zones gives you less opportunity to bring every area on the board up to the windows requested by your paste profile and leaves you more reliant on conveyor speed, which in turn makes it more likely you will break the time rules instead. 6-8 vertical zones is a decent compromise, fits in an oven that is not too ridiculous (4m or less) and this is also where many major brands start with their ranges. My experience when we purchased ours was that of the major brands Heller was easily the value choice, it was cheaper and had more features than its rivals, BTU wasn't far behind. Similar spec'd well known Chinese brands were cheaper (about 25%). Its possible once its 20 years old you'd choose a more expensive brand 2nd user, but Hellers are everywhere. That said I visited a firm that primarily made boards for the European Space Agency and they ran a 3 zone oven, but it was a good one from ERSA.
Pipped and routed panels with a lot of flex will need support, you can do that with pins or systems like grid-lok. I wouldn't underestimate the importance of board support, on our Essemtec, boards are supported by manually adjusted pins, these are hard to place accurately under a double side board, hard to pack densely for a complex flexible panel and hard to set at just the right height. There was grid-lok style option, I wish I had chosen it but at the time it wasn't needed. This can have a radical affect on placement accuracy as the board surface isn't exactly where the machine expects it to be. Some but not all placement machines measure the placement force which would help in these cases or with warpage.
Other toys like X-ray, washing and even AOI are for later when volume, customer or complexity demands them
For ovens I strongly recommend the Essemtec ovens. They were designed to replace 7 zone ovens and have some crazy space age technology in order to flow air properly/evenly across the entire pcb. I have had a number of ovens in my life including a few Hellers (1 brand new), goldstar ovens, chinese ovens, madell, vitronics, manncorp. The only one that gave NO issues were the Essemtecs. Everyone who visits my facility is generally in disbelief that a small oven can do what the bigger Hellers can do and much simpler to use/manage. They are soo good they are used to make clear contact lenses due to the level of swiss precision. In my opinion this is Essemtecs best product. I've owned 8 of them when running small production lines. Heating element never breaks, nor the controller.
The following users thanked this post: ajb
ajb
Re: Setting up a 'real' SMT line - advice?
«
Reply #24 on: June 30, , 04:04:31 pm »
I've seen a demo of Essemtec (Fox if I remember correctly) with paste jetting option which cost (again if remember correctly) ~£20k
The very system we were discussing, in fact. As I understand it, you pay somewhat more than that for the factory-installed optional components to support jetting, and then you pay somewhat more than that again for the field-installable jetting equipment. Given that it adds up to about the cost of a standalone printer, I only see it making sense if you simply do not have the floor space for one. Or I guess if you absolutely need to do same-day prototypes and you're also getting a whole LPKF or whatever setup for in-house board fabrication.
-----
As far as placement machines, it's looking like we either go for cost effectiveness at meeting our near term needs or we go for flexibility for the future. As it stands, and with the caveat that I need to consolidate my notes and review the numbers again, the former looks like Hanwha or Yamaha and the latter looks like Europlacer. The other options generally cost more for nebulous benefit to our current needs, or have some sort of tradeoff in feeder/component capacity.
Europlacer definitely has a higher up front price tag, but that comes with some pretty attractive capabilities. In its favor, it does look like the ii-Feed tape feeders are much less expensive than conventional feeders, even accounting for the more expensive carts and magazines needed to support them. Having more internal tray capacity so we could use cut tapes to stretch beyond feeder capacity is also attractive, and we have in-house CNC machining capability to make trays and fixtures as needed. I'm trying not to get TOO smitten with Europlacer.
Again, that's all based on preliminary notes I need to review, and I have one or two more conversations to follow up on yet. The plan is to get it narrowed down to a couple of options and then try to schedule some hands-on time with the machines before we make a final decision. I appreciate all the input!
Comments
0