A vendor places 100,000 CPH on a requirements sheet, one more cases 115,000 CPH, and unexpectedly the buying discussion ends up being a drag race: which select and place maker has the bigger number?
I think that is the wrong question.
The number you need to appreciate is not the number of elements the maker can theoretically place in an hour. It is the amount of parts– and eventually the number of excellent boards— your certain product mix can relocate through the line during a normal manufacturing change.
Those are very different numbers.
Yamaha, for example, specifies its YRM20 at up to 115,000 CPH with its high-speed RM head, but clearly defines that number as being achieved under maximum problems. Its general-purpose HM setup is specified at 98,000 CPH, while an odd-shaped-component FM configuration drops to 35,000 CPH. Very same platform. Substantially different throughput.
IPC-9850A is a lot more revealing. The industry standard was developed partly because devices providers traditionally used various methods to explain placement performance, making machine-to-machine contrasts challenging. IPC also cautions that its standardized examination boards are less complex than actual production boards which users must anticipate different derating on their own products.
That is the concern.
So prior to I would certainly accept an SMT choice and place maker acquisition, I would determine throughput from the PCB backwards as opposed to from the sales brochure ahead.
Brochure CPH Is Not Pick and Area Maker Throughput
CPH means parts per hour.
Easy sufficient.
However there are a number of variations of that number concealing behind the exact same acronym:
- theoretical positioning CPH,
- standardized standard CPH,
- optimized-program CPH,
- immediate manufacturing CPH,
- sustained production CPH,
- and effective good-output CPH.
Just the last 2 pay the bills.
An equipment rated at 100,000 CPH mathematically implies:
100,000 ÷ 3,600 = 27.78 positionings per second
That appears impressive. Yet a genuine PCB does not present 27.78 the same resistors straight underneath the positioning head every secondly.
The head relocations.
The board indexes.
Fiducials are checked.
Elements are selected from various feeder positions.
Components may need bottom vision.
Big ICs need various velocity accounts.
Nozzles can transform.
A tray element may be a number of hundred millimeters far from the optimum pick-up path.
And manufacturing quits when reels, feeders, nozzles, boards, drivers, software, examination stations, or upstream devices quit complying.
This is why I deal with brochure CPH as a maker ability indicator, not a manufacturing projection.
The difference matters a lot more in high-mix manufacturing. A 2024 Siemens situation including Indonesian EMS producer Sat Nusapersada reported that process-preparation changes boosted general line performance by 33%, cut line-configuration work by 31%, and decreased incoming-project-data prep work time by 92%. That is an explanatory outcome because the gain came from removing process friction, not merely raising mechanical positioning rate.
Speed is only one variable.
Start With the Only Formula That Issues: Boards per Hour
If your consumer purchases assembled boards as opposed to raw positionings, start with boards per hour.
The standard estimation is:
Boards per hour = 3,600 ÷ real board cycle time in secs
After that:
Real positioning CPH = boards per hour × positionings per board
Suppose one PCB contains 800 placements.
During a production-representative test, the device finishes one board every 42 secs.
Your numbers ended up being:
3,600 ÷ 42 = 85.7 boards/hour
After that:
85.7 × 800 = 68,560 CPH
If the device was advertised at 100,000 CPH, your product-specific utilization of the headline positioning price is:
68,560 ÷ 100,000 = 68.6%
That 68.6% number is even more important to me than the glossy 100,000 CPH case.
Yet we are not completed.
Presume the machine loses one more 8% of scheduled production time to feeder replenishment, microstops, nozzle cleansing, driver response, small alarm systems, and various other interruptions.
Then sustained output ends up being roughly:
68,560 × 0.92 = 63,075 CPH
Your expected 100,000-CPH maker is now producing approximately 63,000 continual placements per hour on this job.
That does not mean the equipment is bad.
It implies the original contrast misbehaved.
Construct the Throughput Version From Your Real PCB
This is where a major equipment assessment begins.
Do not send a supplier an element matter and ask, “What CPH will we get?”
Send out the BOM, centroid data, board dimensions, panelization, part packaging, feeder needs, nozzle needs, and target lot dimension.
Porquê?
Porque 1,000 placements per PCB informs me virtually absolutely nothing by itself.
Think about 2 boards.
Board A has 1,000 placements controlled by 0402 resistors and capacitors fed from nearby 8 mm tape feeders.
Board B also has 1,000 placements, yet includes QFNs, BGAs, ports, shield containers, a number of tray-fed ICs, high electrolytics, and parts calling for extra vision processing.
The placement matter equals.
The mechanical issue is not.
For your pick and location device CPH estimation, break the BOM right into practical positioning courses:
| Element course | Typical examples | Throughput influence |
|---|---|---|
| High-speed chips | 0201, 0402, 0603 resistors/capacitors | Highest positioning price |
| Typical ICs | SOIC, SOT, QFN | Modest vision and traveling penalty |
| Precision ICs | BGA, fine-pitch QFP | Greater inspection/accuracy penalty |
| Big elements | Connectors, transformers, electrolytics | Slower acceleration and positioning |
| Tray-fed devices | CPUs, BGAs, specialty ICs | Longer pick-up traveling |
| Odd-form parts | Shields, sockets, unusual connectors | Specialized nozzles/heads might dominate cycle time |
Currently the discussion changes.
You are no more asking, “How quick is the machine?”
You are asking, “How quick is this equipment on my board?”
That is the buying question.
Determine Actual vs Theoretical Positioning Speed With a Derating Version
For early-stage buying, prior to you have a supplier simulation, I advise developing three throughput situations as opposed to pretending you know one exact number.
Call them:
Optimistic. Expected. Tension situation.
For illustration, think a 100,000-CPH ranked SMT choice and place maker.
| Throughput element | Optimistic | Anticipated | Tension case |
|---|---|---|---|
| Sales brochure CPH | 100,000 | 100,000 | 100,000 |
| Product/vision derating | 90% | 78% | 65% |
| Board dealing with performance | 98% | 94% | 90% |
| Feeder/nozzle efficiency | 98% | 95% | 90% |
| Operating availability | 95% | 90% | 82% |
| Estimated continual CPH | 82,100 | 62,700 | 43,200 |
This is not a global derating table. Do not replicate those percentages into a capital request and call the job completed.
The factor is the approach.
The version forces the supplier and customer to identify where the missing CPH goes.
Which conversation is efficient.
IPC-9850A makes basically the exact same theoretical warning from a requirements point of view: standardized placement testing offers usual comparison problems, yet real items generate their very own derating due to the fact that actual boards are extra challenging than the metrology examination automobile.
There is no truthful universal conversion such as:
Ranked CPH × 70% = actual CPH.
Sometimes 70% is traditional.
Sometimes it is fantasy.
Add Every Second the Positioning Head Is Not Placing
This is where lots of ROI spreadsheets quietly break down.
Take a look at the full board cycle.
PCB transfer and clamping
Suppose real placement needs 28 seconds, but discharging the finished PCB, loading the following PCB, placing it, clamping it, and finishing board recognition take in an additional five seconds.
Your actual cycle is not 28 secs.
It is 33 seconds.
That evidently small five-second charge enhances cycle time by almost 18%.
Reconhecimento fiducial
Machines may evaluate international and regional fiducials. The moment can rely on board design, panelization, cam approach, and procedure requirements.
One inspection occasion barely matters.
Repeat it throughout every panel for 2 changes and it does.
Vision handling
A resistor and a large BGA are not comparable positioning events.
Element acknowledgment, orientation adjustment, coplanarity checks, laser measurement, and other examination procedures can modify SMT placement rate significantly.
ASMPT’s 2024 SIPLACE CA2 spec demonstrates this connection plainly: the hybrid system was called capable of as much as 50,000 dies/hour ou 76,000 SMT components/hour, with precision down to 10 µm at 3σ. Different component-handling needs generate different heading rates even on the exact same devices idea.
Feeder travel
Feeder layout matters greater than several purchasers expect.
If high-volume components sit in inadequate feeder locations, placement heads travel further than essential during thousands of repeated cycles.
A creative optimization program can decrease that fine.
A bad configuration can produce squandered motion all the time.
Nozzle modifications
Automatic nozzle change appears irrelevant because each event is brief.
Increase it by a severely optimized job.
Currently it is manufacturing time.
Tray dealing with
Tray-fed ICs can develop long traveling paths, and matrix-tray replenishment may interrupt production differently than smart tape feeders.
Once more: count the seconds.
Do Not Determine CPH Without Computing Product Mix
This is just one of my tougher rules.
If the manufacturing facility runs 30 items, I do not wish to see ROI based upon the fastest one.
That is sales mathematics.
Rather, develop a product-weighted model.
Picture these three regular monthly products:
| Produto | Placements/board | Real cycle | Boards/month | Positioning volume |
|---|---|---|---|---|
| Controller A | 850 | 39 sec | 20,000 | 17,000,000 |
| Power Board B | 430 | 31 sec | 12,000 | 5,160,000 |
| Communications C | 1,250 | 62 sec | 8,000 | 10,000,000 |
The device should be substitute versus all 3.
Then add transition time.
Suppose you can position very quickly once the line runs, however the procedure transforms products 6 times per shift and sheds 14 mins each time.
That is:
6 × 14 = 84 minutes shed per shift
On an eight-hour shift, 84 mins stands for 17.5% of scheduled time.
Currently ask on your own the uncomfortable concern.
Would certainly you instead buy a 120,000-CPH machine with a 14-minute transition or a 95,000-CPH system that regularly transforms over in six mins?
There is no universal solution.
However in high-mix production, the nominally slower maker can easily deliver more boards.
A December 2024 SMT case study reported around 130 mins monthly of labor decrease just by automating line-width changeover throughout three production lines. That is an additional reminder that schedule and changeover losses are worthy of a place in the throughput equation.
Action Continual Throughput, Not a 60-Second Presentation
A machine can look stunning for 5 mins.
I desire the hour.
Better yet, I desire the shift.
Prior to acquiring, define a production-representative approval examination making use of one or more of your actual products.
I would certainly request at the very least:
- your actual PCB or an agreed equal,
- your actual part mix,
- production feeder setup,
- regular vision setups,
- typical placement-accuracy demands,
- reasonable board input and output,
- at least one feeder replenishment event,
- measured board cycle time,
- pickup-error and retry matters,
- placement-error matters,
- machine stops and alarm system duration,
- good boards produced throughout the concurred examination home window.
The outcome needs to include both immediate CPH e continual CPH.
Do not allow those terms come to be compatible.
If the provider tells you the machine attained 96,000 CPH for a brief optimized series yet the total manufacturing run averages 71,000 CPH, your acquiring version should utilize something closer to 71,000.
Not 96,000.
The Most Effective Choose and Area Equipment for Production May Not Have the Greatest CPH
This is where purchase obtains fascinating.
Fuji describes the NXT III in regards to 67,200 CPH per square meter, clearly drawing attention to performance per flooring location rather than merely pricing estimate one outright speed figure.
That is a valuable means to think.
Manufacturing facility business economics involve more than peak placement rate:
- floor-space productivity,
- feeder capacity,
- transition behavior,
- operator requirements,
- programming time,
- component range,
- placement precision,
- fixing response,
- spare-part accessibility,
- energy and compressed-air needs,
- maintenance periods,
- and line balancing.
An SMT choice and location maker going for 85% usage is generally more fascinating to me than a quicker equipment running at 55%.
And do not think the placement device is immediately the line traffic jam.
If your printer, SPI, AOI, reflow process, conveyor reasoning, board loader, or hand-operated assessment procedure restricts the line to 70 boards per hour, purchasing a placement maker capable of 110 boards per hour will certainly not magically create 110 ended up boards.
You acquired idle capacity.
Expensive still capacity.
Maintenance Belongs in the Throughput Estimation
This part obtains neglected because it does not show up in the attractive maker demo.
Mechanical problem impacts repeatability, travel behavior, alarms, maintenance downtime, and inevitably accessibility.
Direct guides, bearings, sphere screws, gliding systems, feeders, and drive systems have upkeep demands. Those requirements belong in the operating-cost model before purchase.
For Panasonic-oriented SMT upkeep, as an example, the site’s Panasonic MP grease for SMT machinery is explained for applications consisting of ball screws, bearings, straight overviews, and sliding components. The noted item information gives a temperature series of − 54 ° C to +177 ° C. Various other maintenance programs might include items such as OKS 470 oil for bearings, round screws, and guides, which is listed for SMT and industrial devices applications, or NSK NF2 industrial grease for bearing and machinery applications.
For heavier-duty equipment elsewhere in an electronic devices facility, OKS 402 commercial oil is another maintenance referral worth separating from precision-SMT-specific lubricant needs.
But do not improvise lubricant selection.
The equipment maker’s upkeep guidebook and accepted lubricating substance specification need to win every disagreement. Compatibility, base oil, thickener chemistry, thickness, NLGI grade, contamination danger, and solution interval issue.
A less costly lubricant that creates an unexpected four-hour standstill is not inexpensive.
The Procurement Examination I Would Place In Composing
Below is the commercial part suppliers often tend to do not like.
Do not buy solely versus rated CPH.
Put the expected manufacturing result into the acceptance requirements.
A severe ask for quotation must include something like:
Target item: PCB-147 Placements: 1,180 per panel Panel measurements: 330 × 250 mm Smallest component: 0201 statistics or defined actual bundle Largest part: defined port measurements Accuracy tools: defined BGA/QFP/QFN plans Target maintained throughput: 65,000 CPH minimum Target board cycle: ≤ 65.4 secs Test duration: agreed production-representative run Maximum pickup-error price: agreed threshold Optimum unintended device downtime: concurred threshold Called for placement top quality: agreed Cpk/accuracy criterion Feeder arrangement: production arrangement, not a demonstration-only layout
Currently both sides know what “quickly” indicates.
And compute the target cycle independently:
If the PCB has 1,180 placements and you call for 65,000 sustained CPH:
Required boards/hour = 65,000 ÷ 1,180 = 55.08
Therefore:
Maximum board cycle = 3,600 ÷ 55.08 = 65.36 seconds
That is a testable purchase need.
“Broadband” is not.
An Easy Pick and Area Device CPH Computation Worksheet
Prior to requesting quotes, load this out for your leading 3 to ten items.
| Input | Instance |
|---|---|
| Placements per PCB/panel | 1,180 |
| Rated machine CPH | 100,000 |
| Simulated positioning time | 50 sec |
| PCB handling + fiducials | 5 segundos |
| Vision/nozzle overhead | 4 sec |
| Complete cycle time | 59 sec |
| Boards/hour | 61.02 |
| Product-specific CPH | 72,000 |
| Running schedule | 90% |
| Sustained efficient CPH | 64,800 |
| Brochure-to-real ratio | 64.8% |
The computation is:
Complete board cycle = positioning + transfer + acknowledgment + tooling/vision expenses
After that:
Boards/hour = 3,600 ÷ complete board cycle
After that:
Product CPH = boards/hour × positionings per board
After that:
Continual CPH = item CPH × operating availability
For the instance:
3,600 ÷ 59 = 61.02 boards/hour
61.02 × 1,180 ≈ 72,000 CPH
72,000 × 0.90 = 64,800 sustained CPH
Currently you have a number that can get in the labor model, ability plan, and ROI estimation.
That is far more valuable than 100,000 printed in 36-point type on a sales brochure.
What Recent Manufacturing Data Claims Concerning Purchasing Ability Very Carefully
There is lots of funding moving into electronics producing.
United state Demographics Bureau information for NAICS 334418– published circuit assembly production– records roughly $27.2 billion in united state sales, deliveries, or revenue for 2023.
At the more comprehensive semiconductor-equipment level, SEMI later on reported that globally manufacturing-equipment billings reached $117.1 billion in 2024, up 10% from $106.3 billion in 2023. Assembly and product packaging tools sales rose 25% year over year.
More financial investment does not make careless capability mathematics more secure.
It makes errors much more pricey.
The purchase team that comprehends actual throughput has an advantage due to the fact that it can distinguish between a really effective machine and a device that occurs to win a criteria.
Regularly Asked Inquiries
What is genuine pick and place maker throughput?
Real choice and location equipment throughput is the sustained variety of elements or finished PCB assemblies an SMT positioning device generates throughout typical manufacturing after making up board transfer, fiducial acknowledgment, vision handling, feeder traveling, nozzle changes, part mix, small quits, replenishment, transitions, operator treatment, and various other manufacturing losses.
It must therefore be gauged over a depictive run rather than presumed straight from the manufacturer’s optimum CPH requirements.
How do you compute pick and location throughput?
Select and place throughput is calculated by splitting 3,600 seconds by the gauged complete board cycle time to get boards per hour, multiplying that figure by the number of put parts per board, and after that using a schedule variable for downtime, feeder replenishment, alarms, upkeep, and typical manufacturing interruptions.
For an 800-component PCB with a 42-second cycle, the preliminary outcome is approximately 68,560 CPH prior to added schedule losses are used.
What is the distinction between actual and academic positioning speed?
Theoretical positioning rate is an equipment’s maximum component-placement rate under defined standard or optimum conditions, while actual positioning rate is the outcome accomplished on a specific manufacturing board after element recognition, mechanical travel, feeder places, board handling, precision demands, nozzle operations, item intricacy, and factory disturbances influence the cycle.
That is why two 100,000-CPH makers can generate significantly different outcome on the exact same PCB.
What portion of ranked CPH should I anticipate from an SMT choice and place machine?
There is no global percent of ranked CPH that properly anticipates genuine SMT result since the derating relies on PCB dimensions, part mix, feeder layout, tray usage, vision needs, positioning precision, nozzle adjustments, software optimization, board handling, changeover frequency, operator methods, and the equipment architecture itself.
Use a supplier simulation followed by a production-representative approval examination as opposed to relying upon an approximate 60%, 70%, or 80% presumption.
Is the fastest choice and place equipment always the most effective equipment for manufacturing?
The fastest pick and place device is not immediately the very best manufacturing device because sustained manufacturing facility result depends on utilization, transition time, component versatility, feeder capability, programming effectiveness, positioning top quality, upkeep demands, line harmonizing, service support, and whether an additional procedure such as printing, inspection, reflow, or hand-operated handling is the actual bottleneck.
A lower-rated platform can therefore create more salable boards per shift in a high-mix setting.
What info should I give a supplier prior to requesting a throughput estimate?
A trusted throughput estimate needs the supplier to obtain production-representative PCB dimensions, panelization, BOM, centroid data, positionings per board, component product packaging, feeder types, tray requirements, smallest and biggest components, precision packages, nozzle needs, target whole lot dimensions, changeover regularity, needed accuracy, expected change pattern, and target boards or elements per hour.
The more completely the vendor designs your real item, the less significant the generic brochure CPH becomes.
Make the Vendor Prove the Number
Prior to acquiring a pick and area device, quit asking which model has the highest possible CPH.
Ask for your CPH.
Give the provider your actual board data. Get a simulation. Recognize positioning time, board-handling time, vision charges, feeder travel, nozzle adjustments, transition losses, and anticipated running availability individually.
After that run the example board.
Measure it.
And put the continual throughput need into the acquisition acceptance terms.
Because the device that markets 115,000 CPH is not always the device that will certainly make you the most cash.
The equipment that accurately creates the required variety of great boards per shift is.
