Homepage » Allgemein »

Good results with all printers

EPP Benchmark-Arena: Comparison test of stencil printers at the SMT 2008
Good results with all printers

During this years SMT show in Nuremberg, the EPP and EPP Europe magazines (Konradin Mediagroup) with support of Mesago Messe Frankfurt (SMT-Organiser) organized a benchmark test of four paste printers. The five daily test runs were a real crowd puller. The visitors used the opportunity for a realistic and real-time comparison of systems from different suppliers. The benchmark focused on in-line-capable mid-range equipment. The official results of this benchmarks are published in this article.

Thorsten Niermeyer, Director of Sales, Koh Young Technology

On the printer side, the companies DEK (Horizon 01i), EKRA (X6 HSI), Essemtec (SP900 USC-V) and Speedprint (SP210 avi), represented in Germany by ANS electronik, entered the comparison, which was evaluated by the 3 D Paste-Inspection system (Aspire) from Koh Young, represented in Germany by PB Tec. All printers were equipped with stencils from Christian Koenen, solder paste F 640 from Heraeus and Sontara under stencil cleaning rolls from Vliesstoff Kasper. The PCBs were supplied by Heraeus as well and were cleaned between the test runs with cleaning agents from Zestron in the Systronic CL 500 cleaning system. All handling units came from Asys, which included loading- and unloading stations and a corner module.
What was being printed?
In each test run, each of the four stencil printer printed ten PCBs. Five test runs were performed each day and these were commented on in front of the interested audience. All 40 PCBs were measured at printing speeds by the 3 D paste inspection system, an extract of the results was presented with the Koh Young SPC software package live to audience. The stencil had a great range of apertures. The smallest depot had a nominal size of 175 µm diameter and the largest one was 2 x 1.5 mm. While the stencil had a thickness of 100 µm, it had a step-down to 70 µm for the smaller deposits. Each printer worked with the stencils in different frames or frameless stencil clamping systems.
What was measured?
The printer suppliers agreed before, that only deposits of real components should be evaluated. These amounted to a total of 2150 paste depots per PCB. Paste structures and test pattern on the PCB that are trying to show the limits of the equipment of the material were not evaluated during this benchmark. Figure 1 shows the measured pads in white colour, while the pads in grey were ignored. This included the bottom row of pads, close to the edge of the PCB, as the printers used different clamping methods and not all of them could reach to the edge of the PCB. DEK used a vacuum tool, EKRA edge snuggers and Essemtec and Speedprint a top-clamping. All ten boards of each test run were used for the measurement and evaluation, even though in a production environment the first three or four boards would not be used until the paste is kneaded a few times and has the right viscosity. Nominal value for the paste height was the stencil thickness, for the area of each depot the respective aperture opening was used and for the nominal paste volume the multiplication of area and height. Area and height are expressed in percent (%) of each respective deposit, while the height was measured in µm.
How was the measurement done?
The 3 D paste inspection system was integrated at the end of the line and measured in real time all printed PCBs. Koh Young’s Aspire system is a highly accurate measurement system with a height resolution of 0,37 µm. It measures the volume, height, area, x/y offset of each paste deposit, as well as performing a bridging check around the deposits. Koh Young claims market leadership in the area of 3 D paste inspection and uses a white light phase shift Moiré principle. All measurement results were stored and used for this evaluation. A visualization of the measurement can be visualized for each deposit, as the example in figure 2 shows.
How good was the printing?
The results of the benchmark can be judged in various categories. In general the average values (x) for paste height, area, volume and offset were considered. The variation of the results s or s (standard deviation) amongst the ten printed PCBs during one test run is an indication for the repeatability of a printer. But average value by itself isn’t sufficient to judge the result, more importantly the standard deviation in each of the above categories has to be considered. This should possibly be small and provides an insight how far the results of all pads are deviating from the nominal or average value. The deviation of height, volume and area is bigger than can be expected in a production environment, as all results were considered including the “Step-down” pads. The representative results for the different days can be viewed in the following images and tables. Figure 3 shows the average volume per PCB of the second test run of the first day in the upper graph, while the lower graph shows the standard deviation per PCB. The results show an oscillation for Speedprint between 90 % and 115 % of the nominal volume, depending on the squeegee direction. The squeegee pressure could not be adjusted independently for the 2 different directions (has been made available since). For each squeegee direction the values are stabile and do not exceed the values of the other printers. The results for Essemtec were varying very much. A defect in the camera was noticed and it was reconfigured in the evening. This can be noticed in the results for the next day, as can be seen comparing figure 3 to 4.
Figure 4 graphs the average volume and the average height for each PCB printed in the fourth test run on the second day. Table 1 shows the respective average values (x or Avg = Average) of the ten PCBs. A meaningful statement about the printer capability is the standard deviation s (sigma), which shows the deviation or process width. While all printers were equipped with stencils from Christian Koenen, it has to be remembered that these are not identical, as each printer had their “own“. The stencils have a manufacturing tolerance of +/- 3 µm. Hence some small deviation between them is possible.
Table 2 documents the second test run of the third day. It can be observed that the results from the printers are very stable and show little or no deviation compared to the day previous. The volume of the paste deposits from table 2 is visualized in figure 4 a. This shows for each printer manufacturer the distribution of all 21.500 paste deposits for this test run (ten PCB with each 2.150 deposits). This distribution is a graphical representation of the small differences between the printers and the respective sigma values.
Defects printed
Every day more than 430.000 solder paste deposits were measured and evaluated. Only very few defects (< 30 ppm) occurred. The defects that were found included blocked apertures that resulted in no paste on the PCB (figure 5), offset of the print (figure 6). Debris or dirt particles that were found on the PCBs (figure 7) were usually not from the cleaning materials, but rather from the environment of the show hall. Depending on the squeegee material and the pressure, the occurance of “tales” or “dog ears” was more pronounced. An example for too high squeegee pressure is shown in figure 8. As the tolerance for the different categories were set quite wide with +/- 50 %, only very few paste deposits were outside these thresholds. The printers with top clamping (Speedprint and Essemtec) in a few cases printed too much paste in the area close to the edge of the PCB. Figure 9 shows a deposit with 180 % of paste of the nominal volume and a paste height up to 196 µm. This could be corrected by choosing a different clamping method, such as a vacuum tool.
Process analysis and process improvement
For three days PCB were printed and paste deposits were measured; the results were not only presented to the public, but also used to optimise the printers. Figure 10 shows the example of the changing the under stencil cleaning cycle in the DEK printer. While on the first day, the stencil was cleaned after every fourth PCB (blue curve), the cleaning was switched off during the print of the ten PCBs on the third day (green curve). The printing results were more uniform during the ten prints, while the average height was still lower on the first PCB, after the under stencil cleaning. This measurement can help to determine the cleaning frequency, which in turn can help to optimize the consumption of cleaning material, as well as improve cycle times. The printer manufacturers used the x/y offset measurement provided by Koh Young continuously to change settings and improve the results. A further analysis shows a difference in height and volume between DEK and the other three printers (figure 4). To establish a possible cause, DEK and EKRA swapped the stencils on the last day for the last test run. Figure 11 has the graph for the average height per PCB with the different stencils in the two machines. It is apparent that stencil A has to be thicker, as the average height is higher on both DEK and EKRA when using this stencil.
Also the uniformity between the ten printed PCBs, it is obvious that stencil A shows less variation. This suggests that not only the printers contribute to the variation in the results, but also the materials such as stencils and their clamping systems.
Summary
Evaluating the results, it is difficult to classify the four printers into first and second best. The printers delivered very comparable results (see table 1 and 2). Nominating a winner or runner up would only be possible, if the categories and the weighting would have been fixed. This was not intended. For individual test runs it can be said, that EKRA showed the smallest variation for height and volume on the second day (table 1), while DEK inspite of a higher offset (x) showed the smallest deviation (s) in X-Y-Direction. During the last test run of the day it was Speedprint who had the smallest deviation on offset (see figure 12). It became transparent that the results of the printers are very close. On the third day (table 2) Essemtec showed a smaller standard deviation for height and volume than EKRA, but it was DEK again, who had the smallest X-Y offset. As this comparison did not include cycle times or cleaning cycles, no final concluding judgment is possible, which would declare a winner. But it became evident that besides the printers, the stencils and other materials such as the paste, squeegee and cleaning materials contribute to the defects during the paste printing process. It has proven further, that permanent process control through real 3 D measurement has to be conducted in order to detect process deviations and correct these.
EPP Europe 437

Table 2

Table 1
Current Issue
Titelbild EPP EUROPE Electronics Production and Test 11
Issue
11.2023
READ
Newsletter

Subscribe to our newsletter now

Webinars & Webcasts

First hand technical knowledge

Whitepapers

Find all current Whitepapers here

Videos

Find all current videos here


Industrie.de Infoservice
Vielen Dank für Ihre Bestellung!
Sie erhalten in Kürze eine Bestätigung per E-Mail.
Von Ihnen ausgesucht:
Weitere Informationen gewünscht?
Einfach neue Dokumente auswählen
und zuletzt Adresse eingeben.
Wie funktioniert der Industrie.de Infoservice?
Zur Hilfeseite »
Ihre Adresse:














Die Konradin Verlag Robert Kohlhammer GmbH erhebt, verarbeitet und nutzt die Daten, die der Nutzer bei der Registrierung zum Industrie.de Infoservice freiwillig zur Verfügung stellt, zum Zwecke der Erfüllung dieses Nutzungsverhältnisses. Der Nutzer erhält damit Zugang zu den Dokumenten des Industrie.de Infoservice.
AGB
datenschutz-online@konradin.de