Homepage » Allgemein »

Automatically better test results

New from F&K Delvotec Austria:
Automatically better test results

F&K Delvotec extended its Series 56xx bonder and tester by an automatic test head with built-in targeting camera. This makes it possible to perform pull and shear tests not only under manual control but also fully automatically.The benefits to the user are twofold – the tests become both simpler and better. Simpler – because there is very little operator interaction required once the test is set up, and better – because there is no more operator influence on the test results.

Siegfried Seidl and Dr. Josef Sedlmair, F&K Delvotec Austria, Braunau

Quality Assurance is important but cumbersome, and continuously monitoring a wirebond production by making repetitive pull- or shear-tests is as time-consuming to the operator as it is monotonous. This was the reason for F&K Delvotec from Braunau, Austria, to extend its Series 56xx bonder and tester by an automatic test head with built-in targeting camera, the 5600C.
Unchanged Hardware Basis
One body, many heads – this Hydra principle has made the 56xx Desktop Micro Factory a real star in the marketplace with more than 200 machines installed within four years in Europe alone. Most customers opt for at least two different bond technologies (such as thin-wire wedge bond and heavy wire), resulting in far more than 500 bond heads delivered. This same base was now extended on the software side to allow fully automatic bond testing. On the hardware side, a new test head 5600C is available which contains a motorized P axis for 360° rotation and a miniaturized color camera with a range of optics depending on the desired resolution.
It can be fitted with a number of different measurement cartridges for the various test ranges in pull-testing (100 cN to 3000 cN) and shear-testing (500 and 5000 cN). The advantage to the user is that the test head is purchased only once, and the test cartridges are easily exchanged within seconds. For those customers who already have a conventional pull- or shear-test head without a camera: not to worry, it will still run on the same machine without any changes.
Software: Simply switch over
The tester software is based on the standard testing software and has the same operator interface as the bonder software. Bond programs made on a 56xx can even be read by the tester, so all the wire positions are already known for products made on the same machine type.
The camera is very helpful even in normal manual operating mode. Using the joystick, the user positions the test head over the desired wirebond at the crosshairs, and rotates the hook symbol visible at the crosshairs at a right angle to the wire direction. When starting the test, the test head then rotates the hook 90° to avoid the wire, moves the head slightly sideways away from the loop, lowers the hook under the loop, rotates it back and pulls upward to perform the test. All the parameters used for this procedure, such as heights, distances from the loop etc. can be pre-set very simply. As usual, the tester records not only the final pull test value but the entire force-distance curve during testing so that anomalies caused by wire kinks, stretching or peeling are visible.
A big advantage of this operating mode is that there is no uncertainty about the hook position due to parallax errors from the standard microscope which always looks at the measuring spot at an angle. This is especially important when wires are close together, and for untrained operators.
Powerful automatic operation
The automatic mode is, of course, what the new test heads are all about. First, all test points are programmed by identifying the source and destination bonds for every wire to be tested. These bond positions are normally situated on two different chips (or one chip and one substrate position) which is important for the position correction later on. The assignment of bonding positions to programmed chips can be checked and changed at any time by referring to the coordinate window.
As per default, the center of a wire bridge is the testing position. This can be modified either globally (in the default settings) or individually per wire. Programmed wires are shown in green with the testing position marked separately. To make sure, the testing process can be simulated without damaging the wire.
Finally, reference positions have to be trained so that the wires are located correctly even when the positions of the chips shift from part to part. Such changes in the relative position of chip and substrate alter the positions, lengths and angles of the wires connecting them, and therefore must be compensated by the tester software. Usually, two reference positions are taught in for the substrate and another two per chip.
During manual testing, getting the hook positioned in the correct rotation is fairly cumbersome, especially when there are few parallel or near-parallel wires. After all, the hook has to be inserted at right angles under the loop or else the results will suffer in reproducibility. Compared to that, the automatic mode is extremely helpful: after finding the source and destination positions, the software calculates the wire direction and the correct hook rotation.
To perform the actual test, the operator only has to mount the part to be tested, adjust the exact positions of the reference points and repeat this for each programmed chip. This adjustment can even be done automatically by a pattern recognition unit (PRU) if the base is equipped with that, just as for the bonder version of the 56xx.
When running the test, the camera image shows the currently tested wire in blue and all tested and destroyed wires in red. For non-destructive tests which are frequently used for heavy wire applications, the wires already tested are shown in green unless they failed the test which will make them show up in red.
Not subject to operator variations
It is well known in practice that it makes quite a difference to the test result where the pull hook is positioned along the wire. For wires connecting bond pads on the same height, the correct position is obviously the center of the loop, while for wires bridging a step the hook has to be closer to the higher pad. This will make sure that the angles at both bond pads are identical and hence the same force is felt by the source and destination bond. Skilled operators know that by putting the hook much closer to one bond, they can shift the pull force almost entirely to that bond only, allowing a weak bond site at the other end to escape detection.
For gold-ball bonds on thick-film hybrids, this is a classical point of concern for the quality manager, because the stitch (or second) bond is always on the thick film pad and is inherently much weaker than the first bond on the chip. To deal with this problem, security balls on top of the second bond, or a base bump under it, or even a combination of both are frequently used. Even that may not be sufficient, and a seasoned bond operator who does not like to have his or her production stopped can fool the system easily. All that is needed is to pull the wire very close to the first bond (where the wire sticks up almost vertically), delivering almost no pull force to the weak second bond. This can have terrible long-term results because sub-standard bonds can escape detection and lead to device failure much later in the field. At that point, quality management is usually called in to investigate, but has no chance to find the root cause, because the original quality data of the parts will show perfectly good pull tests.
The automatic operation puts an end to this manipulation, whether intended or accidental, because the position along the wire will be maintained very precisely.
Shear tests just as simple
Especially for heavy wires, shear tests are more revealing than pull tests because they give an indication of the actual bonding interface. In contrast to pull-tests, however, shear tests do give meaningful results also for non-destructive tests, because the test loads can be much closer to the destructive load without risking damage to the bond. Non-destructive tests are, of course, ideally done on 100% of production which is be completely out of the question for manual shear-testing but is quite feasible for the automatic tester.
Programming the tester is analogous to the pull-test with the added complexity that there may be stitch-bonds with far more than two bonds per wire. Setting these tests up is very simple with the rotation angle of the measuring chisel and the motion direction automatically calculated from the wire direction, even allowing for S-shape bonds.
In customary manual operation, the operator enters a failure code after each tested bond so that lift-offs, heel breaks and wire breaks can be separated out and analyzed. In automatic mode this can be done, if desired, after the test run is over, by automatically positioning the camera over all test positions and allowing the user to enter the relevant codes, or by going only to those wires which failed the test. The same can be done for shear-tests.
The measurement results are stored and evaluated on the tester, but not only there. All data (as well as the test programs) are managed in a standard SQL database which can reside on the tester or on a host server system, or on both. The bond tester provides a standardized evaluation package for quick reports, and more elaborate analysis can be done on any PC connected to the SQL database by using an evaluation package provided with the tester. Depending on the level purchased, there is a variety of test reports which can be re-designed, expanded or altered by the user, and there is a very simple but powerful feature to export data into all kinds of further analysis or presentation software. Finally, given that the data reside in a well-defined SQL database, they can seamlessly flow into proprietary SPC systems developed by many large electronics manufacturers for their QA systems.
The camera can be exploited to record images of the test part before or after testing, independently from testing itself. This can be useful for later quality analysis, for documentation purposes such as creating testing manuals, especially if typical error pictures occur repeatedly and are to be documented. A case in point are shear tests of heavy wires or of dies where it is highly significant what the sheared residue looks like as regards symmetry, bond area coverage, cratering etc.
Coming features
Further useful and comfort-enhancing features are under development, such as monitoring and measuring the loop height for non-destructive bond tests. All of them will help making quality assurance and statistical process control for die- and wirebonding simpler to use and even more powerful: another crucial milestone on the long road to zero-defect production.
EPPE 420

zusammenfassung
F&K Delvotec Austria hat die erfolgreiche Bonder- und Tester-Serie 56xx um einen automatischen Testkopf erweitert. Das vereinfacht prozessintegrierte Testabläufe und ermöglicht gleichzeitig Qualitätsverbesserungen.
F&K Delvotec Austria a ajouté une tête d’essai automatique aux microsoudeuses-testeuses de sa série à succès 56xx. Celle-ci simplifie les opérations de vérification intégrées au processus tout en améliorant la qualité des produits.

F&K Delvotec Austria
F&K Delvotec Austria, located in Braunau in Upper Austria, is an independent company within the F&K Delvotec group and serves the worldwide market of benchtop and laboratory equipment for wirebonding and testing. Thanks to the extremely successful Desktop Micro Factory 56xx series the company has expanded to almost 30 employees within the last few years since taking up business in 1994 under the management of Siggy Seidl and Dr. Farhad Farassat.
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