Lorentz Forces inducing Stress on wirebonds that are ortogonal to the magnetic field
For suggestions and questions send e-mail to:
Gino Bolla
This theory came up here at FCC. The idea is that the Jumper Wirebonds are orthogonal to the magnetic Field and so whenever there is a current flowing they feel the associated Lorentz Force. While the AVDD and AVDD2 bonds have currents that are constant with time the DVDD connection and bonds carry a current that swing at every L1A. In this pdf document the current swing is described and measured. Furthermore this theory can potentially explain even the DOIM failures because even there some of the power bonds are orthogonal to the B field.
The Ipothesys behind this theory is that the continuos movements of the wirebonds can stress them up (fadigue) to a point where they break.
LINK to the pages mantained by Reid Mumford.
Here are the profiles of the Jumper wirebonds as measured by Joel Goldstein at SiDet
Here is the transient behavior during a L1A-DIGITIZE-READOUT cycle
GRAY trace is the current vs time and each mV correspond to 1 mA
YELLOW trace is the DVDD voltage on the PHI side of the ladder
RED trace is the DVDD voltage on the Z side of the ladder
We are pursuing this line of investigation with 2 parallel efforts:
One in CDF with the TOF magnet (up to 2 T) where we blindly put in wirebonds
and we pulse them with a 100 mA swing current at various frequencies.
Joel Goldstein is doing most of the work and he already broke two bonds at frequencies in the 5-20 KHz range.
The other is in the IB1 building at the Technical Division. We are using their Calibration Magnet. The DUT (Device Under Test) is a single Wire-Bond that can be looked at from the top or from the side by using a 45 deg. mirror in connection with a high magnification optic.
| Instrumentation-Movie.mpg 18.5 MB - Here you can see the setup at the technical Division with the Magnet, the optic and the electronic necerssary to pulse the wirebonds |
| Optic-Movie.mpg 48 MB - Here you can see the capabilities of the optic being used |
| Resonance-Movie.mpg
4.7MB - Here you can see a 2 mm bond moving when a resonant frequency is excited |
Here are some of the carachteristics of the magnet we are going to use at the Technical division
From left to right
- Drawings of the magnet
- Magnetic Field with 150 mm POLE CAPS versus COIL current for various GAPs between the POLE CAPs
- Magnetic Field versus position inside the POLE CAPs for the Proposed Magnet Settings for the experiment.
Here are some SEM pictures taken at the Technical Division (once again thanks to Bob Kephart) of details of good and broken bonds. We are now going to prepare on better samples so that this friday we can take pictures with a better angle.
From left to right
- A good bond foot.
- A bond that has been stressed with the Lorentz forces that is deeply cracked but not yet broken
- A picture of a broken bond (the wire is hanging in the air). This failure was induced in the CDF TOF magnet by Joel.
- A broken foot.
- A broken wire.
Here is the documentation collected so far: