TWiki> DetDev Web>TFHybShort (revision 1)EditAttach

Introduction

Four electrical grade wafers were delivered, plus one that was damaged during fabrication. Each wafer has three hybrids, hence the electrical grade wafers have 12 hybrids and two of the hybrids on the damaged wafer are not affected by the damage. Initial connectivity test with probe needles gave the following results

  • All 14 hybrids show a short between power and ground, with a resistance dominated by the contact resistance of the needles.
  • None of the signal lines seem to be shorted to ground (veryfied on one hybrid)
  • A selection of signal lines were probed on one hybrid, all where connected correctly.

Thermal imaging, hybrid 2 wafer 08

This is the middle hybrid of the damaged wafer. The GND and PWR pads on the power end of the hybrid were connected with 17 bond wires and the voltage difference between the GND and PWR plane was monitoried on the signal end of the hybrid. The V/I characteristics and corresponding resistance are shown below.The graph called Vin is the voltage at the input on the power end, and Vmon is the voltage measured at the signal end.

VI_Hyb2_w08.jpgRin_Hyb2_w08.jpg

The current was scanned many times up to increasing maximum values. Two hot-spots were became visible on the hybrid after a few power cycles.

4p5AfromCold_21to32deg_SE_cropped.jpg

After a few power cycles up to maximums between 5 and 10 A, the the lower hotspot disappeared, but the short remained.

10A2862mV_22to60deg_SE_cropped.jpg

These two spots don't correspond to any particular features in the layout.

LayoutScreenShot.jpg

After a further few cycles the short disappeared, and a small crater is visible on the location of the upper hot spot. No damage is visible on the location of the other hot spot.

Hyb2_BurntHotspot.jpg

Hyb2_BurntHotspot_close.jpg

Thermal imaging, hybrid 1 wafer 08

The same measurements were repeated with hybrid 1 on wafer 8, but this time with a solder connection and only on the power end of the hybrid. The V/I characteristics are shown below.

VI_Hyb1_w08.jpgRin_Hyb1_w08.jpg

The difference in measured resistance (77 vs. 238 mOhm) could be due to the contact resistance. The hybrid was powered repeatedly up to 10 A but no hot spots were visible. All heat dissipation seem to be at the power pads of the hybrids.

5A361mV_22to36deg_PE.jpgPowerPads_10A.jpg

Comparison w08, Hybrid 1 & 2

High resolution images of the same location of Hybrid 1 and 2, both with 5 A current.

Hotspot1_5A.jpgSamePosAsHotSpotOnHyb1_5A.jpg

-- LarsEklund - 2010-07-21

Topic attachments
I Attachment History Action Size Date Who Comment
JPEGjpg 10A2862mV_22to60deg_SE_cropped.jpg r1 manage 20.6 K 2010-07-21 - 17:06 LarsEklund  
JPEGjpg 4p5AfromCold_21to32deg_SE_cropped.jpg r1 manage 26.2 K 2010-07-21 - 16:45 LarsEklund  
JPEGjpg 5A361mV_22to36deg_PE.jpg r1 manage 12.4 K 2010-07-21 - 17:10 LarsEklund  
JPEGjpg Hotspot1_5A.jpg r1 manage 9.8 K 2010-07-21 - 17:15 LarsEklund  
JPEGjpg Hyb2_BurntHotspot.jpg r1 manage 84.1 K 2010-07-21 - 16:58 LarsEklund  
JPEGjpg Hyb2_BurntHotspot_close.jpg r1 manage 62.1 K 2010-07-21 - 16:58 LarsEklund  
JPEGjpg LayoutScreenShot.jpg r1 manage 111.0 K 2010-07-21 - 16:52 LarsEklund  
JPEGjpg PowerPads_10A.jpg r1 manage 11.9 K 2010-07-21 - 17:12 LarsEklund  
JPEGjpg Rin_Hyb1_w08.jpg r1 manage 24.3 K 2010-07-21 - 16:36 LarsEklund  
JPEGjpg Rin_Hyb2_w08.jpg r1 manage 21.7 K 2010-07-21 - 16:36 LarsEklund  
JPEGjpg SamePosAsHotSpotOnHyb1_5A.jpg r1 manage 23.2 K 2010-07-21 - 17:15 LarsEklund  
JPEGjpg VI_Hyb1_w08.jpg r1 manage 24.6 K 2010-07-21 - 16:36 LarsEklund  
JPEGjpg VI_Hyb2_w08.jpg r1 manage 23.9 K 2010-07-21 - 16:36 LarsEklund  
Edit | Attach | Watch | Print version | History: r3 < r2 < r1 | Backlinks | Raw View | Raw edit | More topic actions...
Topic revision: r1 - 2010-07-21 - LarsEklund
 
This site is powered by the TWiki collaboration platform Powered by PerlCopyright © 2008-2021 by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding TWiki? Send feedback