Back to DC/DC Homepage

<

Info Source	QML Manufacturer?	Type	Problem Category	Link to more info	Comments
Non-NASA			Misapplication? Quality?		Date: 1994 Converter failed during thermal vacuum testing of the spacecraft due to either: 1) Overheating of the converter lead to thermal instability of the rectifier Schottky diode on the 5 V output side of the transformer. OR 2) an open circuit due to a solder joint failure in the diode. The open circuit would lead to an arc forming between the wire and the edge of the diode. Material melted by the arc could form a short that would remain even after the diode cooled to room temperature. Both scenarios lead to thermal overstress of the converter. (af)
Non-NASA	Yes	28 Vin, Triple, 5V, ±15V, 30W	Element Failure – cracked capacitors and solder joints		Date: 10/95 One sample converter from two vendors was submitted to a preliminary assessment test sequence (thermal shock, vibration, mechanical shock). Both devices showed signs of damage after the test, with the most severe damage shown in the form of detached and cracked capacitor stacks, and cracked solder joints to pin. The project recommendation was to proceed with vendor 2.
Non-NASA			Element Failure – package bottom surface not flat		Date: 1995 Package bottom was not flat causing ineffective installation. (af)
Non-NASA			Insufficient Characterization		Date: 1995 Converter overly noisy for use with A/D converter application. (af)
JPL	No	28Vin, 5Vout	Unexpected Electrical Behavior – By Design		Date: 7/11/96, Seawinds Converter would not start if the input was not first returned to zero. Analysis showed insufficient internal design margin. Report Z23378. Contact: Gary Bivins
Non-NASA	Yes	28 Vin, Triple, 5V, ±15V, 30W	Quality – low bond pull, Performance – TCycle test failures, incoming		Date: May 1997 During environmental testing (endurance, temperature cycling, vibration) and DPA of commercial grade converters. Numerous anomalies reported; · Electrical test failure (device non-functional) and PIND failure during incoming screening · Catastrophic failure during endurance test · Thermal cycled device failed leak test after 200 cycles and failed electrical test after 500 cycles · DPA showed various wires below pull test limits and capacitors below die shear limit
Non-NASA	Yes	Triple	Quality – PIND failures, low bond pull, Performance – Life test failures		Date: May 1997 During environmental testing (endurance, temperature cycling, vibration) and DPA of commercial grade converters. Numerous anomalies reported; · PIND failure during incoming screening · Electrical test failure during endurance test device out of spec at 500hrs and 1000hrs test points, and total failure at 2000hrs) · Electrical test failure (isolation) after vibration · DPA showed various wires below pull test limits and capacitors below die shear limit
Non-NASA	Yes	28 Vin, 5V Dual, 15W, 28 Vin, 15V Dual, 15W	Quality – leak test and bond pull discrepancy		Date: Sept. 19, 1997 leak test failures associated with cracked seals (possible handling problem) and one low pull strength wire bond.
Non-NASA	Yes	28 Vin, 5V Single, 15W, 28 Vin, 5V Dual, 15W, 28 Vin, 15V Dual, 15W	Quality – leak test and die shear discrepancy		Date: Sept 15, 1997 leak test failures associated with cracked seals (possible handling problem) and some low values at die shear test
Non-NASA	Yes	28 Vin, 5V Single, 15W, 28 Vin, 15V Dual, 15W	Radiation – performance reduction		Date: Sept 1, 1997 Test showed devices remain functional upto 50krads but with small increase in output voltage and approx. 20% reduction in efficiency at 16 V input voltage.
Non-NASA	Yes	28 Vin, 5V Single, 15W, 28 Vin, 5V Dual, 15W, 28 Vin, 15V Dual, 15W	Radiation – device failure		Date: Sept 1, 1997 Test showed that the 15V Dual and the 5V Single devices remain functional up to 49krads but with reduction in output voltage and efficiency, and increase in input current. However the 5V Dual device lost functionality at 5.5krad. After 7.5hrs annealing at room temperature device had shown some recovery (device functional at 28 V and 40V but not at 16 V). This anomaly was investigated in analysis RA0102, where device was subjected to radiation at lower dose rate. Device again lost functionality at 3-4 krads and did not recover after 53-hrs room temperature anneal (biased). However after high temperature bake (27hrs @ 100degC) device returned to full functionality.
JPL	No	28Vin, 5Vout	Unexpected Electrical Behavior		Date: 12/12/97, MPL Due to the onset of oscillations, a fuse was blown during box-level burn-in. The ambient temperature was 80C and the loading was “light”. Report P1004. Contact: Gary Bivins
Non-NASA	Yes	120Vin, 50W	Performance – design margin, Quality – lack of Process Control, Quality Control		Date: 1996-1997 encountered many problems particularly for 120 V types: · Major technical problems – the design of the new hybrids had not been sufficiently investigated and verified. Insufficient margin had been included in the design to cope with variation of the characteristics of add-on elements. · Serious manufacturing difficulties – lack of process control, with processes often introduced prior to verification and qualification. Processes not very repeatable with the quality of the final product being very operator dependant · Deficiencies in quality assurance – problems are solved with a view to maintaining production without ensuring problem does not re-occur. NCR process is ineffective. Implementation of new quality procedures was poor. · Bad management;- procurement of add-on parts and materials poorly controlled; program planning ineffective; poor reporting resulted in lack of visibility for customer. All these factors contributed to innumerable deficiencies being encountered with the parts (up to July 98 at least 43 NCR’s had been issued for this procurement), leading to major delays (up to one year) with the delivery of parts
JPL	Yes	28Vin, 3.3Vout	Incorrect Mounting onto chassis		Date: 4/6/98, Seawinds During EMC testing, the device failed the CE and isolation tests. The hybrid case was not isolated from the chassis. Report: Z46807. Contact: Gary Bivins
JPL			Packaging – Workmanship		Date?: Mars98 Workmanship failures. Parts eventually assembled at JPL Contact: Gary Bivins
Non-NASA			Quality		Date: 1998 Failure during thermal vacuum testing caused a system intermittent failure. System-level failure effects analysis pointed to the DC/DC converter as the problem. Damage and workmanship discrepancies, such as component interference, presence of metallic debris and lack of conformal coating, found inside the converter support a conclusion that this failure was created by a short inside the DC/DC Converter.
JPL	Yes	28Vout, +/-15Vin, 5Vin, Triple, 30W	Element Evaluation, Element Performance		Date: 8/28/98, Jason Failure of multiple output capacitors during Hybrid-level life test. Catastrophic failure resulted. Contact: Gary Bivins
Non-NASA			In-Flight failure		Date: 1999 DC/DC Converter on A side failed catastrophically; no output. Lack of blown fuse on the input side indicates that the front end did not have a short. (af)
Non-NASA	Yes	Various (28 V and 120V inputs)	Quality – unresponsive to NCR’s		Date: May 1999 Many problem encountered in getting the vendor to comply with requested changes in manufacturing and quality procedures that had been identified in order to achieve high rel. product.
JPL	Yes	28Vin, +/-15Vout, dual, 30W	Element Evaluation, Element Performance		Date: 6/17/99, SIRTF Catestrophic failure due to failed stacked capacitor on the input during hybrid-level life test. Contact: Gary Bivins
JPL	Yes		Element Evaluation, Element Performance		Date: 1/10/00. GRACE, GALEX PWM controller chip Unitrode 1824 was found to be unstable in the circuit when the operating temperature was above 85ºC causing a reduction in switching frequency. Converter Vout found to droop 200mV to 400mV for low input voltage conditions with full loads. When these converters were power derated (to 75%) for Toperation > 85ºC the performance returned to specified values. Internal design changes were made to “re-center” the switching frequency. Element evaluation is a standard process that can be used to discover internal part limitations or inadequacies prior to their installation in the hybrid. Contact: David Gerke
Non-NASA	Yes		Other – failure at burn-in		Date: March 2000 The DC/DC Converter exhibited catastrophic failures during / after Burn-in. Problem was attributed to wrong loading of the outputs of this converter (zero load inhibits voltage regulation which causes the output voltages to rise beyond limits, thus destroying the capacitors across the output lines). The conclusion at that time was that nothing else than the output capacitors had seen the overvoltage and a repair by the vendor replacing the affected capacitors and re-testing the DC-DC converters was performed. However after rework, parts exhibited some electrical isolation failures, after burn-in, between input and output (which should normally be isolated via the internal transformer). Failure analysis showed totally burnt bobbins. This indicated that the conclusion after the first burn-in failure (electrical over-voltage at output would only affect the components at the output side) had not been backed up by a good failure analysis, and was incorrect. It was evident that during the rework to replace the capacitors, the transformers had not been checked. In fact, it is now also LAA's understanding that when the capacitors broke by the over-voltage, this would produce a low resistive path (short circuit) which then overstressed the DC/DC converter (transformer, possibly other components as well). After these failures the lot was returned to the vendor to be replaced by a new lot.
JPL and GSFC	Yes	28Vin, 30W	Packaging – Design, Process, Radiation – Device Upset	Element Stuck to Lid	Date: 3/16/00 SIRTF at JPL, Image and Gravity Probe B at GSFC Lid stuck to internal devices causing solder joint failure when the lid deflected in a vacuum, stiffener added by users. Pulse width modulator upset caused failure of diode. PWM circuitry changed, diode changed.
JPL			Packaging - Design, Process, Performance – electrical w/temp		Date?, GALEX, Grace, MIRO Shorted capacitors due to excess solder trapped underneath them. Adhesion failure caused an electrical interconnect failure. Burn-in results showed electrical efficiency not to spec over all temp ranges. Conductive epoxy used on substrate caused isolation failures. Contact: Gary Bivins
JPL			Element Failure - Capacitors		Date? GALEX, TES Capacitor failures linked to life test failures. Contact: Gary Bivins
LaRC	Yes		Element Selection – Tin plating		Date? Concerns about optocouplers, tin whiskers. Pre-cap found need for significant rework Contact: John Pandolf
LaRC	Yes	28Vin to 15Vout (SCD spec)	In Flight Failure - Unknown		Date: 2000 CERES (EOS) Loss of regulation after 7 months in service. Loss of CTR in optocoupler providing feedback. Determined to NOT be radiation related. Contact: John Pandolf
JPL	Yes	28Vin, 15Vout, Single, 65W			Date: 4/29/00, SIRTF Catestrophic failure during thermal vacuum testing. Report Z72922. Contact: Gary Bivins
JSC	No	120Vin, 12Vout, Single	Element Failure – MOSFET		Date? SVS LDC 9747, 28Vin/12Vout Shorted MOSFET Contact: Stephen Trifilo
Non-NASA			Quality		Date: 2001 Root cause traced to vendor workmanship issues. Vendor implemented process changes to improve manufacture of device and independent review team approved process changes. (af)
JSC	No	120Vin, 12Vout, Single	Element Failure – Thermal Switch		Date 2001 SVS LDC 9744, 120Vin/12Vout, Failed Thermal Switch Contact: Stephen Trifilo
JSC	No	120Vin, 5Vout, Single	Element Failure – Shorted Diode		Date 2001 SVS LDC 9747, 120Vin/5Vout, Shorted Surface Mount Diode Contact: Stephen Trifilo
JSC	No	28Vin, 12Vout, Single	Element Failure – Shorted Capacitor		Date 2001 SVS LDC9803, 28Vin/12Vout, Shorted Multilayer Ceramic Capacitor, Surface Mount Contact: Stephen Trifilo
JSC	No	120Vin, 5Vout, Single	Element Failure – Shorted Capacitor		Date 2001 SVS LDC9751, 120Vin/5Vout, Shorted Multilayer Ceramic Capacitor, Surface Mount Contact: Stephen Trifilo
JSC	No	120Vin, 12Vout, Single	Element Failure – Shorted Capacitor		Date 2001 SVS 120Vin/12Vout, Shorted Multilayer Ceramic Capacitor, Surface Mount Contact: Stephen Trifilo
GSFC	Yes		Element Selection – non-high rel, Screening – not correct for part type or application conditions, Element Failure – film capacitor		Date: Jan 3, 2001-GLAS Non-established reliability stacked film capacitors used in the converter experienced significant physical changes due to the method in which they were screened, installed and applied. The method used to screen the parts was suspected of aging the parts. The methods used during hybrid build up after the capacitors are installed inadvertently removed material from the parts as well as added foreign material to them. The electrical circuit used applied a signal to the capacitors with a pulse component (100’s of kHz) which was not a screening condition. These factors lead to a very high failure rate for these capacitors after they were installed. Test data to resolve new screening methods and the resulting part reliability was not generated. Contact Mike Sampson, Henning Leidecker
JPL	Yes	28Vin, +/-12Vout, 80W	Unexpected Electrical Behavior		Date: 2/5/01, EMLS No turn-on with a 2.5S load was encountered. Needed to use an external turn-on delay circuit. See JPL report Z70457. Contact: Gary Bivins