Diode burn-in test

Thread Starter

JBizkit

Joined Jul 26, 2019
8
Hi All! Is anyone familiar with a reverse bias burn in test on a diode (HTRB)? I need to set up this test, but I can't seem to find a block diagram or a schematic. It sounds fairly straightforward: apply positive voltage to the cathode and anode to ground. I plan to have use a single voltage source for this and have the the diodes in parallel.

Should I have a resistor included after each DUT in case one devices goes short? That way in the event of one failure the remaining parts will still be stressed?
 

dl324

Joined Mar 30, 2015
10,539
Out of curiosity, why do you want to do burn-in testing? That's usually performed by manufacturers and it's done under multiple conditions.
 

Thread Starter

JBizkit

Joined Jul 26, 2019
8
I am up screening a commercial device to mil temp, performing pre and post burn-in electrical characterization. The burn-in requirement is reverse bias at +125C.
 

dl324

Joined Mar 30, 2015
10,539
I am up screening a commercial device to mil temp, performing pre and post burn-in electrical characterization. The burn-in requirement is reverse bias at +125C.
Are you buying milspec parts? Or are you trying to screen commercial parts? Are you going planning to do 100% testing? Manufacturers can get away with relatively small sample lots because a high burn-in failure percentage would be indicative of a process parameter that was out of tolerance. You can't do that because you don't know what lots your parts are from.
 

Thread Starter

JBizkit

Joined Jul 26, 2019
8
I am screening commercial parts and will be testing 100% to screen out failures.

"You can't do that because you don't know what lots your parts are from. " - Exact reason I am 100% testing.
 

Papabravo

Joined Feb 24, 2006
13,560
You don't understand. You cannot test in reliability. The results of your test will be meaningless because you cannot trace the components back to the manufacturing lots, as they came from the manufacturer. I'm curious about where you got the quaint notion that this would be a reasonable thing to do, Was it your boss? If so what are his credentials? Did he get his degree from Midnight Auto Supply and Cycle Parts?
 

Thread Starter

JBizkit

Joined Jul 26, 2019
8
Up-screening a COTS part is actually quite common, especially when dealing with a part going end of life and/or the original OEM is no longer in existence. Often in these cases manufacturing lot traceability not possible and the next best thing is to procure an equivalent part that is 100% screened for reliability through electrical and mechanical stress testing.

I got the "quaint notion" that this is acceptable from the end users source control document. I am not trying to rebrand commercial parts from Amazon and pass them off as something they're not.

Historically, the company I am with has outsourced this testing to a third party lab and my reason for the post was to learn how to do perform this screening internally.
 

Papabravo

Joined Feb 24, 2006
13,560
Has it occurred to you that such testing may actually subject the parts to life shortening stress? I think you are standing on very thin ice here and don't have a solid understanding of what you need to do and why you need to do it.
 

Reloadron

Joined Jan 15, 2015
5,370
While I never called it a burn in I have done and written procedures for reverse breakdown testing. Been a long, long time ago but testing was on stud mount diodes used in a F-15 Eagle generator. Small lots were pulled from incoming receiving inspection. We tested them on a Tektronix 575 curve tracer. You slowly increased the voltage (X Axis) while watching the current (Y Axis) and as long as the samples broke down below the data sheet reverse breakdown specifications they were considered a pass. However following testing the samples were destroyed since it was unknown if the reverse breakdown testing in any way damaged the parts.

You can apply a reverse voltage slowly increasing it while monitoring the current on each individual diode. You will see current start on some before others. If these are stud mount drill the plate and insert a temperature control sensor. We heated the plates with strip heaters. You can manually test or automate and actually manually or automatically collect the data and store it. Pretty much up to you how far this needs to be taken.

Ron
 
Top