Has anyone here put a product through HALT testing?
Any stories? How many iterations did you do? What broke? What did it cost?
Was it worth it?
--
John Larkin Highland Technology Inc
Precision electronic instrumentation
The "worth" problem is that the results are only meaningful if there's complete survival. If something breaks, have you proved anything? ...Jim Thompson
-- | James E.Thompson | mens | | Analog Innovations | et | | Analog/Mixed-Signal ASIC's and Discrete Systems | manus | | San Tan Valley, AZ 85142 Skype: Contacts Only | | | Voice:(480)460-2350 Fax: Available upon request | Brass Rat | | E-mail Icon at
As usual, you have it completely backwards. The point of HALT testing is to break it, to locate the weakest points.
-- John Larkin Highland Technology Inc www.highlandtechnology.com jlarkin at highlandtechnology dot com Precision electronic instrumentation
What have you proved by breaking something by excess environment... only that you can break it.
Your stuff... it'll probably all just go up in flames.
Though, I must admit, I did completely MELT a Silicon chip once ;-) ...Jim Thompson
--
| James E.Thompson | mens |
| Analog Innovations | et |
| Analog/Mixed-Signal ASIC's and Discrete Systems | manus |
| San Tan Valley, AZ 85142 Skype: Contacts Only | |
| Voice:(480)460-2350 Fax: Available upon request | Brass Rat |
| E-mail Icon at http://www.analog-innovations.com | 1962 |
I love to cook with wine. Sometimes I even put it in the food.
It's supposed to find the weakest point, and to improve that weakest point.
We have serious, complex products with proven field MTBFs in the millions of hours, sometimes tens of millions.
Only once?
I often test parts to destruction, to see what kind of margins they will have in operation. Some parts don't meet their own specs, some are 20x better, some don't even have specs for important things.
-- John Larkin Highland Technology, Inc jlarkin att highlandtechnology dott com http://www.highlandtechnology.com
Wouldn't HASS be more easily done yourself and appropriate for more products? For HALT, how exactly does lab accelerate lifetime translate to actual lifetime? Most items barely make it to warranty expiration, right? And most items are obsolete within two years, right? Military, avionics and automotive products are exceptions. If you have to ask if it pays, then it probably won't.
I blew up a BC108 once, just for the hell of it. Put a car battery across the EB junction. ***BANG***! Surprisingly violent reaction which actually deformed the metal can, putting dome in its little roof. Great fun!
Don't know. I have a customer who wants it done.
We have products that have been in production for 18 years, and still sell.
-- John Larkin Highland Technology, Inc jlarkin att highlandtechnology dott com http://www.highlandtechnology.com
Yep, it seems that some datasheets play fast and loose with the facts. I found it almost impossible to find optocouplers that came even close to their published specs. I've seen "Flame Proof" resistors go up in flames. I know we have embedded products that have been in continuous duty for
21 years. You gotta' break things sometimes to see what the weak points are.
We use it on Military stuff to find the weak point(s) if any...
No. I mean MELT... it turned into a little B-B sized ball of Silicon.
What happened is an alternator regulator failed, and the alternator produced an arc across the chip, melting it ;-) ...Jim Thompson
--
| James E.Thompson | mens |
| Analog Innovations | et |
| Analog/Mixed-Signal ASIC's and Discrete Systems | manus |
| San Tan Valley, AZ 85142 Skype: Contacts Only | |
| Voice:(480)460-2350 Fax: Available upon request | Brass Rat |
| E-mail Icon at http://www.analog-innovations.com | 1962 |
I love to cook with wine. Sometimes I even put it in the food.
Silicone gel filled power semiconductor modules explode leaving that disgusting snot all over the place. I've not had to clean the mess up, but it's supposed to be very un-fun.
I once blew a large die in a TO-3 package so violently it left a square shaped dent in the TO-3 lid ;-)
(BTW we, in the semiconductor industry call that "snot" just that, "greased pig snot". It's main purpose is to contain wirebonds in their proper location during encapsulation; and also keep them from moving around afterwards.) ...Jim Thompson
--
| James E.Thompson | mens |
| Analog Innovations | et |
| Analog/Mixed-Signal ASIC's and Discrete Systems | manus |
| San Tan Valley, AZ 85142 Skype: Contacts Only | |
| Voice:(480)460-2350 Fax: Available upon request | Brass Rat |
| E-mail Icon at http://www.analog-innovations.com | 1962 |
I love to cook with wine. Sometimes I even put it in the food.
What little i saw, was this supposed HALT scheme is vibration oriented. What about electrical testing and monitoring at selected time points at elevated temperature, and looking for drift,etc? Wasn't there something like that done in the 70's or so for the military? Wasn't something called Weibull Curves involved? Or has this all been forgotten and acting stupid in this case a bit non-productive?
That is called MELT testing, not HALT testing... And to answer when the HALT cycle ends,it is defined as "HALT" (or STOP...
Can't give a value assessment, but yes, I've had preproduction and production level assemblies go through HALT testing.
The main issues were seldom electronic, or even vaguely electrical, as the basic designs take environmental ranges into account. It was actually reassuring to find out just what extremes were needed to provoke an alteration in the product's performance.
Real problems were often mechanical or assembly method-oriented. Screws loosening, adhesives or mechanical supports failing.
As a result, retesting regularly on production batches also proved informative in fab process control.
You can't have loose screws or or other damage, simply due to effects regularly occuring in shipment by air, if you are obliged to document ppm failure rates.
RL
Yep. My experience has been all mechanically oriented. Overstressing electrical components proves, at least to me, nothing except that you're good at torching stuff >:-}
...Jim Thompson
--
| James E.Thompson | mens |
| Analog Innovations | et |
| Analog/Mixed-Signal ASIC's and Discrete Systems | manus |
| San Tan Valley, AZ 85142 Skype: Contacts Only | |
| Voice:(480)460-2350 Fax: Available upon request | Brass Rat |
| E-mail Icon at http://www.analog-innovations.com | 1962 |
I love to cook with wine. Sometimes I even put it in the food.
You lucky guy!
BTDT.
If you try to wipe it away, it just rolls into "boogers".
Second most insidious stuff, after Dow Corning 340 thermal compound.
--
"Design is the reverse of analysis"
(R.D. Middlebrook)
Best thing for them...
;-)
--
"Design is the reverse of analysis"
(R.D. Middlebrook)
I have seen funny failures in HALT.
A resistor that only failed open just slightly at about 70 degrees. Temperature sweeps at temperature extremes of minus 40 and plus 125 degrees were ok. I learned from that, that HALT testing with data logging during sweeps can be informative
Other than that, some people think they can test worst case temperature of components, but miss that the HALT fan in blowing air in at high rates, so components only rise little above the HALT chamber temperature setting
We test to find first fault. One must be critical of the findings, since if the immunity level is high, then you don't really need to do anything with the first fault. But, it could be low, so you would need to "fix" subsequent faults also
People are now talking about HALT-EMC, but I think that's mixing just too much to be able to diagnose the fault
Cheers
Klaus
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.