4647	2010-05-06 07:48:36	popmedkoh	When is a junkbox semiconductor too old to use?
Message	Date	From	Subject
I've been in this wonderful hobby of ours for many years. Just how many came to sharp focus when a fellow ham asked me to take a look at his Astron RS-35M 12 Volt 35 Amp linear PS which he had acquired because it had zero output. Fortunately the beefy bits (xfmr, rectifiers, caps, pass transistors) were fine. The 723 (plastic part with 1995 date code) seemed to have failed. Since it was socketed, I re-seated it (having fixed more than one piece of equipment this way) to no avail. Unfortunately, the friendly local Radio Shacks no longer stock this chip. So my dilemma: I can order one for 64 cents plus shipping and handling--and wait OR I can reach into my ever-expanding and ever-aging junk box and pull out one of the following: A Fairchild 723 ceramic dip date code 7137 An ITT 723 ceramic dip date code 7250 An SG (presumably Silicon General) ceramic dip date code 7503 All are unused and assuming they all work initially, which is a better choice: an almost 40 year old silicon IC with better thermal characteristics or a newer plastic part? I don't want to jeopardize a friendship because a part fails in a week but there is the stubborn part of me that feels that the cost of a repair should not exceed the acquisition price of the equipmemt: $0 in this case. Any thoughts? David KE1LY
4648	2010-05-06 08:29:28	Corey Minyard	Re: When is a junkbox semiconductor too old to use?
My understanding is that most electronic parts do not "age" when unused, except for electrolytic capacitors that dry out. Perhaps others suffer similar effect, like paper insulated inductors, though they are pretty rare now days. Semiconductors will suffer from electromigration under use. That's the gradual exchange of ions across the junction region, and it is affected by the size of the junction region, heat, and probably a number of other factors. However, unused, my understanding is that they are unaffected. So you old unused silicon is probably good to go. -corey - AE5KM popmedkoh wrote: > I've been in this wonderful hobby of ours for many years. Just how many came to sharp focus when a fellow ham asked me to take a look at his Astron RS-35M 12 Volt 35 Amp linear PS which he had acquired because it had zero output. Fortunately the beefy bits (xfmr, rectifiers, caps, pass transistors) were fine. The 723 (plastic part with 1995 date code) seemed to have failed. Since it was socketed, I re-seated it (having fixed more than one piece of equipment this way) to no avail. Unfortunately, the friendly local Radio Shacks no longer stock this chip. > > So my dilemma: I can order one for 64 cents plus shipping and handling--and wait OR I can reach into my ever-expanding and ever-aging junk box and pull out one of the following: > > A Fairchild 723 ceramic dip date code 7137 > > An ITT 723 ceramic dip date code 7250 > > An SG (presumably Silicon General) ceramic dip date code 7503 > > All are unused and assuming they all work initially, which is a better choice: an almost 40 year old silicon IC with better thermal characteristics or a newer plastic part? I don't want to jeopardize a friendship because a part fails in a week but there is the stubborn part of me that feels that the cost of a repair should not exceed the acquisition price of the equipmemt: $0 in this case. > > Any thoughts? David KE1LY > >
4649	2010-05-06 08:34:28	k5nwa	Re: When is a junkbox semiconductor too old to use?
I have tunnel diodes made in 1960, 61 and they are still good, the thing is they have about 1000 times more impurities than a regular silicon devices so they would suffer from migration even worse yet they are fully functional. On Thu, 06 May 2010 10:28:56 -0500, Corey Minyard <minyard@acm.org> wrote: > > My understanding is that most electronic parts do not "age" when unused, > except for electrolytic capacitors that dry out. Perhaps others suffer > similar effect, like paper insulated inductors, though they are pretty > rare now days. Semiconductors will suffer from electromigration under > use. That's the gradual exchange of ions across the junction region,and > it is affected by the size of the junction region, heat, andprobably a > number of other factors. However, unused, my understandingis that they > are unaffected. So you old unused silicon is probably goodto go. > -corey - AE5KM > popmedkoh wrote: > > I've been in this wonderful hobby of ours for many years. Just how > many came to sharp focus when a fellow ham asked me to take a look at > his Astron RS-35M 12 Volt 35 Amp linear PS which he had acquired because > it had zero output. Fortunately the beefy bits (xfmr, rectifiers, caps, > pass transistors) were fine. The 723 (plastic part with 1995 date code) > seemed to have failed. Since it was socketed, I re-seated it (having > fixed more than one piece of equipment this way) to no avail. > Unfortunately, the friendly local Radio Shacks no longer stock this chip. > > > > So my dilemma: I can order one for 64 cents plus shipping and > handling--and wait OR I can reach into my ever-expanding and ever-aging > junk box and pull out one of the following: > > > > A Fairchild 723 ceramic dip date code 7137 > > > > An ITT 723 ceramic dip date code 7250 > > > > An SG (presumably Silicon General) ceramic dip date code 7503 > > > > All are unused and assuming they all work initially, which is a better > choice: an almost 40 year old silicon IC with better thermal > characteristics or a newer plastic part? I don't want to jeopardize a > friendship because a part fails in a week but there is the stubborn part > of me that feels that the cost of a repair should not exceed the > acquisition price of the equipmemt: $0 in this case. > -- Using Opera's revolutionary e-mail client: http://www.opera.com/mail/ [Non-text portions of this message have been removed]
4650	2010-05-06 09:00:48	Corey Minyard	Re: When is a junkbox semiconductor too old to use?
Well, the junctions on these devices are so large and the current so low that electromigration is generally not an issue at all. Generally it's thermal cycles that take the biggest toll on electronic systems like these. I work (or worked, I guess, since I'm a Linux hacker for a living now) in telecom, and since they never turn things off, (no power switches :-) and run fairly heavily all the time, electromigration is a concern on modern ICs with small geometries. And capacitors drying out, of course. -corey k5nwa wrote: > I have tunnel diodes made in 1960, 61 and they are still good, the thing > is they have about 1000 times more impurities than a regular silicon > devices so they would suffer from migration even worse yet they are fully > functional. > > On Thu, 06 May 2010 10:28:56 -0500, Corey Minyard <minyard@acm.org> wrote: > > >> My understanding is that most electronic parts do not "age" when unused, >> except for electrolytic capacitors that dry out. Perhaps others suffer >> similar effect, like paper insulated inductors, though they are pretty >> rare now days. Semiconductors will suffer from electromigration under >> use. That's the gradual exchange of ions across the junction region,and >> it is affected by the size of the junction region, heat, andprobably a >> number of other factors. However, unused, my understandingis that they >> are unaffected. So you old unused silicon is probably goodto go. >> > > >> -corey - AE5KM >> > > >> popmedkoh wrote: >> >>> I've been in this wonderful hobby of ours for many years. Just how >>> >> many came to sharp focus when a fellow ham asked me to take a look at >> his Astron RS-35M 12 Volt 35 Amp linear PS which he had acquired because >> it had zero output. Fortunately the beefy bits (xfmr, rectifiers, caps, >> pass transistors) were fine. The 723 (plastic part with 1995 date code) >> seemed to have failed. Since it was socketed, I re-seated it (having >> fixed more than one piece of equipment this way) to no avail. >> Unfortunately, the friendly local Radio Shacks no longer stock this chip. >> >>> So my dilemma: I can order one for 64 cents plus shipping and >>> >> handling--and wait OR I can reach into my ever-expanding and ever-aging >> junk box and pull out one of the following: >> >>> A Fairchild 723 ceramic dip date code 7137 >>> >>> An ITT 723 ceramic dip date code 7250 >>> >>> An SG (presumably Silicon General) ceramic dip date code 7503 >>> >>> All are unused and assuming they all work initially, which is a better >>> >> choice: an almost 40 year old silicon IC with better thermal >> characteristics or a newer plastic part? I don't want to jeopardize a >> friendship because a part fails in a week but there is the stubborn part >> of me that feels that the cost of a repair should not exceed the >> acquisition price of the equipmemt: $0 in this case. >> >> > > > >
4651	2010-05-06 09:07:14	popmedkoh	Re: When is a junkbox semiconductor too old to use?
Thanks for the encouraging words! Hope I won't have to duck if and when this monster of a power supply bounces back to me! I did run into a strange case where my wifes stereo developed noise in both channels at the same time. This was an all germanium transistor unit sold in the late 60's which failed after about 10 years. Turns out that the input stage transistors in both channels had become very noisy and the problem went away with replacement. Likely a bad batch of transistors with even more impurities than usual for germanium devices. I do have some 30+ yo eproms (silic
4652	2010-05-06 09:08:25	michael taylor	Re: When is a junkbox semiconductor too old to use?

4653	2010-05-06 09:11:59	michael taylor	Re: When is a junkbox semiconductor too old to use?

4654	2010-05-06 11:01:52	ajparent1	Re: When is a junkbox semiconductor too old to use?

4655	2010-05-06 11:22:27	Tim	Re: When is a junkbox semiconductor too old to use?
I have a number of power supplies from the 1970's, all with LM723's in them, all continuing to work fine. I see no problem putting in an old one, or new one, or whatever works. I have a big bag of late 60's/early 70's PNP medium-power transistors in TO-39 cans - I think they were originally specced in core memory systems. Never had a problem with any of them. My two cents: if the 723 is bad, it might not be the root cause. If you put a new one in, there's a pretty good chance that it'll get fried by the same root cause. Similar issues with any switching power supply: there's a big transistor and it failed (sometimes spectacularly). Put a new one in, and it'll fail right away too, because the cause is somewhere else in the circuit.
4659	2010-05-06 13:22:52	popmedkoh	Re: When is a junkbox semiconductor too old to use?
The power supply is going through a gentle burn-in period without problems as this is written. The load is only about 2.5 amps (two hi-intensity bulbs in parallel--remember the Tensor lamps?) I have nothing else readily at hand that won't pose a fire hazard or wake up the neighborhood or both. Also, anything close to the rated power of this supply means racking up the kilowatt hour costs! Remember, this is a freebie repair! I went with the Fairchild ceramic dip ic which also happens to be the oldest having been made in the 37th week of 1971. I chose it because it's the only company of the 3 that still makes chips. Well, that's not entirely true. According to the web Silicon General became Symmetricom which then merged with Microsemi which makes devices for ethernet applications among others. The plastic device that failed is a Motorola MC1723CP date code 8941 (not '95 as I originally reported). After 24 hours this Astr
4661	2010-05-07 05:28:02	Alan Melia	Re: When is a junkbox semiconductor too old to use?
Hi David I used to work in semicon reliability in that period. I would say any device with a date code after about 1970 should be reasonably good for many years yet if it has stored in normal "office" conditions. Most failure mechanisms require volts to activate that and even intermetallic formation only go slowly at room temperature. My motto is "stick it in.....if it works, screw the lid on.....if it still works, you are on a winner" :-)) Alan G3NYK
4662	2010-05-07 08:32:44	Gary	Re: When is a junkbox semiconductor too old to use?

4663	2010-05-07 12:16:59	ajparent1	Re: When is a junkbox semiconductor too old to use?
I've done same using Nichrome heating element wire I have a bundle of. IT can be had from appliance repair shops. Mine give about 1ohm per 2ft so if I need a high power load a few lengths looped with wire nuts to some heavy copper wires in a bucket of water works well. I used that to test a 28V 25A supply (700W) at full load. Cauti
4669	2010-05-09 08:39:26	Dave	Re: When is a junkbox semiconductor too old to use?

4670	2010-05-09 22:52:28	popmedkoh	Re: When is a junkbox semiconductor too old to use?

4671	2010-05-10 05:32:20	ha5rxz	Re: When is a junkbox semiconductor too old to use?
I experimented with EPROM erasure using sunlight last Summer after the bulb in my eraser died. After four weeks in the sunlight there was still data in the EPROM but it was badly corrupted. A lamp that provides UVC rather than UVA or UVB is the only way to go. HA5RXZ