[SI-LIST] Re: Reset problem with power supply
- From: Jerry Martinson <jerry_martinson@xxxxxxxxx>
- To: steve weir <weirsi@xxxxxxxxxx>, olaney@xxxxxxxx
- Date: Sat, 6 Dec 2008 18:00:36 -0800 (PST)
There was a brief time several years ago where there was electroytic chemical
providers that had a mistake in their recipe that caused excessove hydrogen gas
production when they were used. Multiple cap component vendors used these
suppliers. This resulted in a lot of failures of these caps industry-wide
regardless of de-rating. I have seen this happen myself in circuits that were
very conservatively designed.
This "cap plague" of the early part of the decade was somewhat of a fluke but
gave electrolytic caps (and some of the prodcuts using them such as compact
flourescents) a worse reputation than they probably deserved.
----- Original Message ----
From: steve weir <weirsi@xxxxxxxxxx>
To: olaney@xxxxxxxx
Cc: hmurray@xxxxxxxxxxxxxxx; si-list@xxxxxxxxxxxxx
Sent: Saturday, December 6, 2008 2:39:39 PM
Subject: [SI-LIST] Re: Reset problem with power supply
Electrolytic caps come along with some engineering requirements. The
first is long term product life: electrolytics have a typical shelf
life of 5-10 years for cheap to very high quality. This is a function
primarily of the seals. However it is not that unusual to find a box
with large body computer grade electrolytics going for more than 20
years. The big issues: are initial conditioning, adequate temperature,
voltage derating, and absolutely : NO EXPOSURE TO HALOGENS. A 10 year
life is readily attainable, and with a lot of derating 20 years can be
had in large body parts. Miniatures are really constrained to about 10
years no matter what is done. Consumer products contain only miniatures
these days, and are so fiercely price sensitive that no one pays for the
kind of derating needed to see long service lives. As a result, the
electrolytics rank #1 to #2 for failure rates in consumer electronics
ahead or behind of the power semiconductors.
Steve
olaney@xxxxxxxx wrote:
> Greetings, Hal. As you have noted, the Schmitt trigger properly handles
> slow power risetimes. A proper design will include an RC time constant
> at the input so that it also handles fast risetimes by including a
> minimum delay before reset releases. As I mentioned, there are decent
> ICs to do this that are lots more convenient than rolling your own
> design. Reset circuits are a specialty in their own right. It's easy to
> slap together a lousy one, tougher to design a bullet-proof one from
> scratch.
>
> The issue with electroytic caps isn't whether they are "modern" but
> whether they are cheap. There are decent Al electrolytics to be had for
> a modest premium (OK, sometimes a big one), but the aluminum caps
> actually used in consumer gear seem as crummy as ever. I've had
> motherboards fail after several years because the caps in the switching
> supply weren't properly rated for the high frequency current and ran too
> hot. When it's really bad you get a brown tint to the circuit board in
> the area around where the caps are mounted because the internal losses
> increase as they age. At the end they run hot as a pistol.
>
> The last time this happened I metered the caps and found that 100uF caps
> had dropped to less than 2uF! Replacing them did not bring the board
> back from the dead because the excessive ripple voltage peaks from
> inadequate filtering had already done their dirty work. For machines you
> really care about, checking the power supply caps once in awhile might be
> a reasonable preventive maintenance procedure. A finger touch while
> under power for a quick temperature check is probably good enough (but
> I'd touch the plastic insulation rather than the bare metal). Tantalum
> caps generally do not have the aging issue, though they still need to be
> rated for the current demand.
>
> Orin
>
> On Sat, 06 Dec 2008 01:34:44 -0800 Hal Murray <hmurray@xxxxxxxxxxxxxxx>
> writes:
>
>> How does a Schmitt trigger solve the problem? I'd expect it to make
>>
> things cleaner if the input signal has a slow rise time, but that's not
> going to solve the problem of reset going away before the power is
> stable.
>
>> Are electrolytics getting old really a serious problem with modern
>>
> caps? I've seen lots of credible reports of very old gear being brought
> back to life by replacing all the old filter caps. I hadn't thought much
> about the implications on gear I was designing. How should I think
> about modern filter caps getting old?
>
>
>>> Cable inductance absolutely should not be an issue. If it
>>>
>> actually
>>
>>> is, be grateful that you found a design flaw before you reached
>>> production! This might be related to the power risetime. Your POR
>>> circuit needs to operate properly even if you slowly bring the
>>>
>> voltage
>>
>>> up manually. Some circuits depend on fast power risetime to
>>>
>> operate
>>
>>> properly, and work OK... most of the time. The worst ones
>>>
>> capacitor
>>
>>> couple the voltage bus into a transistor. Not only do these fail
>>>
>> to
>>
>>> reset on a slow rise, if an aluminum electrolytic cap is used, the
>>> value will drop with age as it dries out and eventually the
>>>
>> circuit
>>
>>> won't work even for normal supply risetime. A Schmitt trigger is
>>>
>> a
>>
>>> good basis for a bullet-proof approach, as are any number of
>>> supervisor ICs that also check for overvoltage, brown-out, and the
>>> like.
>>>
> ____________________________________________________________
> Take a break - you deserve it. Click here to find a great vacation.
> http://thirdpartyoffers.juno.com/TGL2141/fc/PnY6rw2gGUymLyXCUuqFPiRkgWMzRlFZD0pHTeyutmKSJLoJ92e4X/
> ------------------------------------------------------------------
> To unsubscribe from si-list:
> si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field
>
> or to administer your membership from a web page, go to:
> //www.freelists.org/webpage/si-list
>
> For help:
> si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field
>
>
> List technical documents are available at:
> http://www.si-list.net
>
> List archives are viewable at:
> //www.freelists.org/archives/si-list
> or at our remote archives:
> http://groups.yahoo.com/group/si-list/messages
> Old (prior to June 6, 2001) list archives are viewable at:
> http://www.qsl.net/wb6tpu
>
>
>
>
--
Steve Weir
Teraspeed Consulting Group LLC
121 North River Drive
Narragansett, RI 02882
California office
(866) 675-4630 Business
(707) 780-1951 Fax
Main office
(401) 284-1827 Business
(401) 284-1840 Fax
Oregon office
(503) 430-1065 Business
(503) 430-1285 Fax
http://www.teraspeed.com
This e-mail contains proprietary and confidential intellectual property of
Teraspeed Consulting Group LLC
------------------------------------------------------------------------------------------------------
Teraspeed(R) is the registered service mark of Teraspeed Consulting Group LLC
------------------------------------------------------------------
To unsubscribe from si-list:
si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field
or to administer your membership from a web page, go to:
//www.freelists.org/webpage/si-list
For help:
si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field
List technical documents are available at:
http://www.si-list.net
List archives are viewable at:
//www.freelists.org/archives/si-list
or at our remote archives:
http://groups.yahoo.com/group/si-list/messages
Old (prior to June 6, 2001) list archives are viewable at:
http://www.qsl.net/wb6tpu
------------------------------------------------------------------
To unsubscribe from si-list:
si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field
or to administer your membership from a web page, go to:
//www.freelists.org/webpage/si-list
For help:
si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field
List technical documents are available at:
http://www.si-list.net
List archives are viewable at:
//www.freelists.org/archives/si-list
or at our remote archives:
http://groups.yahoo.com/group/si-list/messages
Old (prior to June 6, 2001) list archives are viewable at:
http://www.qsl.net/wb6tpu
Other related posts:
