Ascend Archive
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: (ASCEND) MAX 4000 + 56K modem cards = fire

On Tue, Sep 09, 1997 at 10:41:13AM -0500, Karl Denninger wrote:
> On Tue, Sep 09, 1997 at 10:31:54AM -0500, Jake Messinger wrote:
> > On Tue, 9 Sep 1997 ascend@digistar.com wrote:
> > 
> > > I have a Davis Weather monitor (a weather station) inside the main server
> > > room and looking at the temperature, the average is 62F, the peak being
> > > 66F, the minimum being 61F.  When the MAX 4000s died, the temperature was
> > > well below 70F.
> > 
> > What was the temp inside the max? Id suggest sticking a meat thermometer
> > inside it.
> 
> I'd like to note that a PM3, with a full load of modem cards, will continue
> to operate if you *DISCONNECT* the single fan (which is really there to cool
> the power supply) and it will NOT overheat in an average room under those
> conditions.
> 
> Control of overheating starts by being *very* careful about power consumption
> to begin with.  
> 
> Remember that you pay for every watt consumed *twice* -- once to feed the
> machine, and again to remove it from the room.  There is no such thing as a
> free lunch, and every watt you consume in a piece of computer hardware ends
> up in the room as heat.

Now I'm following up to my own posts.

A *polite* hint to manufacturers:

	If you MUST build hardware which requires forced-air cooling to
	operate without damage (I fully understand that density and
	dissipation requirements sometimes require this) then you have 
	an *OBLIGATION* to include temperature monitoring facilities in
	your harware which generate reportable status -- as well as an
	automated overtemp shutdown.

	Anything less should, IMHO, VOID your UL/CSA certification.

	CISCO does this in their larger router products (which will
	definitely croak if the fans fail or air intake is obstructed).
	ASCEND, at least in the MAX, does not (I don't know about the 
	TNT).  But they should.  How much does a thermocouple and a 
	small board to shut off the power supply cost?  $10?  C'mon 
	folks.

We had significant heat dissipation problems with the MAXes when we had 
them here.  In fact, we had to change our racking plan and remove the 
side panels from both end racks in order to obtain rational flow-through
temperature gradients through the chassis.  With the end panels in place 
the operating temperatures of the MAXes were way beyond my comfort level 
for hardware of this type.

IMHO the thermal engineering of the MAX with V.34 modem boards is marginal 
at best. Using DM12s was better than the original DM8s, but not by a lot.
	 
--
-- 
Karl Denninger (karl@MCS.Net)| MCSNet - Serving Chicagoland and Wisconsin
http://www.mcs.net/~karl     | T1's from $600 monthly to FULL DS-3 Service
			     | NEW! K56Flex modem support is now available
Voice: [+1 312 803-MCS1 x219]| 56kbps DIGITAL ISDN DOV on analog lines!
Fax:   [+1 312 803-4929]     | 2 FULL DS-3 Internet links; 400Mbps B/W Internal
++ Ascend Users Mailing List ++
To unsubscribe:	send unsubscribe to ascend-users-request@bungi.com
To get FAQ'd:	<http://www.nealis.net/ascend/faq>
Follow-Ups: References: