[vdr] Re: Hardware falures

[Stefaan Coddé <Stefaan.Codde@pandora.be>]

I doubt that 35°C is too much for the tuner on a Rev1.3 board
but if the outside coax plug reached that temp , then it sure get's hot
inside
the tuner meaning, higher then those 35°C.
A fan can help but add's extra noise and there's a silent solution
available.

Heat has always been a problem in a PC  but everyone seems to forget the
basic
behaviour of hot air.... it rises and hit's the top of the case.
Now : this tip works only good on desktop cases and sure NO towers.

Drill many air holes thru the desktop cover, right above the tuner. Holes
with a diameter
of 3mm and pitch of 8mm will do just fine.

Second : if You place the unit in a rack, be sure that there's room enough
above the computer
so that the hot air can excape there too and there ya go... no fan needed.

Still getting too hot?  Add a heatsink on the tuner and the DSP and drill
more holes in the
cover.

I've drilled approx 1200 holes in my desktop cover.  This was needed cause
it's water cooled and the radiator
is located inside the chassis and there's no fan blowing air through the
radiator....and... it works.

Tower PC's : too bad, all cards are mounted horizontally which is VERY bad
for heat dissipation.

btw : for those who are interested.  I'mm working on a water cooled ATX
power supply but here too:
there are still ventilation holes needed on top cause *every* component on a
board generates heat.
Some more then others.


----- Original Message -----
From: "Rainer Zocholl" <NoSpam-@zocki.toppoint.de>
To: <vdr@linuxtv.org>
Sent: Monday, April 29, 2002 9:20 PM
Subject: [vdr] Re: Hardware falures


> einStein@donau.de(Christian Berger)  29.04.02 16:09
>
> Once upon a time Christian Berger shaped the electrons to say...
>
> >Servus
>
> >Uhm, are your DVB cards also failing so often? Recently I sent mine to
> >the factory for the second time. I just want to check wether that's
> >just my card or a general problem with Haupauge's cards.
>
>
> Meanwhile i have returned two/three cards
>
> one Nova
> Failure Report: "artefacts"
> "no question asked" full refund after "only" 6 Weeks
>
> I had checked another Nova of the same series:
> Artefacts too.
> (nova did not become hot at all)
>
>
> and one V.2.1 is STILL in the fingers of the "distributor"
> now the second time in the 4th Week.
> The frist time, after 6 Weeks the broken card was returned
> with -believed it- "Tested in 4 PCs, no failure found".
> The problem was:
> If the PC was powered off for more than 4 and less than 24h,
> the card blocks the PCI bus in the boot up.
>
> The fault was perferctly reproducible on my PC
> when i unplug mains entirely for 4..6h and switch
> the PC on by the mains switch! (Yes, the BIOS allows that
> setting: "boot on power return", and that is a MUST for
> timed operation, because the clock interrupt may have occured
> while power was entirely off..)
>
> It makes timed recordings impossible, because every
> evening, i see the computer hung in BIOS. But only when
> that card was inserted. Other cards did not make no such
> problem. It was a clear problem by that card.
>
>
> I would say: every second of my cards were broken/unusable in some way.
>
>
>
>
> If you use the V.1.3 you _must_ place a fan to blow
> between the boards!
> V2.1 seems not to be so temperature sensitive, have higher RF
> sensitiveness(26uV vs 28uV) and are too generating/dissipating
> much less heat.
>
> I messured after a short while approx 35C(!) on the outside koax plug
> of a V1.3, when the cards starts to generate many errors.
> The remaining case was only 28C.
> After adding a 60mm fan for 2 cards, all are 28C ;-)
>
>
>
>