[rescue] Fwd: Re: SPARC10 died today

[john]

-------- Forwarded Message --------
Subject: 	Re: SPARC10 died today
Date: 	Thu, 30 Nov 2017 21:07:19 -0500
From: 	john <jferg977 at aol.com>
To: 	John Hudak <jjhudak at gmail.com>



This is pretty comprehensive and follows my thoughts.B  As it happens, it
does fire-up but doesn't even run the POST.B  it does start the drives
but with no POST, it's pretty hard to ascertain where problem might be.
A while back i had similar problem with toshiba notebook, which would
start to boot flash the screen and lockup.B  problem was corroded
connections on SMD power regulator which I was able to remove, (hot air
gun) replace part and re-tin connections and solder it back down for
another 3 years of constant service. When it did it again, i threw it
out and bought an HP

I bought another 10 from guys in California and am beginning to shop for
modern software which runs on modern machines to replace 1992 stuff I
was running on the Sun.B  I'mB  not sure I'm up to the level of preventive
maintenance you describe.

many thanks for the thoughts,

john ferguson


On 11/30/2017 08:53 PM, John Hudak wrote:
> If one is dependent on machines of this vintage, then certain 
> preventative maintenance areas should be addressed before the machine 
> goes belly up.
>
> First is that the power supply needs to be rebuilt.B  Of course the 
> filter caps are the easy/well known things to replace.B  Everyone seems 
> to anecdotally know that without understanding the designs of the 
> powersupplies.
>
> Switch Mode power supplies (SMPS) should have the caps in the 
> switching circuit replaced as well before they develop a low ESR which 
> can take out most of the semiconductors in the switcher.
> Incoming transient suppression devices (e.g. MOVs) should be replaced.
>
> Safety circuits (OVP, OCP) within the PS should be evaluated and caps 
> and older carbon resistors should be replaced.
> The PS feeds all the chips on the MB, and I/O devices and if the power 
> supply suffers an OV, it feeds that directly to the semiconductors 
> which can destroy them.
>
> If there are well known failure modes of the PS, components associated 
> with the failure modes should be replaced.
> It is generally a good idea to install components that have a higher 
> working voltage or wattage rating than that of the original. Heat 
> build up will not stress oversized components as much as the original 
> speced components.
>
> Consider installing a constant voltage transformer between the power 
> source and the computer.B  It will help reduce the effects of 
> transients and harmonic 'crap' on the incoming line
>
> If the PS is loaded to within 80-90% of its capacity, one may want to 
> consider adding some form of forced cooling (.e.g. fan) to remove heat 
> as heat is the main culprit to breakdown of both passive and active 
> components.
>
>
> Keeping an old PS as a standby is illogical as a means of keeping the 
> machine runningB  because old parts are old parts and a old PS can fail 
> and take out downstream devices. I've seen cases where NOS PS that had 
> been sitting on a shelf forever installed in machines, only to have 
> them fail shortly after being put into service.
>
> Likewise, turning the machines on every 3 months will keep the 
> capacitors from deforming and losing their capacitance but, again, it 
> won't help the other failure modes...old parts are old parts.
>
> Depending on the design approach of the CPU board(s), some may employ 
> bulk electrolytic capacitors in the area of the power supply 
> connectors.B  Those should be replaced.
>
> good luck
> J