George White wrote in a message to Roy J. Tellason:

 GW> Hi Roy,

 GW> On 11-Mar-04, Roy J. Tellason wrote to Greg Mayman:

 GW> 

 GM>> 6 months seems to be very common these days where it used to be
 GM>> from 1 to 5 years. And of course the electronics are
 GM>> unrepairable, you just replace the modules... except that by that
 GM>> time they are no longer made.

 RJT> No,  actually the electronics were repairable.  But they weren't
 RJT> very well made. The cheapest model had a pair of freestanding
 RJT> speaker boxes,  but used only a pair of TDA2040 chips for output
 RJT> devices,  which tended to explode with regularity as they applied
 RJT> 40 volts or so to the supply pin,  which is just at the absolute
 RJT> maximum rating for that part.  Power supply rectifiers were
 RJT> 1N4001! These also often went bad..

 GW> You do realise that the voltage rating on low number 1N400X series
 GW> was pretty well meaningless, don't you? They used to make a batch
 GW> of diodes, then select out for voltage breakdown. Over 1000V and it
 GW> was a 4007, 800V for 4006, etc. Once they had enough of a type all
 GW> the rest would end up in the lower rated types. Anything that
 GW> wasn't short circuit (almost) ended up as 4001, so that the devices
 GW> you got could have been virtually any capability, and would most
 GW> likely have been way over 500V... The most likely killer would have
 GW> been current, or even more likely juging by my experience with
 GW> approving 4004s, the packaging not being properly sealed. Some
 GW> makes were really bad... 

Yeah,  you could be right,  I suppose...

Still,  that strikes me as seriously under-spec'ing a part.  For operation
where the DC output is 40V (to power a couple of power amp chips that have
a 35V maximum rating!) you'd want more than that,  at least the 4002,  if
not higher.

These things were really made to break!

---