Arfa Daily wrote:
"Jamie" t wrote in message
news
Peter Andersen wrote:
I have to NAD T762 on the table. Both with burned out (shorted) 5,1V
zenerdiodes (D718).
I have replaces them with a new 5W type instead of the original 2W.
But the diodes is getting extremely hot and I can't figure out if this is
ok (like the way it was designed). They are sitting directly in series
with a 7812 from a 22V DC supply. The thought behind the design must bee
to divide the power loss between the zenerdiode and the 7812.
The amplifier is working, but I don't like those hot zeners!
Is there anybody out there that have worked on a NAD T762?
Peter
Well,
I have a theory..
If the Zener is in fact in series with the feed to the 7812
from the 22 volt source? May I suggest that maybe 22 volts isn't
the real voltage it should be? If this is the case, it would place
the zener in a high resistive state and it will get hot!..
The reason say this is because the 7812 (last time I checked) can
handle around 36 volts.
Since your Zener is 5.1 volts, that leaves 2 things in conclusion.
#1.
It really isn't being used in Series but in parallel with a feeder
resistor to be used as a reference! In which case, the feeder as reduced
in size of the load from the circuit has been removed!.
#2
The 5.1 volt zener was picked to bring the Vcc source voltage down
a bit with in the safe operating range.
I would be willing to bet that it's suppose to be more than 22 volts.
Well I must say, Jamie, I am having some difficulty following those lines of
reasoning. If the nominal supply is 22v, and a 5v1 zener is glued in series
with that to the input of the 7812, that's going to leave around 17v on the
input to the reg. So why is that a problem in as much as you think that the
starting voltage should be higher than 22v ? I must confess that I don't
really see the point of reducing the input volts by 5 - either 17 or 22
seems like a perfectly valid amount to me, just that at 22v, the reg might
need a bit more heatsinking on it. However, on that point, if the original
zener was rated at 2 watts, then allowing for some margin and derating to
say 1.5 watts, that would allow for a maximum current of 1.5 / 5 or ~330mA,
so that would imply that the designer was not expecting more than this to be
drawn from his 7812, even though these, in their commonest form, are rated
to an amp. Now if that is the case, the 7812 is almost idling anyway with
17v on its input, so I really can't see it caring much, temperature -wise,
about having 22v on it rather than 17v.
However, something else just occured to me. The OP was taking about zeners
in the plural - two I think. So is that two zeners, in series with each
other, and then in series with the input to the reg ? If that is the case,
then the maths certainly doesn't stack up, as 10v will be lost across the
zeners, leaving only 12v on the input to the regulator, which clearly isn't
enough, as Jamie suggests might be the case.
As a matter of interest, as these zeners had apparently failed short
circuit, and we know that the 7812 can take up to 36v on its input, how did
the OP know that there was a problem, given that he says that the amp is
apparently working ok ?
Arfa
Ok, I'll add to that. What if, there were filter caps that reduce the
ripple of the 22 volt source and the cap's were defective? This in turn
wouldn't give you a constant voltage and thus, your meter would not be
giving you the actual where it should be if working caps were in there.
So lets assume for the moment, 22 * 1.414 = 31.108;
Lets say we place a 5.1 in series, net results would be ~25 V
At this level, the zener's resistive path would be much lower
and thus, less heat dissipation. And ofcourse, bringing the input
voltage down a bit for the post regulator will help reduce the
dissipation.
Not knowing exactly what the device is.
May I suggest that it's a small device being operated via an incorrectly
supplied wall wart or a supply unit that has failed caps.
Just an analogy !
Ok, Bye fer now!
--
"I'm never wrong, once i thought i was, but was mistaken"
Real Programmers Do things like this.
http://webpages.charter.net/jamie_5