How often does a zener have a failure mode where it acts like a good
(not leaky) much lower voltage zener? I usually see leaky or shorted
ones rather than substantial drift...

Fault was 185v line in 'scope low at 170v and slight ripple, failure was
24v zener which was acting like a good 3.3v zener.
'Scope was still working, just had slightly reduced intensity and
thought it better to check voltages before twiddling the sub-intensity
pot

Lee

Lee wrote:
How often does a zener have a failure mode where it acts like a good
(not leaky) much lower voltage zener? I usually see leaky or shorted
ones rather than substantial drift...

Fault was 185v line in 'scope low at 170v and slight ripple, failure was
24v zener which was acting like a good 3.3v zener.
'Scope was still working, just had slightly reduced intensity and
thought it better to check voltages before twiddling the sub-intensity
pot

Lee

That is common for a zener.. excessive heat usually does that.

From what I understand the zener is fabbed a little differently so
not to destroy itself when doing the break down. Normal diodes tend
to not like being used as zeners for any length of time. I guess
it could be possible you may had a zener that had some impurities in
the process that could of allowed it to break down like non-zener semi's
or, it just got wacked some how. If this reference is supplying some
out side device as a source of energy or ref on the scope? It's possible
it got whacked.

Also, if only supplies internal references, check to make sure the
connecting circuit is still using it. If not, the zener could be
clamping too much and operating hot..

Normally the evidence is showed with a over heated board that it's
attached to.


Jamie

On 01/06/2013 18:20, Jamie wrote:

Also, if only supplies internal references, check to make sure the
connecting circuit is still using it. If not, the zener could be
clamping too much and operating hot..

Normally the evidence is showed with a over heated board that it's
attached to.


Thanks.

I'm not sure I understand the reasons it's used in this application
either, it's a 5W part and it seems to be used to create a 185v supply
from a 245v tap (+ bridge obviously) on the mains transformer. This
185v supply is then regulated by a comparator/series pass transistor to
160v. The weird thing is that even with the zener failed, the 160v
supply was still just about right, (it was at 163v).
Replacing the zener has put the 185v rail back to the proper value and
the 160v rail is now solid as well, minus the ripple.

The zener was mounted off-board with fibreglass sleeves on its leads so
they expected it to get hot Hasn't burnt the board though.
Replacement runs hot as well, but not uncomfortably so.

Lee

Ah, penny drops. I had to model the circuit in Spice before I understood
it
It's used to create a stable 160v for the positive end of the resistor
divider at the input to the comparator, the negative end of the chain is
connected to a separately regulated -12v (from a different transformer
tapped supply) and the comparator (other input is ground) and series
pass transistor then regulates the 245v supply from the bridge to 185v.
Which in turn is dropped by the zener to 160v

If the zener is shorted then the 160v supply stays at 160v, but with
lots of ripple. The 185v supply isn't used for anything else other than
as a marked test point and to supply the 160v rail via the zener. That's
the thing that confused me originally Why have a series pass
transistor but then drop that supply through a zener?

I guess there is a good reason to do it this way, but it confuzzled me

Lee

Lee wrote:

On 01/06/2013 18:20, Jamie wrote:

Also, if only supplies internal references, check to make sure the
connecting circuit is still using it. If not, the zener could be
clamping too much and operating hot..

Normally the evidence is showed with a over heated board that it's
attached to.



Thanks.

I'm not sure I understand the reasons it's used in this application
either, it's a 5W part and it seems to be used to create a 185v supply
from a 245v tap (+ bridge obviously) on the mains transformer. This
185v supply is then regulated by a comparator/series pass transistor to
160v. The weird thing is that even with the zener failed, the 160v
supply was still just about right, (it was at 163v).
Replacing the zener has put the 185v rail back to the proper value and
the 160v rail is now solid as well, minus the ripple.

The zener was mounted off-board with fibreglass sleeves on its leads so
they expected it to get hot Hasn't burnt the board though.
Replacement runs hot as well, but not uncomfortably so.

Lee
sounds like a poor design.. maybe it should have a heat sink hanging
off it.

Jamie

Lee wrote:

Ah, penny drops. I had to model the circuit in Spice before I understood
it
It's used to create a stable 160v for the positive end of the resistor
divider at the input to the comparator, the negative end of the chain is
connected to a separately regulated -12v (from a different transformer
tapped supply) and the comparator (other input is ground) and series
pass transistor then regulates the 245v supply from the bridge to 185v.
Which in turn is dropped by the zener to 160v

If the zener is shorted then the 160v supply stays at 160v, but with
lots of ripple. The 185v supply isn't used for anything else other than
as a marked test point and to supply the 160v rail via the zener. That's
the thing that confused me originally Why have a series pass
transistor but then drop that supply through a zener?

I guess there is a good reason to do it this way, but it confuzzled me

Lee

Possibly two things..

1. The series transistor is under switch control so that it could shut
off the source or, it could be a current limited.

2. The series transistor could be stepping down the voltage so that
zener shunt does not exert so much current.. Just enough to maybe smooth
out the ripple..

Jamie

"Lee" wrote in message
...
Ah, penny drops. I had to model the circuit in Spice before I understood
it
It's used to create a stable 160v for the positive end of the resistor
divider at the input to the comparator, the negative end of the chain is
connected to a separately regulated -12v (from a different transformer
tapped supply) and the comparator (other input is ground) and series pass
transistor then regulates the 245v supply from the bridge to 185v. Which
in turn is dropped by the zener to 160v

If the zener is shorted then the 160v supply stays at 160v, but with lots
of ripple. The 185v supply isn't used for anything else other than as a
marked test point and to supply the 160v rail via the zener. That's the
thing that confused me originally Why have a series pass transistor but
then drop that supply through a zener?

I guess there is a good reason to do it this way, but it confuzzled me

Lee


It's not an 'active smoothing' (sometimes also known as electronic
smoothing) circuit is it by any chance ? I came across this principle some
years ago in some piece of test equipment that I owned. Come to think of it,
it might even have been in the Marconi lab AF generator that I still have.
As I recall, it used a tetrode or pentode valve as the series pass element
in this system. Basically, the circuit emulates a large smoothing cap by
feeding ripple back in reverse phase to the control element. If this was not
working correctly in your circuit due to the faulty zener, it might account
for why the rail was still basically of the correct voltage, but with excess
ripple ?

Arfa

On 02/06/2013 02:24, Arfa Daily wrote:




It's not an 'active smoothing' (sometimes also known as electronic
smoothing) circuit is it by any chance ? I came across this principle
some years ago in some piece of test equipment that I owned. Come to
think of it, it might even have been in the Marconi lab AF generator
that I still have. As I recall, it used a tetrode or pentode valve as
the series pass element in this system. Basically, the circuit emulates
a large smoothing cap by feeding ripple back in reverse phase to the
control element. If this was not working correctly in your circuit due
to the faulty zener, it might account for why the rail was still
basically of the correct voltage, but with excess ripple ?

Arfa

Interesting idea, does sound like they could be doing that
The comparator is a TL072 and the reference voltage is fed into the
inverting input...
With the faulty zener there was a varying 2-3v of 100hz positive-going
ripple on the 160v line, looked like a failed smoothing cap. With the
replacement it's a few millivolts.

Lee

On Jun 2, 3:03*am, Lee wrote:
On 02/06/2013 02:24, Arfa Daily wrote:







It's not an 'active smoothing' (sometimes also known as electronic
smoothing) circuit is it by any chance ? I came across this principle
some years ago in some piece of test equipment that I owned. Come to
think of it, it might even have been in the Marconi lab AF generator
that I still have. As I recall, it used a tetrode or pentode valve as
the series pass element in this system. Basically, the circuit emulates
a large smoothing cap by feeding ripple back in reverse phase to the
control element. If this was not working correctly in your circuit due
to the faulty zener, it might account for why the rail was still
basically of the correct voltage, but with excess ripple ?

Arfa

Interesting idea, does sound like they could be doing that
The comparator is a TL072 and the reference voltage is fed into the
inverting input...
With the faulty zener there was a varying 2-3v of 100hz positive-going
ripple on the 160v line, looked like a failed smoothing cap. With the
replacement it's a few millivolts.

Lee- Hide quoted text -

- Show quoted text -

I would surely try to put someting on the replacement zener to cool
it, a fin, or maybe clamp the body to a piece of electricallly
isolated metal. Even mounting the zener with as short leads as
possible if connected to something on a pwb that would act as a heat
sink, or maybe even solder some copper tape wings onto the zener
leads, anything to help it run cooler.

wrote:

On Jun 2, 3:03 am, Lee wrote:

On 02/06/2013 02:24, Arfa Daily wrote:








It's not an 'active smoothing' (sometimes also known as electronic
smoothing) circuit is it by any chance ? I came across this principle
some years ago in some piece of test equipment that I owned. Come to
think of it, it might even have been in the Marconi lab AF generator
that I still have. As I recall, it used a tetrode or pentode valve as
the series pass element in this system. Basically, the circuit emulates
a large smoothing cap by feeding ripple back in reverse phase to the
control element. If this was not working correctly in your circuit due
to the faulty zener, it might account for why the rail was still
basically of the correct voltage, but with excess ripple ?

Arfa

Interesting idea, does sound like they could be doing that
The comparator is a TL072 and the reference voltage is fed into the
inverting input...
With the faulty zener there was a varying 2-3v of 100hz positive-going
ripple on the 160v line, looked like a failed smoothing cap. With the
replacement it's a few millivolts.

Lee- Hide quoted text -

- Show quoted text -


I would surely try to put someting on the replacement zener to cool
it, a fin, or maybe clamp the body to a piece of electricallly
isolated metal. Even mounting the zener with as short leads as
possible if connected to something on a pwb that would act as a heat
sink, or maybe even solder some copper tape wings onto the zener
leads, anything to help it run cooler.

The little tap wings works, I've done things like that. But, I also
do this;
I keep a roll of copper tape with adhesive around. Cut in narrow
strip and make little fan folds, stick to the diode body, solder it
together around the body..

Use precision tweezers to make the folds. You shouldn't need to solder
to a lead. You could put a dab of glue on there if that makes you feel
better.

Jamie

Jamie t wrote in message
...
wrote:

On Jun 2, 3:03 am, Lee wrote:

On 02/06/2013 02:24, Arfa Daily wrote:








It's not an 'active smoothing' (sometimes also known as electronic
smoothing) circuit is it by any chance ? I came across this principle
some years ago in some piece of test equipment that I owned. Come to
think of it, it might even have been in the Marconi lab AF generator
that I still have. As I recall, it used a tetrode or pentode valve as
the series pass element in this system. Basically, the circuit emulates
a large smoothing cap by feeding ripple back in reverse phase to the
control element. If this was not working correctly in your circuit due
to the faulty zener, it might account for why the rail was still
basically of the correct voltage, but with excess ripple ?

Arfa

Interesting idea, does sound like they could be doing that
The comparator is a TL072 and the reference voltage is fed into the
inverting input...
With the faulty zener there was a varying 2-3v of 100hz positive-going
ripple on the 160v line, looked like a failed smoothing cap. With the
replacement it's a few millivolts.

Lee- Hide quoted text -

- Show quoted text -


I would surely try to put someting on the replacement zener to cool
it, a fin, or maybe clamp the body to a piece of electricallly
isolated metal. Even mounting the zener with as short leads as
possible if connected to something on a pwb that would act as a heat
sink, or maybe even solder some copper tape wings onto the zener
leads, anything to help it run cooler.

The little tap wings works, I've done things like that. But, I also
do this;
I keep a roll of copper tape with adhesive around. Cut in narrow
strip and make little fan folds, stick to the diode body, solder it
together around the body..

Use precision tweezers to make the folds. You shouldn't need to solder
to a lead. You could put a dab of glue on there if that makes you feel
better.

Jamie


Another option, topology dependent, is using the full length replacement
zener leads, scraping off any coating to the traces and curving the leads
to match the traces and soldering the leads to the traces for an inch or
more