not totally repulsive

[[email protected]]

Prior to that, how many existing boards do you need to use up, and how
many customers can you afford to loose when they do not work

About as many as he can afford to tighten, I suppose.

 

[Rich Grise]

.

turbofans. I'll take the extra supplies as the lesser-of-two evils.

If Satan and Lucifer showed up at the inn, and the proprietor rented
them each a room, would he be the lessor of two evils? ;-)

Cheers!
Rich

 

[John Larkin]

If Satan and Lucifer showed up at the inn, and the proprietor rented
them each a room, would he be the lessor of two evils? ;-)

Cheers!
Rich

Or, as Captain Jack Aubrey said, the proper British sailor always
chooses the lesser of two weevils.

John

 

[John Larkin]

About as many as he can afford to tighten, I suppose.

The chances of this not working are nil; I'd be more worried about
meteor damage in shipping. But in fact my customers are great: when we
do have a problem, we tell them the truth, we work with them to fix
it, we ship them replacement products, whatever it takes. They know
that complicated things sometimes go wrong. What they appreciate is an
honest, energetic fix.

John

 

[John Devereux]

The chances of this not working are nil; I'd be more worried about
meteor damage in shipping.

In fact I would think it is more likely that the "proper" LDO solution
will e.g. start oscillating, for some reason.

 

[Spehro Pefhany]

In fact I would think it is more likely that the "proper" LDO solution
will e.g. start oscillating, for some reason.

For example, if it prefers black capacitors over orange ones.


Best regards,
Spehro Pefhany

 

[Andy]

In fact I would think it is more likely that the "proper" LDO solution
will e.g. start oscillating, for some reason.

If the LDO starts oscillating "for some reason", it is either a faulty
(i.e. not meeting its specs) component, or the design was incorrect
(failing to take into account the entire range of operating
conditions, tolerances, tempcos, drift/aging, etc.) which is much more
common, particularly among advocates of using forward biased zener
diodes for power supplies.

The diode circuit cannot be designed correctly (because it is being
operated outside its specifications).

If you are saying it is easier to "get it working" for a diode circuit
than an LDO, you may be correct, under the right conditions (which
generally disappear the moment the product is shipped).

As mentioned earlier, continuously forward biasing a diode designed to
operate reverse-biased may cause long term problems.

Andy

 

[Spehro Pefhany]

If the LDO starts oscillating "for some reason", it is either a faulty
(i.e. not meeting its specs) component, or the design was incorrect
(failing to take into account the entire range of operating
conditions, tolerances, tempcos, drift/aging, etc.) which is much more
common, particularly among advocates of using forward biased zener
diodes for power supplies.

The diode circuit cannot be designed correctly (because it is being
operated outside its specifications).

If you are saying it is easier to "get it working" for a diode circuit
than an LDO, you may be correct, under the right conditions (which
generally disappear the moment the product is shipped).

As mentioned earlier, continuously forward biasing a diode designed to
operate reverse-biased may cause long term problems.

Andy

What makes you think zeners should not be operated in forward bias?
The normal zener supply off line voltage switches between forward and
reverse bias at line frequency. Do you have any reason to suspect
there would be problem?


Best regards,
Spehro Pefhany

 

[Spehro Pefhany]

What makes you think zeners should not be operated in forward bias?
The normal zener supply off line voltage switches between forward and
reverse bias at line frequency. Do you have any reason to suspect
there would be problem?

^^^ any


(Maybe I'm dealing with too many people these days who are not
fortunate enough to have the Queen's English as their mother tongue).
;-)


Best regards,
Spehro Pefhany

 

[John Larkin]

If the LDO starts oscillating "for some reason", it is either a faulty
(i.e. not meeting its specs) component, or the design was incorrect
(failing to take into account the entire range of operating
conditions, tolerances, tempcos, drift/aging, etc.) which is much more
common, particularly among advocates of using forward biased zener
diodes for power supplies.

There are lots of tricky/flakey/poorly specified LDOs being sold in
volume. Many/most are very sensitive to the capacitive+esr load
environment, and many of the datasheets are fuzzy about that. Ditto
switchers.

The diode circuit cannot be designed correctly (because it is being
operated outside its specifications).

What a bizarre thing to say. I'm an electrical engineer, and I do
stuff like this all the time. The behavior of forward-biased PN
junctions is fairly well known by now. And as Austin pointed out, the
actual operating range of Vccaux is huge.

If you are saying it is easier to "get it working" for a diode circuit
than an LDO, you may be correct, under the right conditions (which
generally disappear the moment the product is shipped).

As mentioned earlier, continuously forward biasing a diode designed to
operate reverse-biased may cause long term problems.

Why would it do that? What would be the failure mechanism? What would
be the failure mode?

Incidentally, zeners are use in the forward direction all the time,
like in bidirectional clippers and transzorbs.

John

 

[Tom Del Rosso]

Dang, the power supplies are more trouble than the FPGAs.

In the FPGA the engineering is already done for you!

 

[Tim Williams]

The point is current spreading. Because they aren't intended for handling
large forward currents, their junctions aren't designed to handle the
thermal effects. Like an SCR's dI/dt rating, local heating can cause
failure not expected for that current level.

Tim

 

[John Larkin]

The point is current spreading. Because they aren't intended for handling
large forward currents, their junctions aren't designed to handle the
thermal effects. Like an SCR's dI/dt rating, local heating can cause
failure not expected for that current level.

Tim

I'm not all that concerned about dissipating 40 milliwatts in a 1-watt
zener.

John

 

[Andy]

What a bizarre thing to say. I'm an electrical engineer, and I do
stuff like this all the time. The behavior of forward-biased PN
junctions is fairly well known by now.

You have a degree in Electrical Engineering. So do I.

When you design electronics that has to work right every time or the
wrong people will die, you learn to do it right.

The forward biased behavior of PN junctions designed to be operated
primarily (albeit not exclusively) in the negative biased mode is not
nearly so simple. I'm not an expert on forward biased zener diodes,
and I wouldn't use it that way unless I was. Apparently, from other
posts on this, there are potential problems that you and I are not
fully aware of. I won't design a product that way. I may design an
experiment that way, but I will specify additional analysis,
screening, and source control measures to ensure that, once proven to
be effective over the entire operating range, a product will continue
to operate correctly over its entire range of operating conditions,
over the duration of its production and useful life. Anything less is
hacking, not engineering. And the cost of those additional measures
nearly always exceeds the cost of doing it right in the first place.

Andy

 

[John Larkin]

You have a degree in Electrical Engineering. So do I.

When you design electronics that has to work right every time or the
wrong people will die, you learn to do it right.

The forward biased behavior of PN junctions designed to be operated
primarily (albeit not exclusively) in the negative biased mode is not
nearly so simple. I'm not an expert on forward biased zener diodes,
and I wouldn't use it that way unless I was. Apparently, from other
posts on this, there are potential problems that you and I are not
fully aware of. I won't design a product that way. I may design an
experiment that way, but I will specify additional analysis,
screening, and source control measures to ensure that, once proven to
be effective over the entire operating range, a product will continue
to operate correctly over its entire range of operating conditions,
over the duration of its production and useful life. Anything less is
hacking, not engineering. And the cost of those additional measures
nearly always exceeds the cost of doing it right in the first place.

Andy

I didn't design the diode in originally; it was a hack to fix a
problem on these boards. People who are "not an expert on forward
biased zener diodes" are conjecturing problems without suggesting what
they might actually be, except prissy comments about "not the intended
use", as if the silicon cares.

All rectifier diodes are zener diodes, and they seem to work in the
forward direction for a long time.

The only thing remotely like this that I've ever heard of is the fact
that longterm zenering of a transistor b-e junction can reduce beta.
I've never heard that forward biasing a diode can damage its
forward-bias performance. [1]

Hell, I'll probably use the diode next rev, too. I'm starting to like
it.

John

[1] except for GaAs tunnel diodes.