1% zener diodes?

[Michael]

Hey there - is there such a thing as a 1% zener diode? They all seem
to be about 5%. I was looking to make a 10V vref and the best I was
able to find was the Digi-Key EDZTE6110BCT-ND, which is accurate to
2.3%. This just strikes me as odd - and makes me wonder if I'm looking
in the wrong place or something.

I did, however, discover that there are actual 10V vref chips out
there - so I'm planning on going with one of those.

-Michael

 

[Rene Tschaggelar]

Hey there - is there such a thing as a 1% zener diode? They all seem
to be about 5%. I was looking to make a 10V vref and the best I was
able to find was the Digi-Key EDZTE6110BCT-ND, which is accurate to
2.3%. This just strikes me as odd - and makes me wonder if I'm looking
in the wrong place or something.

I did, however, discover that there are actual 10V vref chips out
there - so I'm planning on going with one of those.

You're assuming that these devices start to conduct
at 10.00V sharp ? This is not the case anyway. The
specified conduction voltage is defined at 1mA.

And yes, a 10V reference is a different device.

Rene

 

[Jan Panteltje]

Hey there - is there such a thing as a 1% zener diode? They all seem
to be about 5%. I was looking to make a 10V vref and the best I was
able to find was the Digi-Key EDZTE6110BCT-ND, which is accurate to
2.3%. This just strikes me as odd - and makes me wonder if I'm looking
in the wrong place or something.

I did, however, discover that there are actual 10V vref chips out
there - so I'm planning on going with one of those.

Zener diodes are actually in a breakdown region.. so unreliable,
not accurate etc by nature.

10V reference? A LM317 with 2 resistors or a 10 turn poti will
make any reference you like, for less money with better temp stability,
then a zener, possibly for less money too.
All that with more power....

 

[Michael]

You're assuming that these devices start to conduct
at 10.00V sharp ? This is not the case anyway. The
specified conduction voltage is defined at 1mA.

And yes, a 10V reference is a different device.

Rene

Well, my knowledge of them was that you would always put a resistor
inline with them to give it a proper bias current so that you would
hit the exact right voltage. I need the 10V to connect to the non-
inverting input of an Op-amp (OP747). Thus I'm not worried about a
current drain on the line.

-Michael

 

[whit3rd]

Hey there - is there such a thing as a 1% zener diode?

No. Zener diodes are mass-produced and sorted into 10% and 5%
bins, but if you require 1% accuracy you need a laser-trimmed
resistor (at least) in addition to bias current regulation and
other niceties. 10V 1% references are trimmed ICs, not
discrete zeners.

REF-01 is a typical such part; see

http://www.analog.com/en/prod/0,2877,REF01,00.html

 

[D from BC]

Hey there - is there such a thing as a 1% zener diode? They all seem
to be about 5%. I was looking to make a 10V vref and the best I was
able to find was the Digi-Key EDZTE6110BCT-ND, which is accurate to
2.3%. This just strikes me as odd - and makes me wonder if I'm looking
in the wrong place or something.

I did, however, discover that there are actual 10V vref chips out
there - so I'm planning on going with one of those.

-Michael

Even if 1% tolerance did exist, I think the thermal drift (due to self
heating and a range of ambient temperatures) would make the 1%
pointless.
D from BC

 

[Jim Thompson]

Even if 1% tolerance did exist, I think the thermal drift (due to self
heating and a range of ambient temperatures) would make the 1%
pointless.
D from BC

TL431 ??

...Jim Thompson

 

[James Arthur]

Hey there - is there such a thing as a 1% zener diode? They all seem
to be about 5%. I was looking to make a 10V vref and the best I was
able to find was the Digi-Key EDZTE6110BCT-ND, which is accurate to
2.3%. This just strikes me as odd - and makes me wonder if I'm looking
in the wrong place or something.

I did, however, discover that there are actual 10V vref chips out
there - so I'm planning on going with one of those.

-Michael

The 10V reference chip is a better choice, but I once had a set of
precision zeners at either 1 or 2% increments, IIRC.

Once upon a time zeners were king as reference diodes. The 1n829a,
for example: 6.2V, 5ppm/C.

Cheers,
James Arthur

 

[[email protected]]

The 10V reference chip is a better choice, but I once had a set of
precision zeners at either 1 or 2% increments, IIRC.

Once upon a time zeners were king as reference diodes. The 1n829a,
for example: 6.2V, 5ppm/C.

Only at 7.5mA - to get all the precision that you paid for with the !
N829, your needed a precision resistor biased from the reference
voltage to keep the current stable at just the right value. See

http://www.ortodoxism.ro/datasheets/motorola/1N823.pdf

and note that the temperature stability is specified at 7.5mA
+/-0.01mA - the better the diode the closer you've got to hold the
current to 7.5mA to maintain that stability.

 

[Don Lancaster]

Well, my knowledge of them was that you would always put a resistor
inline with them to give it a proper bias current so that you would
hit the exact right voltage. I need the 10V to connect to the non-
inverting input of an Op-amp (OP747). Thus I'm not worried about a
current drain on the line.

-Michael

In general, zeners make piss poor regulators.




--
Many thanks,

Don Lancaster voice phone: (928)428-4073
Synergetics 3860 West First Street Box 809 Thatcher, AZ 85552
rss: http://www.tinaja.com/whtnu.xml email: [email protected]

Please visit my GURU's LAIR web site at http://www.tinaja.com

 

[Spehro Pefhany]

Hey there - is there such a thing as a 1% zener diode? They all seem
to be about 5%. I was looking to make a 10V vref and the best I was
able to find was the Digi-Key EDZTE6110BCT-ND, which is accurate to
2.3%. This just strikes me as odd - and makes me wonder if I'm looking
in the wrong place or something.

Precision voltage reference zeners are pretty much a thing of the
past. You were limited to a certain voltage (around 6.2V) and forward
current (quite a few mA, IIRC) to get zero (nominal) tempco-- as
determined by the physics. ICs are much better for most applications
requiring much accuracy, and are reasonably priced.

I did, however, discover that there are actual 10V vref chips out
there - so I'm planning on going with one of those.

-Michael

The LM4040 is an inexpensive shunt (two active pins, just like a
zener) reference that might do your job, and it's available in 10V
with tolerancces of 2%, 1%, 0.5%, 0.2% and 0.1%, and spec'd at 150uA.
Tempco is nothing special (100 or 150ppm/K for all tolerances, but
still far better than what you'd get with an uncompensated 10V zener.

Of course if you can use a series reference that might save you a
part, but they tend to be a bit more expensive.


Best regards,
Spehro Pefhany

 

[Joerg]

Only at 7.5mA - to get all the precision that you paid for with the !
N829, your needed a precision resistor biased from the reference
voltage to keep the current stable at just the right value. See

http://www.ortodoxism.ro/datasheets/motorola/1N823.pdf

and note that the temperature stability is specified at 7.5mA
+/-0.01mA - the better the diode the closer you've got to hold the
current to 7.5mA to maintain that stability.

In them days I preferred the uA723 when I needed precision. Sometimes I
still do that

 

[[email protected]]

Precision voltage reference zeners are pretty much a thing of the
past. You were limited to a certain voltage (around 6.2V) and forward
current (quite a few mA, IIRC) to get zero (nominal) tempco-- as
determined by the physics. ICs are much better for most applications
requiring much accuracy, and are reasonably priced.



The LM4040 is an inexpensive shunt (two active pins, just like a
zener) reference that might do your job, and it's available in 10V
with tolerancces of 2%, 1%, 0.5%, 0.2% and 0.1%, and spec'd at 150uA.
Tempco is nothing special (100 or 150ppm/K for all tolerances, but
still far better than what you'd get with an uncompensated 10V zener.

At Cambridge Instruments, twenty years ago, we liked the Linear
Technology voltage references

http://www.ortodoxism.ro/datasheets/lineartechnology/1019fcs.pdf

5ppm/C and available with tight tolerances. Farnell doesn't seem to
stock the 10.00V part, but does have the 5.00V voltage part for about
$15 each in small quantities.

When I last looked, Thaler could do better, but their parts were
appreciably more expensive.

http://www.thaler.com/thcpdf/vre310.pdf

 

[Jamie]

In them days I preferred the uA723 when I needed precision. Sometimes I
still do that

I found the 723 seem to have a lot of failed units out there for
what ever reason?
THey just seem to fail and go into full output..
that's not very good if you don't have crowbar protection.

 

[Joerg]

I found the 723 seem to have a lot of failed units out there for
what ever reason?
THey just seem to fail and go into full output..
that's not very good if you don't have crowbar protection.

I usually do have a crowbar but I have never seen that phenomenon, and
it's been decades. What happened a lot in my teenage days was that chips
showed up from "re-labelers" and other shady sources, often at regular
electronics stores. Never seen a uA723 from there but I did end up with
a batch of uA741 that were totally out of spec. This is very frustrating
for a teenager who isn't a chip expert yet.

The 723 is alive and kicking, and very reliable. Look at any linear
brick supply from Condor or PowerOne. I just bought a few for a
prototype and sure enough they still use the good old 723. By default
they usually have a crowbar but not so much out of concern for the 723,
it's to protect against a pass transistor that shorts out.

 

[Rich Grise]

On Tue, 24 Jul 2007 06:10:10 -0700, the renowned Michael


Precision voltage reference zeners are pretty much a thing of the
past. You were limited to a certain voltage (around 6.2V) and forward
current (quite a few mA, IIRC) to get zero (nominal) tempco-- as
determined by the physics. ICs are much better for most applications
requiring much accuracy, and are reasonably priced.

I once read somewhere that you could put an ordinary forward-biased
silicon diode in series with a zener and their tempcos would tend to
cancel out, albeit they didn't say how well.

Cheers!
Rich

 

[Michael A. Terrell]

I usually do have a crowbar but I have never seen that phenomenon, and
it's been decades. What happened a lot in my teenage days was that chips
showed up from "re-labelers" and other shady sources, often at regular
electronics stores. Never seen a uA723 from there but I did end up with
a batch of uA741 that were totally out of spec. This is very frustrating
for a teenager who isn't a chip expert yet.

The 723 is alive and kicking, and very reliable. Look at any linear
brick supply from Condor or PowerOne. I just bought a few for a
prototype and sure enough they still use the good old 723. By default
they usually have a crowbar but not so much out of concern for the 723,
it's to protect against a pass transistor that shorts out.

The only 723 failures I can recall resulted in low or no output.
I've only seen a few bad ones, but loads of them in equipment and
modular power supplies.

i agree about the crowbar. Anyone who builds that type of supply
needs OVP.

--
Service to my country? Been there, Done that, and I've got my DD214 to
prove it.
Member of DAV #85.

Michael A. Terrell
Central Florida

 

[Winfield]

... I once had a set of precision zeners at either
1 or 2% increments, IIRC.

I also have a set of 1% zener diodes, but for real-world
use they're nearly useless, because they have to be used
at the exact specified test current. That's why I still
have 'em, never found a good way to use them!

 

[Jim Thompson]

I also have a set of 1% zener diodes, but for real-world
use they're nearly useless, because they have to be used
at the exact specified test current. That's why I still
have 'em, never found a good way to use them!

Use the zener to reference an OpAmp style current mirror when sources
current to the zener.

The "magic" is in how to get it to start ;-)

I built a chip long ago which did that using the E-B junction as the
zener. Not for much accuracy, but for significant PSRR in the bias
system.

...Jim Thompson

 

[D from BC]

TL431 ??

...Jim Thompson

That is an IC full of transistors.
D from BC