Zener Diode Dilemma

[jjk]

Zener Diode Dilemma

I purchased 50 1N4728 3.3v Fairchild zener diodes and trying to test them before installing in a circuit I'm building.

The diode test function on my meter shows 0.6v forward biased and .9v reversed biased. Performing the same tests on a 5.6v zener, I get 0.6v forward biased and when reversed biased I see the same reading as not connected to the diode at all - which is what I would expect.

Further, I inserted the diodes in a test circuit described below.

vcc -> 560 ohm resistor -> milliamp meter -> zener cathode, zener anode to ground.

When vcc voltage was applied, current was measured as shown below with the diode conducting all the time, and I could never get the voltage across thezener to reach 3.3v without smoking the resistor.

1.5v 0.1mA
2.0v 0.6mA
2.5v 1.1mA
2.7v 1.5mA
2.9v 1.7mA
3.0v 1.8mA

All of the 3.3v parts I tested exhibited the same behavior. Suspecting a bad lot of diodes I purchased a few 3.3v diodes made by NTE. They all too exhibit the same behavior - conducting well before their breakdown voltage.

Trying the above with a 5.6v zener shows no current flow until the voltage across the diode reached the device rated breakdown voltage. I also put a scope on the power supply to verify it was stable.

Any guidance as to what I may be doing wrong would be appreciated.

Thanks.

 

[Tom Biasi]

Zener Diode Dilemma

I purchased 50 1N4728 3.3v Fairchild zener diodes and trying to test them before installing in a circuit I'm building.

The diode test function on my meter shows 0.6v forward biased and .9v reversed biased. Performing the same tests on a 5.6v zener, I get 0.6v forward biased and when reversed biased I see the same reading as not connected to the diode at all - which is what I would expect.

Further, I inserted the diodes in a test circuit described below.

vcc -> 560 ohm resistor -> milliamp meter -> zener cathode, zener anode to ground.

When vcc voltage was applied, current was measured as shown below with the diode conducting all the time, and I could never get the voltage across the zener to reach 3.3v without smoking the resistor.

1.5v 0.1mA
2.0v 0.6mA
2.5v 1.1mA
2.7v 1.5mA
2.9v 1.7mA
3.0v 1.8mA

All of the 3.3v parts I tested exhibited the same behavior. Suspecting a bad lot of diodes I purchased a few 3.3v diodes made by NTE. They all too exhibit the same behavior - conducting well before their breakdown voltage.

Trying the above with a 5.6v zener shows no current flow until the voltage across the diode reached the device rated breakdown voltage. I also put a scope on the power supply to verify it was stable.

Any guidance as to what I may be doing wrong would be appreciated.

Thanks.

First, the diode function on your multimeter will not give you any
useful information about zeners.
Second, Your test looks strange. Is your supply current limiting perhaps?

 

[Fred McKenzie]

Zener Diode Dilemma

I purchased 50 1N4728 3.3v Fairchild zener diodes and trying to test them
before installing in a circuit I'm building.

The diode test function on my meter shows 0.6v forward biased and .9v
reversed biased. Performing the same tests on a 5.6v zener, I get 0.6v
forward biased and when reversed biased I see the same reading as not
connected to the diode at all - which is what I would expect.

Further, I inserted the diodes in a test circuit described below.

vcc -> 560 ohm resistor -> milliamp meter -> zener cathode, zener anode to
ground.

When vcc voltage was applied, current was measured as shown below with the
diode conducting all the time, and I could never get the voltage across the
zener to reach 3.3v without smoking the resistor.

1.5v 0.1mA
2.0v 0.6mA
2.5v 1.1mA
2.7v 1.5mA
2.9v 1.7mA
3.0v 1.8mA

All of the 3.3v parts I tested exhibited the same behavior. Suspecting a bad
lot of diodes I purchased a few 3.3v diodes made by NTE. They all too exhibit
the same behavior - conducting well before their breakdown voltage.

Trying the above with a 5.6v zener shows no current flow until the voltage
across the diode reached the device rated breakdown voltage. I also put a
scope on the power supply to verify it was stable.

JJK-

Plotting your Voltage vs current data suggests you may be measuring the
forward-bias curve of several diode junctions connected in series. (It
looks more like a constant-Voltage curve than constant-current.)

The only 3.3 Volt Zener in my collection is a 1N5575A. It checks good
as a diode on the diode ranges of a Fluke 8020A multimeter. I did not
run your current vs Voltage test, but would expect it to behave as well
as your 5.6 Volt test.

I wonder if both your source of the 1N4728 diode and NTE, get their
diodes from the same factory? There could have been a batch that were
mis-labeled, or they could have fallen victim to counterfeit parts.

Fred

 

[Tom Biasi]

JJK-

Plotting your Voltage vs current data suggests you may be measuring the
forward-bias curve of several diode junctions connected in series. (It
looks more like a constant-Voltage curve than constant-current.)

The only 3.3 Volt Zener in my collection is a 1N5575A. It checks good
as a diode on the diode ranges of a Fluke 8020A multimeter. I did not
run your current vs Voltage test, but would expect it to behave as well
as your 5.6 Volt test.

I wonder if both your source of the 1N4728 diode and NTE, get their
diodes from the same factory? There could have been a batch that were
mis-labeled, or they could have fallen victim to counterfeit parts.

Fred

The curve is linear until he gets to 2.7 volts then there is a bump
(reading error?). This looks like a forward conducting diode and the
readings are across the resistor.

 

[jjk]

The curve is linear until he gets to 2.7 volts then there is a bump

(reading error?). This looks like a forward conducting diode and the

readings are across the resistor.

Thanks to all who replied.
There is definitely very wrong. I just took delivery of 50 Fairchild replacement zeners and reran my tests to verify, this time using a 330 ohm resistor. The results are below for both 3.3v and 5.6v parts. The 5.6v part worksas expected and doesn't conduct until the avalanche voltage is approached whereas the 3.3v part is still exhibiting the wrong behavior. I had to raise vcc to 25v to see 3.3v at the cathode, then the magic smoke came out of the resistor.

vcc -> 330 ohm resistor -> milliamp meter -> zener cathode, zener anode to ground.
330 ohm resistor measures 327 ohms

1N4728A 3.3v 1W Zener diode

Iz
vcc (mA) Vzener
1.5 0.2 1.42
2.0 0.8 1.72
2.5 1.7 1.91
3.0 2.8 2.05
3.5 4.1 2.16
4.0 5.3 2.24
4.5 6.6 2.3
5.0 8.0 2.37
5.5 9.4 2.43
6.0 10.8 2.48
7.0 13.6 2.56
8.0 16.5 2.62
9.0 19.4 2.69
10.0 22.4 2.73
20.0 55.2 3.1
25.0 71.3 3.3


1N752 5.6v zener diode

Iz
vcc (mA) Vzener
1.5 0 1.5
.... ... ...
4.0 0 4.0
4.5 0 4.49
5.0 .1 5.0
5.5 .2 5.43
6.0 1.0 5.67

 

[JW]

Thanks to all who replied.
There is definitely very wrong. I just took delivery of 50 Fairchild replacement zeners and reran my tests to verify, this time using a 330 ohm resistor. The results are below for both 3.3v and 5.6v parts. The 5.6v part works as expected and doesn't conduct until the avalanche voltage is approached whereas the 3.3v part is still exhibiting the wrong behavior. I had to raise vcc to 25v to see 3.3v at the cathode, then the magic smoke came out of the resistor.

vcc -> 330 ohm resistor -> milliamp meter -> zener cathode, zener anode to ground.
330 ohm resistor measures 327 ohms

1N4728A 3.3v 1W Zener diode

Iz
vcc (mA) Vzener
1.5 0.2 1.42
2.0 0.8 1.72
2.5 1.7 1.91
3.0 2.8 2.05
3.5 4.1 2.16
4.0 5.3 2.24
4.5 6.6 2.3
5.0 8.0 2.37
5.5 9.4 2.43
6.0 10.8 2.48
7.0 13.6 2.56
8.0 16.5 2.62
9.0 19.4 2.69
10.0 22.4 2.73
20.0 55.2 3.1
25.0 71.3 3.3

Where are you getting these diodes?

 

[Tom Biasi]

Thanks to all who replied.
There is definitely very wrong. I just took delivery of 50 Fairchild replacement zeners and reran my tests to verify, this time using a 330 ohm resistor. The results are below for both 3.3v and 5.6v parts. The 5.6v part works as expected and doesn't conduct until the avalanche voltage is approached whereas the 3.3v part is still exhibiting the wrong behavior. I had to raise vcc to 25v to see 3.3v at the cathode, then the magic smoke came out of the resistor.

vcc -> 330 ohm resistor -> milliamp meter -> zener cathode, zener anode to ground.
330 ohm resistor measures 327 ohms

1N4728A 3.3v 1W Zener diode

Iz
vcc (mA) Vzener
1.5 0.2 1.42
2.0 0.8 1.72
2.5 1.7 1.91
3.0 2.8 2.05
3.5 4.1 2.16
4.0 5.3 2.24
4.5 6.6 2.3
5.0 8.0 2.37
5.5 9.4 2.43
6.0 10.8 2.48
7.0 13.6 2.56
8.0 16.5 2.62
9.0 19.4 2.69
10.0 22.4 2.73
20.0 55.2 3.1
25.0 71.3 3.3


1N752 5.6v zener diode

Iz
vcc (mA) Vzener
1.5 0 1.5
... ... ...
4.0 0 4.0
4.5 0 4.49
5.0 .1 5.0
5.5 .2 5.43
6.0 1.0 5.67

Do you have a different power supply to try?

 

[jjk]

Where are you getting these diodes?

I received a package of 50 Fairchild devices from Jameco.com.
Then I purchased 5 NTE devices from a local electronics store.
All of the ones I have tested from the above failed.
I contacted Jameco and they sent 50 replacements and all that I tested failed in the same way. I've been in contact with the Jameco technical rep and they are looking into this and trying to obtain devices from another lot from their supplier.

 

[jjk]

On 10/20/2013 2:04 PM, Fred McKenzie wrote:




























Do you have a different power supply to try?

I also tried a 9v battery connected to a 100K pot, wiper connected to a 10kresistor to the zener cathode, and the anode connected to the other end ofthe pot and to the battery negative post.

+9v
|
100K pot ---10K ----cathode
| |
gnd------------------|

Works fine with all of my 5.x zeners.
The 3.3v zeners never get above 1.8v at the cathode when the pot wiper is aat 9.2v.

This feels like a bad dream where you run and not get anywhere.
I sure would like to get this behind me and continue with the larger project I'm working on.

 

[Fred McKenzie]

I also tried a 9v battery connected to a 100K pot, wiper connected to a 10k
resistor to the zener cathode, and the anode connected to the other end of
the pot and to the battery negative post.

+9v
|
100K pot ---10K ----cathode
| |
gnd------------------|

Works fine with all of my 5.x zeners.
The 3.3v zeners never get above 1.8v at the cathode when the pot wiper is a
at 9.2v.

This feels like a bad dream where you run and not get anywhere.
I sure would like to get this behind me and continue with the larger project
I'm working on.

JJK-

I did not see how you were using the 3.3 V Zeners. If you are
generating 3.3 Volt power, you might consider using the LM-317. They
are available in several packages including surface mount, and can be
adjusted from 1.2 to 39 Volts using a pair of resistors.

Fred

 

[Tom Biasi]

I also tried a 9v battery connected to a 100K pot, wiper connected to a 10k resistor to the zener cathode, and the anode connected to the other end of the pot and to the battery negative post.

+9v
|
100K pot ---10K ----cathode
| |
gnd------------------|

Works fine with all of my 5.x zeners.
The 3.3v zeners never get above 1.8v at the cathode when the pot wiper is a at 9.2v.

This feels like a bad dream where you run and not get anywhere.
I sure would like to get this behind me and continue with the larger project I'm working on.

It is not likely that you are receiving so many defective parts. I
suggested a different power supply, now I suggest a different meter.
I do this only to eliminate the possible variables. The next is to
eliminate you. I mean give them to someone else to test. Please don't be
insulted.

Tom

 

[jjk]

jjk wrote:













JJK-



I did not see how you were using the 3.3 V Zeners. If you are

generating 3.3 Volt power, you might consider using the LM-317. They

are available in several packages including surface mount, and can be

adjusted from 1.2 to 39 Volts using a pair of resistors.



Fred

Thanks for your comments. The potentiometer circuit I described was for testing an alternate power supply source to rule out a problem with my bench supply.
I intend to use the zener for over voltage pin protection on 3.3v devices.

 

[jjk]

On 10/22/2013 8:59 AM, jjk wrote:

On Sunday, October 20, 2013 2:27:14 PM UTC-4, Tom Biasi wrote:

On 10/20/2013 2:04 PM, Fred McKenzie wrote:








It is not likely that you are receiving so many defective parts. I

suggested a different power supply, now I suggest a different meter.

I do this only to eliminate the possible variables. The next is to

eliminate you. I mean give them to someone else to test. Please don't be

insulted.



Tom

Tom, LOL. No insult taken. I got a chuckle from your process of eliminationcomment.

I agree having that many failures is suspect and is why I posted to this group for possible help. I tried 3 meters and 2 power supplies. In every caseall of my 5.* volt zeners work perfectly as expected. All I do is swap outthe 5v zener for the 3.3v device and see the problem. I even reversed thedirection of the 3.3v zener thinking the cathode may have been marked incorrectly. The voltage at the anode in this case is an expected 0.6v.

If anyone reading this group would like to test a few of these, please sendme an email at jjk439 at gmail dot com and I would be glad to send them toyou.

 

[jjk]

On 10/23/2013 6:52 PM, jjk wrote:

On 10/22/2013 8:59 AM, jjk wrote:

On 10/20/2013 2:04 PM, Fred McKenzie wrote:

Tom, LOL. No insult taken. I got a chuckle from your process of elimination comment.



I agree having that many failures is suspect and is why I posted to this group for possible help. I tried 3 meters and 2 power supplies. In every case all of my 5.* volt zeners work perfectly as expected. All I do is swap out the 5v zener for the 3.3v device and see the problem. I even reversed the direction of the 3.3v zener thinking the cathode may have been marked incorrectly. The voltage at the anode in this case is an expected 0.6v.



If anyone reading this group would like to test a few of these, please send me an email at jjk439 at gmail dot com and I would be glad to send them to you.

I am now relieved that my sanity on this issue is intact, but still frustrated at not being able to find any functional zener diodes. I received a phone call from a tech at Jameco Electronics stating they took delivery of another 50 3.3v zener diodes, and they all failed their tests the same way they failed mine.
They concur there is a problem with these devices.

Thanks to all who replied to my post.

 

[Tom Biasi]

I am now relieved that my sanity on this issue is intact, but still frustrated at not being able to find any functional zener diodes. I received a phone
call from a tech at Jameco Electronics stating they took delivery of another 50 3.3v zener diodes, and they all failed their tests the same way they
failed mine.
They concur there is a problem with these devices.

Thanks to all who replied to my post.

I guess Jameco doesn't have any incoming QA inspection.

 

[Andrew Gabriel]

There is definitely very wrong. I just took delivery of 50 Fairchild replacement zeners and reran my tests to verify, this time using a 330 ohm resistor. The results are below for both 3.3v and 5.6v parts. The 5.6v part works as expected and doesn't conduct until the avalanche voltage is approached whereas the 3.3v part is still exhibiting the wrong behavior. I had to raise vcc to 25v to see 3.3v at the cathode, then the magic smoke came out of the resistor.
vcc -> 330 ohm resistor -> milliamp meter -> zener cathode, zener anode to ground.
330 ohm resistor measures 327 ohms
1N4728A 3.3v 1W Zener diode
Iz
vcc (mA) Vzener
1.5 0.2 1.42
2.0 0.8 1.72
2.5 1.7 1.91
3.0 2.8 2.05
3.5 4.1 2.16
4.0 5.3 2.24
4.5 6.6 2.3
5.0 8.0 2.37
5.5 9.4 2.43
6.0 10.8 2.48
7.0 13.6 2.56
8.0 16.5 2.62
9.0 19.4 2.69
10.0 22.4 2.73
20.0 55.2 3.1
25.0 71.3 3.3

This exactly matches the datasheet, which says it needs at least 76mA
in order to act as a zener voltage reference.

This is a common issue with low voltage zeners - the current they
need to operate is high (compared with higher voltage zeners).

If you want a low voltage reference at low current (e.g. for battery
operated equipment), an LED can often be used (forward biased) at a
much lower operating current (1mA or less), although not quite as
stable a reference.

 

[jjk]

There is definitely very wrong. I just took delivery of 50 Fairchild replacement zeners and reran my tests to verify, this time using a 330 ohm resistor. The results are below for both 3.3v and 5.6v parts. The 5.6v part works as expected and doesn't conduct until the avalanche voltage is approached whereas the 3.3v part is still exhibiting the wrong behavior. I had to raise vcc to 25v to see 3.3v at the cathode, then the magic smoke came out of the resistor.

vcc -> 330 ohm resistor -> milliamp meter -> zener cathode, zener anodeto ground.

330 ohm resistor measures 327 ohms

1N4728A 3.3v 1W Zener diode

vcc (mA) Vzener

1.5 0.2 1.42

2.0 0.8 1.72

2.5 1.7 1.91

3.0 2.8 2.05

3.5 4.1 2.16

4.0 5.3 2.24

4.5 6.6 2.3

5.0 8.0 2.37

5.5 9.4 2.43

6.0 10.8 2.48

7.0 13.6 2.56

8.0 16.5 2.62

9.0 19.4 2.69

10.0 22.4 2.73

20.0 55.2 3.1

25.0 71.3 3.3

This exactly matches the datasheet, which says it needs at least 76mA

in order to act as a zener voltage reference.



This is a common issue with low voltage zeners - the current they

need to operate is high (compared with higher voltage zeners).



If you want a low voltage reference at low current (e.g. for battery

operated equipment), an LED can often be used (forward biased) at a

much lower operating current (1mA or less), although not quite as

stable a reference.



--

Andrew Gabriel

[email address is not usable -- followup in the newsgroup]

Thanks for your comments Andrew.
I'm wondering why the zener would conduct at all at voltages as low as 1.4v?
My goal was not to use the zener as a regulator, but to protect 3.3v devices from potentially external over voltage events (microcontroller driving 5vcircuits).

 

[[email protected]]

I guess Jameco doesn't have any incoming QA inspection.

Doesn't Jameco still sell floor sweepings? I sure wouldn't trust any
component bought from them.

 

[Tom Biasi]

Doesn't Jameco still sell floor sweepings? I sure wouldn't trust any
component bought from them.

I don't know their practices.

 

[Andrew Gabriel]

Thanks for your comments Andrew.
I'm wondering why the zener would conduct at all at voltages as low as 1.4v?

Zener voltage regulation happens in reverse breakdown mode.
All zeners leak reverse current at voltages below the breakdown voltage.
With low voltage zeners, this leakage current is very high, up to 76mA
for this part number. This makes low voltage zeners unsuitable for
many applications, where you might expect a theoretically perfect
zener to work.

My goal was not to use the zener as a regulator, but to protect 3.3v devices from potentially external over voltage events (microcontroller driving 5v circuits).

If the 50mA or so leakage at 3V is an unacceptable issue for you,
you might instead consider a potential divider across the 3V supply,
with the centerpoint connected to an SCR gate to crowbar the supply.
Adjust the potential divider ratio to give the right tripping voltage.
A red led (forward biased) added at the top of the potential divider
might make it more accurate/sensitive.

I haven't actually tried this at such a low voltage, but I suspect it
may work better than a zener.