high-capacitance zener diode

W

Winfield Hill

Guest
I've often pointed out that zener diodes are NOT slow devices,
and in fact are very fast acting when a breakdown voltage is
presented, sub-ns in fact. But zener diodes they do have high
capacitance, which could lead to an assumption of slow operation.

How high? I just measured the Crydom 5KP6.0A Transient Voltage
Suppression (TVS) zener diode. This is a 6V TVS that starts
breaking down at about 7V (50mA) and limits to 10.3V with 485A
of current flowing. It's not a very large part (about 0.3" dia
by 0.3" long, with 0.050"-thick axial leads), but it can handle
5000W peak during a single 10us risetime, 1000us falltime pulse.

Crydom's curves go off the scale at 100,000pF, and don't show
the capacitance for diode voltages below 12V. So I measured a
5KP6's 0V capacitance and saw 1.6uF. :>) That's pretty high!

Thanks,
- Win

whill_at_picovolt-dot-com
 
On 15 Mar 2004 08:52:28 -0800, the renowned Winfield Hill
<Winfield_member@newsguy.com> wrote:

I've often pointed out that zener diodes are NOT slow devices,
and in fact are very fast acting when a breakdown voltage is
presented, sub-ns in fact. But zener diodes they do have high
capacitance, which could lead to an assumption of slow operation.

How high? I just measured the Crydom 5KP6.0A Transient Voltage
Suppression (TVS) zener diode. This is a 6V TVS that starts
breaking down at about 7V (50mA) and limits to 10.3V with 485A
of current flowing. It's not a very large part (about 0.3" dia
by 0.3" long, with 0.050"-thick axial leads), but it can handle
5000W peak during a single 10us risetime, 1000us falltime pulse.

Crydom's curves go off the scale at 100,000pF, and don't show
the capacitance for diode voltages below 12V. So I measured a
5KP6's 0V capacitance and saw 1.6uF. :>) That's pretty high!
Click to expand...
Youch!
http://www.littelfuse.com/PDFs/Products/5KP.pdf Fig 3.

The curves go up to 100,000pF for an ~7.5V device with zero bias.

It's a log-log graph..

Best regards,
Spehro Pefhany
--
"it's the network..." "The Journey is the reward"
speff@interlog.com Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog Info for designers: http://www.speff.com
 
Spehro Pefhany wrote...
On 15 Mar 2004, the renowned Winfield Hill wrote:

I've often pointed out that zener diodes are NOT slow devices,
and in fact are very fast acting when a breakdown voltage is
presented, sub-ns in fact. But zener diodes they do have high
capacitance, which could lead to an assumption of slow operation.

How high? I just measured the Crydom 5KP6.0A Transient Voltage
Suppression (TVS) zener diode. This is a 6V TVS that starts
breaking down at about 7V (50mA) and limits to 10.3V with 485A
of current flowing. It's not a very large part (about 0.3" dia
by 0.3" long, with 0.050"-thick axial leads), but it can handle
5000W peak during a single 10us risetime, 1000us falltime pulse.

Crydom's curves go off the scale at 100,000pF, and don't show
the capacitance for diode voltages below 12V. So I measured a
5KP6's 0V capacitance and saw 1.6uF. :>) That's pretty high!

Youch!
http://www.littelfuse.com/PDFs/Products/5KP.pdf Fig 3.
The curves go up to 100,000pF for an ~7.5V device with zero bias.
It's a log-log graph..
Click to expand...
Yep. That's a similar curve to one in the Crydom the data sheet.
It appears the 0V curve is increasing its slope as it goes off
the graph at 0.1uF for roughly 8V devices. Whatever, I measured
1.6uF for the Crydom device. If one starts at 0V, a 1A current
only has a 600mV/us positive slew rate.

Thanks,
- Win

whill_at_picovolt-dot-com
 
On 15 Mar 2004 08:52:28 -0800, Winfield Hill
<Winfield_member@newsguy.com> wrote:

I've often pointed out that zener diodes are NOT slow devices,
and in fact are very fast acting when a breakdown voltage is
presented, sub-ns in fact. But zener diodes they do have high
capacitance, which could lead to an assumption of slow operation.

How high? I just measured the Crydom 5KP6.0A Transient Voltage
Suppression (TVS) zener diode. This is a 6V TVS that starts
breaking down at about 7V (50mA) and limits to 10.3V with 485A
of current flowing. It's not a very large part (about 0.3" dia
by 0.3" long, with 0.050"-thick axial leads), but it can handle
5000W peak during a single 10us risetime, 1000us falltime pulse.

Crydom's curves go off the scale at 100,000pF, and don't show
the capacitance for diode voltages below 12V. So I measured a
5KP6's 0V capacitance and saw 1.6uF. :>) That's pretty high!

Thanks,
- Win

whill_at_picovolt-dot-com
Click to expand...

The only zorb I have handy is a little bidirectional 600 watt 6.8 volt
puppy from Digikey, their number P6KE6.8ADICT-ND (the "ND" purports to
mean "No Discount") which measures 2800 pF on both my AADE meter and
on a Boonton 72, zero bias.

1.6 uF is impressive, especially if you plan to use this to clamp
signal lines. If you knew the junction area, you could calculate the
effective plate separation.

A *very* low capacitance zener might be handy at times.

John
 
On Mon, 15 Mar 2004 10:17:59 -0800, John Larkin
<jjlarkin@highSNIPlandTHIStechPLEASEnology.com> wrote:

(snip)
A *very* low capacitance zener might be handy at times.

John
Click to expand...
If you are thinking about protection circuits for high frequency
signals, this is typically implemented by placing a low capacitance
rectifier in series with the zener or tranzorb.


================================

Greg Neff
VP Engineering
*Microsym* Computers Inc.
greg@guesswhichwordgoeshere.com
 
I read in sci.electronics.design that John Larkin <jjlarkin@highSNIPland
THIStechPLEASEnology.com> wrote (in <5bsb505e3278lum7ihpaa4ll1l17gc2c6m@
4ax.com>) about 'high-capacitance zener diode', on Mon, 15 Mar 2004:

1.6 uF is impressive, especially if you plan to use this to clamp signal
lines. If you knew the junction area, you could calculate the effective
plate separation.
Click to expand...
Feerguy could use a million to power his electric car!
--
Regards, John Woodgate, OOO - Own Opinions Only.
The good news is that nothing is compulsory.
The bad news is that everything is prohibited.
http://www.jmwa.demon.co.uk Also see http://www.isce.org.uk
 
Winfield Hill wrote:
Spehro Pefhany wrote...

On 15 Mar 2004, the renowned Winfield Hill wrote:


I've often pointed out that zener diodes are NOT slow devices,
and in fact are very fast acting when a breakdown voltage is
presented, sub-ns in fact. But zener diodes they do have high
capacitance, which could lead to an assumption of slow operation.

How high? I just measured the Crydom 5KP6.0A Transient Voltage
Suppression (TVS) zener diode. This is a 6V TVS that starts
breaking down at about 7V (50mA) and limits to 10.3V with 485A
of current flowing. It's not a very large part (about 0.3" dia
by 0.3" long, with 0.050"-thick axial leads), but it can handle
5000W peak during a single 10us risetime, 1000us falltime pulse.

Crydom's curves go off the scale at 100,000pF, and don't show
the capacitance for diode voltages below 12V. So I measured a
5KP6's 0V capacitance and saw 1.6uF. :>) That's pretty high!

Youch!
http://www.littelfuse.com/PDFs/Products/5KP.pdf Fig 3.
The curves go up to 100,000pF for an ~7.5V device with zero bias.
It's a log-log graph..


Yep. That's a similar curve to one in the Crydom the data sheet.
It appears the 0V curve is increasing its slope as it goes off
the graph at 0.1uF for roughly 8V devices. Whatever, I measured
1.6uF for the Crydom device. If one starts at 0V, a 1A current
only has a 600mV/us positive slew rate.

Thanks,
- Win

whill_at_picovolt-dot-com
Click to expand...
is some kind of miller effect going on?
can you get the energy back?


--
local optimization seldom leads to global optimization

my e-mail address is: <my first name> <my last name> AT mmm DOT com
 
On Mon, 15 Mar 2004 13:45:16 -0500, Greg Neff <greg@microsym.com>
wrote:

On Mon, 15 Mar 2004 10:17:59 -0800, John Larkin
jjlarkin@highSNIPlandTHIStechPLEASEnology.com> wrote:

(snip)

A *very* low capacitance zener might be handy at times.

John


If you are thinking about protection circuits for high frequency
signals, this is typically implemented by placing a low capacitance
rectifier in series with the zener or tranzorb.
Click to expand...

And DC bias the junction quiescently?

John
 
"John Woodgate" <jmw@jmwa.demon.contraspam.yuk> schreef in bericht
news:TALRGhJvNgVAFww1@jmwa.demon.co.uk...
I read in sci.electronics.design that John Larkin <jjlarkin@highSNIPland
THIStechPLEASEnology.com> wrote (in <5bsb505e3278lum7ihpaa4ll1l17gc2c6m@
4ax.com>) about 'high-capacitance zener diode', on Mon, 15 Mar 2004:

1.6 uF is impressive, especially if you plan to use this to clamp signal
lines. If you knew the junction area, you could calculate the effective
plate separation.

Feerguy could use a million to power his electric car!
Click to expand...
You snipped the bit from Win, that he measured it at 0V.
Now we need a motor that delivers some torque at 0V.

;-)


--
Thanks, Frank.
(remove 'x' and 'invalid' when replying by email)
 
I read in sci.electronics.design that Frank Bemelman <f.bemelmanx@planet
..invalid.nl> wrote (in <40560807$0$90733$c3e8da3@news.astraweb.com>)
about 'high-capacitance zener diode', on Mon, 15 Mar 2004:
"John Woodgate" <jmw@jmwa.demon.contraspam.yuk> schreef in bericht
news:TALRGhJvNgVAFww1@jmwa.demon.co.uk...
I read in sci.electronics.design that John Larkin <jjlarkin@highSNIPland
THIStechPLEASEnology.com> wrote (in <5bsb505e3278lum7ihpaa4ll1l17gc2c6m@
4ax.com>) about 'high-capacitance zener diode', on Mon, 15 Mar 2004:

1.6 uF is impressive, especially if you plan to use this to clamp signal
lines. If you knew the junction area, you could calculate the effective
plate separation.

Feerguy could use a million to power his electric car!

You snipped the bit from Win, that he measured it at 0V.
Now we need a motor that delivers some torque at 0V.

Left as an exercise for the student.
Click to expand...
--
Regards, John Woodgate, OOO - Own Opinions Only.
The good news is that nothing is compulsory.
The bad news is that everything is prohibited.
http://www.jmwa.demon.co.uk Also see http://www.isce.org.uk
 
On Mon, 15 Mar 2004 11:42:59 -0800, John Larkin
<jjlarkin@highSNIPlandTHIStechPLEASEnology.com> wrote:

On Mon, 15 Mar 2004 13:45:16 -0500, Greg Neff <greg@microsym.com
wrote:

On Mon, 15 Mar 2004 10:17:59 -0800, John Larkin
jjlarkin@highSNIPlandTHIStechPLEASEnology.com> wrote:

(snip)

A *very* low capacitance zener might be handy at times.

John


If you are thinking about protection circuits for high frequency
signals, this is typically implemented by placing a low capacitance
rectifier in series with the zener or tranzorb.



And DC bias the junction quiescently?

John
Click to expand...
I like to if I build a discrete solution, but most integrated
tranzorb+rectifier offerings don't provide for this. For examples
check out low-capacitance products offered by:

http://www.protekdevices.com/dave_application.htm
http://www.semtech.com/products/



================================

Greg Neff
VP Engineering
*Microsym* Computers Inc.
greg@guesswhichwordgoeshere.com
 
On Mon, 15 Mar 2004 11:42:59 -0800, the renowned John Larkin
<jjlarkin@highSNIPlandTHIStechPLEASEnology.com> wrote:

On Mon, 15 Mar 2004 13:45:16 -0500, Greg Neff <greg@microsym.com
wrote:

On Mon, 15 Mar 2004 10:17:59 -0800, John Larkin
jjlarkin@highSNIPlandTHIStechPLEASEnology.com> wrote:

(snip)

A *very* low capacitance zener might be handy at times.

John


If you are thinking about protection circuits for high frequency
signals, this is typically implemented by placing a low capacitance
rectifier in series with the zener or tranzorb.



And DC bias the junction quiescently?
John
Click to expand...
I've seen this circuit used (and am currently looking at similar
issues for a bus protection circuit). Seems like it might thoroughly
booger up the first couple of edges after the bus has been sitting
idle.

Best regards,
Spehro Pefhany
--
"it's the network..." "The Journey is the reward"
speff@interlog.com Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog Info for designers: http://www.speff.com
 
"John Woodgate" <jmw@jmwa.demon.contraspam.yuk> schreef in bericht
news:VgkQ+GKRygVAFwT+@jmwa.demon.co.uk...
I read in sci.electronics.design that Frank Bemelman <f.bemelmanx@planet
.invalid.nl> wrote (in <40560807$0$90733$c3e8da3@news.astraweb.com>)
about 'high-capacitance zener diode', on Mon, 15 Mar 2004:
"John Woodgate" <jmw@jmwa.demon.contraspam.yuk> schreef in bericht
news:TALRGhJvNgVAFww1@jmwa.demon.co.uk...
I read in sci.electronics.design that John Larkin
jjlarkin@highSNIPland
THIStechPLEASEnology.com> wrote (in
5bsb505e3278lum7ihpaa4ll1l17gc2c6m@
4ax.com>) about 'high-capacitance zener diode', on Mon, 15 Mar 2004:

1.6 uF is impressive, especially if you plan to use this to clamp
signal
lines. If you knew the junction area, you could calculate the
effective
plate separation.

Feerguy could use a million to power his electric car!

You snipped the bit from Win, that he measured it at 0V.
Now we need a motor that delivers some torque at 0V.

Left as an exercise for the student.
Click to expand...
I *knew* you had an answer ;)

--
Thanks, Frank.
(remove 'x' and 'invalid' when replying by email)
 
On Mon, 15 Mar 2004 20:46:19 +0100, "Frank Bemelman"
<f.bemelmanx@planet.invalid.nl> wrote:


Now we need a motor that delivers some torque at 0V.

;-)
Click to expand...
There's absolutely nothing keeping you from making such a motor.

John
 
"John Larkin" <jjlarkin@highSNIPlandTHIStechPLEASEnology.com> a écrit dans
le message news: 2m7c50l63ob7t3v3ic7smu96p9u8o9e1i4@4ax.com...
On Mon, 15 Mar 2004 20:46:19 +0100, "Frank Bemelman"
f.bemelmanx@planet.invalid.nl> wrote:


Now we need a motor that delivers some torque at 0V.

;-)

There's absolutely nothing keeping you from making such a motor.

John
Click to expand...
Right. There's (almost) no pb having torque at 0V. Having torque and speed
is another matter.

Thanks,
Fred.
 
One can only imagine what the output waveform might be like after feeding a
resistor in series with one of those things to ground, with a few volts of
sinewave at some frequency where the RC is significant.

--
Regards,

Adrian Jansen
J & K MicroSystems
Microcomputer solutions for industrial control
 
"Spehro Pefhany" <speffSNIP@interlogDOTyou.knowwhat> wrote in message
news:5d4c50l0c97jueo06nbgh0mllva5ane7fc@4ax.com...
On Mon, 15 Mar 2004 11:42:59 -0800, the renowned John Larkin
jjlarkin@highSNIPlandTHIStechPLEASEnology.com> wrote:

On Mon, 15 Mar 2004 13:45:16 -0500, Greg Neff <greg@microsym.com
wrote:

On Mon, 15 Mar 2004 10:17:59 -0800, John Larkin
jjlarkin@highSNIPlandTHIStechPLEASEnology.com> wrote:

(snip)

A *very* low capacitance zener might be handy at times.

John


If you are thinking about protection circuits for high frequency
signals, this is typically implemented by placing a low capacitance
rectifier in series with the zener or tranzorb.



And DC bias the junction quiescently?
John

I've seen this circuit used (and am currently looking at similar
issues for a bus protection circuit). Seems like it might thoroughly
booger up the first couple of edges after the bus has been sitting
idle.

Best regards,
Spehro Pefhany
Click to expand...
yep, can and does. If your comms protocol has some preamble it may not be an
issue though. some USB transorbs do have the TS pinned out (eg USB6B1) to
allow for dc bias.

I shot myself in the foot once with an IGBT gate driver, using a similar
type of transorb. In that case, I changed to a lower capacitance transorb -
with 600nF or so of gate capacitance (nasty powerex/bitsashitty 1200V 600A
IGBT) I could put up with quite a bit of stray C, but it knocked the hell
out Tr, Tf when driving a much smaller IGBT with correspondingly larger Rg.

the problem with a diode + dc bias (eg 100k) is finding a diode that is also
OK with 500A current pulses. this sort of problem is a lot easier to solve
when say protecting a USB port.
 
On Wed, 17 Mar 2004 18:55:45 -0800, the renowned "Terry Given"
<the_domes@xtra.co.nz> wrote:
yep, can and does. If your comms protocol has some preamble it may not be an
issue though. some USB transorbs do have the TS pinned out (eg USB6B1) to
allow for dc bias.
Click to expand...
With differential (eg. RS-485) it looks subtle- on the bench and with
no common mode difference between the signals the diodes will allow
the reciever to get into the normal signal range with no problem. Not
sure that still follows under more extreme conditions.

I shot myself in the foot once with an IGBT gate driver, using a similar
type of transorb. In that case, I changed to a lower capacitance transorb -
with 600nF or so of gate capacitance (nasty powerex/bitsashitty 1200V 600A
IGBT) I could put up with quite a bit of stray C, but it knocked the hell
out Tr, Tf when driving a much smaller IGBT with correspondingly larger Rg.

the problem with a diode + dc bias (eg 100k) is finding a diode that is also
OK with 500A current pulses. this sort of problem is a lot easier to solve
when say protecting a USB port.
Click to expand...
Yes, though USB2.x can get pretty fast compared to RS-485, which is
closer to audio than UHF. A cruddy old slow rectifier diode is quite
rugged.


Best regards,
Spehro Pefhany
--
"it's the network..." "The Journey is the reward"
speff@interlog.com Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog Info for designers: http://www.speff.com
 
"Spehro Pefhany" <speffSNIP@interlogDOTyou.knowwhat> wrote in message
news:95kh5059bjeehgfnlanvsso1r2d66s6b2n@4ax.com...
On Wed, 17 Mar 2004 18:55:45 -0800, the renowned "Terry Given"
the_domes@xtra.co.nz> wrote:

yep, can and does. If your comms protocol has some preamble it may not be
an
issue though. some USB transorbs do have the TS pinned out (eg USB6B1) to
allow for dc bias.

With differential (eg. RS-485) it looks subtle- on the bench and with
no common mode difference between the signals the diodes will allow
the reciever to get into the normal signal range with no problem. Not
sure that still follows under more extreme conditions.

I shot myself in the foot once with an IGBT gate driver, using a similar
type of transorb. In that case, I changed to a lower capacitance
transorb -
with 600nF or so of gate capacitance (nasty powerex/bitsashitty 1200V
600A
IGBT) I could put up with quite a bit of stray C, but it knocked the hell
out Tr, Tf when driving a much smaller IGBT with correspondingly larger
Rg.

the problem with a diode + dc bias (eg 100k) is finding a diode that is
also
OK with 500A current pulses. this sort of problem is a lot easier to
solve
when say protecting a USB port.

Yes, though USB2.x can get pretty fast compared to RS-485, which is
closer to audio than UHF. A cruddy old slow rectifier diode is quite
rugged.


Best regards,
Spehro Pefhany
--
Click to expand...
if youre mad enough, RS485 works at 10Mbps (and I use USB transorbs like the
USB6B1.....)

if youre stupid too, you'll roll your own protocol, thats edge triggered
(dont blame me, I just made the stupid thing go)

Regards,
Terry
 
You must log in or register to reply here.

Welcome to EDABoard.com

Sponsor

Back
Top