which diode for 55 Khz switching supply??

[albert]

Good Day All.

I am building a cockcroft-walton voltage multiplier. It will have 10
stages, with each stage having a 100 volt output, so the total output
is ~1000 volts. It will operate at 55 Khz and only needs to produce
small currents (microamps). I want to have minimal switching diode
losses since I will have lots of diodes:>:

I need a fast diode with small input capacitance. I need to have it
withstand 200 volts at 100 microamps.

Which diode should I use??????

All suggestions are appreciated.

Al

 

[John Larkin]

Good Day All.

I am building a cockcroft-walton voltage multiplier. It will have 10
stages, with each stage having a 100 volt output, so the total output
is ~1000 volts. It will operate at 55 Khz and only needs to produce
small currents (microamps). I want to have minimal switching diode
losses since I will have lots of diodes:>:

I need a fast diode with small input capacitance. I need to have it
withstand 200 volts at 100 microamps.

Which diode should I use??????

All suggestions are appreciated.

Al

Take a look at MMBD2004, aka Central CMPD2004S. Dual 250 (or 300) volt
diode in a SOT-23. Central also has a triple HV diode in a tiny
package.

John

 

[Mark Jones]

In news:[email protected] (John Larkin):

Take a look at MMBD2004, aka Central CMPD2004S. Dual 250 (or 300) volt
diode in a SOT-23. Central also has a triple HV diode in a tiny
package.

John

Also you'll probably find that there will be other losses like diode drop,
ESR, parasitic inductance, etc. Plan on over-driving the input to some
degree to get the required output. If you need a form of regulation, you
could create a low-current resistor divider on the output, fed to an op-amp,
which adjusts the input voltage... that said, it would probably be more
effecient if you could reduce the number of multiplier stages as much as
possible. Even in HV flyback supplies, typically only one or two multipliers
are used; the input voltage is typically ~9000v.

Mark

 

[albert]

Also you'll probably find that there will be other losses like diode drop,
ESR, parasitic inductance, etc. Plan on over-driving the input to some
degree to get the required output. If you need a form of regulation, you
could create a low-current resistor divider on the output, fed to an op-amp,
which adjusts the input voltage... that said, it would probably be more
effecient if you could reduce the number of multiplier stages as much as
possible. Even in HV flyback supplies, typically only one or two multipliers
are used; the input voltage is typically ~9000v.

John and Mark, Thanks!

It's for a photomultiplier supply, so it needs to have 10 stages. Each
stage connects to a dynode element, so there is no way around the
necessity of having many stages. The good news is that the upper
stages draw pico amps (or less) and the lower voltage stages supply
the brunt of the current.

I recently discovered that many have problems with the concept of
drawing power from each stage-they generally think of a a voltage
multiplier as a means to achieve a single (but very high) output
voltage.

If I go diode hunting, what should I look for in the spec sheets? Many
have a Trr (reverse recovery time) and it's also common for them to
spec the capacitance of the junction.

Thanks again,

Al

 

[John Stewart]

John and Mark, Thanks!

It's for a photomultiplier supply, so it needs to have 10 stages. Each
stage connects to a dynode element, so there is no way around the
necessity of having many stages. The good news is that the upper
stages draw pico amps (or less) and the lower voltage stages supply
the brunt of the current.

I recently discovered that many have problems with the concept of
drawing power from each stage-they generally think of a a voltage
multiplier as a means to achieve a single (but very high) output
voltage.

If I go diode hunting, what should I look for in the spec sheets? Many
have a Trr (reverse recovery time) and it's also common for them to
spec the capacitance of the junction.

Thanks again,

Al

You will need very good regulation of the voltage. Gain of any PM
I ever worked with was a very steep function of the operating voltage.

Good Luck, John Stewart

 

[albert]

You will need very good regulation of the voltage. Gain of any PM
I ever worked with was a very steep function of the operating voltage.

Good Luck, John Stewart

I might need the luck, thanks for your good wishes.

I know about the regulation requirements, would like to spice model
the pm tube I intend to use with the power supply, but I think I can
use multiple ramping current sources to do the job in spice.

Regards,

Al

 

[Oppie]

when I worked with PMTs, we always had a regulated HV supply which drove the
dynode chain resistors. One good technique was to put a 100V zener in place
of the resistor between the photocathode and the first dynode. this forces a
high fixed gain from the first stage (so you don't just multiply noise).
Also, the dynode chain resistors were chosen such that the voltages didn't
re-distribute too much as the anode current rises. Otherwise it causes
linearity problems. In general, gain is proportional (almost exponential) to
the voltage between dynodes. Too low a voltage and the tube saturates. Too
high and the noise increases.

Also a shutdown provision for the high voltage upon sensing an abnormally
high photocurrent is a good idea to protect an expensive tube.

Back to the original question, I have a 5kv rated unit I use from EDI
(electronic Devices Inc) #DR500 and another one from HV Components #G5FS.
Both companies make lower voltage units.

Oppie
http://www.e-edi.com/
http://www.electronicdevices.com/

 

[Rene Tschaggelar]

It's for a photomultiplier supply, so it needs to have 10 stages. Each
stage connects to a dynode element, so there is no way around the
necessity of having many stages. The good news is that the upper
stages draw pico amps (or less) and the lower voltage stages supply
the brunt of the current.

I recently discovered that many have problems with the concept of
drawing power from each stage-they generally think of a a voltage
multiplier as a means to achieve a single (but very high) output
voltage.

If I go diode hunting, what should I look for in the spec sheets? Many
have a Trr (reverse recovery time) and it's also common for them to
spec the capacitance of the junction.

posted to too many newsgroups.

I tend to doubt you get the setup stable.
The voltage on the upper stages depends on the current drawn
from the lower stages. The gain, aka current is voltage dependent.
What is wrong with a multiplier chain for each stage ?
A few diodes more spent ... plus a regulator for each stage.

Rene

 

[albert]

I tend to doubt you get the setup stable.
The voltage on the upper stages depends on the current drawn
from the lower stages. The gain, aka current is voltage dependent.
What is wrong with a multiplier chain for each stage ?

If you use resistors, of course this is true...if the lower dynodes
suck to much current, then the voltage to the middle and lower dynodes
drops.

I am doing one chain for each stage!

If I could afford to burn 200-300 milliwatts for my portable (battery
powered) photomultiplier, I'd run resistors and and be a happy camper.

A few diodes more spent ... plus a regulator for each stage.

Send me a diagram for a regulator for each stage that allows me to use
4 or 5 milliamps of supply current from my battery and I'll run it.
How does one regulate each stage separately without wasting alot of
input power?

Good luck.

 

[Rene Tschaggelar]

If you use resistors, of course this is true...if the lower dynodes
suck to much current, then the voltage to the middle and lower dynodes
drops.

I am doing one chain for each stage!

If I could afford to burn 200-300 milliwatts for my portable (battery
powered) photomultiplier, I'd run resistors and and be a happy camper.


Send me a diagram for a regulator for each stage that allows me to use
4 or 5 milliamps of supply current from my battery and I'll run it.
How does one regulate each stage separately without wasting alot of
input power?

Hmmm, battery powered ...
I'd have to think about that one.

Rene

 

[albert]

Hmmm, battery powered ...
I'd have to think about that one.

Not only battery powered, but I want good regulation, low ripple and I
want it to draw lass than 4 ma at 12v in....the Hamamatsu hc120 is
rated for 7 ma.

So, it's a tall order:>:

I believe it can be done with low power op amps, the TI ucc3977 driver
chip, a piezotransformer and an optimized cockcroft-walton multiplier
with a micropower voltage reference chip. I'm using a spice simulator
to work on the multiplier now and the driver chip and the piezoceramic
transformer are on order.

When/IF I make it work, I hope to make the plans available on the
internet for free-my little jab at Hamamatsu for their lack of
support>:>

Keep makin' trouble and good luck to all.

A

 

[John Larkin]

I believe it can be done with low power op amps, the TI ucc3977 driver
chip, a piezotransformer

Why piezo? You only need about 100 volts p-p to drive the C-W chain,
and that's easy with a real transformer, or even a single-inductor
flyback circuit. I thought piezo transformers weren't very efficient.

A real transformer will probably be more flexible as regards
oscillator frequencies and pulse widths. If a piezo is resonant,
you're stuck at an operating point that may not be optimum.

When/IF I make it work, I hope to make the plans available on the
internet for free-my little jab at Hamamatsu for their lack of
support>:>

I wonder way all the Hamamatsu stuff is so frightfully expensive.

John

 

[Watson A.Name - Watt Sun, Dark Remover]

Also you'll probably find that there will be other losses like diode drop,
ESR, parasitic inductance, etc. Plan on over-driving the input to some
degree to get the required output. If you need a form of regulation, you
could create a low-current resistor divider on the output, fed to an op-amp,
which adjusts the input voltage... that said, it would probably be more
effecient if you could reduce the number of multiplier stages as much as
possible. Even in HV flyback supplies, typically only one or two multipliers
are used; the input voltage is typically ~9000v.

John and Mark, Thanks!

It's for a photomultiplier supply, so it needs to have 10 stages. Each
stage connects to a dynode element, so there is no way around the
necessity of having many stages. The good news is that the upper
stages draw pico amps (or less) and the lower voltage stages supply
the brunt of the current.

I recently discovered that many have problems with the concept of
drawing power from each stage-they generally think of a a voltage
multiplier as a means to achieve a single (but very high) output
voltage.

If I go diode hunting, what should I look for in the spec sheets? Many
have a Trr (reverse recovery time) and it's also common for them to
spec the capacitance of the junction.[/QUOTE]

The PMTs I've seen use a resistive divider, not a C-W V multiplier.
There's nothing wrong with using a resistive divider, after all, you
said that it only needs picoamps.

The 1N4148 signal diode series seems to have one that's rated for a
couple hundred volts, it's the 1N4155.

The 1N4005 series will have too much losses at 55 kHz. You should
consider using regular 60 or 50 Hz; with that low a current the caps
would be reasonably small and you could use 1N4005 rectifiers.

Thanks again,

Al

--
@@F@r@o@m@@O@r@a@n@g@e@@C@o@u@n@t@y@,@@C@a@l@,@@w@h@e@r@e@@
###Got a Question about ELECTRONICS? Check HERE First:###
http://users.pandora.be/educypedia/electronics/databank.htm
My email address is whitelisted. *All* email sent to it
goes directly to the trash unless you add NOSPAM in the
Subject: line with other stuff. alondra101 <at> hotmail.com
Don't be ripped off by the big book dealers. Go to the URL
that will give you a choice and save you money(up to half).
http://www.everybookstore.com You'll be glad you did!
Just when you thought you had all this figured out, the gov't
changed it: http://physics.nist.gov/cuu/Units/binary.html
@@t@h@e@@a@f@f@l@u@e@n@t@@m@e@e@t@@t@h@e@@E@f@f@l@u@e@n@t@@

 

[John Woodgate]

Send me a diagram for a regulator for each stage that allows me to use 4
or 5 milliamps of supply current from my battery and I'll run it. How
does one regulate each stage separately without wasting alot of input
power?

There is another multiplier circuit that has push-pull drive to feed
each multiplier stage in parallel instead of in series. It has many
advantages over Cockroft-Walton, including much lower source impedance
and higher efficiency. I think you should do a Google search for it.

 

[albert]

The PMTs I've seen use a resistive divider, not a C-W V multiplier.
There's nothing wrong with using a resistive divider, after all, you
said that it only needs picoamps.

While the tube draws picoamps at the upper dynodes, the resisitive
divider is a constant load on the power supply...regardless of the
current drawn by the tube itslef.

If you think this is OK, well, I suppose we have different
standards::> But, typical values for the divider resistor are 330K
(each), in series this is (about) 3.3 meg ohms (assuming 10 resistors
are needed for the divider string). This 3.3 meg total resistance
appears across the 1KV supply whether the tube is conducting or not.
In fact, you can take the tube away completely and those resistors
still load the power supply and dissipate the same amount of heat into
the air.

Resistive dividers dissipate big amounts of heat and are cheap. In
some aps, they are clearly the way to go.

For battery powered portable operation the cockcroft-walton is the way
to go.

You see resistive dividers because your aps aren't don't need a high
efficiency power supply and becasue your aps can tollerate the heating
from the dissipation of the resistors.

The 1N4148 signal diode series seems to have one that's rated for a
couple hundred volts, it's the 1N4155.

The 1N4005 series will have too much losses at 55 kHz. You should
consider using regular 60 or 50 Hz; with that low a current the caps
would be reasonably small and you could use 1N4005 rectifiers.

I had found the 4148, they are cheap and available. They might be ok.
There is also the 1N3060, used by Hamamatsu and others...it's more
expensive and less 'available' tho.

I did some in depth researching yesterday. Some higher voltage diodes
have high leakage (which is devastating for the upper dynodes
efficiency when driven from a CW supply. Still others have 1.25 volt
forward drop...which adds up when you have 22 of them in a CW
multiplier! Clearly, (as one of you pointed out earlier) there is no
easy answer and there must be a compromise in the selection process.


Regards,

Al

 

[albert]

Why piezo? You only need about 100 volts p-p to drive the C-W chain,
and that's easy with a real transformer, or even a single-inductor

Well, to some extent, the piezo has never been used for this before.

But, the piezo has some big advantages and lend themselves to
miniaturization (unlike transformers). A small piezo can be 98 percent
efficient. Larger ones, such as those used in backlight inverters do
90 percent easily. This is for an ac input and does not include driver
losses (if a driver is used).

Driving them is no simple matter, as you pointed out.

I'm going to use a ucc2977 by TI. It is set up for constant current
output, but the sensing circiut is external, so running it in a
constant voltage mode looks doable with minor changes. The ucc3977
uses 1 ma from the supply so it is quite power stingy. I'm hoping to
use another ma for the voltage sensing circuit and another ma to drive
bipolar (extrnal) switching transistors. Based on spice simulations,
another 1 or 2 ma will be needed to light (and keep lit) the cw
supply. This means I might end up with 5 ma drawn from the supply
(when the tube is not illuminated).

You might want to take a spin over to TI and take a look at the chip
and 2 very nice ap notes they have! The chip works by varying the
frequency to the piezo, hence it controls the output nicely. Since the
piezo looks like a series resonant circuit, it draws LESS power when
driven away from it's resonant frequency...so controlling the
frequency of the input signal is much more efficient than you might
think.

I wonder way all the Hamamatsu stuff is so frightfully expensive.

I have the same problem, and they are my least preferred vendor. My
effort to make a practical cw based supply is inspired by THIER lack
of willingness to supply the CW supply they make as a building block
component (they will only sell it encapsulated with a fixed socket
type).

Like I said, I am very willing to do all this and then make the plans
available to all who might use them-ideally, via the internet.

Keep smiling:>:

Al

 

[albert]

There is another multiplier circuit that has push-pull drive to feed
each multiplier stage in parallel instead of in series. It has many
advantages over Cockroft-Walton, including much lower source impedance
and higher efficiency. I think you should do a Google search for it.

Yes, I have looked atthat circuit also. It is commonly referred to as
a CW also, but it isn't!

The CW is better suited for higher voltages at lower currents.

The OTHER multiplier needs different component values with respect to
voltage ratings whereas the CW doesn't have this restriction.

Honestly, I haven't modeled this alternative yet, but the upper stages
would require greater than 1 KV PIV, and these types of diodes are
very lossy. Even 500 volt PIV dioded have dismal switiching losses
and the middle stages of the circiut you sugeest would need to have
those diodes.

The CW is much better suited of low load current such as the PM tube
draws.

Have a great day and a better tomorrow.

A

 

[John Popelish]

Yes, I have looked atthat circuit also. It is commonly referred to as
a CW also, but it isn't!

The CW is better suited for higher voltages at lower currents.

The OTHER multiplier needs different component values with respect to
voltage ratings whereas the CW doesn't have this restriction.

Honestly, I haven't modeled this alternative yet, but the upper stages
would require greater than 1 KV PIV, and these types of diodes are
very lossy. Even 500 volt PIV dioded have dismal switiching losses
and the middle stages of the circiut you sugeest would need to have
those diodes.

The CW is much better suited of low load current such as the PM tube
draws.

Have a great day and a better tomorrow.

I don't understand why you need 500 volts out of any single stage.
Can't you stack the supplies in series to make the dynode voltages?
Each stage should be about 100 volts. Of have I missed something
important?

 

[John Larkin]

Well, to some extent, the piezo has never been used for this before.

But, the piezo has some big advantages and lend themselves to
miniaturization (unlike transformers). A small piezo can be 98 percent
efficient. Larger ones, such as those used in backlight inverters do
90 percent easily. This is for an ac input and does not include driver
losses (if a driver is used).

Driving them is no simple matter, as you pointed out.

I'm going to use a ucc2977 by TI. It is set up for constant current
output, but the sensing circiut is external, so running it in a
constant voltage mode looks doable with minor changes. The ucc3977
uses 1 ma from the supply so it is quite power stingy. I'm hoping to
use another ma for the voltage sensing circuit and another ma to drive
bipolar (extrnal) switching transistors. Based on spice simulations,
another 1 or 2 ma will be needed to light (and keep lit) the cw
supply. This means I might end up with 5 ma drawn from the supply
(when the tube is not illuminated).

You might want to take a spin over to TI and take a look at the chip
and 2 very nice ap notes they have! The chip works by varying the
frequency to the piezo, hence it controls the output nicely. Since the
piezo looks like a series resonant circuit, it draws LESS power when
driven away from it's resonant frequency...so controlling the
frequency of the input signal is much more efficient than you might
think.



I have the same problem, and they are my least preferred vendor. My
effort to make a practical cw based supply is inspired by THIER lack
of willingness to supply the CW supply they make as a building block
component (they will only sell it encapsulated with a fixed socket
type).

Like I said, I am very willing to do all this and then make the plans
available to all who might use them-ideally, via the internet.

Keep smiling:>:

Al

I'm currently working on a HV driver for an optoelectronic device, as
a spare-time project for a friend who's trying to start up a business
around his new e-o effect. I did one version a while back using a tiny
pot-core stepup transformer (a real pain to wind) and a 4-stage C-W
using four of the MMBD2004 dual diodes. That worked pretty well (got
900v from 5v in); I deliberately loaded it to speed up the discharge
time constant - the HV controls optical attenuation, and we need
control speed - so it wasn't very efficient. I'm going to do another
generation, using a stock center-tapped surface-mount torroid as a
flyback-autotransformer sort of thing, driving two 3-stage C-W
thingies, one for + and one for -, shooting for 1200-1400 volts total
between supplies. The bipolar thing avoids some breakdown issues in
the e-o package.

My control circuit is a sot-23 micropower opamp as the error sensor, a
sot-23 "tiny logic" schmitt trigger as the oscillator/duty cycle
generator/fet driver, and a small n-channel mosfet to drive the
inductor.

If the piezo works as well as you suggest, that might be interesting
too. Got any links to parts?

John

 

[albert]

Hi John,

I'm currently working on a HV driver for an optoelectronic device, as
a spare-time project for a friend who's trying to start up a business
around his new e-o effect. I did one version a while back using a tiny
pot-core stepup transformer (a real pain to wind) and a 4-stage C-W
using four of the MMBD2004 dual diodes. That worked pretty well (got
900v from 5v in); I deliberately loaded it to speed up the discharge
time constant - the HV controls optical attenuation, and we need
control speed - so it wasn't very efficient. I'm going to do another
generation, using a stock center-tapped surface-mount torroid as a
flyback-autotransformer sort of thing, driving two 3-stage C-W
thingies, one for + and one for -, shooting for 1200-1400 volts total
between supplies. The bipolar thing avoids some breakdown issues in
the e-o package.

WOW, I'm willing to bet that your opto-related power supply is a near
perfect ap for the piezo device!!!!! These things excel at moderate to
high resistance load currents!

Before starting, please keep an eye peeled on the TI website. They are
going to release a new driver chip any day now. It's a 20 pin chip,
however it has lower power consumption and much more flexibility than
the present 8 pin drivers. It should be possible to convert the chip
from a constant current source to a constant voltage source.

My control circuit is a sot-23 micropower opamp as the error sensor, a
sot-23 "tiny logic" schmitt trigger as the oscillator/duty cycle
generator/fet driver, and a small n-channel mosfet to drive the
inductor.

If the piezo works as well as you suggest, that might be interesting
too. Got any links to parts?

There is a wealth of technical info on the web, but there are few
suppliers of actual parts. In the US, there is only one vendor for the
actual parts. NO ONE MAKES A LOW POWER UNIT, they are all in the 2-6
watt output range.

CTS used to make them, the are out of the piezotransformer market now.

Steminc in Florida makes them, but refused to answer questions about
availability and second sourceing. They have been selling them on
EBAY. They are presently shut down for vacation till th e5th of
January.

As far as other vendors, there are quite a few. But, they seem to be
all Asian and Digikey doesn't carry these devices from any of the
manufacturers! They are going into Asian contract manufacturing houses
I think:>:

Anyway, the Asian vendors are Panasonic, Fuji and a Korean company. I
forget the anme of the Korean vendor. NEC/Tokin and Tamura make
complete convertes based on PZT's. Another one is Hitachi. Piezo
Systems Inc (MA) are interested in PZT's and might make them someday.

Hope this helps. I have about 10 MB of misc technical docs, spec
sheets and various related items. I'm happy to email this stuff to you
if you like.

If you only want a primer, the 1999 EDN article on these devices is a
good start, let me know and I'll email the PDF to you and you can
decide if you want to look further>:>

Be warned however, driving these little buggers ain't easy. Although
they have Q's from 500 to 1500 and should oscillate, they don't work
well in a power oscillator circuit:>: There are mode changes that are
load dependent, th emechanical resonance and the electric resonance
frequencies are different and you can't run them without load-they
easily go to 10KV output without a load and the crystal is fractured.

Driving them with a VCO and having the output voltage determine the
vco's frequencyt is the way to go......when driven slightly off
resonance, they do not draw much power since they look like a series
resonant circuit.

Anyway, have fun:>:

Regards,

Al