1. Home
  2. Electronics Design
  3. Zener protection for 10kV pulses

Zener protection for 10kV pulses

I have the following circuit for one section of a single atom mass spectrometer.

G _____|\\__________ | |/ | Vac | | _|_ Gnd C --- | +---------+ | | | _____ R | 1_| 2_ | | DAC |-----\\/\\/--+-------|_| |_|---|--+ |_____| | | | | R 3_ 4_ | | |--\\/\\/--+-------|_| |_|---+ | _|_ | | C --- +------------+ | | gnd

where Vac is ~1.1MHz, at ~ 200Vpp after G (ENI RF amplifier). The DAC is a standard cpu controlled National instruments PXI card, so that I can output +/-10V through resistors R = 100kOhm onto four rods, shown here as 1,2,3,4. These rods are about 1 cm long stainless, and are contained in a vacuum system, separated from each other by ~1cm. Rods

1 and 4 are shorted together, and get RF+DC on them. Rods 2 and 3 are shorted together, and only get a DC field on them. The resistors block the RF from going back into the DAC, and the C = 0.1uF to ground is a current path to ground for RF. The C = 0.1uF between the RF supply and the rods blocks DC from getting into the RF amplifier G. Single atoms are trapped by the RF and DC fields in the center of the rod structure (this is a common design of a mass spectrometer). We use this to capture single atoms to do laser spectroscopy in gas (~1e-3 torr He, Ar, Xe).

This system is repeated 16 times (same RF, but different DC channels), which you guys helped me design almost a year ago, in case anyone remembers, and we can now trap and image single Ba atoms. In case you're interested in a picture of the setup, take a look at

formatting link
You can see a section of the trap itself at
formatting link
to get an idea of what I'm talking about.

Now, I have a simple question related to this. We're going to place a nano-tip (Tungsten, with a radius of about 50nm) near the front of this single atom trap, and pulse it at about 10kV in order to try to shoot single ions of Barium through it. I have a high voltage op-amp that can do this with a bandwidth of about 10kHz, but don't want to destroy the NI DAC if there happens to be any breakdown. The system will look something like this from the side S1 S2 S16 _______ _______ _______ (______()(______()... (______() Tip (______()(______() ...(______()

----->

_______ _______ _______ (______()(______()... (______() (______()(______() ...(______()

Where the Tip will be pulsed with a square wave at 10kV. If there is any electrical breakdown to the segments, which are connected to the DAC and computer, how can I make sure nothing dies (either by arcing which should be taken care of by the resistor, or by any sorts of EMF that are created by pulsing this tip at 10kV)? I'm going to put a fairly large resistor inline with the tip, maybe 100MOhm, rated to 20kV or so. In addition, I assume that I should install zeners in the circuit above - probably between the resistors and the DAC? I wanted to ask to make sure I don't do something stupid.

Also, should I replace the caps to ground withe something better? I'm not sure what the pulsing is going to affect or not.

Reply to
jwodin
Loading thread data ...

sorry, my links were bad.

part of the trap is at:

formatting link
the vacuum chamber that holds it is at:
formatting link
and an EMCCD photo of a single atom (1 sec. exposure) is at:
formatting link

in case anyone is interested...

Jesse

snipped-for-privacy@gmail.com wrote:

Reply to
jwodin

I'd start with a simple beefing up the robustness of the path from the rods to the DAC, as per the modification below.

. G . _____|\\_______________ . | |/ | . Vac | . | _|_ . Gnd C --- . | ,---------, . | | | . _____ Rx R | 1_| 2_ | . | DAC |--\\/\\/--+--\\/\\/--+-------|_| |_|---|--, . |_____| | | | . | R 3_ 4_ | | . +--\\/\\/--+-------|_| |_|---+ | . _|_ _|_ | | . C --- C --- '------------' . | | . gnd gnd

This way a spark discharge to the rod sees multiple filters on the way to the DAC. I'd make the resistors -\\/\\/- from a proper HV part, or multiple ordinary resistors in series. Finally, I'd add the zener you spoke of, just after Rx, going toward the DAC, plus another resistor to the DAC. That should do it. Next, start worrying about creaming your ENI amplifier. Keep in mind Zener and TVS diodes have lots of capacitance.

--
 Thanks,
    - Win
Reply to
Winfield Hill

okay, thanks. I'll try it. We're going to try to start pulsing in a few weeks, so I'll let you know if I obliterate my lab and all electronics therein

jesse

W> >

Reply to
jwodin

Jesse, you didn't explain how you're going to protect your ENI 1.1MHz RF power amplifier! We use step-up transformers for that function, BTW, resonating, if IIRC.

--
 Thanks,
    - Win
Reply to
Winfield Hill

On Mon, 05 Jun 2006 16:35:12 -0700, jwodin wrote: ...

Transzorbs. They're zeners, but much faster and able to dissipate amazing amounts of energy, as long as it's quick. I once worked at a place where high-KV electron guns were routine, and they used to use MOVs until we discovered transzorbs.

formatting link

It depends on what you have now, and what kind of configuration. Theoretically, you could make a vacuum cap and integrate it into your chamber... But for some reason a doorknob capacitor comes to mind, if there even is still such a thing. :-)

Good Luck! Rich

Reply to
Rich Grise

yes, at this point, I have no idea how to protect the amplifier. I'm going to go talk to some electrical engineers up at the linear accel (SLAC) up the road, and hopefully have them tell me I'm not nuts (besides, I'm using their 10kV opamp).

Jesse

W> >

Reply to
jwodin

I agree they're fast, sub-nanosecond (but so are zeners), and they happily handle high currents for a sizable time (due to their internal copper-mass blobs, which zeners don't have). But sadly they have lots of capacitance, which makes a real issue for protecting the 1.1MHz RF amplifier. But perhaps they can play a role in the solution.

--
 Thanks,
    - Win
Reply to
Winfield Hill

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.