Should the opening and closing performance of the tristate ion shutter be tested at the ms level?

[kbdydd]

I want to test the opening and closing performance of the tristate ion shutter. When testing the opening or closing time of the shutter, should the opening and closing time be at the ms level?

[dkayser]

When I tested I built a simple particle-based controller where I could change the gate opening time between 500 and 150 micro seconds by pressing a button. 500uS for seeing if there is anything there, and then 150uS to get sharper peaks. Unless you work in vacuum you probably do not need faster gate open times.

[bhclowers]

Dirk is correct. Even a modulation duty cycle of 50% should show you very broad on/off cycling of the ions. Depending on the performance of the amplifier you are using, you can have a period of 50 ms or more with a 50% duty cycle. You are looking for broad swings before tightening the gate pulse width. Amplification should be around 10^9 V/Amp and a target rise time of 30 us or so. Remember you are unlikely to see a signal with a single pulse as the system will be quite noisy. Multiple averages will be needed to reduce the random noise and reveal a peak.

[kbdydd]

When I tested I built a simple particle-based controller where I could change the gate opening time between 500 and 150 micro seconds by pressing a button. 500uS for seeing if there is anything there, and then 150uS to get sharper peaks. Unless you work in vacuum you probably do not need faster gate open times.

Thank you for your answer
In this system, I used a vacuum ultraviolet lamp as the ionization source with a drift voltage of only 3KV. However, I encountered some problems while testing the shutter;
(1) When I apply a 130V pulse to the tristate ion shutter, a phenomenon similar to high-voltage breakdown occurs at the position of the shutter, with electric sparks accompanied by a sizzling sound. However, I am certain that the shutter is completely insulated from the electrode and the three shutters (using a spacer made of tetrafluoroethylene, with a thickness of 0.3mm)

(2) During the testing (with the shutter open), the signal baseline of this system was very strange. I used an amplifier with a gain multiple of 10 ^ 10V/Amp, but its initial baseline was -80pA. Using acetone as the sample for injection, the measured signal unexpectedly rose to 130pA, which is completely opposite to the report in the literature.

(3) The shutter seems to be not working; When I connect the shutter to the pulse generator, the initial baseline of this system directly rises to the 130pA position in question (2). Afterwards, even though I inject the sample, the detector does not generate any signal until I disconnect the pulse generator from the shutter, and even when I touch the wire connected to the shutter, its baseline will rise to 130pA, so that I cannot truly observe whether the shutter is working.

(4) When the system baseline rises to 130pA, disconnect the pulse generator from the shutter. It takes several minutes for the baseline to decrease to -80pA, and the descent is very slow, looking very much like a capacitor charging and discharging.

(5) Should the negative electrode (ground) of the shutter pulse generator be connected to the negative electrode (ground) of the system high voltage (voltage applied to the solvent and drift regions)? Or connect the negative electrode (ground) of the shutter pulse generator to the middle of the two resistors on the three shutters mentioned in the literature?

My thoughts on these issues;
For question (1), is it the reason for the close distance between the shutters?
For questions (2), (3), and (4), does the electric field in the solvent or drift region affect the electrons in the metal shutter, resulting in the shutter not working?

[kbdydd]

Dirk is correct. Even a modulation duty cycle of 50% should show you very broad on/off cycling of the ions. Depending on the performance of the amplifier you are using, you can have a period of 50 ms or more with a 50% duty cycle. You are looking for broad swings before tightening the gate pulse width. Amplification should be around 10^9 V/Amp and a target rise time of 30 us or so. Remember you are unlikely to see a signal with a single pulse as the system will be quite noisy. Multiple averages will be needed to reduce the random noise and reveal a peak.

Thank you very much for your explanation. When verifying Mr. Drik's ideas, I encountered some problems. If you have time, I hope you can help me answer them. Thank you.

[dkayser]

(1) When I apply a 130V pulse to the tristate ion shutter, a phenomenon similar to high-voltage breakdown occurs at the position of the shutter, with electric sparks accompanied by a sizzling sound. However, I am certain that the shutter is completely insulated from the electrode and the three shutters (using a spacer made of tetrafluoroethylene, with a thickness of 0.3mm)

Maybe high humidity? Stacked drift tubes are very hard to get air tight, try FKM gaskets and run a pressure check with the drift tube under water at 1,5 Bar or so. PTFE gaskets are too hard and seal much worse. No air should escape or you will ruin your ionization. You need to dry all gases with 3A and 4A mol sieve, and activated coal helps if you build an internal air loop. It makes a huge difference.

Also electrode hole size and electrode distance are related.

Maybe sharp edges on the shutter grids from manufacturing? Do the gates bend inwards when you apply pressure from bolting the drift tube?

(2) During the testing (with the shutter open), the signal baseline of this system was very strange. I used an amplifier with a gain multiple of 10 ^ 10V/Amp, but its initial baseline was -80pA. Using acetone as the sample for injection, the measured signal unexpectedly rose to 130pA, which is completely opposite to the report in the literature.

How much did you inject? How much sampling are you doing (try 512x)? Or are you just measuring ion current? In that case measure the ion current of your source before injecting sample. While you are at it, check whether your gate voltage is sufficient.

Regarding your other questions:

• There are many different approaches to connecting a tri state grid. And even more failure modes. Maybe your voltages are not well regulated? Does your signal generator provide enough current for driving your gate capacitance?
• Do not underestimate how much the gas preparation makes a difference. Very dry and airtight is what you want.

==> Figure out all high voltage discharges first. There should be no sizzling sounds, let alone sparks.
==> Check for air ingress, especially near the ionization region. Check if you overtighten the bolts, get FKM gaskets or O-Rings
==> Build a simple setup with your transimpedance amplifier to verify ion current

Good luck!