Reusing a Helicoflex gasket

Print Friendly, PDF & Email

Can you re-use a Helicoflex gasket?  The answer is yes, sometimes.

A Helicoflex gasket uses the concept of plastically deforming to seal between two flat and polished surfaces instead of the much more common knife edge seal used with copper gaskets and CF flanges.

On Physical Electronics X-ray Photoelectron spectrometers Helicoflex gaskets are used on the monochromator, the SCA (spherical capacitive analyzer) and the LS specimen stage.

Here is a link to information on how a Helicoflex gasket works –  http://technetics.com/products/sealing-solutions/metal-seals/helicoflex-delta/

First of all, Helicoflex gaskets are designed to be used only once.   However,  since these gaskets are expensive it may be worth trying to reuse it as long as you have a new gasket on hand in case the trick I will show  you in this blog post does not work.  My success rate with this trick is about 50%.  RBD Instruments provides most of the Helicoflex gaskets used on PHI XPS systems.

For this example, I will attempt to make it possible to reuse a monochromator gasket as shown in the picture below.

monochromator_gasket

mono-chromator gasket

Note that once compressed, the center surface of the gasket is flattened out.

better_view_of_flat_area

Step one is to uncompress the gasket.   To do that, find a spot on the outside edge of the gasket where you can insert a small screw driver and gently spread open the two lips of the gasket enough to get a large flat head screwdriver inside.

insert_small_screwdriver

Then, carefully work the large flat head screw driver around the perimeter of the gasket until the entire gasket is un-compressed.  Try not to deform the edges of the slit.

insert_large_screwdriver

During this process the gasket will become deformed.  Use a round surface such as a Philips screwdriver to smooth out the bumps without re-compressing the gasket.

smooth_out

Next, using a very fine emery cloth (I used some 30uM 3M sand paper) smooth out the edges of the flat surface on the gasket.  You just want to break the edges, not make it round.

break_edge_on_flat

Clean the gasket with isopropanol or methanol to remove the small particles.

clean_with_IPA

Finally, use a small amount of Apiezon L or other HV or UHV vacuum grease to the surface of the gasket (on the flat area).   You want to use a very small amount and spread it out evenly.  The vacuum grease will help to fill in any small scratches.   Wipe off the inside and outside of the gasket.   You only want a small amount on the flat area.

small_amount_of_vacuum_grease

You are now ready to install the gasket.   Good luck!

The gasket that I used in this blog post did seal fine, so I got lucky.  🙂

 

 

Homemade Titanium Sublimation Pump

Print Friendly, PDF & Email

In this post I show how we made a small homemade titanium sublimation pump for an 8” Kimball Physics spherical octagon UHV vacuum chamber.

Our little chamber has a 60 l/s ion pump, but even with baking (both IR and UVC), we were able to get only into the low 10-8 to high 10-9 Torr range.  However, using the little titanium sublimation pump, which we “Frankensteined” together using parts we had readily available, allowed us to get in the low 10-9 to high 10-10 Torr range, a factor of ten improvement.

A titanium sublimation pump works by heating a titanium filament wire to about 1300 degrees C. That is hot enough to create titanium gas molecules (sublimate) but not so hot that the filament wire melts.  The sublimated titanium deposits on the wall of the chamber (or preferably on a shield wall) and forms a thin film. This layer of titanium is very reactive and will bond with other molecules in the vacuum chamber such as CO and O2. Disassociated hydrogen and water vapor also diffuse into the titanium layer.

The reactivity of the titanium film is increased with lower temperatures, but most titanium sublimation pumps are operated at room temperature. Over time, the titanium film will become coated and need to be replenished. All commercial titanium sublimation pumps have 3 to 4 filaments so that when one filament burns out you can switch to another. Those filaments are also relatively thick in diameter at 12 gauge (.080”) and need about 50 amps of current to operate.

For our homemade titanium sublimation pump, we used 24 gauge (.020”) so that we could operate at a much lower current of 4 amps.  We also only have one filament.

Before I show how we made our homemade titanium sublimation pump, here are links to some videos on how TSPs work:

https://www.youtube.com/watch?v=j5Y7m2ZJfgg

https://www.youtube.com/watch?v=9vJedaxRsxI

The first thing that we needed was a 2-pin electrical feedthrough on a 2.75” CF flange.  For that we used a Getter pump flange from a PHI 04-303 ion gun as shown below.

filament_wire_connections

filament_wire_connections

We then needed to somehow support and electrically isolate the TSP filament wire. To do that we used a coupler and some little shoulder washers.

Ceramic_support

Ceramic_support

The top part of the getter pump assembly is conveniently designed to allow gas molecules to pass through but also block direct deposition of titanium into the vacuum chamber.

blocks_direct_deposition

Next we added a few turns into the titanium wire so that it would have a little bit of a spring to it. Then we connected the wire to the flange and support assembly. We have only one filament and so by effectively doubling the length of the wire we could also double the amount of titanium that we would be sublimating.

filament_wire

filament_wire

For a chamber wall we used a 2.75” nipple that has a tube ID of 1.6” and a length of 4”.  The larger the surface area the better, but for the size chamber that we have, we are limited to a small 2.75” nipple. We mounted this nipple on our chamber horizontally so that any flakes that form will not get into the chamber or ion pump.

2.75_inch_CF_nipple

2.75_inch_CF_nipple

For a power supply, we used a 30 volt 5 amp Lavolta.

30V_5amp_power_supply

30V_5amp_power_supply

After installing our homemade titanium sublimation pump into the chamber we pumped down and were ready to operate the TSP.

To operate our titanium sublimation pump, we slowly increased the power supply current while observing the color of the light coming off the TSP filament and also monitoring the chamber pressure.  The filament needs to be orange for the titanium to sublimate. Too hot and the lifetime of the filament will be reduced. Too low and the pumping effect is reduced. By experimenting we determined that about 3.8 amps DC was the correct amount of current.  Once that was determined, we could just periodically turn the TSP on for about 2 minutes at a time.  We did that 3 times over a 6-hour period and then let the chamber pump overnight. The next morning we were in the high 10-10 Torr range.  Success!

tsp_filament_in_operation

tsp_filament_in_operation

Conclusions:

  1. It is possible to make a small titanium sublimation pump using off-the-shelf components that will operate with less than 5 amps of DC current.
  2. Adding a titanium sublimation pump to a small chamber can help to get from HV to UHV.

 

 

 

 

Resolving USB Conflicts with Virtual COM Ports

Print Friendly, PDF & Email

If you run multiple USB devices that operate as virtual RS232 COM ports (the ubiquitous serial port standard) on Windows, you may have run into problems with conflicts between devices. An application may connect to the appropriate device when it’s the only one connected, only to “get confused” if there is another device sharing the PC. Happily, there are a few simple things you can try that will often resolve the problem.

9103s and Arduinos Playing Nice Together

9103 and Arduino USB Devices

A 9103 Picoammeter and Arduino

To most Windows applications, virtual COM ports (VCPs) all look the same. An application can open a port and and attempt to communicate with the connected device, but since there’s no fixed protocol – each device speaks its own “language”, any message sent can have undetermined effects if the device you’re communicating with is not the one your were expecting. Some applications simply connect to the first COM port available, other’s may provide a way to select the COM port your device is connected to – but you’re still responsible for figuring that out.

When manufacturers produce hardware for PCs they can apply for unique vendor ad product IDs for their device, and there are ways for applications to safely query these. But that only solves part of the problem. Many devices use third-party USB chips and drivers from companies like FTDI, so they share the same IDs. These devices look the same to a Windows client application, or to a person perusing the Device Manager in Control Panel.

RBD’s own 9103 Picoammeter utilizes FTDI’s popular USB VCP chips, as do many versions of the popular Arduino microcontroller boards, so these two sets of devices can be confused by client applications when used on the same PC. And s it turns out, they are often used together. Here are a few tricks for getting these device to play nice together.

Solution 1: Connect Each Device and Run Each Client in Order

Many applications require you to specify the port for the selected device. Others (like Actuel for the 9103) poll the COM ports in numerical order and check and connect to the first available. If these devices first check the vendor and product ID (like the 9103), they will at least skip ports that do not match. But they cannot distinguish between two devices using the same USB chip (like FTDI’s). Setting up a device connection / application order can solve this.

In the case of a 9103 / Arduino conflict, remove all other devices, then plug in the 9103 and power it on. Next run the Actuel software. The software will find and take control of the 9103 port, and once assigned, you can safely plug in the next device and run its client.

Another order might make more sense for your particular application. Experiment with your configuration, and there are more than two devices, try getting two working first. Document the process and just make sure it’s followed anytime you reboot / power-on.

Solution 2: Change the COM Port Number for a Particular USB Port

You can force Windows to use a different COM port number than the one automatically assigned. This may help with applications that select the lowest numbered port.

For example, if the 9103 is connected to COM4 and another FTDI device is on COM3, the 9103 client software may incorrectly select the device on COM3. Setting the 9103 to COM2 may allow you to now connect the devices and run the client applications in any order, depending on how those other devices / applications behave. Some experimentation may be necessary.

With the 9103 connected and turned on, run Control Panel / Device Manager, and find the selection for “Ports (COM and LPT)”, click and you should see an entry for “USB Serial Port (COM4)” (the COM# may be different of course). Double-click for properties.

The COM port settings for the 9103 USB VCP

The COM port settings for the 9103

Now select the “Port Settings” tab, and click the “Advanced…” button. From this window you can select a new COM port assignment:

Choosing a COM port for the 9103

Choosing a COM port for the 9103

Keep in mind that plugging a device into different USB port will change the COM port assigned to it.

More Info

Of course, you’ll want to ensure you have the latest drivers installed. For FTDI, they can be found here:

http://www.ftdichip.com/FTDrivers.htm

If you’re thinking of programming your own serial port application, here’s a quick tutorial at the API level. Many popular languages include code for VCP programming, and third-party libraries are available:

http://xanthium.in/Serial-Port-Programming-using-Win32-API