Difference between revisions of "Equipment/High Vacuum Station"

From London Hackspace Wiki
Jump to navigation Jump to search
m (Updating photo)
Line 180: Line 180:
Inline temperature sensor, flow-sensor, pumps, etc.. all available from PC water-cooling suppliers.
Inline temperature sensor, flow-sensor, pumps, etc.. all available from PC water-cooling suppliers.
The implosion-sheild is done. This is both to protect users in the case of catastrophic failure of the glass bell-jar, and protect the bell-jar from users and passers-by in the mixed usage workshop.
The bell-jar will be padlocked down as an access-control measure.
We're still in need of vacuum gauges, feedthrus and vacuum hose to connect the backing pump. Getting close now!

Revision as of 11:37, 14 May 2015

High vacuum station
20150513 214222.jpg
Model Unknown
Sub-category Physics
Status Under construction
Consumables Pump oil
Accessories Unknown
Training requirement Yes
Training link Unknown
Owner LHS
Origin Unknown
Location Basement, ante-room beside lift
Maintainers Sci


The HVS is an in-progress equipment project to provide LondonHackspace members with access to a multi-purpose high vacuum system.

Proposed Uses

  • Creating x-ray sources, CRTs & other electron-beam experiments
  • Vacuum glasswork (EG; thermionic valve making/repair)
  • Surface coatings (optical coatings, photocathodes, metalic finishes, titanium oxide coatings)
  • Electron microscopy
  • Outer-space environment test chamber

Design concept

To create an ultra-high vacuum source capable of being purposed to a wide range of high-vacuum related tasks.

  • Combining a resin degassing chamber/line into the design has now been abandoned as there are enough lower-spec rotary pumps available now to build a dedicated station for that purpose.

Backing/roughing pump

The backing pump is a rotary pump that should be capable of meeting or exceeding the backing pressure and maximum output of the diffusion pump.

There is a pledge for a suitable pump, as none offered so far have sufficient ultimate pressure and/or pumping speed. http://wiki.london.hackspace.org.uk/view/Pledges/Vacuum_Pump The pledge was successful and a backing pump has been aquired.

Diffusion pump

The diffusion pump donated by James Collings is an Edwards E04. It's ultimate vacuum capacity is 5x10^-7 mbar (0.0000005mbar) with a pumping speed of air of ISO 600l/s^-1 (hydrogen: ISO 800l/s^-1). It has a maximum starting pressure of 300 mbar and requires a backing pump capable of keeping the backing pressure at about 1mbar (assuming DC704EU pump fluid). Recommended backing pumps are the Edwards RV8 or RV12. The RV8 has a displacement of 9.7m^3h^-1 and an ultimate pressure of 2x10^-3. The RV12 is advised for fast roughing, has a displacement of 14.2m^3/h^-1 and an ultimate pressure of 2x10^-3 mbar.

The diff pump operates by boiling a special silicone oil to vapour and forcing it out through special baffles onto a water-cooled jacket. There the oil re-condenses and trickles down along with any air molecules it grabbed on the way. It sets up a pressure differential across the pump, where at its exist it should be at a high enough pressure the backing pump will be able to pump it away.

Current prices for the compatible fluids (as of October 6th, 2011, direct from Edwards Vacuum) are;

  • DC702 fluid - £117, 500ml (critical backing pressure 1.3 mbar, ultimate pressure 5x10^-6 mbar)
  • DC704EU fluid - £134.10, 500ml (critical backing pressure 1.0 mbar, ultimate pressure 5x10^-7 mbar)
  • DC705 fluid - £211.50, 500ml (critical backing pressure 0.9 mbar, ultimate pressure 5x10^-7 mbar)
  • Santovac 5 fluid - £852.30, 500ml (critical backing pressure 0.8 mbar, ultimate pressure 5x10^-7 mbar)
  • Edwards L9 fluid - Unknown price, possibly discontinued (critical backing pressure 0.9 mbar, ultimate pressure 5x10^-7 mbar)

On this I would recommend DC704EU fluid. For £17 extra you can achieve a magnitude better vacuum than the absolute cheapest. The spec sheet also describes it as a good durable and general-purpose fluid.

Cold baffle

The cold baffle is a small plate in the pump stack with water-cooled plates. These cold plates minimise any pump fluid vapour that might otherwise deflect upward and contaminate the vacuum processes above.

Pump stack

Above the diffusion pump are a series of manifolds for various purposes, including the cold baffle mentioned as part of the diffusion pump above.

To fulfil all the uses proposed, placing an extension piece with a branch next with the main plate valve above it will allow the full power of the diffusion pump to be accessed by a side-valve without having to pump down the main chamber. This high-vacuum access valve could then be used for attaching vacuum tubes, glass-forming manifolds, electron microscopes, fusors or any other number of parts. A foreline vacuum pressure sensor will be necessary here.

The valve currently need a slightly wider packing piece to be used in the stack at this location. The current one is only 3/4" thick, and in this position fouls the valve control gearbox.

Vacuum chamber

The vacuum chamber is a recessed base, likely for a glass bell-jar. However once fitted with a new O-ring gasket it will likely be fine for initial use with a simple flat cover. 25mm polycarbonate will probably do.

The chamber has three ports for vacuum feedthrus, though one is modified and may limit use. There are a number of threaded mounting points and, most notably, a large angled port at the circumference. This large port will require a special blanking plate making, but after that can probably be used for any custom or special-use interfaces required.

Support equipment

Both diffusion pump and cold baffle require a self-contained cooling loop. Possibly even separate radiators if one effects the critical temperature of the other.

The parts from James contained four Pirani Gauge heads and two Penning Gauge heads. These will need gauge controllers/readouts finding or making for them.

Some surface-coating processes work best if a glow-discharge plasma is used to remove surface oxidisation before coating. Sci has a 5.5Kv mains transformer that should be capable of producing this in vacuum.

Surface coatings also require a tungsten crucible or wire to be heated to vaporise metals. A high-current variable PSU or Variac would be used here.

To keep all the component parts secure and in-line, as well as keeping dangerous parts from being exposed, building a steel framed trolley to house it all seems like a sound idea.

Servicing and care

A wire mesh over the pumping hole of the vacuum chamber should help a lot to prevent any debris reaching the pump. After any use of the chamber for deposition purposes, it should be cleaned with fibre-less cloth and isopropyl alcohol.

(proposed) usage rules

  • Do not use without being trained.
  • No mercury or mercury-containing products should be allowed in the chamber, due to potential vaporisation in backing-pump exhaust, pump oil/fluids contamination and use of aluminium manifolds which mercury will rot.
  • No marshmallows. Or any other foodstuffs. Use the degasser.

Parts needed

Some additional parts will be needed.

  • Water pump (for cooling loop).
  • 1.5-2" thick, ~4" wide sturdy plastic (delrin?) to turn an large threaded fitting for the chamber.
    • I have a 100m dia, 45mm thick lump of nylon and a fair bit of 50mm delrin if it's any use. artag.
  • 20mm+ thick polycarbonate sheet (will probably order cut to size).

Cooling system

The E04 diffusion pump uses an 850Watt heater. The typical surface temperature will reach upwards of 200c (maximum 250c for pump body, max 400c for pump base).

The pump technical data specifies the maximum water pressure should be 5bar/72.5psi. Minimum water supply temperature should be 10c, and maximum return temperature should be 35c. (presuming return in this case means "water coming back out of the pump cooling jacket") The minimum flow-rate should 75l/h^-1.

As the baffle and valve assembly as also have cooling-loops, some additional capacity should be factored in. No numbers exist for these parts however. As they are indirectly heated however and part of reflux prevention, they perhaps should have a secondary cooling loop or auxiliary radiator between them.

Estimating any radiator capable of handing about 1Kw should be fine. Several PC water-cooling pumps are within capacity, and variable-speed.

Due to likely enclosed location, heat exhaust should be ducted away.

Progress reports


All components are currently being stored in the Out Of Order workshop toilet on Jontys suggestion. This should keep them away from being damaged or contaminated by workshop dust. All manifolds seem in good order and have been in use until recently. Diffusion pump heater and thermal cut-out both test as good with a meter. Wiring should be replaced however. Pump should be drained & cleaned as a precaution. Fluid should be measured and filtered/inspected. A temperature sensor could be installed on the drain port. ~ Sci


Post-xmas checkup on stored parts. Took some measurements of pump stack parts. Short video here: http://www.youtube.com/watch?v=3U_D1sNpXWc ~ Sci


Finished taking measurements of pump-stack parts. Inventoried vacuum pipe fittings. Have added scrap industrial control panel to parts pile (sounder works at 5v) as well as pressure-switch and 3 castors (4th currently errant). All parts so far are in imperial dimensions, large manifolds are all SpeediVac and smaller ones are GeneVac. Several of the parts have been modified by the previous owner, so I'm inclined to look for additional parts. Though others have made do with copper pipe, I'd be concerned about trace-gas leakage through such thin material. Also we may need an additional bellows pipe section to get complete vibration-isolation from the backing pump (maybe. And only crucial for electron-microscope purposes). Will see if I have any compatible parts spare, and keep my eye on ebay. Also took dimensions of the storage-toilet as I presume, being the closest thing to a "clean room" available, it'll be where the pump system will live. So will constrain station dimensions to ensure it'll fit.

Considering use of a large steel pipe-T (water main size, ~10" ID) to serve as a bell-jar/chamber with port for a large observation window.

Chamber has some very simple adaptive manifold connectors it should be fairly easy to replicate/customise, as well as add more custom ports as need be.


Have collected a Pirani/Penning gauge controller. Purchased from ebay for £5, plus packet of chips for ride to collection. Edwards Model 4, approximate age of existing gauge heads. Should be compatible. Supports two pirani and one penning head. Not rack/panel-mount, but should do the job very nicely. Will need compatible mains plug.

Likewise purchased some mixed Speedivac fittings (£11 including postage) that should make building up the vacuum pipework a lot nicer (avoiding using the home-made joints is a bonus).

Also donated a water-flow meter. Currently runs on 24v relay, but has fitting for hall-sensor control, so suggest gutting of stock electronics and wiring directly into control systems.


Most of the welding is done, but we ran out of welding gas. Need more before transformer mounts can be done. Dry-fitting has begun, though will need short 5/16ths 24tpi Whitworth bolts to connect to chamber base.

Backing pump has been mounted on anti-vibration feet.

Welds will need grinding flat and re-doing in places once dry-fit of all parts is done and setup proved.


Long-overdue update to the wiki page.

The diffusion pump has been serviced and cleaned. The cooling jacket is leak-free. It appears to have previously gone UTA catastrophically while at operating temperature, as it contained glass debris and burnt-on pump oil residue. It has been temporarily sealed for dust protection and returned to the equipment trolley. The next step is to complete the radiator & fan enclosure to get the pump cooling-loop operational.

To go to a bell-jar based design, circles of billet aluminium 13" dia are available from Alauminium warehouse, but will need facing on a large lathe. A used gas cylinder has been donated to serve as bell-jar. Suitable seals are approximatly £30. The trolley will likely need re-designing eventually.

Inline temperature sensor, flow-sensor, pumps, etc.. all available from PC water-cooling suppliers.


The implosion-sheild is done. This is both to protect users in the case of catastrophic failure of the glass bell-jar, and protect the bell-jar from users and passers-by in the mixed usage workshop.

The bell-jar will be padlocked down as an access-control measure.

We're still in need of vacuum gauges, feedthrus and vacuum hose to connect the backing pump. Getting close now!