Accurate needle valves

For posts specifically relating to fusor design, construction, and operation.
Post Reply
guest

Accurate needle valves

Post by guest »

Is it of any advantage to have a precise needle valve?. I mean one with a very fine scale of adjustment. I have made miniature needle valves in the past and it occurred to me that if you had a needle about 150 mm (6") long made out of spring steel, the expansion coefficient would be such that you should get slightly over a millimetre of length change with a heater giving 100 deg C temperature change. So the plan would be to have a completely sealed valve unit with a couple of feedthroughs connected to a small heater element which would have the voltage varied to vary the temperature of the needle: the higher the voltage the more the valve shuts off. I would hope it would be a lot easier to set a given voltage for the heater than to screw a valve knob to X number of turns
DaveC
Posts: 2346
Joined: Sat Jun 30, 2001 1:13 am
Real name:

Re: Accurate needle valves

Post by DaveC »

A thermally expanding needle is an interesting concept which might solve one major problem with needle valves: They turn on way too fast for good control....at most pressures. Because of this some folks use two valves in series, as has been discussed here. But making this actually work is tricky.

I have used a number of so-called high vacuum leak valves when I was doing work with a Mass Spectrometer.. P ~ 10^-6 Torr or below. I think at one time I had one of every type adjustable leak valve on the market. Several were in the $3-5K apiece price range and all were manual. One major problem with the needle valves in the zero-to-just-on leak range, is the end play in the barrel threads. This type of leak valve is usually built around a 0 - 1" micrometer. While you can barely feel any play or backlash between the thimble thread and the nut, it seemed always to be enough to make the turn-on rather abrupt. Once flow was established, then control was usually very nice.

Tom Dressel's idea of using the capillary throttle ahead of the valve seems real good, too.

I would suggest trying your idea...I don't remember seeing anything like it on the market...but I'm not sure.
guest

Re: Accurate needle valves

Post by guest »

About the only problem I can think of with the idea, is that it would be open full throttle when cold/off, but it should be possible to modify the geometry to reverse that effect.
DaveC
Posts: 2346
Joined: Sat Jun 30, 2001 1:13 am
Real name:

Re: Accurate needle valves

Post by DaveC »

I think this would be a minor a inconvenience. It is usually necessary to pump both sides of the valve or open it full first, to clear out the atmospheric gasses.

Then it would natural to energize the heating element to cause the valve needle to expand and seal the valve tightly. To create the controlled leak, you would just reduce the heater power somewhat... (how much would take a little practice) to begin the flow.

Sounds plausible to me.

Dave Cooper
guest

Re: Accurate needle valves

Post by guest »

I've given the matter some more thought: picture a metal tub say 10 inches long with flanges and fittings on both ends, inside the tube you have a needle valve seat at one end, at the other end, you have the base of a needle valve shaft attached. These are both fitted so that the needle is tightly seated at room temperature (brazed?) and no gas flows. By fitting an external heater to the outside of the tube wall you can make the tube lengthen and pull the needle by it's base away from the seat. The only thing to watch out for would be the needle base must not get too much heat transferred from the tube via it's attachment or via radiation/conduction from the tube wall. The advantages would be the valve is off when cold/no power applied, and you do not have an electric heater inside a sealed chamber with hydrogen gas.
guest

Re: Accurate needle valves

Post by guest »

The only thing I can think of is that you will have to acount for transient behavior. Once steady state is achieved, you should get nice control. The needle or tube, which ever component you heat, will have to be well insulated to eliminate changes in voltage/flow calibration due to ambient temperature changes. Neat Idea. I have had problems with needle valves in some refrigeration experiments that I did in graduate school. The only problem for my aplication would have been the heat transfer to the gas being metered.
guest

Re: Accurate needle valves

Post by guest »

After thinking a little more it occured to me that what you need is not to insulate the valve, but supply a constant heat transfer coefficient at a constant temperature if you want to control the energy input to the heater. Insulation would act like a damping device. You need to be able to remove heat to a constant temperature at a constant rate in order to get good response from the valve.

It would probably be easier to maintain constant energy input to the heater, and vary either cooling air or water flow to the needle. Unless of course you build some type of fancy adaptive control that takes the room temp into account. This should be fairly easy to do with a microcontroller and thermocouple.

If all you want to do is read a flow gauge and change the energy to the heater manually, or in an automated fashion, this is all needless complication. If you need repeatable flows for a given wattage input it will require some tinkering.
guest

Re: Accurate needle valves

Post by guest »

You are right about not insulating the valve Monty. I'm working on a quick prototype with an aluminium tube wall if the solder I'm using is ok for vacuum work (18%tin, 80% lead ,2% silver). I was all set to wrap the whole assembly in fibreglass cloth once finished. I was worried about insulating the needle from the heated tube but so long as the expansion coefficients are reasonably different it should work ok.
Tom Dressel
Posts: 287
Joined: Fri Jun 22, 2001 4:44 pm
Real name:

Re: Accurate needle valves

Post by Tom Dressel »

It is a creative approach to the problem. But if the goal is to set the knob on a variac and expect to get a consistant and reproducable chamber pressure, the device described probably will not do it. I have found that the leak rate needed to provide a given pressure depends on a number of factors. Probably the most important, is where the chamber has been in terms of pressure. That is, if I recently pumped down from air, the leak to maintain say 5 microns, is much less than if I have been valved off all night and recently pumped down from 50 microns.

In general a medium sized leak, on the air side of a capillary tube works well over a fairly wide range of pressures (3 to 50 microns) and chamber "operating history". For the capillary tube I use a 24" length of 0.005" OD stainless steel, purchased from Small Parts Inc.
guest

Re: Accurate needle valves

Post by guest »

I'm hoping that I'll end up with a valve with fairly fine adjustments compared to a mechanical type and that is also easy to seal. I seriously doubt it'll give repeatable results under all conditions as there are all sorts of thermal "can of worms", like how long will it take flow to follow the heater settings, not to mention how long it will take to reach equilibrium as the needle begins to heat up. I hope a unit made as small as possible with a feedback loop temperature controller as standard, will give a needle valve that can be manually throttled back to a flow rate suitable for a fusor. Eventually if it gave fine enough flows, you could have it under micro control with a look up table to eliminate any nasty quirks.
Post Reply

Return to “Fusor Construction & Operation (& FAQs)”