Shop OBEX P1 Docs P2 Docs Learn Events
Idea on measuring vapor? — Parallax Forums

Idea on measuring vapor?

James LongJames Long Posts: 1,181
edited 2009-04-02 06:52 in General Discussion
Ok, so I'm working on new machine, and need to measure a vapor height.

Does anyone have an idea on how to do it?

Here is some specs:

The vapor will be at 260C.

The vapor height will vary.

The vapor will displace the air. The vapor is heavier than air, so the vapor will be contained in a blanket in the bottom of a container.

I'm thinking a capacitance tube type sensor, but want opinions before I go and design the senor. It will need to be of stainless or something similar.

Any input is appreciated,

James L

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
James L
Partner/Designer

Lil Brother SMT Assembly Services
«1

Comments

  • Mike2545Mike2545 Posts: 433
    edited 2009-03-21 01:38
    How sensitive does the measurement have to be?

    Can you see the vapor?

    Is it caustic?

    What is the specific gravity of the vapor?

    Can you get something to "float" on it?

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Mike2545

    This message sent to you on 100% recycled electrons.
  • Erik FriesenErik Friesen Posts: 1,071
    edited 2009-03-21 02:14
    You could have a photodiode looking down into the tube. At intervals there could be led's shining onto the vapor. Each led could be running at a set frequency.
  • PhilldapillPhilldapill Posts: 1,283
    edited 2009-03-21 03:18
    How about a PING sensor? The speed of sound through the vapor will be higher than through the air. If the distance between the PING))) and the bottom wall is fixed, you could calculate distance based on the time it takes for the signal to get back... That's if the vapor blanket is like 6" thick, and not 6mm...
  • James LongJames Long Posts: 1,181
    edited 2009-03-21 03:29
    Ok,

    I'll try to answer the previous questions in turn:

    Mike:The distance is pretty critical in this application, a few mm is the resolution needed. I do not know the specific gravity of the vapor, the data sheet of the material doesn't say. The material is inert other than the heat it will transfer to the sensor. The vapor is basically transparent, but it could be considered a very very light"fog". The vapor will not support anything that I know of, maybe aerogel.

    Erik: The temperature could be a problem. Because the vapor is at 260C it will condense onto the tube and transfer it's heat, causing the tube to be at 260C. Also, I'm not sure how different LED's could help in this situation.

    Phil: I do not think the speed would vary a great deal. It too probably wouldn't give the resolution needed.

    I appreciate all the input, I just can't think of any idea other than a number of thermocouples lining the wall in a vertical fashion. With the temperature, and it being vapor, I have run out of ideas.

    James L

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    James L
    Partner/Designer

    Lil Brother SMT Assembly Services
  • Mike2545Mike2545 Posts: 433
    edited 2009-03-21 03:50
    Not knowing exactly what you are trying to accomplish its hard to come up with a workable solution, if you need a certain level for a process I would suggest a dish within a dish so any overflow would be contained. If it's a volume then maybe scuppers of sorts could come into play. Maybe you could meter the flow somehow, if this vapor comes from a reaction/physical change then I'd say measure the solid/liquids that make up the vapor, at this point all we can do is guess at the possibilities until say "That's the one"

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Mike2545

    This message sent to you on 100% recycled electrons.
  • Beau SchwabeBeau Schwabe Posts: 6,572
    edited 2009-03-21 05:04
    James Long,

    If the air and the vapor could be considered a "binary gas mixture", then something along the lines of a acoustic gas analyzer might work for you.

    Reference:
    scitation.aip.org/getabs/servlet/GetabsServlet?prog=normal&id=RSINAK000057000011002837000001&idtype=cvips&gifs=yes

    My understanding of an acoustic gas analyzer is that a frequency is swept and applied to a fixed volume column of the gas mixture. Depending on the mixture ratio, the resonate frequency of the gas column will be affected.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Beau Schwabe

    IC Layout Engineer
    Parallax, Inc.
  • James LongJames Long Posts: 1,181
    edited 2009-03-21 07:40
    Mike2545 said...
    Not knowing exactly what you are trying to accomplish its hard to come up with a workable solution, if you need a certain level for a process I would suggest a dish within a dish so any overflow would be contained. If it's a volume then maybe scuppers of sorts could come into play. Maybe you could meter the flow somehow, if this vapor comes from a reaction/physical change then I'd say measure the solid/liquids that make up the vapor, at this point all we can do is guess at the possibilities until say "That's the one"

    Mike,

    The sensor is for a vapor phase reflow system I am designing. The system will be vapor injected (some what different that previous designs)and controlled for a certain rise of the vapor (more heat, more fluid, thicker blanket rising from the bottom). Because the vapor is the transporter of energy for reflow, it's level will be important to control. This system will do everything from preheat the board to a total reflow of the solder.

    Vapor fluid is a PFPE fluid which is inert, but has a high boiling point (260C). Its condensation point will be at the same temperature. We are planning to use Galden HT260 if you want to look up it's data sheet. That may provide some information if your interested.

    Because the vapor is the method of energy transference, the height (when close to the board) will be critical. It's rise speed will be controllable, but needs to be known. If the rise is too fast, the board will be heated too fast (above the 3 deg C/second for most designs).

    I have other ideas for monitoring it, but still the rise speed of the vapor is a good indication. We will need to slow down the vapor rise once we get close to the board.

    I am going to use a non-touch temperature sensor (Parallax unit probably) to monitor the PCB for it's temperature rise, but in the beginning it will be hard to calibrate the system without a system to monitor the vapor rise.

    I have thought of different ideas, but still I can not come up with an idea that helps. Maybe two metal strips extremely close to each other. Because the vapor will condense on these as it heats them up, I could use the liquid to short the strips together, moving vertical as the vapor climbs. I'm not sure how close they would need to be, but it would be in the thousands of an inch range.

    That should help everyone understand the exact need.

    Beau: Not sure if that would work or not. I imagine it would, but cost could be a determining factor for using the acoustic gas analyzer.

    James L

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    James L
    Partner/Designer

    Lil Brother SMT Assembly Services
  • James LongJames Long Posts: 1,181
    edited 2009-03-21 10:30
    Ok,

    How about a capacitance sensor, but with a twist.

    Since everything below the vapor line (top barrier) would come to the vapor temp of 260C. I come up with an idea, but want you guys to suggest problems with it.

    How about two identical copper plates that are tapered width wise from top to bottom (top being the widest part). These plates would act as two plates of a capacitor. Now here is where it gets weird.

    The plates would be spaced very close together for the vapor to condense causing a film to form between them (don't know hot to space them in a 260C environment at this point). The film would only be present in a short space at and just below the vapor barrier (another assumption). The lower part of the tapered plates (being at the vapor temp) would prevent the condensate from running to the bottom (I'm assuming at this point).

    Then we could use the plates as a capacitor. With a few trial runs, the level could be figured out.

    The plates would need to be connected to the side wall of the reflow chamber to help them release their thermal energy and not let the thermal energy travel above the vapor barrier (preventing the condensation of the vapor).

    Here are some numbers for anyone who knows how to figure out the capacitance of two plates, and can theorize on the idea.

    The fluid has the following properties:

    Dielectric Strength
    kV (2.54mm gap)

    40

    Dielectric Constant

    2.1

    Volume Resistivity
    Ohm-cm

    1015

    I can live without a level sensor, but the first state of the heating process will be problematic without it.

    Any one have any other ideas that do not cost a fortune. Thanks Beau, just seems a costly method.

    James L

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    James L
    Partner/Designer

    Lil Brother SMT Assembly Services
  • Mike2545Mike2545 Posts: 433
    edited 2009-03-21 13:09
    Instead of closley monitoring the vapor level, why not have·a pool of vapor, one that you can keep full and dip the boards in it till they reach temperature. You could use a platform to raise and lower the board into the vapor while monitoring the temperature.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Mike2545

    This message sent to you on 100% recycled electrons.
  • kwinnkwinn Posts: 8,697
    edited 2009-03-21 15:59
    How about a thermocouple mounted on a carriage that moves up and down on a threaded rod with a servo circuit to maintain it at a constant temperature? The position and rate of movement should provide the data you need.
  • LeonLeon Posts: 7,620
    edited 2009-03-21 16:02
    Many vapours absorb IR much more than air, you could probably use that property.

    Leon

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Amateur radio callsign: G1HSM
    Suzuki SV1000S motorcycle
  • James LongJames Long Posts: 1,181
    edited 2009-03-21 16:28
    OK, another round of answers

    Mike: because a PCB shouldn't increase in temperature more than 3 deg C per second, it is very hard to control the board temperature rise with the dip method. Also the carriage adds a lot of moving parts (precision) which are present in the high temperature zone. The level (flatness) of the board would be critical as well.

    Kwinn: I have thought about that, but a thermocouple (purchased K type) typically has a metal mass protecting the junction. It would heat well, but cool slow. It would need a heat sink mounted to it to help reduce it's temperature (or give off thermal energy). I think it would be hard to chase the fluctuation.

    Leon: I'm just not sure. You have me on that information. I just don't know much about vapor IR absorption. I would be afraid the IR sensor would be saturated from the vapor IR. Would the vapor give off IR?


    At this point, I think the copper plates would be the easiest test. I'm not sure how to space the plates, for most insulation materials (very thin materials) do not agree with the high temperature. I'm assuming with the width change causing more dielectric between the plates as the trapped film rises, will give a variable reading.

    We are not to the point of testing yet, there are other parts to be made for the chamber before testing can commence. Also I must spring for the $700.00 gallon of fluid. Yes, it is very expensive.

    So far the design is pretty unique and should be a viable solution for small production companies, considering the next similar system costs $35,000.00

    Because we are doing something unique, I guess saving money theorizing is not going to work. It will be trial and error to find a solution that works well.

    James L

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    James L
    Partner/Designer

    Lil Brother SMT Assembly Services

    Post Edited (James Long) : 3/21/2009 4:36:23 PM GMT
  • Mike2545Mike2545 Posts: 433
    edited 2009-03-21 17:02
    James Long said...
    OK, another round of answers

    Mike: because a PCB shouldn't increase in temperature more than 3 deg C per second, it is very hard to control the board temperature rise with the dip method. Also the carriage adds a lot of moving parts (precision) which are present in the high temperature zone. The level (flatness) of the board would be critical as well.

    How are you going to control the board temp by raising the level of vapor vs dippping the board? Isn't it essentially the same?

    I just thougth it might be a less challanging way to contol the process if you were to dip the boards rather than flood a chamber with vapor.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Mike2545

    This message sent to you on 100% recycled electrons.
  • kwinnkwinn Posts: 8,697
    edited 2009-03-21 17:02
    James, you have a valid point on the cooling time of the thermocouple. I wonder if something like nichrome wire or something similar would work as a temperature sensor. As for thin insulation materials how about mica? Not sure what max temperature it is good for, but it is fairly high.

    Checked mica melting point, and it is over 1000 degrees C. Also Nichrome and similar heater wires have a temperature coefficient of resistance that should be simple enough to measure if you decided to go that route.

    Post Edited (kwinn) : 3/21/2009 5:33:22 PM GMT
  • kelvin jameskelvin james Posts: 531
    edited 2009-03-21 17:28
    If the vapour is dense enough, a laser beam would become visible passing through it. Possibly a photo-diode or a something like a camera sensor could work as the light detector, which could be mounted over the tank. Whether the laser diode could take the heat may be a problem. Just a thought that may be worth exploring.


    **
  • Erik FriesenErik Friesen Posts: 1,071
    edited 2009-03-21 20:01
    I hadn't noticed the 250C temp so my led idea would hardly work. One thing though, I have one of those stanley laser tapes that have 1/8" resolution. The stickler is that it can't measure any less than 2 feet. Would the vapor be dense enough for optical sensing or are you still in the think stage and haven't seen the vapor?

    You will have to try the cap sensing idea but I would hazard a guess that there will be calibration issues on each use.
  • PhilldapillPhilldapill Posts: 1,283
    edited 2009-03-21 20:07
    Mike, I believe the difference with dipping vs. vapor, is that the vapor has a lower heat transfer coefficient. Hot air heats things much more slowly than hot water(I know water isn't the thing you are dipping it in, but the point is the same).
  • Mike2545Mike2545 Posts: 433
    edited 2009-03-21 20:30
    Phill, put something in a tub and fill it with water or lower something into a tub full of water, same difference...It just might be easier to lower the board into a pool of vapor vs flooding the area with vapor.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Mike2545

    This message sent to you on 100% recycled electrons.
  • PhilldapillPhilldapill Posts: 1,283
    edited 2009-03-21 21:27
    Mike, the two are very different... If you lower an object into a tub of water, there is a huge amount of thermal mass present to keep "feeding" the object with heat. If you have an object in a tub, and slowly start adding water, there is little thermal mass, causing the object to heat more slowly. What I was implying, however, was that the vapor itself has a lower specific heat than another substance like liquid solder, or flux or whatever.
  • Mike2545Mike2545 Posts: 433
    edited 2009-03-21 21:33
    If the time taken to lower the object is the same as filling the tub, I think you will find its about the same. I never said to drop it in.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Mike2545

    This message sent to you on 100% recycled electrons.
  • James LongJames Long Posts: 1,181
    edited 2009-03-21 23:03
    Ok,

    Since we are not face to face I will try to explain the difference of dipping vs vapor rise.

    With dipping the vapor is present with all of it's energy. When you dip you will need to control the descent of the board very precisely. This takes mechanical devices to control the board from swinging, and a system to control the board descent with thousands of inch resolution. The vapor will be constantly renewed from the power required to keep the blanket at a certain level (container will be condensing the vapor some). Although the vapor level will drop (because of the added thermal mass robbing vapor heat), the board will need to chase it at the rate needed to keep the temperature rise constant. This is simple, until designing the mechanical portions of the device. There are not many things which like the temperature of the chamber. I just feel the materials, mechanics, and controls are much more complicated with the dip method. Also you will need a way to monitor the board temperature. With a non-touch thermal sensor, the board being in motion can cause problems. If it moves with the board, the added complexity is just not my idea of fun.

    With injection, you can control the height with only a pump which pumps the fluid through a temperature controlled element and into the chamber. The pump is much easier to speed up or down to control the amount of vapor (height) in the chamber (remember that a constant supply of vapor is needed to maintain a certain height, because the chamber is condensing some of the vapor). If you need more vapor, just pump more fluid through the heating element, need less slow the pump down. This takes a huge amount of the hot zone mechanics out of the equation. There are more complications because the heating element can be only so hot, but it will be controlled separately to maintain the correct temperature (this is the case no matter the height method).

    Both methods are now in use for production (there are commercial machines which use both methods). I just can't justify $35,000.00 for a machine that only does what has been previously discussed. I know the process is complicated, but still doesn't justify that amount of money. I am putting my own twist to the design, but the principle is the same.

    From a board perspective, the methods are identical if the speed of heat transfer is maintained in compliance. It is the mechanical complexity which prevents the dip method from being the most practical.

    I hope that clears up the debate of dip vs injection.

    James L

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    James L
    Partner/Designer

    Lil Brother SMT Assembly Services

    Post Edited (James Long) : 3/21/2009 11:09:15 PM GMT
  • Mike2545Mike2545 Posts: 433
    edited 2009-03-21 23:49
    Yeah, I get it, I have used an insecticide fogger that put out a vapor, less sophisticated than you are talking about doing, I just remember flow control was difficult. Thought a different method might be move the part instead of adjust the smoke, as it were.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Mike2545

    This message sent to you on 100% recycled electrons.
  • $WMc%$WMc% Posts: 1,884
    edited 2009-03-22 02:14
    James Long

    Have You thought of hi-freq RF like microwaves? RF frequency above 1000MHz are really affected by water vapor in the atmosphere, (Doppler effect) the cure is to crank up the power to cut through this vapor in the commutations business.
    My Satellite connection has 1 Watt @ 1250MHz, I loose My signal to the Satellite in a good rain.or in heavy fog.

    I think a uWatt microwave transmitter and receiver might work for you.
    With no vapor 100% signal----as the vapor is applied the signal will degrade----Pure vapor 25% signal
    25%·signal from the receiver as A "zero" will verify that the transmitter is working·and aid in Quality control

    Since its a uWatt device it should consume Little power!

    ____Just a thought____$WMc%______uWatt = microwatts

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    The Truth is out there············································ BoogerWoods, FL. USA

    Post Edited ($WMc%) : 3/22/2009 3:25:14 AM GMT
  • Carl HayesCarl Hayes Posts: 841
    edited 2009-03-22 02:37
    No one seems to have suggested optical methods.

    (1)· Do you know the index of refraction of the vapor?· It's not very hard to measure.· If it is different from that of air, a light beam will bend when transiting the interface between the vapor and the air above it.· It most likely is different from that of air if the vapor is much denser than air.· Then you can measure the angle through which the light beam bends, or you can calculate the critical angle for complete reflection and detect that.

    (2)· Does the vapor rotate the polarization of a light beam passing through it?· Most don't, but some may be dextro- or levorotatory if the vapor is composed of a single species (levo or dextro) of a compound that has two mirror-image molecular forms.

    (3)· Does the vapor absorb radiation (light, radio, xray, whatever) differently, with respect to frequency (wavelength, or color), from the way air absorbs it?· That is, is there a pair of frequencies of which, say, air absorbs more of the first frequency and the vapor absorbs more of the other?· You could measure the position of the interface by comparing the absorbtion of the two when a ray passes vertically through the container.

    Non-optical methods:

    Number (1) above applies equally to sound waves, which will be refracted when passing through a boundary between gases that conduct sound at different speeds (just as light is refracted when passing through a boundary between gases that conduct light at different speeds (that is, gases·that differ in refractive index).

    Another way:· depending upon the depth you're measuring and upon the density of the vapor, the pressure at the bottom will vary with depth.· This pressure variation will be predictable if the vapor is all at the same temperature.· You could use a manometer, or other pressure sensor.· Or you could just weigh the container.

    Yet another:· what's the dielectric constant of the vapor?· If it's diffeent from that of air, you could make a capacitor whose capacitance varies with the depth, and measure the frequency change in an oscillator.

    Another:· what's the dielectric strength of the vapor, in KV/mm for example?· You could pass a spark through it and measure the peak breakdown voltage, perhaps in KV.

    That's all the storming my brain is up to at the moment, but there are probably thousands of ways to sense this.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    · -- Carl, nn5i@arrl.net
  • Beau SchwabeBeau Schwabe Posts: 6,572
    edited 2009-03-22 05:26
    I second Carl's suggestions. I had a similar thoughts using his spark method, and optical refraction description.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Beau Schwabe

    IC Layout Engineer
    Parallax, Inc.
  • potatoheadpotatohead Posts: 10,261
    edited 2009-03-22 10:29
    I'm up late musing over things.

    What about resonance?

    Higher frequency audio would have standing harmonics that would be disturbed by the vapor, differently than just plain air. Perhaps there is a fairly direct relationship between these and the size of the chamber and the desired height of the vapor.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Propeller Wiki: Share the coolness!
    Chat in real time with other Propellerheads on IRC #propeller @ freenode.net
    Safety Tip: Life is as good as YOU think it is!
  • Phil Pilgrim (PhiPi)Phil Pilgrim (PhiPi) Posts: 23,514
    edited 2009-03-22 17:20
    Can you see the rising vapor cloud from outside the enclosure, perhaps through a window on the front or side? Can it be backlit, perhaps through a window on the opposite side? Does it have sufficient opacity to create a visible demarcation between cloud and non-cloud in that circumstance?

    -Phil
  • W9GFOW9GFO Posts: 4,010
    edited 2009-03-22 17:52
    kwinn said...
    Also Nichrome and similar heater wires have a temperature coefficient of resistance that should be simple enough to measure if you decided to go that route.

    Wouldn't measuring the resistance of a vertically mounted coil of such wire work for what you want. It would have a very low thermal mass, is very cheap and it's response time should be near instantaneous.

    Rich H

    Post Edited (W9GFO) : 3/22/2009 6:03:34 PM GMT
  • James LongJames Long Posts: 1,181
    edited 2009-03-22 19:48
    I appreciate all the different ideas.

    Because there are many facets of proposals, I figured some reference materials were needed. Below is a website that you can take a look at the vapor cloud. This will be good reference to the vapor and it's properties. Because I have no idea the values when the pictures were taken, I can not say if the vapor is thick or thin. The following is just a reference for the vapor, and has very little in common with my design. The vapor is about the only item that is the same.

    http://www.ibrtses.com/projects/vapourphasesoldering.html

    Some of the questions you ask, I have no answer for, but I do appreciate the ingenuity shared on the ideas. I welcome anyone who would like to help find a solution.

    There are many things left to finish before I will have vapor to measure, but the project is progressing.

    I have some ideas to try, but I think Rich H. has the best method so far. I didn't think about the resistance chance of a heating element. I'm only worried about the heat transfer of the wire itself.

    James L

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    James L
    Partner/Designer

    Lil Brother SMT Assembly Services
  • Carl HayesCarl Hayes Posts: 841
    edited 2009-03-22 20:08
    $WMc% said...
    RF frequency above 1000MHz are really affected by water vapor in the atmosphere, (Doppler effect) ...
    It is certainly true that water vapor increases absorption at certain frequencies, but the Doppler effect·is a completely unrelated phenomenon.· No connection at all.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    · -- Carl, nn5i@arrl.net
Sign In or Register to comment.