Dear Peter, We don't understand the plots you sent in last mail. They are definitely interesting, but one should know the experiment conditions to understand them. I attached my calilbration comparison. With regard to long term measurement with stable sensor temperature, I afraid we have no time for that. In that case we need: 1. some PT-100 sensors to measure Humidity sensor temperature and ambient temperature also. 2. a heater to stabilize the HS temperature (fortunately I have additional DAC channels to do it on our new control board) 3. temperature dependence data of the calibration. This is much easier to stabilize HS at the temperature higher than hall T, then to keep it at 21 degC, as it was calibrated. So we need a formula to convert dewpoint from 21 degC to e.g. 35 degC. Now we plan to measure the moisture at the Pd filter output. Theoretically it should have a zero humidity. Could you please update your addressbook? I'm basically using pkravt@gmail.com address now. It is much better then my old servers because it filters spam very good and mailbox is 2Gb. Best wishes, Peter. Wednesday, June 29, 2005, 11:00:29 PM, you wrote: > Dear Peter et al., > These are interesting plots in the attachment (though I don't > understand the differences between the individual curves, different > sensors?) I suggest to try a long term measurement > at the same sensor temperature, to observe the recovery. > Peter, could you send me your spreadsheet. Perhaps our data is the same, > hard to see on a log plot. > Best regards > Peter > ---------- Forwarded message ---------- > Date: Wed, 29 Jun 2005 12:49:04 -0400 > From: info@kahn.com > To: Peter Kammel <kammel@npl.uiuc.edu> > Subject: Re: Humidity sensor (fwd) > Dear Peter: > I apologize for the day delay in responding. > We have observed on many occasions that when a sensor and sampling system is > exposed to > a temperature transient there is a two-phase reaction. The first is an > impulse spike in > moisture resulting from the disturbance to the system equilibrium. This can > be a positive > excursion on heating (outgassing) but also, interestingly, a negative > excursion (moisture > retention) on cooling. If the temperature is than allowed to either return > to “normal” or to > stabilize and hold at the new value, this impulse reaction will subside as > the moisture level > corrects itself dynamically (time frame depending on the dewpoint temperature > being > measured but at low levels typically 16-24 hours). You will then see a final > steady state offset > between the original steady temperature moisture sensor output and the final > steady state > output at T’ > It is the difference between these two signals that is a measure of the > sensors true > temperature dependence. Typically this will be in the region of 0.1 deg C / > deg C as specified > in the product literature. > The attached chart shows the full data acquisition from a 10 deg C to 20 deg > C ambient > temperature variation at a steady moisture level of –90 deg C dp. As you can > see, the impulse > part of the reaction is as much as 14 deg C or thereabouts. However the > steady state values > are only a few deg C higher than the original starting conditions. > If you would like to discuss this matter further, I would suggest a > conference call with Mr. Andrew Stokes, the Technical Director at Michell > Instruments, who is most knowledgeable. We could potentially have this > discussion on Thursday morning at 9:00 am CDT (although I would have to check > on Mr. Stokes' availability) or next week. > I look forward to hearing from you. > Sincerely yours, > Dave Alan > KAHN INSTRUMENTS, INC. > Peter Kammel wrote: >> Dear Dave, >> >> I tested the calculator and got exactly the result I expected: >> ppmv @ x bar = ppmv @ 1bar /x. >> Example: >> frostpoint (dewpoint) = -50 C, corresponds to 38.84 ppm @ 1 bar and >> to 3.88 ppm @ 10 bar. >> >> Thanks for the attached article, which is educational. >> >> I asked my colleagues at PSI about their measurements. They measured the >> T dependence by slowly cooling the sensor and warming it up again. >> There was no hysteresis observable. Initially the noticed the different >> readings, when the door of the experimental hall was opened and the >> overall temperature changed. At constant temperature the sensor is very >> stable. >> >> We are waiting for your advice. >> >> Thanks >> >> Peter >> >> On Mon, 27 Jun 2005 info@kahn.com wrote: >> >> > Dear Peter: >> > >> > Attached are the following documents: >> > >> > * An article providing practical suggestions on improving response >> > speed at low dewpoints >> > * A humidity calculator >> > >> > I will call you tomorrow with comments regarding temperature effects on >> > the Pura sensor. >> > >> > Sincerely yours, >> > Dave Alan >> > KAHN INSTRUMENTS, INC. >> > >> > >> > Peter Kammel wrote: >> > >> > >Dear Dave, >> > > >> > >That's the first report I got from our collaborators in Switzerland. >> > >I will call you soon and ask for your advice. >> > > >> > >Best regards >> > > >> > >Peter >> > > >> > >Peter Kammel / pkammel@uiuc.edu >> > >Department of Physics, Loomis Laboratory >> > >University of Illinois at Urbana-Champaign >> > >1110 West Green Street, Urbana, IL 61801 >> > >Tel (217) 333-5424 / Fax (217) 333-1215 >> > > >> > > >> > >---------- Forwarded message ---------- >> > >Date: Mon, 27 Jun 2005 17:11:40 +0200 >> > >From: Peter Kravtsov <pkravt@gmail.com> >> > >To: Peter Kammel <kammel@npl.uiuc.edu> >> > >Cc: Claude Petitjean <claude.petitjean@psi.ch>, >> > > Malte Hildebrandt <malte.hildebrandt@psi.ch>, >> > > Francoise Mulhauser <Francoise.Mulhauser@psi.ch>, >> > > Bernhard Lauss <lauss@berkeley.edu> >> > >Subject: Humidity sensor >> > > >> > >Dear Peter and colleagues, >> > > >> > >Attached is our news and questions about the humidity sensor. >> > > >> > > >> > > >> > -- Best regards, Peter mailto:pkravt@gmail.com
Attachment:
HS_cal_test.xls
Description: MS-Excel spreadsheet