Does anyone have an Agilent 16452A Liquid Test Fixture?
|
On Fri, 22 Oct 2021 at 20:00, Askild <megafluffy@...> wrote:
Hi, Thank you very much for the very? kind offer.? Any photos and physical measurements you can take would be very helpful There are 4 SMA connectors, which connect to Hcur, Lcur, Hpot & Lpot. Can you measure with a multimeter and determine which pieces of the fixture connect to each of the 4 connectors? It looks as though Lpot and Lcur connect to one side, and Hpot and Hcur connect to the other. But where do each go? What pieces are electrically isolated from all 4 connectors, but electrically connected to the chassis? The manual says that there are various values of capacitance, depending on what size plates are connected. Would I be right in assuming that if you assemble this, but don’t fully tighten the thumb screws, that those capacitance values decrease, since you are increase the spacing between the plates? Am I right in assuming that the gap the manual speaks of, is the gap between the electrodes when fully assembled? I assume that the ceramic is fixed to each side, and can not easily (if at all) be removed. Can you see how it is fixed to the body? I assume that ceramic sits a certain distance below the top surfaces of each side. What is that distance? Is the ceramic tapped, so you screw into that when you insert the electrodes? If so, how far approximately do the screws go in? Can you measure the depth of where it's tapped? I am obviously seriously confused about some aspects of the assembly.? I can't understand why increasing the thickness of the electrodes increases the gap. I would assume if you increase the thickness of the plates, that would decrease the gap between the electrodes, not increase it. Let's assume hypothetically the ceramic is 4 mm below the surface on each side, so the space between the two pieces of ceramic are 8 mm. Then if you insert 3 mm thick electrodes each side, you take up 2 x 3 = 6 mm with metal, leaving a gap between the electrodes of 8-6=2 mm. That would match one of the data pints given in the manual.
3.0 mm thickness, 2 mm gap Now, if you use thinner electrodes (say 1.3 mm) I would expect the gap to increase to 8 - 2 x 1.3 = 8 - 2.6 = 5.4 mm. But instead the? manual says
1.3 mm thickness, 0.3 mm gap. How can using thinner electrodes result in decreasing the gap? I would have expected the dead opposite! I'm obviously totally mis-understanding the construction there. There appears to me to be 3 connectors for liquid. 1) One of the side, at an angle, where liquid is inserted. 2) One at the bottom, where liquid is drained. 4) MAYBE one at the top, where I have put a question make on this drawing, I can understand that trying to put liquid into this structure, with the bottom sealed, would be impossible unless the air could get out. So I assume at the top is an air-vent, where I have put a question mark on the attached diagram. Is that true? There appears to be a lip on the electrodes. Am I right in assuming that the lip is placed in such a way that it goes over the ceramic? I assume that it perhaps helps centre the electrodes on the ceramric. As you can see, I am a bit confused about the construction of this! Dave |
|
On Fri, 22 Oct 2021 at 20:00, Askild <megafluffy@...> wrote:
One other thing. All the other fixtures I have come across that connect to the LCR meter do so with BNC connectors, as that's what the LCR meter uses. Yet this fixture is different. Given the LCR meters uses BNC connectors, why do you think HP used SMA connectors on the unit, then supplied adapters to BNC? Why did they not just fit BNCs? Could it be that the connectors need to be hermetically sealed, and that it's impossible to buy hermetically sealed BNCs, but it is SMAs? I know hermetically sealed Ns and SMAs are available, but I have never looked for BNCs, but it would not totally surprise me if they don't exist. Dave |
|
Looks relatively simple to me.? Knowing the area of the plates and the spacing between them, you can easily calculate the dielectric constant of the material between the plates.? You have a fairly good picture of the disassembled fixture, and more details can be found here:
That paper tells us that the electrodes are nickel plated cobalt (Fe 54%, Co 17%, Ni 29%), the insulator is Alumina (Al2O3), and the O-ring is Viton (Fluro rubber). There are a number of papers available, both on Keysight's website and other places, on measuring dielectric constant of water and other liquids, including papers from NBS (now NIST).? There are published tables of dielectric constants of dozens of materials.? I would probably start by reading all the available literature (there's a fair bit on the Keysight website, and likely more available in old HP literature and elsewhere), and then I would build a fixture and start making measurements.? That, of course, is where you would start finding the "unknown unknowns" (a term used by engineers long before anybody ever heard of Donald Rumsfeld).? As problems show up, you simply attack them, one by one.? An engineer I once worked with called this, appropriately, "peeling the onion."? Just poking around the web, I see that the dielectric constant of some liquids varies with temperature and pressure, so you'd want to factor that into your measurements.? (That would be one layer of the onion.) I was once involved in the development of a soil moisture meter.? Most soils have a dielectric constant around 4, while water has a dielectric constant of 80 at 20° C, as I recall.? The dielectric constant of the soil/water matrix is therefore mildly dependent upon the type of soil, and highly dependent upon the concentration of water in the matrix.? Fascinating stuff. Do let us know if you find anything interesting. |
|
On Thu, 21 Oct 2021 at 09:22, Jim Allyn - N7JA <jim@...> wrote:
There’s a document from NPL that discourages the use of the term “dielectric constant”, since it is not a constant, but as you say, varies with temperature, pressure etc. You can’t have a variable constant! The two words are mutually exclusive.? The correct term to use is permittivity, or relative permittivity. Note it’s a complex number, with a real and imaginary part, with the imaginary part being the loss. Er = Er’ - j Er” ((Note the dash and double dash). Yes, I understand the basic way they Keysight fixture works, but I don’t understand how the thickness of the electrodes changes the spacing the way it does. I can’t understand how thicker electrodes results in a wider gap. ? Dave Dr. David Kirkby, Kirkby Microwave Ltd, drkirkby@... Telephone 01621-680100./ +44 1621 680100 Registered in England & Wales, company number 08914892. Registered office: Stokes Hall Lodge, Burnham Rd, Althorne, Chelmsford, Essex, CM3 6DT, United Kingdom |
|
On Sun, 24 Oct 2021 at 18:46, Askild <megafluffy@...> wrote:
Yes, I understand that now. I assumed that this would have been made from stainless steel for maximum corrosion resistance. But I see it is 100% nickel, which is blinkin expensive! I could not actually find anyone with suitable material online, but I did not look that hard. I think I would probably get it made from a grade of stainless steel with a high nickel content, as that's much more affordable. I don't have any specific use cases for it - the one time I did want something similar for oil, I found out from someone at NPL that you can't measure the loss of low-loss liquids using an LCR meter. If making one, I would bury a couple of pt100 platinum resistance thermistors in it, so the temperature can be measured accurately. Dave . |
|
On Sat, Oct 23, 2021 at 06:35 PM, peter bunge wrote:
Why the Cobalt alloy?? It may as well be solid nickel or nickel plated copper.? I see there are at least a few hams here, so undoubtedly you will know about radio repeaters, and the duplexers (filters) they use to allow simultaneous transmission and reception with only one antenna.? The tuning rod in duplexer cavities is made of invar, which is 36% nickel and 64% iron.? Invar has a low thermal coefficient of expansion, and is used so the frequency of the cavity won't drift as the temperature changes.? I know of mountaintop repeater sites locally where the ambient temperature can vary at least 50° C. with the seasons.? There are other similar alloys such as Inovco and FeNiCo that also have low coefficient of expansion.? HP says the electrode is nickel plated cobalt (Fe 54%, Co 17%, Ni 29%).? I'd guess thermal coefficient of expansion is one of the reasons for this choice, perhaps combined with corrosion resistance, thermal conductivity, electrical resistivity, and HP-only-knows what else. |
|
On Mon, 25 Oct 2021 at 06:00, John Lyles <jtml@...> wrote: When I worked for E. I. DuPont Co. 30 years ago, we had a coaxial fixture made from brass, that had a connection on the bottom (banana pins?) that plugged into the top of a Hp 4342A Q meter. With a standard coil to establish resonance, we could measure low loss dielectric fluids quite accurately. ? I spoke to someone at NPL who said that to measure low-loss liquids, you do indeed need a resonance technique, and look at the Q with and without the liquid. The Q would be calculated from the 3 dB bandwidth. The basic point I was trying to make is that you can't measure the low-loss liquids by measuring the impedance with an LCR meter. There's a large document from NPL I have cited before on this topic. Dave |
|
On Sat, Oct 23, 2021 at 05:03 PM, Dr. David Kirkby, Kirkby Microwave Ltd wrote:
"These chips Are fries, This queue A line, This tap A faucet, Wardrobe A closet, Vacation Holiday, Underground Subway, Chemist A drug store, Autumn Is fall, A garden Is yard, You see now It's not hard" From "English to American" video: https://www.youtube.com/watch?v=5Na7b5MiVak |
|
On Tue, 26 Oct 2021 at 02:21, Jim Allyn - N7JA <jim@...> wrote: On Sat, Oct 23, 2021 at 05:03 PM, Dr. David Kirkby, Kirkby Microwave Ltd wrote: "The United States and Great Britain are two countries separated by a common language." The term "dielectric constant" is widely used in the UK, so it's not just the USA which uses it. However, if you think about it logically * Pi is a mathematical constant (3.14159..). * c is a physical constant (299792458? m/s, which is approximately 3 x 10^8 m/s) * something that changes can not be a constant. Science and engineering advances, and just because we have always done it one way, does not mean we should continue to do it that way. NPL's argument for discouraging the use of the term dielectric constant is to discourage people from thinking they are measuring a constant. If you accept that relative permittivity is a better term, you can then, when appropriate, mention it is actually a complex number. The relative permittivity Er of a material is a complex number, normally written with a dash and double-dash like Er = Er' - j Er'' where the Er'' is the loss term ? Just because * you Americans have used the term "dielectric constant" for years * I myself have used the term "dielectric constant" for use * many Brits have used the term for years does not mean we should continue to use such a term, when someone points out that it is not a constant but a variable. Feel free to continue to call this variable a constant if you wish - many Brits will no doubt do it too. But this particular British person is not a stick in the mud. Dave G8WRB |
|
On Tue, Oct 26, 2021 at 12:33 AM, Dr. David Kirkby, Kirkby Microwave Ltd wrote:
If you accept that relative permittivity is a better termOh, I agree that relative permittivity is probably a better term, and it is even possible that I will get used to it in the 20 or 25 years I may have left in this life.? I did, after all, get used to saying "Hertz" instead of cycles per second (or cps), "megahertz" instead of megacycles, and gigahertz instead of kmc.? But it's tough to stop using a term you have used for decades, especially when everybody around you still uses the old term. I try to think in metric sometimes, and it's getting a little more comfortable - very slowly.? Perhaps some time we will happen to be in the same country at the same time, and we can discuss it over a pint or two, which I understand is still the only legal measure used for beer in the UK.? In the meantime, remember: “Science advances one funeral at a time.” - Max Planck |
|
On Fri, 22 Oct 2021 at 20:00, Askild <megafluffy@...> wrote:
Hi, Can you determine how the electrodes are fixed into the outer case? I understand that there's one electrode on each side. The electrodes must be electrically isolated from the main body, so I assume they sit on? the ceramic insulator. But what holds them to the ceramic? What holds the ceramic in the case? I'm trying to work out all 3 of these conditions can be simultaneously satisfied. 1) Wires connect the electrode to the SMA connectors. 2) Electrode attached on an insulator 3) Insulator attached to the body. Apart from using an adhesive, which I doubt is used, I can't see how everything can be fixed in place. One other thing. There's an O-ring. Why not two, given there's a spacer? I would have thought there was a potential for leaks both sides of the spacer, so two o-rings would be needed. Dave |
Dr. David Kirkby, Kirkby Microwave Ltd
Jim Allyn - N7JA