Have you tried Quartz? Depending on structure (crystalline or "fused"), it can have 3 to 10 times the thermal conductivity of "normal" glass. Higher power flashlamps are quartz-made for just that reason.
As another possibility, can you coat the perimeter walls of the cell to make them IR-opaque (unless the cell material is IR-opaque anyway) and use radiated heating instead? This would also increase the distance between the cell and the heater (with its unavoidable magnetic fields).
You can make the "unavoidable" magnetic fields pretty small by twisting the heater wires. Twisted pair isn't as perfect as coaxial cable, but it's a lot better than a plain coil.
It does suffer from another problem - it isn't cheap. George's stuff seems to be aimed at teaching labs, and keeping the price reasonable is part of his job.
change in heat input, you are generally better off if the time constant is long - external changes are kept well away from the area you are controllin g.
ture sensor is appreciably separated from your heating element, life get mo re complicated. If you are using a PID algorithm to control the temperature , the proportional gain then has to be kept pretty low to prevent oscillati on, and you rely on the integral term to get decent regulation.
Hi Bill, This was a bit of a throw away question. (Well it's a concern, b ut I don't have time to fix it.) Here are some pics of the heater.
formatting link
The thermal couple is in the black heat shrink and stuffed up next to the Rb cell.
So because this is only a heater there are two thermal time constants. There is the heating time which is determined by how much power I have and the heat capacity of the gizmo. And then there is the cooling time which is the same heat capacity times so me thermal resistance to the outside world.
If there is too much insulation then any over temperature situation could take a long time to come back down.
I'm not sure exactly how to balance this, but I feel that at the maximum te mperature I'd like to be using 'most' (maybe 60-70%??) of the available pow er. (I'll run it up to 50 and 100 C and report needed powers.)
In the 3rd pic, is the heater the wires in the outer, amber band? And is that insulated from the Rb cell by the white stuff? That will have a huge time delay, and the loop dymanics will be tricky.
Can you attach the heater directly to the cell, or at least through a better thermal path?
Yes, two large thermal masses, lightly coupled.
--
John Larkin Highland Technology, Inc
picosecond timing laser drivers and controllers
jlarkin att highlandtechnology dott com
http://www.highlandtechnology.com
a change in heat input, you are generally better off if the time constant is long - external changes are kept well away from the area you are control ling.
erature sensor is appreciably separated from your heating element, life get more complicated. If you are using a PID algorithm to control the temperat ure, the proportional gain then has to be kept pretty low to prevent oscill ation, and you rely on the integral term to get decent regulation.
, but I don't have time to fix it.) Here are some pics of the heater.
=0
Yeah... the amber is kapton tape holding the heater wire in place. Making it even worse the wire is on the outside of a glass cylinder, so the heat has to go through the glass, then through the air, and then to the ce ll.
There is a constraint that I didn't mention. The cell is full of Rb, and I want to keep the Rb off of the front face of the cell... Rb on the glass i s a mirror and reflects the light. So I want the heat to go into the cell through the exposed front faces.
(In a later similar design I put a cold finger that connects the Rb cell to the outside world... again kinda weakly.)
I measured the power (voltage) vs temperature set point. Maximum voltage is ~26 volts.
Temperature Votlage 50 C 10.5 70 C 13.5 90 C 18.0 100 C 20 v
The max Temp is ~120 C So I don't think I have much wiggle room.
If they are only one-offs then the cost of the system is not an issues. If he were making something which a1000 labs were going to use, it MIGHT have a bit more significance.
And IR thermometry is a lot cheaper than it was back when you set foot in a lab.
Wind turbines are unlikely to kill anyone except birds and small-plane pilots, but if the gearboxes fail after 5 years or so, the economics will get trashed. I've seen references to a megabuck to replace the gears on a big turbine, more if it's offshore.
--
John Larkin Highland Technology, Inc
picosecond timing precision measurement
jlarkin att highlandtechnology dott com
http://www.highlandtechnology.com
As a pilot, I would suggest nobody is going to fly close enough to the ground to hit a turbine blade. You can see them on Altamont pass near Livermore. One sighting is enough to convince you to climb to a safer altitude:
formatting link
Helicopter pilots are the exception. They often fly into high voltage transmission lines in the fog at night.
Molten Salt Reactors are the solution. Thorium will last forever. Thay can eliminate hazardous waste from pressurized water reactors. Respond automatically to changes in load. Cannot melt down - already melted. They are walk away safe. Miniscule amount of radioactive residuals, safe after hundreds instead of hundreds of thousands of years. Already proven in operation. Can operate in small local reactors instead of multimegawatt operations that lose power transmitting over long distances. Dependable, predictable 24 hour operation instead of relying on weather and sunshine. Can operate anywhere. Much higher efficiency. Thermal to electrical efficiency of 45% instead of 32-36%. Low pressure operation increases safety. No need for lakes and streams for cooling. Can be can be air- cooled at little loss in efficiency.
formatting link
Many significant advantages that cannot be duplicated in existing technologies. All we need is a small portion of the investment in solar, wind and existing nuclear technologies to solve the remaining problems and bring this technology to fruition.
You are trolling. You cannot believe that burning trillions of barrels of gas and oil is good for the planet. Among other things, this produces CO2, which results in:
Methane emissions from Siberia permafrost and Arctic seafloor resulting from increase in temperature. Methane is a much more powerful greenhouse gas than CO2. This is positive feedback in operation.
Ocean acidification along with temperature increases are moving food species further north and killing coral, which is home to many food sources.
Ocean level is rising, which will flood New York and most of the East seacoast. Sandy shows how vulnerable the area is.
Thorium reactors could help solve this. Unfortunately politics will ensure nothing is done. Energy companies own the government, which prevents any threat to their future profits.
The current methods of producing energy are not sustainable. The risks are clear. There is a cheaper solution that is readily available.
Regardless if you believe in global warming or not, Thorium reactors are cheaper, safer and more reliable than the alternatives. Reducing CO2 emissions is just a welcome byproduct.
55.8 million years ago and lasted for about 170,000 year. There was a big n egative spike in the carbon isotope ratio's in the carbonate laid down then , and a runaway methane release seems to have been involved.
There's about 10 metres of sea level rise tied up in the Greenland and West Antarctic ice sheets. They'd take many centuries to melt in place, but if they slide off into the ocean - as the Laurentian ice sheet seems to have d one at the end of the most recent ice age, the sea level rise could happen a lot faster, and rather less predictably.
Thorium reactors may turn out to generate cheaper electricity than solar pa nels, and their waste may turn out to be easier to handle than the radioact ive waste from current reactors, but any claims about price, safety and rel iabilitity are just claims until somebody builds a second one - the one tha t the AEC ran from 1965 to 1969 for a total of 15000 hours was just a proof
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.