coppercone - 3-7-2018 at 07:02
What uses are small microwave reactors? I know something you can buy is roughly 20,000$, 300w, stainless steel with internal pressure and temperature
sensors plus camera to monitor the reaction.
What uses do these have? I cant imagine it would cost more then 5000$ to make one that runs in the 5hhz band with a 300w magnetron, or like 1500$ for
a 2.4ghz one, and i am thinking of an electrically precise one, not something slapped together out of sheet metal waveguide.
DrP - 3-7-2018 at 07:20
You might want to break a bond at a targeted frequency that wouldn't break/react under normal conditions? That's my guess anyway. That and heating
mixtures.... and warming soup maybe.
coppercone - 3-7-2018 at 08:00
Im not sure about the frequency thing. A broad band high powrr microwave source requires pretty much solid state amplifiers, alot of them in parallel.
To get 300w across 2-18ghz you would need like 50000$ of mmic chips. To do it with tubes you would be limited to atound an octave, so like 2-4,
4-8,8-16 and so on, and it would cost like 30g for the tube.
Im pretty sure the microwave reactors for sale use fixed frequency like a microwave oven.
Also dont you need like +20ghz to resonate any molecular scale bonds?
A single mmic that does like 5 watts between 2 to18ghz is around 700$ from most common manufacturers. So 1400 for10, 14000 for 100w, 42k for 300w. You
would also have to make custom waveguide couplers or use high power coaxial which is non ideal. I dont know if you can use the same cavity for all the
reactions either. Dont know enough about microwaves yet. But it would also require a big rackmount power supply and water cooling i think. It would be
half the size of a rackmount box at least i think.
[Edited on 3-7-2018 by coppercone]
coppercone - 3-7-2018 at 08:15
http://www.analog.com/en/products/amplifiers/rf-amplifiers/p...
450 for 8 watts. So 30 grand in amplifier chips only.
Then broadband power rf design to 20ghz.
It would cost at least 150k to make manufacturing reasonable in 1000 a year quantity i think, if tge company is run by benevolent communists. Since
its a heater you might not need to put such a good power supply and noise design etc since your just using it as a energy source. So say 10k for the
electronics.
What frequency range would be useful for lab chemistry thats not weird long chain molecules or something?
[Edited on 3-7-2018 by coppercone]
coppercone - 3-7-2018 at 08:36
https://www.google.com/url?sa=t&source=web&rct=j&...
It looks like you are describing the athermal effects which are under debate.
Has anyone here used a microwave reactor? Is it worth developing?
Sulaiman - 3-7-2018 at 08:46
Or you could use LDMOS power transistors,
e.g. 250W rms (3x 120W rms) @ 2.1 - 2.2 GHz GBP126 incl. p&p ? https://www.ebay.co.uk/itm/AMPLIFIER-LINEAR-TRANSISTOR-250W-...
coppercone - 3-7-2018 at 16:06
Why not just use a magnetron if you don't care about frequency? Its cheaper and more robust typically, especially if you dial down the output power
and cool it good. No reason why you can't use a microwave oven one in that case.
I was explaining what a variable frequency one would require. It would be difficult to make one using discrete transistors anyway because of phase
margin I think, which is why they make MMIC for wideband applications.
This would be if you want to investigate the so called athermal effects.
[Edited on 4-7-2018 by coppercone]
[Edited on 4-7-2018 by coppercone]
coppercone - 4-7-2018 at 07:57
Actually thinking about it, it might be easier to use a solid state amplifier so long you can make oscillator source, electrical control should be
easier. I was imagining using a filtered drive to feed the input voltage on a the HV transformer, in a very crude scheme.
You would need to measure the voltage standing wave ratio to make sure you are not reflecting alot of energy back into the oscillator though, I think
with a magnetron it would just get hot (so you derate it), the solid state amplifier might not like it, but they have solid state microwave ovens now
so it might be possible and even favorable to use their control circuit.
[Edited on 4-7-2018 by coppercone]