In my opinion it will be very unstable.
NSF-1701 Update - Thanks to Shell, I received the defective magnetron, took the connector off and soldered an SMA pigtail to it. Works great! Also attached very first VNA return loss scan. Note it centers right up where it is supposed to. I'll take some data and give you the Q in a few minutes.Center frequency = 2.432 GHz with 11.6dB return lossQ per Yang method = 821Q per Nasa method = 437My Qr factor = 1.88
Quote from: rfmwguy on 10/25/2015 07:01 pmNSF-1701 Update - Thanks to Shell, I received the defective magnetron, took the connector off and soldered an SMA pigtail to it. Works great! Also attached very first VNA return loss scan. Note it centers right up where it is supposed to. I'll take some data and give you the Q in a few minutes.Center frequency = 2.432 GHz with 11.6dB return lossQ per Yang method = 821Q per Nasa method = 437My Qr factor = 1.88Glad to help! You think the Q could be low because of? Where are you inserting the maggie in the frustum? You have other points to check?Nice Slaughtering. Don't you hate soldering to a heavy heat sink?
Quote from: SeeShells on 10/25/2015 07:39 pmQuote from: rfmwguy on 10/25/2015 07:01 pmNSF-1701 Update - Thanks to Shell, I received the defective magnetron, took the connector off and soldered an SMA pigtail to it. Works great! Also attached very first VNA return loss scan. Note it centers right up where it is supposed to. I'll take some data and give you the Q in a few minutes.Center frequency = 2.432 GHz with 11.6dB return lossQ per Yang method = 821Q per Nasa method = 437My Qr factor = 1.88Glad to help! You think the Q could be low because of? Where are you inserting the maggie in the frustum? You have other points to check?Nice Slaughtering. Don't you hate soldering to a heavy heat sink?I have my trusty BMF soldering iron...no worries. Think lower Q is almost all attributed to mesh sides as opposed to solid copper. Now that I know these dimensions give a great resonance and thermal lift is going to be there anyway, Phase II testing will probably be done with a new solid copper frustum. Bucyrus, here I come.Edit - lower Q might also relate to the 50 ohm mismatch to magnetron radome. I was pleased to see 1.8:1 vswr overall, but it is not as good as a monopole or loop tuned for 2.4 GHz.
Just about ready to put all the parts together....Pics and Centerfolds (clean centerfolds)http://s1039.photobucket.com/user/shells2bells2002/media/CE%20Electromagnetic%20Reaction%20Thruster/EM%20thruster%20016_zpsdj5mg8vf.jpg.html?sort=3&o=0
NSF-1701 Update - Thanks to Shell, I received the defective magnetron, took the connector off and soldered an SMA pigtail to it. Works great! Also attached very first VNA return loss scan. Note it centers right up where it is supposed to. Last pic is screenshot of another can with Phase overlay.Center frequency = 2.432 GHz with 11.6dB return lossQ per Yang method = 821Q per Nasa method = 437My Qr factor = 1.88
Quote from: rfmwguy on 10/25/2015 07:55 pmQuote from: SeeShells on 10/25/2015 07:39 pmQuote from: rfmwguy on 10/25/2015 07:01 pmNSF-1701 Update - Thanks to Shell, I received the defective magnetron, took the connector off and soldered an SMA pigtail to it. Works great! Also attached very first VNA return loss scan. Note it centers right up where it is supposed to. I'll take some data and give you the Q in a few minutes.Center frequency = 2.432 GHz with 11.6dB return lossQ per Yang method = 821Q per Nasa method = 437My Qr factor = 1.88Glad to help! You think the Q could be low because of? Where are you inserting the maggie in the frustum? You have other points to check?Nice Slaughtering. Don't you hate soldering to a heavy heat sink?I have my trusty BMF soldering iron...no worries. Think lower Q is almost all attributed to mesh sides as opposed to solid copper. Now that I know these dimensions give a great resonance and thermal lift is going to be there anyway, Phase II testing will probably be done with a new solid copper frustum. Bucyrus, here I come.Edit - lower Q might also relate to the 50 ohm mismatch to magnetron radome. I was pleased to see 1.8:1 vswr overall, but it is not as good as a monopole or loop tuned for 2.4 GHz.Interesting. It seems Q factor was really low with the copper wire mesh, which could explain the low thrust. It now seems evident copper mesh wasn't the best option after all. But if you got evidence of anomalous thrust even in those conditions, I wonder what a certifiably high Q frustum would do with your same setup...Best of luck!
Quote from: rfmwguy on 10/25/2015 07:01 pmNSF-1701 Update - Thanks to Shell, I received the defective magnetron, took the connector off and soldered an SMA pigtail to it. Works great! Also attached very first VNA return loss scan. Note it centers right up where it is supposed to. Last pic is screenshot of another can with Phase overlay.Center frequency = 2.432 GHz with 11.6dB return lossQ per Yang method = 821Q per Nasa method = 437My Qr factor = 1.88Nice work Dave. Good to see the SPR inspired spreadsheet's dimensional data got your frustum's resonance into the maggies output freq range.Now replace the lossy sides with solid polished copper and see if that increases your loaded Q.Please remember Rogers advise to run these scans as slow as you can so to give the frustum enough time to properly fill the cavity (at least 10x the expected cavity time constant dwell time at each freq step) or the reported loaded Q value may be lower and the VSWR higher than reality.It might also be interesting to see the result if the antenna is mounted on the frustum side wall as Iulian did. You could try quite a few mounting sites and record any changes to resonance and VSWR.
That's right Phil, the spreadsheet was spot-on at resonance predictions. Congrats on creating that useful builder tool. Yes, solid copper for phase II is the plan.
Thoughts??? I'm considering on putting a top cap to deflect hot air 900 away from the frustum. The beryllium gasket seal between the ceramic top plate and tuning chamber will leak hot air.
Quote from: SeeShells on 10/26/2015 01:51 amThoughts??? I'm considering on putting a top cap to deflect hot air 900 away from the frustum. The beryllium gasket seal between the ceramic top plate and tuning chamber will leak hot air.Forcing hot air horizontally will eventually find an edge and rise. But..if it could be cooled enough before edge was reached...hmmmm
Perhaps it makes more sense to create a bad vacuum (e.g. one tenth of atmospheric pressure) with a simple and cheap pump in an enclosure with the test article in it? We just want to increase SNR considerably, right? I don't think that a hard vacuum is needed. We just want to get rid of most of the buoyancy for now. My 2 cents .Edit: Actually, it should easily be possible to create a pretty hard vacuum the cheap way:1) 3D-print a metallic enclosure with cooling channels for liquid nitrogen in the walls, perhaps even just the bottom/floor of the enclosure.2) Put an automated, complete test article in it (sorta like a space probe)3) Shut the enclosure and fill it with pure CO2, so that all other gases are pushed out of the enclosure4) Seal the enclosure and start pumping liquid nitrogen through the wall channels. The CO2 freezes out, until there's only solid dry ice left (maybe best only on bottom/floor of enclosure)5) You got vacuum What do you guys think about this method?