Here Shawyer compares 7 devices, some classic EMDrives, some with just a dielectic & the superconducting LN cooled Cannae drive.
It is important to note the direction of generated thrust as dielectrics apparently generate thrust in the opposite direction to that of EMDrives.
I do note the EW test of a Cannae like device was not tested at cryo temp nor had superconducting interior lining and had an added dielectic. So was not a test of a true Cannae device.
From Shawyer, EW should not expect to see any thrust from an EMDrive like test device when tested fixed, not allowed to accelerate, without added dielectic. Which is what was observed. Static thrust was not measured, with the EW EMDrive like device until a dielectic was added.
Appears EW have discovered a new way to gen thrust in a fixed device. Use a dielectric. From Shawyer device summary, line 1, it appears all that may be needed is a short section of resonate pipe with a dielectic stuck in one end. IE EW Cannae test device minus the cavity.
However as the dielectric thrust is weaker than the EMDrive thrust and in the opposite thrust direction to the classic EMDrive thrust direction, putting a dielectric into an accelerating EMDrive may reduce the overall delivered thrust and effective cavity Q as would be seen by the classic EMDrive operational mode.
Is Shawyer giving away spoilers?

Thank you for your post, as it gives me the opportunity to ask again a couple of questions that are unanswered so far. Perhaps you can answer them:
1) Shawyer reported in the above graph, that his Demonstrator engine is the only EM Drive so far that has shown forces in BOTH directions, towards the small end as well as towards the large end.
1a) What does Shawyer mean by this? Does he mean that the Demonstrator engine displayed, and he measured, forces simultaneously in both directions? But as the forces in both directions are almost equal, that would mean practically no net force. That doesn't seem to make sense as the Demonstrator engine is the only one that had a force large enough that it could move the whole assembly as shown in a video:
1b) Did Shawyer measure a force towards the big end in some tests and a force towards the small end in other tests? If so, what made the difference between the tests? Is the change random, unpredictable and unexplained? (Hopefully not). If the change in force direction can be deliberately controlled, how was Shawyer able to change the force direction for the Demonstrator engine? Did he have to move something? (like the location of the dielectric)? Was there a dielectric in the Demonstrator engine? Or did he just accomplish this change of force direction by changing the exciting frequency (and thereby changing the mode shape)?
2) Shawyer defines the measured "thrust force" in the opposite direction to the direction of motion of the EM Drive. This is completely the opposite of NASA Eagleworks that defines the thrust force in the same direction as the motion of the EM Drive. Can you reconcile and make sense of these two opposite, contradictory definitions and measurements?
Thanks
The device you pictured is the 1st Dynamic Test unit (data apparently not in the summary):
http://emdrive.com/dynamictests.htmlHis portable Demonstrator device is here:
http://emdrive.com/demonstratorengine.htmlWhy he shows thrust in both direction is unknown to me.
I assume he can control the Demonstrator device in such a way to shift the thrust direction. It does have the ability to dynamically control cavity length.
Likewise his Flight Thruster is here:
http://emdrive.com/flightprogramme.htmlThe Shawyer EMDrive thrust direction is consistent in one mode of the line 4 device (Demonstrator), in line 5 (Chinese device), 6 (Flight Thruster).
The 3 dielectric devices and the Cannae cryo cooled, superconducting device generate thrust in the same direction and opposite to the classic EMDrive.
Additionally from the EW tests, apparently the addition of a dielectric to their EMDrive like device and the Cannae like device allows weak thrust to be generated without needing the device to accelerate.
Would be interesting to see EW chop off the Cannae test device and test it again with just the RF feed pipe, with of course the dielectric. Could be a very simple and quick test device to build and if it can generate thrust without moving (as it seem to do in the EMDrive/Cannae like devices), so much the better.
I really like the Shawyer Demonstrator device (attached). As an engineer I can appreciate designing, building and testing it plus the hours and money involved. It is a serious and professionally built device. It is not a toy but a real working thruster that can be taken anywhere to have additional tests done.
Maybe if EW asked Shawyer nicely, they could test it? At least then they have an established test data base and working device to work from.