Pierre's 170W in 1600W out Looped Very impressive Build continued & moderated

[jerdee]

Luc, can you post a schematic for how you are driving your coils, as well as the  propellor connections/code. It's hard to figure out how things are connected and biased, just listening to your explanation.
PmgR

Code might be a bit confusing, as it is all in binary.  So posting the code will most likely be more confusing for most.

We are trying to focus on creating the strongest field on the stator windings that has the least amount of cancellation.

H-Bridges are configured this way all the way around in series.
H-bridge 1 beginning of coil 1 to the end of 5
H-bridge 2 beginning of coil 2 to the end of 6.
etc...

The LOOP in laymen terms is this...

REPEAT
  NORTH is created on coils 2 through 6 while SOUTH is created on coils 17 through 21
  NORTH is created on coils 3 through 7 while SOUTH is created on coils 18 through 22
  NORTH is created on coils 4 through 8 while SOUTH is created on coils 19 through 23
  SOUTH is created on coils 2 through 6 while NORTH is created on coils 17 through 21
  SOUTH is created on coils 3 through 7 while NORTH is created on coils 18 through 22
  SOUTH is created on coils 4 through 8 while NORTH is created on coils 19 through 23

This test is explaining our issues of why we are not getting enough flux across the armature.

We need to isolate the poles.  This test is to eliminate the other poles as much as possible before separating the series coils arrangement.  So only two poles are active at one time in A/C mode.

I don't see how moving forward with 5 coils in series overlapping per pole helps.  I believe the test is proving this.

Hope this helps,
Jerdee

[partzman]

6 Coil Test AC vs DC: https://youtu.be/hrMSWq6iZcE

Luc,

You must have missed my post on OUR on 2018-04-17 regarding this very topology so I will re-post here-

"Much confusion seems to exist on OU with Pierre's device and it's operation regarding the switching of his 36 serially connected stator coils.  The output levels of the builder's attempts thus far are low which indicates there is a flaw in their thinking.  So, in an attempt to resolve the problem, I simulated the power transformer feeding the supercap (SC) pack while supplying the running stator without the rotor inserted.  The idea was to replicate Pierre's SC loaded output voltage with the known input voltage and current to the power transformer under these conditions.  No real magic here but a few parameters had to be assumed such as coupling factor, etc.

Considering the assumed and accepted switching connections by the replicators plus their posted coil dcrs, a relay switched arrangement should present a load in the ~.5-1 ohm range to the SC pack.  When these load numbers were used in the sim, the results were way off from Pierre's measurements and I posted this info on OU to no avail.  IMO, the assumed switch connections are wrong and some basics are being overlooked that is, the voltage and current waveform generated by passing a PM over a coil.  So, this post represents my thoughts on the matter.

Attached is an 18 slot sim for simplification to get the point across.  The current measurements for each coil used are offset for clarity.  It consists of 2 poles that are physically and electrically opposite from each other namely L1,2,3 and L10,11,12.  All other connected coils have zero current and the reason why is apparent if the circuit is studied.  This is the key because if we now do average measurements on the voltages and currents, we arrive at a net load resistance in the range that allow the original power transformer sim to work.  IOW, not all coils are energized all the time.  The switching to move the fields around the stator should be academic from here based on the number of poles desired and the slot pitch.

Any recovery from the inductive switching can now be visualized from the moving electromagnetic field over the stationary loaded rotor similar to a PM passing over coil."

The referenced sim is attached.

Regards,
Pm

[pmgr]

Code might be a bit confusing, as it is all in binary.  So posting the code will most likely be more confusing for most.

We are trying to focus on creating the strongest field on the stator windings that has the least amount of cancellation.

H-Bridges are configured this way all the way around in series.
H-bridge 1 beginning of coil 1 to the end of 5
H-bridge 2 beginning of coil 2 to the end of 6.
etc...

The LOOP in laymen terms is this...

REPEAT
  NORTH is created on coils 2 through 6 while SOUTH is created on coils 17 through 21
  NORTH is created on coils 3 through 7 while SOUTH is created on coils 18 through 22
  NORTH is created on coils 4 through 8 while SOUTH is created on coils 19 through 23
  SOUTH is created on coils 2 through 6 while NORTH is created on coils 17 through 21
  SOUTH is created on coils 3 through 7 while NORTH is created on coils 18 through 22
  SOUTH is created on coils 4 through 8 while NORTH is created on coils 19 through 23

This test is explaining our issues of why we are not getting enough flux across the armature.

We need to isolate the poles.  This test is to eliminate the other poles as much as possible before separating the series coils arrangement.  So only two poles are active at one time in A/C mode.

I don't see how moving forward with 5 coils in series overlapping per pole helps.  I believe the test is proving this.

Hope this helps,
Jerdee

Luc/Jerdee,


What timings are you using and overlap time? This is important as it will effect current draw.


Also, for this configuration, turn up the frequency and see what you get. I predict that the output coil voltage won't increase much more once you go past the stator/rotor max frequency, which is around 50-60Hz. Please see if you can confirms this.


If so, it simply means your rotor is too thick and was collapsing poles in your earlier measurements and your frequency was too high to see any effect with the thinner rotor.


Also repeat your current measurement with the thinner rotor at low frequency and at high frequency. Then you will now what frequency you need to aim for and if you need to make another rotor with an intermediate thickness.


Lastly, see what happens if you extend your single North and South poles in your code above to span half of your rotor, so bias coils 1-15 North, 16-30 South and rotate around.


PmgR

[jerdee]

What timings are you using and overlap time? This is important as it will effect current draw.

I've found it the easiest to use two "wait" or "pause" commands.  So we can always adjust overlap and hold times.  You are correct, the more overlap, the more demand.  Not good. We've found that most of the overlap is not needed.  It's waisted. 

Also, for this configuration, turn up the frequency and see what you get. I predict that the output coil voltage won't increase much more once you go past the stator/rotor max frequency, which is around 50-60Hz. Please see if you can confirms this.

Higher freq...provides less current, so you have to input more voltage.  Nothing unusual.

Also repeat your current measurement with the thinner rotor at low frequency and at high frequency. Then you will now what frequency you need to aim for and if you need to make another rotor with an intermediate thickness.

Our mains focus at this moment is getting as much of the rotating magnetic field as possible.  We are both in agreement that isolation of the pole fields are needed.

Lastly, see what happens if you extend your single North and South poles in your code above to span half of your rotor, so bias coils 1-15 North, 16-30 South and rotate around.

Yes, this makes sense.  This requires the slot arrangment to be reconfigured.  Not 5 slots, but 15 slots.  Also this currently requires 15 times 3 pins from MCU.  Sadly we can't do this yet.  Our gates came in today.  So we have more pinout control from MCU.  Otherwise we run out quickly in pin control from MCU.  I do not think we need to go to higher pin out from MCU to get this device to work.  We are going to be using XOR gates to drop our pin control down by one third. You can run multipled poles in parallel, and save pin count as well.

Isolation of the pole fields is key.   If you can't get the flux to "re-connect" in the armature...you have little results.  This is how I see it at this point.

Jerdee

[gotoluc]

Luc,

You must have missed my post on OUR on 2018-04-17 regarding this very topology so I will re-post here-

"Much confusion seems to exist on OU with Pierre's device and it's operation regarding the switching of his 36 serially connected stator coils.  The output levels of the builder's attempts thus far are low which indicates there is a flaw in their thinking.  So, in an attempt to resolve the problem, I simulated the power transformer feeding the supercap (SC) pack while supplying the running stator without the rotor inserted.  The idea was to replicate Pierre's SC loaded output voltage with the known input voltage and current to the power transformer under these conditions.  No real magic here but a few parameters had to be assumed such as coupling factor, etc.

Considering the assumed and accepted switching connections by the replicators plus their posted coil dcrs, a relay switched arrangement should present a load in the ~.5-1 ohm range to the SC pack.  When these load numbers were used in the sim, the results were way off from Pierre's measurements and I posted this info on OU to no avail.  IMO, the assumed switch connections are wrong and some basics are being overlooked that is, the voltage and current waveform generated by passing a PM over a coil.  So, this post represents my thoughts on the matter.

Attached is an 18 slot sim for simplification to get the point across.  The current measurements for each coil used are offset for clarity.  It consists of 2 poles that are physically and electrically opposite from each other namely L1,2,3 and L10,11,12.  All other connected coils have zero current and the reason why is apparent if the circuit is studied.  This is the key because if we now do average measurements on the voltages and currents, we arrive at a net load resistance in the range that allow the original power transformer sim to work.  IOW, not all coils are energized all the time.  The switching to move the fields around the stator should be academic from here based on the number of poles desired and the slot pitch.

Any recovery from the inductive switching can now be visualized from the moving electromagnetic field over the stationary loaded rotor similar to a PM passing over coil."

The referenced sim is attached.

Regards,
Pm

Thanks for sharing your simulation Partzman

I have not been following the topic at Over Unity Research (OUR)   kind of busy with just this topic here.

If anyone is interested, here is the link to the first page:  http://www.overunityresearch.com/index.php?topic=3599.0
Kind regards
Luc