PopularFX
Home Help Search Login Register
Welcome,Guest. Please login or register.
2026-01-29, 10:12:53
News: Check out the Benches; a place for people to moderate their own thread and document their builds and data.
If you would like your own Bench, please PM an Admin.
Most Benches are visible only to members.

Pages: 1 [2] 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 ... 31
Author Topic: partzmans board ATL  (Read 36325 times)
Topic Options img
img  partzman
Re: partzmans board ATL « Reply #25 on: 2019-08-13, 17:19:20 »
Group: Moderator
Hero Member
*****

Posts: 2232
Hi Smudge,

I agree with your calcs and overall analysis!  However, it is my fault for not being more clear on what I was proposing.  My terms are somewhat confusing as I reread that August 8th post.

So to clarify, I meant to have L2 represent motional emf for L1 from a moving PM even though it appears as induction emf with a coupling of .9 .  I'm not sure how accurate this sim representation really is to the real world so I will build a device to see what gives.  In this case, even if we throw away any stored or increased energy in L1, we would still have a gain even though considerably less. 

At this point in time, I have not succeeded in creating OU with any induced emf generating means.

Thanks for taking the time to analyze this.

Regards,
Pm
   
img  Smudge
Re: partzmans board ATL « Reply #26 on: 2019-08-13, 17:40:10 »

Group: Professor
Hero Member
*****

Posts: 2332
Quote from: partzman on 2019-08-13, 17:19:20
Hi Smudge,

I agree with your calcs and overall analysis!  However, it is my fault for not being more clear on what I was proposing.  My terms are somewhat confusing as I reread that August 8th post.

So to clarify, I meant to have L2 represent motional emf for L1 from a moving PM even though it appears as induction emf with a coupling of .9 .
I think the same thing occurs with regard to forward and back mmf except in this case that forward increasing mmf is the magnet getting closer to L1.  If you work out the direction the flux in L1 changes you find that the energy stored in L1 will decrease and the magnet will have to do work to achieve that, whereas your analysis showing apparent energy gain wrongly assumes the energy in L1 will increase.  Methinks the work done by the moving magnet plus the energy loss in L1 will fully account for the energy gain in L3. 
Smudge
   
img  partzman
Re: partzmans board ATL « Reply #27 on: 2019-08-14, 00:32:20 »
Group: Moderator
Hero Member
*****

Posts: 2232
I have to chuckle a little here because I translated a paper from Abramovich on the working of the TPU sometime back but I really didn't pay any attention to it at first.  I just now began to read it and when I got to the 3rd or 4th paragraph I began to understand that he is talking about a constant current load on the secondary!!!

Anyway, I've attached it below for your reading enjoyment.

Regards,
Pm
   
img  Smudge
Re: partzmans board ATL « Reply #28 on: 2019-08-14, 16:52:22 »

Group: Professor
Hero Member
*****

Posts: 2332
Quote from: partzman on 2019-08-14, 00:32:20
I have to chuckle a little here because I translated a paper from Abramovich on the working of the TPU sometime back but I really didn't pay any attention to it at first.  I just now began to read it and when I got to the 3rd or 4th paragraph I began to understand that he is talking about a constant current load on the secondary!!!
He is talking about a secondary that has a DC output.  His primary coil arrangements produce a longitudinal flux in the core plus a cross flux that rotates.  That cross flux exits the core through a conductive cylinder that surrounds the core and he looks at it like a radial vector that rotates, hence creating a motion induced E field along the cylinder.  He calls that cylinder the secondary, and such a vector would produce DC.  See first image below.

However I think he is wrong in that it is not a radial field, it is a cross field that also exits diametrically opposite, see image 2.  Note that here the E field is in the opposite direction.  Of course instead of a cylinder you could have a split cylinder or even a wire top and bottom, then series them to get AC induction.  Will that form of secondary lead to OU?  I doubt it because when it carries load current it creates a transverse or cross field that will couple to the angled primary coils.
Smudge
   
img  partzman
Re: partzmans board ATL « Reply #29 on: 2019-08-14, 19:00:17 »
Group: Moderator
Hero Member
*****

Posts: 2232
Quote from: Smudge on 2019-08-14, 16:52:22
He is talking about a secondary that has a DC output.  His primary coil arrangements produce a longitudinal flux in the core plus a cross flux that rotates.  That cross flux exits the core through a conductive cylinder that surrounds the core and he looks at it like a radial vector that rotates, hence creating a motion induced E field along the cylinder.  He calls that cylinder the secondary, and such a vector would produce DC.  See first image below.

However I think he is wrong in that it is not a radial field, it is a cross field that also exits diametrically opposite, see image 2.  Note that here the E field is in the opposite direction.  Of course instead of a cylinder you could have a split cylinder or even a wire top and bottom, then series them to get AC induction.  Will that form of secondary lead to OU?  I doubt it because when it carries load current it creates a transverse or cross field that will couple to the angled primary coils.
Smudge

Agreed.  However, that is why I persist in that the load must stay as constant current as possible to keep the counter emf and therefore the lenz effect as low as possible.

It may be interesting to experiment with a rotating magnetic field as in a 3-phase motor field stator to provide the motional emf while a large fixed inductor is placed in the center of the arrangement.

Pm   
   
img  Centraflow
Re: partzmans board ATL « Reply #30 on: 2019-08-14, 19:02:12 »

Group: Experimentalist
Hero Member
*****

Posts: 3914


Buy me a beer
I thought I will put my 2 pence worth which please think hard about it.

If you are looking at types of transformers where the secondary does not affect the primary, I think you will be losing every time.

I think you have to look at a way of using a DC current to move a magnetic field or a charge around a none magnetic core in the form of a loop. If you move a charge between two plates a magnetic field is created.

The primary is a coil wound around a two-plate capacitor, a charge is placed on the capacitor, then a DC current is applied to the coil to charge it, then you discharge that coil, the polarity of the coil changes and the current continues in the same direction as the charge. If this is done in a loop and with several coils around this loop shaped capacitor, the charge in the capacitor will move around that capacitor loop. Moving charge creates a magnetic field which moves at the same rate as that moving charge, like this we create a moving magnetic field which does not affect the coils around the outside (primary) in a negative way.

The coils have to be sequential in their charge and discharge in a special way, SM has hinted at this and the frequency will dictate the speed of that moving charge/magnetic field. The output is DC with a slight blip caused by the loop charge current, it is the loops capacitance which smooths the output to DC and naturally an inductive load can not be connected without isolating the output circuit.

If this is not appropriate for this thread PM, please delete it, but I really think it is valid.

Regards

Mike 8)


---------------------------
"All truth passes through three stages. First, it is ridiculed, second it is violently opposed, and third, it is accepted as self-evident."
Arthur Schopenhauer, Philosopher, 1788-1860

As a general rule, the most successful person in life is the person that has the best information.
   
img  partzman
Re: partzmans board ATL « Reply #31 on: 2019-08-14, 22:43:06 »
Group: Moderator
Hero Member
*****

Posts: 2232
Quote from: Centraflow on 2019-08-14, 19:02:12
I thought I will put my 2 pence worth which please think hard about it.

If you are looking at types of transformers where the secondary does not affect the primary, I think you will be losing every time.

I think you have to look at a way of using a DC current to move a magnetic field or a charge around a none magnetic core in the form of a loop. If you move a charge between two plates a magnetic field is created.

The primary is a coil wound around a two-plate capacitor, a charge is placed on the capacitor, then a DC current is applied to the coil to charge it, then you discharge that coil, the polarity of the coil changes and the current continues in the same direction as the charge. If this is done in a loop and with several coils around this loop shaped capacitor, the charge in the capacitor will move around that capacitor loop. Moving charge creates a magnetic field which moves at the same rate as that moving charge, like this we create a moving magnetic field which does not affect the coils around the outside (primary) in a negative way.

The coils have to be sequential in their charge and discharge in a special way, SM has hinted at this and the frequency will dictate the speed of that moving charge/magnetic field. The output is DC with a slight blip caused by the loop charge current, it is the loops capacitance which smooths the output to DC and naturally an inductive load can not be connected without isolating the output circuit.

If this is not appropriate for this thread PM, please delete it, but I really think it is valid.

Regards

Mike 8)

Hi Mike,

Your comments above are certainly appropriate for this thread and thanks for the input.  I have many experiments on my "yet to do list" which involve techniques similar to what you describe but time seems to go so quickly!  Anyway, I will consider your suggestions and perhaps there may be some combination that will work, we shall see.

Regards,
Pm
   
img  partzman
Re: partzmans board ATL « Reply #32 on: 2019-08-14, 23:20:36 »
Group: Moderator
Hero Member
*****

Posts: 2232
Here is a sim using RLE in a periodic application with a square wave drive source.

Referring to the schematic, C1 is in slightly off-resonance with the leakage inductance of T1 such that the input energy is mostly reactive.  The current inductor L3 is synchronously switched by S1,S2, S5, and S6 such that the ends VL3 and VL4 of L3 are alternately connected between VL2 and ground.  With proper timing, this arrangement builds a unidirectional current in L3 at a near linear rate over time as is seen in plot trace I(L3).

The circuit's operation is allowed to run for 4.9ms to stabilize all the waveforms before the plot data starts.  After the plot starts, at the 100us mark L3 is connected to the secondary L2 for a period of 60us.  This is seen in the first plot along with the peak current of 83.9ma reached in L3 at the 160us mark.  This equates to an energy of (.0839)^2*.1/2 = 351.9uJ.  Please note there is no pre-bias in L3 as there was previously.

For reference, plot #2 shows the input energy of 33.304uJ drawn from the pulse generator V2 for the first 100us period.

Plot #3 shows the input energy of 42.756uJ drawn from the input during the 60us time period the current inductor L3 is being charged.  From this data we can now calculate the apparent COP = 351.9/42.756 = 8.23.

The gain is achieved by the fact that the relatively constant current load in the secondary of T1 produces very little counter emf to the primary.

Regards,
Pm

Edit:
« Last Edit: 2019-08-16, 13:50:47 by partzman »
   
img  partzman
Re: partzmans board ATL « Reply #33 on: 2019-08-16, 02:57:58 »
Group: Moderator
Hero Member
*****

Posts: 2232
Regarding my previous post, it was a little premature as the "apparent" gain is really supplied by the stored energy in L2 and C1 so the end result is COP<1.

Pm
   
img  partzman
Re: partzmans board ATL « Reply #34 on: 2019-08-27, 19:08:17 »
Group: Moderator
Hero Member
*****

Posts: 2232
The following are the results of using RLE or constant current inductor loading in a special built ferrite 'E' cored transformer with three windings.  The simulation uses a model of the transformer made with "H" sources in a gyrator-capacitor method.

The schematic may be difficult to follow for those not familiar with such modeling but basically taking the P1 primary for example, H1 represents the electrical while H2 the magnetic characteristics of that winding.  P1 in this case represents the permeance or the reciprocal of the reluctance of that section of core.  This allows nearly any core configuration to be modeled rather accurately.

A list of the various combinations of connections is shown with their resulting inductance values and is important for determining the outcome of the transformer's OU potential.  For example, note the inductance of 5.31mH when all windings are connected in a buck mode as compared to 4.88mH when S1 is driven with a bucking polarity to P1 and P2 that are shorted in a buck mode as a load.  For all practical purposes this indicates a slight parametric change depending on charging and discharging connectivity.  This is not a sim anomaly but rather accurately represents the actual bench transformer.

So, in this example, we have a full bridge switching arrangement that alternately connects the secondary S1 between the 50v dc supply V10 and ground.  The primaries P1 and P2 are connected in a buck configuration with the current inductor L1 connected as a load.  This load network of P1,P2, and L1 is initially charged to a bias of 50ma prior to the start of the sim.

Power is applied to S1 for 15us which produces a positive current ramp in S1 resulting in a current increase in the secondary network.  At the end of the 15us period, switch S1 shorts L1 and holds the stored current.  The polarity of S1 is also reversed and the collapsing field in S1 now supplies current and therefore energy to the supply V10.  At the end of 30us, we take a tally of all the energies involved minus core losses and would make ready for the next cycle.

From the plot we see the net input energy to be 7.2561uJ taken from the dc supply V10.  At the 15us point in time from cursor #1 we see the current in L1 has reached a peak of 53.602ma.  Therefore, the energy gained in L1 is ((.053602^2)-(.05^2))*.1/2 = 18.659uJ.  The P1 and P2 buck inductance is 5.12mH and the energy needed to independently charge this pair to 50 ma is .05^2*.00512/2 = 6.4uJ.

So, the total input energy is 7.2561uJ + 6.4uJ = 13.656uJ.  Therefore, the apparent COP = 18.659/13.656 = 1.37.

Regards,
Pm     
   
img  partzman
Re: partzmans board ATL « Reply #35 on: 2019-09-06, 21:49:22 »
Group: Moderator
Hero Member
*****

Posts: 2232
Here is a test using a GE ECM 2.3 motor connected in series with a 500mh inductor.  The ECM is an electrically commutated 3-phase PM motor used in HVAC blowers in the US and Canada that I am aware of.

The test schematic is shown below and basically one phase is positioned to attract at switch closure and there is rotor movement.  The scope traces indicate what energy is consumed and available during this event over a specific amount of time.

As seen, over 51.01ms, 4.872 watts or 249mJ are input to the device.  At the end of this period, we see the current in L1 has reached 1.584 amps peak for an output energy level = 1.584^2*.5/2 = 627mJ.  The apparent energy gain is 627/249 = 2.52 .

Regards,
Pm

Edit: This test is invalid due to saturation of the 500mh inductor which results in erroneous results.  The COP <1.
« Last Edit: 2019-09-07, 18:34:45 by partzman »
   
img  partzman
Re: partzmans board ATL « Reply #36 on: 2019-09-21, 18:39:21 »
Group: Moderator
Hero Member
*****

Posts: 2232
[Re-posted]

Continuing on with this line of research, an example is given below of a simulation that "un-folds" or discharges two identical windings on transformer T1 with a coupling of K=.9 connected in a bucking mode.  From the table we see the net inductance of this arrangement is 586uH.

Pre-existing conditions for this simulation are: L1 and L2 are charged to 100ma, L3 (constant current inductor) is charged to 100ma, and nodes VL1 and VL2 are clamped near zero.  Vs is a DC supply that receives the stored energy in the device as the field in L2 collapses. 

The sim begins by releasing VL1 and VL2 and L2 then begins to dump it's energy into Vs via D1.  Owing to the fact of the relatively constant current in L1 via L3, L2 will collapse while L1 remains charged.  The plot is stopped at the period in time when the current in L2 reaches zero.  At this point with L2 discharged, L1 now represents a 2.9mh inductor in series with L3 rather than the original 586uH with the bucking L1 and L2.  It is this parametric mechanism with the constant current source for L1 that produces the high efficiency and if it were not for the voltage drop across L3 thus rendering it a less than a perfect current source, the apparent gains would be higher. 

Measurements are taken for various voltages from 100-250vdc for Vs and are displayed in the table shown.  As can be seen, there is apparent gain with the higher voltages for Vs due to the shorter discharge times resulting in lower resistive losses.  The plot shown is with Vs=250vdc.

Starting energies are as follows: L3 = .1^2*.1/2 = 500uJ, L1 buck L2 = .1^2*586e-6/2 = 2.93uJ.  The ending energy in the plot example of the series connection of L3 and L1 = .097462372^2*.1029/2 = 488.72uJ.  This is subtracted from the starting L3 energy leaving 11.28uJ consumed for a total input energy of 11.28+2.93 = 14.21uJ .  The energy stored in Vs is seen to be 14.905uJ for an apparent COP = 14.905/14.21 = 1.049 .

Small gains but a POC via simulation.

Regards,
Pm

Edit: There are no mistakes in this post as I had thought.
« Last Edit: 2019-09-23, 16:02:31 by partzman »
   
img  partzman
Re: partzmans board ATL « Reply #37 on: 2019-09-23, 18:05:27 »
Group: Moderator
Hero Member
*****

Posts: 2232
This is a followup to my previous post and simulation and will demonstrate the potential gain with a simple transformer's asymmetry.  In this case, the comparison is between the value of the bucking inductance of 296uH  with a K=.95 with L1 and L2 and the individual inductance of L1.  The theoretical max gain is L1/(L1 buck L2) = 2.9e-3/296e-6 = 9.8 which is much greater than the theoretical max of 2 when independently charging each winding via RLE.

The sim and schematic are attached below and the sim shows the plot math results with Vs = 100vdc.  The difference with this device is that the 100nH 100mH current inductor L3 has an added 100uH winding L4 that is driven by a current source I1 that is ramping from 0-100ma in a given period.  This creates a relatively constant current over the period of operation in L3.  The energy consumed by I1 is measured by V(VL4)*I(I1) in the plot and will be shown as a negative value due to the fact that we are supplying a positive current to a small positive voltage across L4.  I consider this as a positive energy however because in a real device, this current would come from a positive source and would in reality be greater by an amount dependent on the efficiency of the source itself.  The energy in any case is reasonably small due to the turns ratio of L3 to L4 of 31.6:1 resulting in a relatively low voltage at VL4.

I won't detail the math in this case but I will add there is a caveat.

Regards,
Pm

Edit: Corrected 100nH to 100mH.   
« Last Edit: 2019-09-24, 16:37:24 by partzman »
   
img  Orthofield
Re: partzmans board ATL « Reply #38 on: 2019-10-09, 00:01:14 »
Sr. Member
****

Posts: 361
Hi Partzman,

My ISP crashed so I'm unable to respond to your recent letter. I'll contact you when I have access again.

Fred
   
img  partzman
Re: partzmans board ATL « Reply #39 on: 2019-10-09, 16:30:07 »
Group: Moderator
Hero Member
*****

Posts: 2232
Quote from: orthofield on 2019-10-09, 00:01:14
Hi Partzman,

My ISP crashed so I'm unable to respond to your recent letter. I'll contact you when I have access again.

Fred

Fred,

OK.

Pm
   
img  partzman
Re: partzmans board ATL « Reply #40 on: 2019-10-30, 15:51:40 »
Group: Moderator
Hero Member
*****

Posts: 2232
I notice that this thread has been read 518 times as of this moment!  Quite a few IMO for so few members.  Hmmmm...

Anyway, continuing on-

This simulation example involves a single linear variable inductor as a parametric element that utilizes the RLE concept.  The result is a parametric generator with apparent gain.

The gyrator/capacitor core model used for L1 was taken from measurements of a bench transformer using a ferrite 3019P pot core.  The inductance is varied by the means described on the schematic.  A voltage controlled voltage source B2 is used to model the variable permeance that creates the model's inductance change.  B2 allows for any math function for control, but in this case the control means is simply the piece wise linear step change in V1.

Prior to the simulation start, the current inductor L2 in series with L1 are pre-charged to 150ma and at this point in time, the inductance of L1 = 536uH.  This action requires .15^2*.100536/2 = 1.131mJ neglecting losses.

The inductance is changed during the 10us "on" period and is now 2.03mH.  Due to the polarities used, an increase in the L1/L2 current is the result and is seen to be 154.89ma clamped at 12us.  The ending energy therefore is .15489^2*.10203/2 = 1.224mJ .

The energy drawn from Vss over the 10us conversion period is 76.576uJ.  Therefore, the apparent COP = (1.224e-3-1.131e-3)/76.576e-6 = 1.21 .

Pm       
   
img  partzman
Re: partzmans board ATL « Reply #41 on: 2019-10-31, 14:14:17 »
Group: Moderator
Hero Member
*****

Posts: 2232
This is a variation of the previous post in that the constant current inductor L2 is reversed as shown.  This results in a positive current flow into the supply Vss during the 10us parametric inductance change period which results in a negative power in Vss.  The downside is that the voltage across L2 with the polarity change now causes a decrease in the 200ma constant current as compared to an increase as previously seen.

The pre-charge energy of L1/L2 prior to the simulation start is .2^2*.100536/2 = 2.01mJ .

After the 10us transition period, the ending current in L1/L2 is seen to be 197.53ma for an energy of .19753^2*.10203/2 = 1.99mJ resulting in a loss of 2.01e-3 - 1.99e-3 = 20uJ .

The input energy from Vss is seen at -49.638uJ for an apparent COP = 49.638/20 = 2.48 .

Regards,
Pm
   
img  partzman
Re: partzmans board ATL « Reply #42 on: 2019-11-02, 13:47:52 »
Group: Moderator
Hero Member
*****

Posts: 2232
Some other member had posted recently about viewing a "guest" that was viewing "nothing or nothing you can see"!  I just looked an there are two 'guests' with that viewing status at this moment. 

Until suitable answers are forthcoming on this matter, I will no longer be posting on this private thread.

Regards,
Pm
   
img  Centraflow
Re: partzmans board ATL « Reply #43 on: 2019-11-02, 15:40:25 »

Group: Experimentalist
Hero Member
*****

Posts: 3914


Buy me a beer
Quote from: partzman on 2019-11-02, 13:47:52
Some other member had posted recently about viewing a "guest" that was viewing "nothing or nothing you can see"!  I just looked an there are two 'guests' with that viewing status at this moment. 

Until suitable answers are forthcoming on this matter, I will no longer be posting on this private thread.

Regards,
Pm

And one "unknown action" which is obviously a hack in the act

regards

Mike 8)


---------------------------
"All truth passes through three stages. First, it is ridiculed, second it is violently opposed, and third, it is accepted as self-evident."
Arthur Schopenhauer, Philosopher, 1788-1860

As a general rule, the most successful person in life is the person that has the best information.
   
img  Peterae
Re: partzmans board ATL « Reply #44 on: 2019-11-02, 17:47:13 »

Group: Administrator
Hero Member
*****

Posts: 4637


Buy me some coffee
There's 18 members so 559 reads is not a lot is it?
I read it twice a day, if I log out and back in I can up the number by 1.

As for guests reading that's impossible, SMF does multiple checks, log out yourself, you won't see it,
As for guests doing things that are not recognized, just means they've been redirected to an invalid thread or area of the forum.

While I was here it has gone up 3 counts, 2 for me and 1 for Smudge.
   
img  Grumage
Re: partzmans board ATL « Reply #45 on: 2019-11-03, 10:45:58 »

Group: Renaissance Man
Hero Member
*****

Posts: 3099


Buy me a cigar
Quote from: Peterae on 2019-11-02, 17:47:13
There's 18 members so 559 reads is not a lot is it?
I read it twice a day, if I log out and back in I can up the number by 1.

As for guests reading that's impossible, SMF does multiple checks, log out yourself, you won't see it,
As for guests doing things that are not recognized, just means they've been redirected to an invalid thread or area of the forum.

While I was here it has gone up 3 counts, 2 for me and 1 for Smudge.

Thanks Peter.

Your explanation makes sense to me! Particularly the guest redirect.   O0

Cheers Graham.


---------------------------
Nanny state ? Left at the gate !! :)
   
img  partzman
Re: partzmans board ATL « Reply #46 on: 2019-11-03, 15:09:26 »
Group: Moderator
Hero Member
*****

Posts: 2232
Peter,

Thanks for the explanation.

Regards,
Pm
   
img  Peterae
Re: partzmans board ATL « Reply #47 on: 2019-11-03, 17:31:09 »

Group: Administrator
Hero Member
*****

Posts: 4637


Buy me some coffee
I understand there can be a whole number of reason for the unknown actions, most of the guests are bots spidering all over the forum they will goto places they are not meant to but won't see anything other than an error or refusal to show a private thread, we are also using https now so we have end to end encryption as well, posts cannot be intercepted.

What's up has some of your private work show up else where?
   
img  partzman
Re: partzmans board ATL « Reply #48 on: 2019-11-03, 22:16:45 »
Group: Moderator
Hero Member
*****

Posts: 2232
Quote from: Peterae on 2019-11-03, 17:31:09
I understand there can be a whole number of reason for the unknown actions, most of the guests are bots spidering all over the forum they will goto places they are not meant to but won't see anything other than an error or refusal to show a private thread, we are also using https now so we have end to end encryption as well, posts cannot be intercepted.

What's up has some of your private work show up else where?

No, not that I'm aware of.  I just found it curious that there were guests that appeared to be in apparent private areas of the forum.

Thanks again for your response.

Regards,
Pm
   
img  partzman
Re: partzmans board ATL « Reply #49 on: 2019-11-04, 16:04:14 »
Group: Moderator
Hero Member
*****

Posts: 2232
All,

I guess another aspect to my questioning the number of reads which is now at 604 as compared to a membership of 18 is, why are there so few comments and those were early on?

Pm 
   
Pages: 1 [2] 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 ... 31
« previous next »


 

Home Help Search Login Register
Theme © PopularFX | Based on PFX Ideas! | Scripts from iScript4u 2026-01-29, 10:12:53