PopularFX
Home Help Search Login Register
Welcome,Guest. Please login or register.
2024-11-26, 08:37:52
News: If you have a suggestion or need for a new board title, please PM the Admins.
Please remember to keep topics and posts of the FE or casual nature. :)

Pages: 1 2 3 [4] 5 6 7 8 9 10 11 12 13 14 15 16 17
Author Topic: The Non-Sense Pulse Motor.  (Read 164088 times)

Group: Elite Experimentalist
Hero Member
*****

Posts: 4728


Buy me some coffee
So i was thinking a little more about the coil typology.
What are the effects of having the PM as the core?

Well my thoughts are as follows--but may also be incorrect.

In a coil with a steel or ferrite core--
When we send a current through a coil,a magnetic field starts to build from 0. This creates an EMF that apposes the current flow from the source. Second-when that current flow is interupted,the magnetic field collapses back through the windings,back to 0. This creates our inductive kickback,where the voltage polarity inverts across the coil,but the current continues to flow in the same direction through the coil.

Now our coil with the PM as the core--

We send a current through the coil so as the field generated is opposite to that of the PMs field.
This means that the field produced from the coil dose not start at 0,as there is an existing field of the opposite polarity already present. So at T=0-the moment of current flow,the induced magnetic field being built by the coil must first pull the existing opposite field down to 0,and then start building its own field,which is opposite to that of the PMs field.
Will this make a difference to the current wave form ?

Second--the inductive kickback.
In a coil with a steel or ferrite core,the field collapses back to 0.
In our coil,the field collapses back to 0,and then inverts back to the field of the PM.
I believe this would result in more energy being released during the inductive kickback cycle.

Any thoughts on this is welcome.


Brad


---------------------------
Never let your schooling get in the way of your education.
   

Group: Elite Experimentalist
Hero Member
*****

Posts: 4728


Buy me some coffee
In regards to my last post,i have decided to build another coil the same as the one with the magnet core,but with a steel core,and do a side by side comparison.


Brad


---------------------------
Never let your schooling get in the way of your education.
   

Group: Tinkerer
Hero Member
*****

Posts: 1770
In regards to my last post,i have decided to build another coil the same as the one with the magnet core,but with a steel core,and do a side by side comparison.


Brad
That would be surprising to me and bloody annoying if you're right. My limited knowledge thought that our iron held the polarity of the mag and when saturated by the coil, switched polarity and then switched back. If it holds no polarity from the start don't we miss something? And I have no idea on the effect of my massive neo rotor  on what I'm seeing.

It will be annoying because of the amount of times I tested motors with ferrite cores back in the day. I had even less of an idea back then and no scope.
   

Group: Tinkerer
Hero Member
*****

Posts: 1770
Great job on the spools Jim.


Brad
Thanks, they're a little rough. I designed them that way for my tiny printer. It's probably an easier job to print a rotor with hollow threaded poles and separate threaded capped spools. I'm still learning the tricks but tinkercad had been the most intuitive for me. That's where you design then you get in print ready in Cura. Then you print from your posh wifi connection.
   

Group: Elite Experimentalist
Hero Member
*****

Posts: 4728


Buy me some coffee
Thanks, they're a little rough. I designed them that way for my tiny printer. It's probably an easier job to print a rotor with hollow threaded poles and separate threaded capped spools. I'm still learning the tricks but tinkercad had been the most intuitive for me. That's where you design then you get in print ready in Cura. Then you print from your posh wifi connection.

I am going o have to learn to use tinker cad,or some form of design program.
You may have to run me through it Jim lol.


---------------------------
Never let your schooling get in the way of your education.
   

Group: Elite Experimentalist
Hero Member
*****

Posts: 4728


Buy me some coffee
Ok,one for the power measuring guru's

Do we use RMS values,or average values when calculating P/in on a DC system,but where the current has a sawtooth waveform?.

Below i have supplied the circuit with scope probe placement,along with the scope shots that include-RMS value's,average values,and math calculated by the scope.

Can one of the EE guys here please calculate the correct values for each.
What i want to know is the power flowing into the cap from the source,and the power flowing through the coil during the on time.

Thanks in advance  O0

P.S
I have added the full size scope shot's,as the resized ones are not very clear.


Brad


---------------------------
Never let your schooling get in the way of your education.
   

Group: Elite Experimentalist
Hero Member
*****

Posts: 4159

Hi Brad,

neither being a Guru nor an EE, but i think all power measurements via the scope math require the input
voltage / current to be in rms, and the outcome in mean (average).

Itsu
   
Group: Experimentalist
Hero Member
*****

Posts: 1808
Brad,

I'm with Itsu but I will recommend what is needed but not being familiar with your scope, this will possibly require help from someone who has the same model as yours to set it up.

To determine the accurate Pin let's say in your circuit, every instantaneous value of voltage v and current i must be sampled and multiplied together with each product stored in a table.  After a known period of time t, all the products accumulated in the table during t are added together and averaged which will represent the average power consumed at the input.  Fortunately, the Math channel on your scope does this process for us but the result must be displayed in Mean or average output.  In your particular case, this process can be repeated for the output measurement.

The time period t your Math channel will use is created by taking measurements of your waveforms between two vertical cursors that are precisely positioned so that a complete cycle or number of cycles are included between them.  The time period you are now using is from screen edge to screen edge which will get you close but is not the most accurate. 

I'll assume your scope has 8 bit vertical resolution so the maximum accuracy in this axis is with the waveform peaks reaching as close to full screen deflection as possible without clipping.

For maximum accuracy in the horizontal axis, select the highest number of horizontal samples or sample rate available which will give the highest number of products to average.

Hope this makes sense and is helpful.

Pm 
   

Group: Professor
Hero Member
*****

Posts: 1940
Hi Brad,
I come from the age of analogue oscilloscopes, but I do appreciate that modern digital scopes take samples of the analogue waveforms. Thus you have the capability of doing math on individual samples.  The correct method is to multiply each voltage sample by the current sample occurring at the same point in time and then to take the average of all those multiplications over one or more complete cycles.  I think the scope should be able to do that.  If you wish to multiply the rms value of current by the rms value of voltage and then take the average this must also be done over complete cycles.  Not having a digital scope I can't advise any further.
Smudge.
Edit PM posted while I wrote this.
   
Group: Elite Experimentalist
Hero Member
*****

Posts: 1399
... .-.. .. -.. . .-.
Good news on the Anet A8, it seems to be printing fine now, on the 4th try.
DIsregard my Youtube question Jim, as the forum came back up :)
I added 'supports' in Cura and the printer seems to be fine. Will find out in about another 30 minutes. 51 minute print btw, on a 3mm rough/fast setting.

Can't seem to edit the Tinkercad STL though...I might well be missing something.
You only seem to get grief from me lately  :-*


Update: Printed fine
You might be able to see how support has been added under the left hand side threads and between the coil ends. It just breaks off, as it's only 1 layer. Not a bad print for 3mm, now to clean it up.
« Last Edit: 2019-01-07, 21:28:47 by Slider2732 »


---------------------------
ʎɐqǝ from pɹɐoqʎǝʞ a ʎnq ɹǝʌǝu
   

Group: Elite Experimentalist
Hero Member
*****

Posts: 4728


Buy me some coffee
Brad,

I'm with Itsu but I will recommend what is needed but not being familiar with your scope, this will possibly require help from someone who has the same model as yours to set it up.

To determine the accurate Pin let's say in your circuit, every instantaneous value of voltage v and current i must be sampled and multiplied together with each product stored in a table.  After a known period of time t, all the products accumulated in the table during t are added together and averaged which will represent the average power consumed at the input.  Fortunately, the Math channel on your scope does this process for us but the result must be displayed in Mean or average output.  In your particular case, this process can be repeated for the output measurement.

The time period t your Math channel will use is created by taking measurements of your waveforms between two vertical cursors that are precisely positioned so that a complete cycle or number of cycles are included between them.  The time period you are now using is from screen edge to screen edge which will get you close but is not the most accurate. 

I'll assume your scope has 8 bit vertical resolution so the maximum accuracy in this axis is with the waveform peaks reaching as close to full screen deflection as possible without clipping.

For maximum accuracy in the horizontal axis, select the highest number of horizontal samples or sample rate available which will give the highest number of products to average.

Hope this makes sense and is helpful.

Pm

Mmm
Well this is why i asked.
I assumed that the math function on the scope was designed to do all this for us,and calculate the actual (real) power being used,taking phase angle also into account.

None the less,i also measured the temperature across both CVR 1 and CVR2. CVR2s temperature is higher than that of CVR1s temperature. This is another indication that the average current flowing through the coil is higher than that flowing in from the source,where the voltage remains almost constant.

I would have posted the video last night,but the forum went down--something about maximum band width been exceeded.
But i will post that video here tonight,or you can just go watch my latest video.

In the last video,i also measure the power being dissipated by the 12 volt LED,which for memory was another 128mW.

Unless some one can point out any mistake in the power measurements,i would say that this particular coil arrangment and simple circuit is running at an efficiency of 200% +.


Brad


---------------------------
Never let your schooling get in the way of your education.
   
Group: Elite Experimentalist
Hero Member
*****

Posts: 1399
... .-.. .. -.. . .-.
Question - are you guys using 1/2" plumbing pipe?
Am having a time of it trying to find any...I blame Pex  >:-)
A thought was of using the metal stakes from solar garden lights, but they're aluminium.

I do have a lot of zinc plated steel in a sheet.
Could a section be cut out and rolled? 


---------------------------
ʎɐqǝ from pɹɐoqʎǝʞ a ʎnq ɹǝʌǝu
   

Group: Elite Experimentalist
Hero Member
*****

Posts: 4728


Buy me some coffee
Question - are you guys using 1/2" plumbing pipe?
Am having a time of it trying to find any...I blame Pex  >:-)
A thought was of using the metal stakes from solar garden lights, but they're aluminium.

I do have a lot of zinc plated steel in a sheet.
Could a section be cut out and rolled?

The stakes on the cheap garden solar lights over here are SS-304 id say.

Yes,the PVC i am using is class 6 13mm--or 1/2 inch if you prefer.

Brad


---------------------------
Never let your schooling get in the way of your education.
   
Full Member
***

Posts: 123
@JimBoot
I have been using a Robo 3D and design on Solid Works. I have never been able to edit an .stl with it. Might be good to upload the dwg file if you are ok with edits.
By the way, DO NOT GET THIS PRINTER. Support is nonexistent and am sure I received a refurb for a new one. Case broke in transit and will not even talk to me. I am ordering a new one, Creality CR-10S to replace it.

@TinMan
   How are you calculating the winds necessary to offset the magnet insert? I am trying to figure out how to offset armature reaction in a gen head just enough to null it. Would be a great help.

thay
   

Group: Elite Experimentalist
Hero Member
*****

Posts: 4728


Buy me some coffee
The stakes on the cheap garden solar lights over here are SS-304 id say.

Yes,the PVC i am using is class 6 13mm--or 1/2 inch if you prefer.

Brad

So i had an idea in regards to this power measurement issue.

My claim is that there is more power flowing through the coil during each pulse than there is being delivered by the source-that being my power supply.

So far i have shown this by way of scope measurements,and heat dissipated by the CVRs,where CVR1 dissipates less heat than CVR2,meaning more current flowing through CVR2,and thus the coil.

Now,what if i was to put a small HHO cell in series with the coil,so as it is between the positive of the cap and input to the coil. I then run the system,and calculate the amount of HHO produced in 5 minutes. I then drive that HHO cell with my power supply only,at the input power it was when driving the circuit,and once again calculate the volume of HHO produced in 5 minutes. What would it mean if more HHO was produced by the coil circuit than that produced by running the HHO cell straight of the power supply at the same amount of power?.
Would this convince you that the coil and circuit is producing its own energy ?
If not,then why are we getting the results we have ?


Brad


---------------------------
Never let your schooling get in the way of your education.
   

Group: Elite Experimentalist
Hero Member
*****

Posts: 4728


Buy me some coffee
@JimBoot
I have been using a Robo 3D and design on Solid Works. I have never been able to edit an .stl with it. Might be good to upload the dwg file if you are ok with edits.
By the way, DO NOT GET THIS PRINTER. Support is nonexistent and am sure I received a refurb for a new one. Case broke in transit and will not even talk to me. I am ordering a new one, Creality CR-10S to replace it.

@TinMan
   How are you calculating the winds necessary to offset the magnet insert? I am trying to figure out how to offset armature reaction in a gen head just enough to null it. Would be a great help.

thay

It is not so much-how many turns,it's more-how much current required to flow through the coil with a set amount of turns  before a neutral field is had.


Brad


---------------------------
Never let your schooling get in the way of your education.
   

Group: Elite Experimentalist
Hero Member
*****

Posts: 4728


Buy me some coffee
OK,here is a little something else for you.

This test at 24 volts,4% duty cycle,@ 86 Hz

Below is a scope shot(with circuit and scope probe positions attached) showing the inductive kickback cycle,and also shows that when the transistor is on,there is 24 volts across the coil(minus the V drop across CVR2). CH1 is on the positive side of the coil,and CH3 is on the negative side. As soon as the transistor conducts,we see CH3 drop to 0 volts-minus the voltage drop across CVR2. This tells us we have close to 23 volts across the coil during the current flow period.

The second thing it shows us,that stands out,is the duration of the inductive kickback cycle  O0
Have you ever seen the duty cycle of the inductive kickback exceed that of the input pulse by this much--or at all ?.

Remember my theory i posted-->reply 75 Quote:
We send a current through the coil so as the field generated is opposite to that of the PMs field.
This means that the field produced from the coil dose not start at 0,as there is an existing field of the opposite polarity already present. So at T=0-the moment of current flow,the induced magnetic field being built by the coil must first pull the existing opposite field down to 0,and then start building its own field,which is opposite to that of the PMs field.
Will this make a difference to the current wave form ?

Second--the inductive kickback.
In a coil with a steel or ferrite core,the field collapses back to 0.
In our coil,the field collapses back to 0,and then inverts back to the field of the PM.
I believe this would result in more energy being released during the inductive kickback cycle.


Well lets have a look at the second scope shot i have posted below.
We are looking at CH4-the blue chanel.
This is the waveform created across a second coil placed face on to our drive coil-->our coil under test.
The second coil has a 220 ohm resistor across it as a load.
What do you see ?  8)


Brad


---------------------------
Never let your schooling get in the way of your education.
   

Group: Experimentalist
Hero Member
*****

Posts: 782
Believing in something false doesn't make it true.
Question - are you guys using 1/2" plumbing pipe?
Am having a time of it trying to find any...I blame Pex  >:-)
A thought was of using the metal stakes from solar garden lights, but they're aluminium.

I do have a lot of zinc plated steel in a sheet.
Could a section be cut out and rolled?

I picked up some black iron pipe nipples from Lowes.  They call it black steel pipe but I am pretty sure it is identical to the black iron pipe we used to sell at the hardware store I once worked in.  It seems to be working ok with the tests I have done so far.  I turned the inside and outside on a lathe to true it up for a better fit over the plastic sleeve and a smoother surface for wrapping tape on before adding the windings.  I guess it was made in China because the inside and outside were pretty irregular.

Carroll



---------------------------
Just because it is on YouTube does not make it real.
   

Group: Elite Experimentalist
Hero Member
*****

Posts: 4728


Buy me some coffee
I picked up some black iron pipe nipples from Lowes.  They call it black steel pipe but I am pretty sure it is identical to the black iron pipe we used to sell at the hardware store I once worked in.  It seems to be working ok with the tests I have done so far.  I turned the inside and outside on a lathe to true it up for a better fit over the plastic sleeve and a smoother surface for wrapping tape on before adding the windings.  I guess it was made in China because the inside and outside were pretty irregular.

Carroll

Hi Carroll

When you get the time,can you set up a circuit the same as mine(posted above),and see if you get the same result's--just for confirmation  O0

Thanks


Brad


---------------------------
Never let your schooling get in the way of your education.
   
Sr. Member
****

Posts: 286
OK,here is a little something else for you.

This test at 24 volts,4% duty cycle,@ 86 Hz

Below is a scope shot(with circuit and scope probe positions attached) showing the inductive kickback cycle,and also shows that when the transistor is on,there is 24 volts across the coil(minus the V drop across CVR2). CH1 is on the positive side of the coil,and CH3 is on the negative side. As soon as the transistor conducts,we see CH3 drop to 0 volts-minus the voltage drop across CVR2. This tells us we have close to 23 volts across the coil during the current flow period.

The second thing it shows us,that stands out,is the duration of the inductive kickback cycle  O0
Have you ever seen the duty cycle of the inductive kickback exceed that of the input pulse by this much--or at all ?.

Remember my theory i posted-->reply 75 Quote:
We send a current through the coil so as the field generated is opposite to that of the PMs field.
This means that the field produced from the coil dose not start at 0,as there is an existing field of the opposite polarity already present. So at T=0-the moment of current flow,the induced magnetic field being built by the coil must first pull the existing opposite field down to 0,and then start building its own field,which is opposite to that of the PMs field.
Will this make a difference to the current wave form ?

Second--the inductive kickback.
In a coil with a steel or ferrite core,the field collapses back to 0.
In our coil,the field collapses back to 0,and then inverts back to the field of the PM.
I believe this would result in more energy being released during the inductive kickback cycle.


Well lets have a look at the second scope shot i have posted below.
We are looking at CH4-the blue chanel.
This is the waveform created across a second coil placed face on to our drive coil-->our coil under test.
The second coil has a 220 ohm resistor across it as a load.
What do you see ?  8)


Brad

This reminds me of the Robert Adams Pulse Motor in the sense that he found a sweet spot of operation that sounds similar to what you are describing.

I purchased all of the Adams material from Nexus if you are interested in PDF copies.


Dave
   

Group: Experimentalist
Hero Member
*****

Posts: 2982


Buy me a beer
Hi Brad

At first look I thought I was looking at a boost circuit, then I realised how you have wired it. ???

My take on this is :-

The transistor switches on, the coil charges compressing the fixed (PM) field into the coil. The transistor switches off and the charged current in the coil wants to run in the same direction as the charge, but this time the charge of the coil is going against the incoming supply, no ground. So what is creating the current? Well my guess is the decompression of the PM magnetic field is the only power source adding to the battery voltage. It is a magnetic spring but without resonance, one shot compress and expand, then back to normal.

That is my first thought, and may not be right.

Regards

Mike 8)

PS. doing something similar but with only coils.

Just looked again, no all the power is coming from the compression of the PM, no power can go beyond the coil unless it comes from the coil, due to the diode being where it is :D


---------------------------
"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.
   
Group: Elite Experimentalist
Hero Member
*****

Posts: 1399
... .-.. .. -.. . .-.
Thanks Brad, Carroll.
Have just checked their website and may have to go in person to our local store. A search for black pipe seems to come up with PVC products.
Am wondering if they think people will use real iron for car lot/road rage fights.

Got a coil wound, using the steel sheet. I had to use foam for the ends and they kept coming off, even with glue, but it has 200 turns. Ceramics inside, while I wait for the 1/2" x 2" N48 neo's. Have ordered 2. Just a test coil, something to get started with.



---------------------------
ʎɐqǝ from pɹɐoqʎǝʞ a ʎnq ɹǝʌǝu
   

Group: Experimentalist
Hero Member
*****

Posts: 782
Believing in something false doesn't make it true.
Hi Carroll

When you get the time,can you set up a circuit the same as mine(posted above),and see if you get the same result's--just for confirmation  O0

Thanks


Brad

Hi Brad,

I did put together the test circuit like you have shown.  My test coil is NOT exactly the same as yours.  When you posted you thought we should probably cut a slit down the side of the pipe I had already finished my coil.  So my pipe does not have a slit down the side to prevent it from acting like a shorted turn.  When I did some tests to see how much current it would take to neutralize the magnetic attraction it took several amps.  I am thinking now it is because of the pipe not having the slit cut in it.  So I rewound the coil with 275 feet of wire.  By the way 200 turns comes out to be 51 feet of wire if you wind it on the pipe of the size you recommended.  So I have increased the length of the wire by a factor of over 5 times.  I now believe you probably wanted the shorter wire so that you could hit the coil with a very short high amperage pulse.  You can clarify that if you wish.  Anyway I am now planning to make a new coil with the split in the side of the iron pipe and the correct number of turns of wire.

Okay, now for some test results.  I have not had time to take pictures or a video of what I did but hopefully anyone that has been following your tests will understand my results.

I connected my scope leads just as you have shown in your schematic.  I was using my Rigol DS1102E.  Channel one across the supply and channel 2 first across CR1 and then across CR2.  I did not bother to use the math functions for two reasons.  I haven't learned how yet.  LOL And since channel one across the supply never changed I realized all I needed to do was look at what I could see on channel 2.

I measured both CR1 and CR2 with a precision meter for measuring ohms into the milliohm range.  Neither of my resistors were rated as precision resistors but amazing they both measured the same at 10.18 ohms.

Since my coil has a lot more wire on it I had to tune the function generator to find the sweet spot for the frequency.  I kept the pulse width at the 5% you recommended.  I found the best transfer of energy to the second cap to be when my frequency was 22khz.

I measured the voltage across CR1 and CR2 a couple of ways and got some strange results.  When I measured the RMS values I got 1.44 volts across CR1 and 1.88 volts across CR2.  I was using a regulated power supply set to 9.02 volts instead of a battery as I wanted to keep the input power steady.  I believe these results are similar to what you were getting with a higher RMS value across CR2 than across CR1.

However what seems really strange is that my scope showed the average voltage across the resistors to be the opposite with the drop across CR1 being 550 mv and the average across CR2 being 235 mv.  Of course the much lower value for the average voltage across the resistors is because of the very short on time of the pulses.  But why the average value shows the opposite of the RMS value is something someone much smarter than me will have to figure out.  Maybe you can do the test again and see what you scope shows for the average value of voltage drop across the resistors.

Take care,
Carroll


---------------------------
Just because it is on YouTube does not make it real.
   

Group: Tinkerer
Hero Member
*****

Posts: 1770
Good news on the Anet A8, it seems to be printing fine now, on the 4th try.
DIsregard my Youtube question Jim, as the forum came back up :)
I added 'supports' in Cura and the printer seems to be fine. Will find out in about another 30 minutes. 51 minute print btw, on a 3mm rough/fast setting.

Can't seem to edit the Tinkercad STL though...I might well be missing something.
You only seem to get grief from me lately  :-*


Update: Printed fine
You might be able to see how support has been added under the left hand side threads and between the coil ends. It just breaks off, as it's only 1 layer. Not a bad print for 3mm, now to clean it up.

Ok here you go - https://www.tinkercad.com/things/iDOajWO2B2y-non-sense-rotor-poles-20mm you should be able to tinker directly and ungroup. the thread support sucks to get off which is why I printed lying them down. Got 4 grandkids here for the week plus back at work, so spread a bit thin this week. :)
   

Group: Experimentalist
Hero Member
*****

Posts: 782
Believing in something false doesn't make it true.
Well I decided to do the tests again.  I realized after watching Brad's video again that I was using way too high of a frequency.  So I redid the test starting out with a frequency of 10 hz.  I also raised my power supply voltage to 12.41 volts.  When I got my frequency up to 200 hz that seemed to give the best transfer of energy to the second cap with the Led across it.

Checking my scope readings again I got 1.85 volts RMS across CR1 and 8.95 volts RMS across CR2!  A big change in difference between the two.  When I checked the average voltage across CR1 I got 1.75 volts and across CR2 I got 2.42 volts.  So this time the scope agreed that more energy was going through CR2 than CR1 with both the RMS values and the average values.

My only conclusion from all this is that these digital scopes have problems with short pulse duration pulses.  Anyone else have any ideas?

Carroll


---------------------------
Just because it is on YouTube does not make it real.
   
Pages: 1 2 3 [4] 5 6 7 8 9 10 11 12 13 14 15 16 17
« previous next »


 

Home Help Search Login Register
Theme © PopularFX | Based on PFX Ideas! | Scripts from iScript4u 2024-11-26, 08:37:52