Join us in the search for Free Energy. Share your experiments and discoveries, post your build logs, and discuss.
We have a strict No-Troll policy. So you can post without fear of being ridiculed.

New Members- Check Your Spam Folder For Activation Link

Please read our Rules. Any problems or suggestions- Contact Us


Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
[split] EEG EM Basic Circuits
#11
Microprocessors and H-Bridges

The following are simply suggestions and provide some insight into one approach
that I use when developing a variety of driver circuits for devices. Maybe a little
different prospective but it does provide another brief insight.

Mouser and STMicrosystems are the focus:

https://www.mouser.com  - worldwide supplier of electronic and related stuff

https://www.st.com - major integrated circuit and related manufacturer

Use the search at these sites to find specific information and devices, etc.

Microprocessor Development

Briefly: ST's series of STM32 32-bit ARM processors are my "go to" choice. STM
has a large variety of processors which also come mounted on PCB's with
connectors, etc. in Nucleo and Discovery configurations at a very reasonable
cost. Peripherals include nearly everything on-board, including Touch LCD's
and attached or built-in USB programming features (no external device required).

Many of these development boards also include an Arduino connector which
provides the ability to directly interface to Arduino, Nucleo and Discovery PCB's
using the included  connectors.

A variety of design/development software is also free and covers a bunch
of tools to auto-configure the processors/boards and write/dowload code
for the their various chips. CubeMX is the primary software tool plus
an integrated TouchGFX provides for somewhat easy professional Graphics.

IMHO, it's worth a day or two looking into the STM environment.
Mouser has nearly all the Nucleo, Discovery and Peripheral boards in stock
at a decent price (much better than Amazon).

Full H-Bridge Peripherals

STM, through Mouser, also has a number of H-Bridge devices mounted on
an Arduino header (plugs into Arduino Uno and most Nucleo/Discovery
boards). 

One example is the "X-NUCLEO-IHM12A1" which is a "Low voltage
dual brush DC motor driver expansion board based on STSPIN240 for STM
Nucleo" : Search Mouser for the part number. There are a variety of these
development boards with various motor drivers. (The connectors on the
board are worth more than the board price ~$11 US)

Way too much info I know, but there are some good resources available!
Anyway, Enjoy Your Designing...

SL

Full Disclosure: I have no affiliation what ever with STMicro or Mouser other than
being a satisfied user of their parts and services.
Reply
#12
Dear SL,

Ansys Simplorer also could perform the job , as is being done in Maxwell Cricut (arbitrary defined current with dataset), 

I could configure/ get same results of two similar setup in both sides, one in Maxwell Circuit and the other one in Simplorer  ( with exactly same Current form and configuration, material, number of turns, frequency ,...),

 I did this, as I was curious to measure the total required power for whole excitations and circuits and finally comparing the P-out with P-in, '

The key thing in here was current adjustment: I did the experiment with a current ( Dataset 200mA r.m.s) in Maxwell circuit and then tried to provide and apply same current for coils in Simplorer precisely, output result was same for both setup ( 165W ), 

Then I measured consumed total power in Simplorer , consumed power was 188 W!

One thing ; excitation of 16 coils in series is impossible, as it will need high voltage level , so rather parallel excitation is look like more practical and rational, I did this with 8 H-bridge , one for each pole.

I can share more detail and evidence, if required.
Reply
#13
Hi AR_AH,

Glad to hear you got Simplorer working and got the same results as Maxwell Circuit.

If I understand correctly - you drove the coils at 200mA r.m.s and got 165W output for both Maxwell 
and Simplorer; then measured the total comsumed power in Simplorer at 188W.

Does that mean the circuit had a loss of 188W - 165W, or -23W?   Hope I'm interpreting that wrong...! 

Just curious - what metal are you using for the Poles?  And, how many turns were you using per coil?  
Good point about driving all 16 coils in series.

Yea, if you don't mind sharing more details, that would be quite interesting, for me anyway.

Good stuff - you're making a lot more headway than I am lately - working on temperature treating metal, 
and trying to find a distributer (no body seems to handle small quantities of the high permeability metal)
and finding someone to do lamination Laser Cutting, Stacking and Bonding [hey, this stuff is expensive 
beyond belief]. Also, I'm quickly concluding annealing is more an Art Form than it is a Science!

Might have to buy a 100 LB roll and send it to the fabricators - would like to do a prototype Proof-of-Concept
first however - don't want to pay for a roll of exotic material and find out it's not that good or the device
doesn't work the way I think it should!

Great work!

SL
Reply
#14
Hi Again, 

Your guess was correct, unfortunately the consumption is higher than production, 

this is really disappointing Guests cannot see images in the messages. Please register at the forum by clicking here to see images.  ; I hope that that i have mistaken somewhere.

attached are the result and circuit setup, 

I did not bring the Maxwell circuit here as I believe it's easy to be configured.

about your other queries: China Steel 35cs210 is used as core, with 150 turns. frequency 1KHZ, pulse current excitation 200mA r.m.s


Attached Files
.pptx   Simplorer EE-TFG.pptx (Size: 306.11 KB / Downloads: 11)
Reply
#15
Hi AR_AH,

Still not sure what's happening with the simulation but my initial "Wild Ass Guess" is
what we discussed earlier - somehow the circuit simulator (Simplorer) is obeying
"Kirchoff's Law", that is; the simulator has to "balance the circuit loop" so it adds 
"Voltage to the Input" to do that (in this case 113.9V).

Not sure what your Input Voltage is, or if it can even be set to a fixed value, but my
guess is you did not set it to 113.9 Volts. 

The rest of the circuit appears OK (200mA).  Nice work on the "Full H-Bridge" design! 

Now we just have to figure out a way to limit, or "fix to a specific value" for the 
input voltage.

Logic tells me that a 3.3VDC SuperCapacitor should be able to push at least 200mA
into a bridge if the SuperCap is continually fed from a source Supply (Vcc). That
should limit, or "fix", the Bridge voltage to less than 3.3VDC (Note, the internal 
resistance of the SuperCap might limit the current output to the Bridge).

When your "fooling around" with SuperCaps and short one, it can give a pretty
good punch (many amps and it will fry your wires) plus they seem to charge
back up very fast (unlike regular capacitors which charge pretty slowly)  Guests cannot see images in the messages. Please register at the forum by clicking here to see images. .

Anyway, sorry for the long discussion but it's at the "core" of determining if this
EE_TFG device has any merrit.

BTW, I'm tempted to spend the >$5K to have some prototype Poles built by
a professional fabricator just to "test the Hypothesis"; but that's a lot of BEER!

Thanks for the information, really appreciate it - don't be discouraged,
figuring it all out using high-tech tools is just a part of the FUN!

Hey, you know what the "experts" say - You can't get more OUT than you put
IN - BUT in this case "you can't PUT MORE IN than you actually PUT IN...
and I don't think you put in 113.9V at 200mA - right?


SL

Just thinking about the Simplorer analysis and, if I recall correctly, they
have components for Batteries and SuperCapacitors - that just
may be the "Fix"!

Might be hard to get 113.9V out of a 12VDC or 3.3VDC Battery!

Just another WAG....

SL
Reply
#16
I hope this thread is okay to post non-simulation stuff in regarding the builds.

I've already tried to build this thing about nine times with transformer steel, solid welding steel, ferrite, and some mixed metal which I'm not sure what it is.

I tried many of the winding configurations on the initial PDF analysis. What I've noticed was that I simply could not get my input anywhere near the input that the PDF specified.  To get the volts that low, I had to use a lot more amperage. Or to get the amperage that low I had to use a lot more voltage.

So I didn't post any of these results because I didn't want to jump the gun and ask these questions until the time arose. But this seems to relate to the subject at hand at this time.

I swept through the frequencies anywhere from about 60 hertz to about 35 kilohertz to see if I could zone in on a frequency that allows me to get close to the input specifications.

Maybe my setup was not ideal, or maybe other errors crept in. But I patiently await more information on this before I try again.
Reply
#17
Hi Jim and AR_AH,

Appreciate your feedback and comments. BTW, no problem where you post at, started this fork to just kind-of
seperate the CAE/Physical stuff from the General/Specific Electronics stuff.

Good to hear you're trying a few things but bare in mind this is still at the early development stage - only about
a month since the birth of the concept and pretty much all the work so far is just theories and ideas. Took over
a year to come up with something that was workable with the LinGen and it still needs more polishing to be
useful. 

However, the EE_TFG might move along more quickly since there is some 'documented experiences' found
in a variety of papers regarding the TFG, using permanent magnets, in recent windmill generators.

So far all the CAE work is "ideal," that is, no accounting for losses or practical implentations has entered the
equation or design. Still a ways away from a "proof-of-concept" prototype build - in the process, however.

Attached an "Electrical Steel B-H Curves - TFG_zxx.pdf" which charts the "steel" aspects. Note the variations
around the "H (amp/m)" vs "B" for, example 10 and 100 for the different materials; then add the annealing, etc..

There were scores of analysis/simulation iterations done just trying to get a "feel" for what was happening - looks
like your experiencing the same thing; but a few did show promise, so I'm still slugging away at it.

Recall that one of the early ideas of this thread was to provide a method (free) of checking the "B Fields" in different
magnetic materials using various coil windings and currents by way of AEDT Student. You may have to enter the
B-H Curve information manually but it's do-able and the results can yield a whole lot of insight with respect to your
intended (hoped for) outcome.

You might even be able to work it backwards - determine the theoretical "B" value and then check it with your Tesla
or Gauss Meter. Both the Magnetostatic and Eddy Loss capabilities of AEDT Student can help.

Also attached a "Coil Winding Wire Resistance.pdf" sheet that might be of general interest -
but it's mostly for AR_AH's reference...

AR_AH - please check my math, I've been known to really screw this stuff up in the past!  Guests cannot see images in the messages. Please register at the forum by clicking here to see images.

SL


Attached Files
.pdf   Electrical Steel B-H Curves - TFG_Zxx.pdf (Size: 936.03 KB / Downloads: 25)
.pdf   Coil Winding Wire Resistance.pdf (Size: 55.29 KB / Downloads: 15)
Reply
#18
Hi SL,

Hope things goes well ,

about the calculation, 

in addition to resistance part (which you sent the result and almost is negligible) , you need to consider the inductance of the coil too, 
which most portion of drop comes from this part due to "rate of change of current during the time varying (pulse width )." 
 

with our example 200mA, and f=1khz (pulse width of 0.5ms): the rate of change of current or di/dt is =200mA/0.5ms =>400 and this should also be multiplied to L too, 

L can be extracted from Ansys matrix table, which shows 600mH (the highest L variable of one of coil's inductance),

So, therefore  the total drop would be  V=400*600mH= 240 volts! (please don't' ask what bout 114 volt as i do not know where this came from in Ansys)

I hope this could clarify, why such voltage drop is required to pass even few milliamps !
Reply
#19
I believe you're referring to inductive reactance. It has to be considered when building such a machine.

   

If you are aiming for relatively high frequency then I think you need to try to minimise inductance, otherwise you will not be able to pass a large current through the coil.

If you are aiming for a low to modest frequency then you can of course aim to maximise inductance.

The above applies if you are aiming to pass a large current through a coil, however if you are trying to minimise the current from the source then I think it's best to aim for as high an inductance as possible.
Reply
#20
Some Very Good points fella's,

Ahhh - The Plot Thickens!

Now you are entering the detailed Engineering Circuit Design Chapter of the EE_TFG. 

After reviewing the following you might quickly appreciate why this "stuff" would be difficult,
if not impossible, to go through in any depth on a text type thread.

Electrical Engineering Khan Academy - Inductor equations: (video)
Inductor Equation and Videos
 
(Note the example - the pulse waveform used - a +/- 2V 4mSec pulse and a 10mH inductor)
Also, on the right hand side see more text and video pages.

Recall: For Inductors in series the inductance is added [ Ltotal = L1 + L2 ], for inductors
in parallel they are divided and indcutance becomes less [ Ltotal = ((L1/L2)/L1 + L2)) ].

Optimization of the # of Turns, Current required, and wiring are a part of the development process.
All the design "trade-offs," including Pole material, will hopefully yield a viable device - TBD...

Even though it may appear simple "on it's face" - the problem is - it can get complicated quite fast!

Hopefully a lot of the actual detailed design will be included in a future paper (if this scheme works).

SL


Attached Files Thumbnail(s)
   
Reply


Forum Jump:


Users browsing this thread: Tisserand, 1 Guest(s)