My Asymmetric Electrodynamic Machines

[Hitby13kw]

Lightworker1, Very impressed with your

JohnStone, PID

Machinealive, Quad pulser now functional - Nice to see you back in the game, looking forward to your Imperial run





Hitby13kw

[Midaztouch]

@Dana

Wow man!

That "LION" is purr-ing like a kitten.. and tamed! Awesome


Keep it Clean and Green

Richie

[Lightworker1]

Lightworker1, Very impressed with your

JohnStone, PID

Machinealive, Quad pulser now functional - Nice to see you back in the game, looking forward to your Imperial run





Hitby13kw

Hello Hitby13kw, many thanks

warmest regards

light

[prochiro]

Hertz

Hello all,
I am slowly gathering information to use to control the Imperial when the generator is loaded down. This process must be done first before load testing can be completed because the speed reduction is instantaneous and there is no time to adjust pots before a stall has pulled the system to below controllable levels. Now, we have noted that we want to monitor generator HZ rather than motor RPM. They are the same thing.

1. • 1 Determine the number of poles the generator has. If you are unsure, consult the "Technical Specifications" section of the manufacturer's website. Also, the number of poles essentially refers to the number of individual windings that are arrange perpendicularly around the generator's rotor. This number will always be a multiple of 2.
     
     
   • 2 Multiply the rotational speed of the generator (in revolutions per minute) by the number of poles (from Step 1).
     
     
   • 3 Divide the result from Step 2 by 120. This will give you the frequency (in hertz or "Hz") of the generator's electrical output.
     
     

So what we need is a simple hall sensor on the shaft as that is as simple as it gets. Arduino can count and adjust duty to make the necessary changes instantly. To make this work, we must have an abundance of voltage available at battery end so out pots and power will not top out. Running the Imperial on the Minimum voltage to get to target RPM is not enough to run a load. More testing now has to be done to determine just how much voltage is enough so we have this reserve to use when needed. We need to know what are the voltage limits of Imperial but with the new arduino program that should be easier.

Is this correct so far.???????????????????

Dana

[Ufopolitics]

That is ALL Correct Sir Dana...

Hello all,
I am slowly gathering information to use to control the Imperial when the generator is loaded down. This process must be done first before load testing can be completed because the speed reduction is instantaneous and there is no time to adjust pots before a stall has pulled the system to below controllable levels. Now, we have noted that we want to monitor generator HZ rather than motor RPM. They are the same thing.

1. • 1 Determine the number of poles the generator has. If you are unsure, consult the "Technical Specifications" section of the manufacturer's website. Also, the number of poles essentially refers to the number of individual windings that are arrange perpendicularly around the generator's rotor. This number will always be a multiple of 2.
     
     
   • 2 Multiply the rotational speed of the generator (in revolutions per minute) by the number of poles (from Step 1).
     
     
   • 3 Divide the result from Step 2 by 120. This will give you the frequency (in hertz or "Hz") of the generator's electrical output.
     
     

So what we need is a simple hall sensor on the shaft as that is as simple as it gets. Arduino can count and adjust duty to make the necessary changes instantly. To make this work, we must have an abundance of voltage available at battery end so out pots and power will not top out. Running the Imperial on the Minimum voltage to get to target RPM is not enough to run a load. More testing now has to be done to determine just how much voltage is enough so we have this reserve to use when needed. We need to know what are the voltage limits of Imperial but with the new arduino program that should be easier.

Is this correct so far.???????????????????

Dana

Hello Dana,


Yes, that is all correct.

A Two Pole Generator will require 3600 RPM's to keep the 60 Hz...while a Four Pole will only require 1800 RPM's to deliver the 60 Hz...*

So far the ones we are using are Two Pole Rotor/Stator...

*Please note all this numbers are applicable to U.S and Canada...when it comes to Europe and South America...and other countries it goes for 50 Hz or 3000-1500 RPM's as described above.


Dana, if Arduino will work much better ( as response speed) with a Hall sensor...then -by all means- go for it and write code accordingly.

Imperial does not have a "limit" on Voltage...unless we are talking about excessive HV set.


Related to the drop down RPM's when load applied...depends on choosing a Higher Idling RPM, based on Voltage Feed/Signal Input...you have from 3600 to 3800 to choose from...


Right now, you could do all this "Manually"...as you add loads...Start with 3600 Idling, and write down required RPM's Increase per Loaded Watts...do as many loads as you can ALWAYS waiting for a "stabilization time" between machines (meaning DO NOT RUSH loads one after the other)...so we can then create a "RATIO OF DEMAND-SUPPLY"...or Watts Load per RPM's.

This data will help you to write codes later on...based on whatever Parameter you choose to keep stabilized...RPM's or Output Hz...


Warm Regards


Ufopolitics

EDIT 1: Try NOT to use Loads that change Amp Draw over time...like an Air Conditioning Cycling Compressor/Motor On Off...as this will completely throw out/disturb your readings.

[machinealive]

Hi guys

With a constant dc, the imperial dropped about 500 rpm per 1000w load.
I would have had to be doing about 6000 rpm to have loaded 5000 w, and only drop to 3600 rpm. I would think it would be the same, pulsed, but I'm not sure.
I think were going to want enough voltage to spin imperial 6500 rpm with unloaded gen.

I fried two fets last night, I ordered some more today. hopefully I can find some tonight at home, but I have more ordered at least.

Which hall sensors should I order. ( I see lightworker has already given this info.)
@ lightworker. Thanks light, I guess I missed your post.

[JohnStone]

With a constant dc, the imperial dropped about 500 rpm per 1000w load.
I would have had to be doing about 6000 rpm to have loaded 5000 w, and only drop to 3600 rpm. I would think it would be the same, pulsed, but I'm not sure.
.....

YES Machine that is the correct view. We are not able to get a fully dynamic design just out of the box - and it is not essential for now.

Any major problem needs to be chiseled to smaller parts in order to digest it. Keeping it simple we will crack the problem by steps:

Step 1:
Your suggstion above. Decent achievment! Due to temperature and other influences we will not get same rpm at different time and we need to get different presets for different loads.
This state is good for intro!

Step 2:
In order to get rid of drawbacks from step 1 we will add slow control of rpm (Hz) by Arduino changing PWM. Slow is essential in order to prevent dynamic oscillations caused by regulation.
This procedure will give a PI regulation scheme => slow adaption but PRECISE.
For measurement: change condition, wait for run in, take values

Step 3:
Tune regulation delay (I-part) in order to speed up run-in at load change - just below system oscillating deliberately. This value will vary from system to system - it's individual

Step 4:
Tuning for dynamic performance. Lot of math required and takings of system behaviour! This is a major effort ond needs to be adapted to individual system properties, voltages ..... It is not essential for portotype and proof of concept but desireble for daily useablity as home generator or vehicle drive.

In order to envision the final and dynamic D-component of regulation you might check the differnce of rmp changing from reading to reading, comparing differences..... If there is a small difference only you might continue with single step increment/decrement.
If the difference of rpm readings increase you might decide to do double increment/decrement. In true life there are complex formulas available and a dedicated science.

-----------
All of you being involved in programming now please consider step 2 and make it as simple as possible.
- Start with absolute safe values i.e. check rpm every second and increment / decrement duty time by one single step.
- Run the system and watch behaviour if changing gen load by 100W, 200W 500W, 1000W. Take values like run-in time, voltage, amperage ....
- You may decrease delay time to 500ms and check dynamic behaviour again.

Step 2 will enable all of us to learn more about the system: i.e. check for performance and COP with higher voltage and lower duty cycle.....

We still do not know where the stable sweet spot of our setups is. It is essential to know it in order to check for further enhancements.
Just playing with parameters is like blind flight. Theroetically and statistically there will be one safe landing possible. But how many lifes and planes do we want to waste?

Keep on good work, friends!
JS

[Ufopolitics]

Driving Imperial...Cruise Control.

Hello Guys,

Good post Sir John Stone!

Definitively we have to work hard here to find that sweet spot...

Now, I believe that we eventually will need to be able to Higher Voltage per On Time...in order to run System with even 72 Volts...but NOT delivering Full Voltage at every on time (Voltage Potentiometer Parameter Regulation)

On the other hand, We need to drive Imperial at "Idling Speed" (whichever speed is required on further testing) WITHOUT using too much Duty Cycle percentage...other words...like 40-60%...that way we have a "CUSHION" for acceleration when dealing with Generator Output/Surges.


@ Machine, You better believe that by pulsing WILL reduce draw...and still be able to withstand Generator demand...way different than linear feed my friend!
Are You using the FET's as per JS Monster spec's?
\
Now related to loading Generator: Remember the Manual/Instructions for Generators...even with a Gas Engine.

1-You are NOT supposed to load "at once" a big Wattage consumption load. As You are required to wait a time between loads...so the Gas Engine "keeps up" with demand.
2- Set Loads on both sides of Generator Output (when applicable)

And so on...


Regards ...Have to go


Ufopolitics

[prochiro]

@ JohnStone
The hall sensor has a reputation of giving a wave rather than a true on/off. A reed switch has low life expectancy. I am thinking that if a hall sensor is used, the signal should be sent thru a schmitt trigger.

Arduino text:
(Instead, if you wish to use an actual Hall Effect sensor, I'd suggest feeding it into a Schmitt trigger to convert it to a digital (on/off) signal at a calibrated level that indicates "directly under the magnet." Additionally, depending on the level of hysteresis implemented in the Schmitt trigger, you may need to do some de-bouncing that would change the rate of de-bounce based on current speed. Then you could treat it like a normal Reed relay.)

Maybe you have a better idea. I know that you are time limited but if you have time, could you give us some insight about that RPM trigger and what it should be. Thanks
Dana

[iankoglin]

Winding question

Lightworker1, Very impressed with your

JohnStone, PID

Machinealive, Quad pulser now functional - Nice to see you back in the game, looking forward to your Imperial run






Hitby13kw

G'day Hitby13kw,UFO
When winding the 19AWG UFO said to wind one pair first then the opposite pair then run to test if OK then unwind and start winding in the series format as per UFO drawing.

Is it OK to wind first pair and then wind the other pair like above then wind each opposing pair and continue to test after each winding of opposing pairs in this way the armature would be proved OK at each level and probably be more balanced?
If not OK Is there a reason to wind in succession or series?

It's good to be part of this forum where everyone is interested in helping one another succeed no bickering etc.

I am determined to push ahead with my P56 motors else I will be left behind
I have already purchased an Arduino Nano it's on it's way but today I received this

Buy Processor & Microcontroller Development Kits Semi, Dev board, RX63N Gadget Renesas GR-SAKURA-FULL online from RS for next day delivery.

Perhaps it is cheaper if compatible

Kindest Regards

Running to keep up

[Hitby13kw]

Another idea to maintain speed.

Now, we have noted that we want to monitor generator HZ rather than motor RPM. They are the same thing.

Since there is a direct relationship between generator HZ output and motor RPM (good insight Dana) rather than use a hall sensor counting pulses on the shaft perhaps it would be simpler to look at the generators output with a zero crossing voltage sensor like this http://www.edn.com/design/analog/4368740/Mains-driven-zero-crossing-detector-uses-only-a-few-high-voltage-parts


Regards team

Hitby13kw

[Hitby13kw]

I have already purchased an Arduino Nano it's on it's way but today I received this
Buy Processor & Microcontroller Development Kits Semi, Dev board, RX63N Gadget Renesas GR-SAKURA-FULL online from RS for next day delivery.

G'day Hitby13kw,UFO
When winding the 19AWG UFO said to wind one pair first then the opposite pair then run to test if OK then unwind and start winding in the series format as per UFO drawing.

Is it OK to wind first pair and then wind the other pair like above then wind each opposing pair and continue to test after each winding of opposing pairs in this way the armature would be proved OK at each level and probably be more balanced?
If not OK Is there a reason to wind in succession or series?

It's good to be part of this forum where everyone is interested in helping one another succeed no bickering etc.

It is my understanding that by winding the Imperial in the pair sequence 1,2,3, . . . to 28 we end up with


Hitby13kw

[iankoglin]

Winding Sequence



G'day Hitby13kw,UFO
What I mean to say is to wind each pair of coils and connecting them to the commutator as per UFO's diagram for winding but winding each opposing pair to keep the weight as even as possible making sure they are wired to the correct pos on the commutator it should fire up the right coil in the correct sequence its just IMHO that as you wind the armature in series like in p1 ,p2, p3 etc would not the windings gradually get larger because of the extra diam as they are wound over the other windings

Kindest Regards

Just trying to get my head around things

[Lightworker1]

Hi guys

With a constant dc, the imperial dropped about 500 rpm per 1000w load.
I would have had to be doing about 6000 rpm to have loaded 5000 w, and only drop to 3600 rpm. I would think it would be the same, pulsed, but I'm not sure.
I think were going to want enough voltage to spin imperial 6500 rpm with unloaded gen.

I fried two fets last night, I ordered some more today. hopefully I can find some tonight at home, but I have more ordered at least.

Which hall sensors should I order. ( I see lightworker has already given this info.)
@ lightworker. Thanks light, I guess I missed your post.

Hello @ Machine, Dana, UFO, All

Going back to March 2013, I started to work on the Allegro A1302 Ratio Matric Hall sensor.

https://www.dropbox.com/s/t1oa24z01sy2qhj/shot0011.png

https://www.dropbox.com/s/cf9zkzxo8n1iwip/shot0001.png

At the time, I was still thinking about using Arduino.

Anyway as the video will show later, I just used my half-working scope and frequency meter.
I used the following circuit to condition Hall-Sensor's raw output signal.

https://www.dropbox.com/s/0caacefu81gjo4a/A1302%20Hall%20Sesor%20Circuit%20Optimized.png

The potentiometer was only included later during the signal optimization process to improve the RPM measurement results.
The was pot adjusted until the good results showed.
At that point it appeared that both red and green LEDS were almost with equal brightness (however this is bit of a subjective and viszual test).

The output signal is a pulsed square wave of a kind.

https://www.dropbox.com/s/s38ym4q069ace2y/shot0013.png

https://www.dropbox.com/s/ccl5y45lc7cmbio/shot0014.png

My scope is not in best of its SCOPE-HEALTH.

Here is a table of results that relates Hz reading output from the Allegro A1302 RPM System to RPM readings and confirming those reading with hand-held RPM Meter.

https://www.dropbox.com/s/xd1hf4wy4koktkp/Testing%20Allegro%20Ratio%20Metric%20Hall%20sensor%20A1302%20and%20its%20associated%20circuit.png

Infect, the primary output from this Allegro A1302 sensor system is really Hz.
We will able to use it directly to control other parameters.

Arduino later on just did the conversionfrom Hz to RPM

Here is a screen-shot of, eagle cad in use.

https://www.dropbox.com/s/eu4zz0hiql71pms/Screenshot%20of%20eagle%20of%20Allego%20A1302%20development.png

Here is the small complete video: Construction of PCB, using freeware eagle cad software, testing the Allegro A1302 sensor.

This was all without Arduino.

https://www.dropbox.com/s/4ezwnugecxoqkyy/Allegro%20A1302%20UFO%20RPM%20Hall%20Sensor%201%20-%20NO%20ARDUINO%20Ver.mp4

Later came Arduino, and used dedicated LCD Screen.
Here I just did the conversion from Hz to RPM

https://www.dropbox.com/s/7j9fj7kwljyxmyj/Test%20rig%20phone%20IMG_1118.MOV

Here I had used UNO, since I was using I2C Bus. It is just a breadboard Version. The Leonardo did not allow me to do that.

Details here: Post# 4643

http://www.energeticforum.com/231184-post4643.html

The RPM sketch is interrupt driven. The following is just for reference to give an idea. I need to clean up the code in my final version.

https://www.dropbox.com/s/q86izco2qq6lglp/sketch_may07a_volt_rpm_from_timeAve_UNO_INT0.ino

Today I just finished part of the IR Sensor mount.
It is late now.

https://www.dropbox.com/s/p31mpkinnbuuymy/Temp%20sensor%20stage2%20%20pic%201.jpg

https://www.dropbox.com/s/4bjf3uub7d6ugsh/Temp%20sensor%20stage2%20%20pic%202.jpg

@ JohnStone. Yes you are 100% correct in stating that automatic 60 Hz must be done in some orderly manner.
I remember many many years ago, the engine management electronics hardware (all analogue hybrid op amps) for the SST Concorde had hybrid circuit modules that were controlling negative velocity feed back loops, over and under acceleration limiting circuits, circuits to spool up the engines so on. We are talking 15,000 to 20,000 RPMs all exact figures I am not sure now.

@ UFO at about same era I also was listening to Pink Floyed, starting with Atom Heart Mother LP. Well good music.

After this post I am going to hit the bed.

If there any info I might have missed please let me know.

Warmest regards






light

[Lightworker1]

Since there is a direct relationship between generator HZ output and motor RPM (good insight Dana) rather than use a hall sensor counting pulses on the shaft perhaps it would be simpler to look at the generators output with a zero crossing voltage sensor like this http://www.edn.com/design/analog/4368740/Mains-driven-zero-crossing-detector-uses-only-a-few-high-voltage-parts


Regards team

Hitby13kw

Hello Hitby13kw, good idea.


warmest regards

light