Muller Dynamo

[joefr]

Hi Kone

I looked again at your original arduino-timed circuit with jsut the backemf/recoil into capand noticed that mosfet right off of the pins 6 and 7 of the driver - whats this switch there for? Are you "switching" in the backemf to that cap then?
Kone I used this circuit from web page http://www.totallyamped.net/adams/ Page 3 Fig. 8
I just upgraded this circuit to use Mosfet TC4421 driver and later add Opto isolator to control circuit with arduino.

If so when is this timed and whats the pulse width?
The Hall1 goes to Arduino input and is timed 1 to 1
Is there overlap in time with the motor coil pulse or is it right after?  Does it turn ON simulataneous to motor coil pulse being turned ON?
When South pole of passing magnet turns on Hall sensor, the Arduino output turns on TC4421 driver and Q1 Mosfet turns on and engage the drive Coil L1
So it is working the same if I have Hall sensor directly connected to TC4421 driver pin 2.

Also in your SSR 1 and SSR 2, when SSR 2 dumps cap to load, (ON) SSR will be OFF - but what is the pulse width of each compared to other?
Are they simulatenous or do you have a longer OFF pulse to SSR 1 as compared to SSR2 ON pulse to make sure it is totally OFF when SSR 2 turns ON such as to allow for switching time-lag?
This is what I am working on now. I am controlling SSR1 and SSR2 depending on voltage in CAP C1 .
When lets say voltage on CAP reaches 100V, SSR1 disconnects C1 from D1 and SSR2 connects CAP C1 to load.
When voltage on CAP C1 decreases to 50V, SSR2 disconnects load from C1 and SSR2 connects CAP to D1.
I can set voltages for CAP C1 from 0-200V in arduino code.
I have to refine the code so that the SSR1 stays off longer when SSR1 is dumping voltage from CAP C1 to load.

Regards JoeFR

[joefr]

Hi All

I posted new video on youtube about different speed of charging 680uF 200V Cap with BEMF from Pulse motor

http://www.youtube.com/watch?v=I94JezZkul4

JoeFR

[konehead]

Hi Joe

I saw yor newest video - cant see the votlage and amps input on that blue screen at but nice video and setup and experiment....

9 times at a peak might be too many times for the coil shorting, not really permitting the coil to collapse, plus all those shorts make for long time to be lurking over the peak period too...

I think you need to have "fairly" larger pulse width to the coil-short event, in order for the coil to collapse all the way down to nothing (and beyond)  so it springs back stronger with really strong kick to fill caps....

Ismael does 5 times at the peaks and from what I understand from him,  there is a single intial coil short that creates those ringing oscillations, and then the next 4 shorts actually are timed to occur at the peaks (both pos and neg peaks) of the actual oscillations themselves - if you can do that, then you wil really be in fat city is what I think.
If this is right, then the first coil short would have different pulse width to it, then the next four... and timing too would be different - not just 50-50 all the way across.
The whole idea is for the ringing osciallitons to EXPAND IN VOLTAGE , not decay...when they decay as they will do,  only the first couple will actually fill the cap up - and the rest is wasted doing nothing in way of filling cap since cap voltage is already higher than they are from the first coouple oscillations.

Anyways I think from your new video demo, you are getting twice the power out now compared to no coil shoritng of the backemf - as the cap fills twice as fast, so that means you could have twice the pulse rate from it to load, so you would have twice the power...

I used that cap-discharge formula calcutaitng the INPUT of my RomeorUKvariant motor yesterday - I was using drained down car battery as the power feed and it reads only 10.25VDC

What I did was install another halleffect that turns on a 1405 mosfet right inbetween the motor coil pulses, never at the same time, and this 1405 mosfet turns on only ONCE a revolution, and its purpose is to connect that battery to a capacitor (6800uf 50V) and fill up the capacitor with a quick pulse once a revolution from the motor.

This way the batteyr never actually powers the motor coils, the capacitor does. I call this a cap tank-circuit but dont know if that is good term for it.

So the cap has only 8.7VDC in it (a voltameter across cap says anyways) while it runs the motor at 1440rpm...
Analog ammeter says 150ma from battery to cap...

so now the motor runs on cap-discharge, so I wanted to verify that the cap discharge formula matches up with what the meters say ( meters say from battery it draws 150ma using analog meter and 1.6VDC for:
1.59 WATTS INPUT at 1440rpm)

put scope across cap and first calibrated scope with 9VDC battery to see where that "fixed" voltage lined up too - anyways I figure its 9.2VDC in cap when cap fills from battery (pulse widht to fill cap from battery through the 1405 mosfet is very fast - more pulses or longer ones could have the votlage higher but its not right now - jsut 9.2V in cap reading with scope)

And this drops to 7.6VDC the scope says...and there are 24 of these cap-fillign pulses per revoltuion of motor per seocnd too ( 1440rpm / 60 = 24)

cap is 6800uf so now got all the numbers for that formula:

9.2 x 9.2 = 84.64
7.6 X 7.6 = 57.76
84.64 - 57.76 = 26.88

.0068 / 2 = .0034  X  26.88 X  24 = 2.193 WATTS input

so its pretty close to what the meters say...maybe digital ammter would show more than the analog type so it's very near the same thing...anywas I woudl say that formula is fairly accurate from this....

[joefr]

Hi Kone

Nice post

I have managed decrease time to charge Cap C1 680uF to 200V in 10seconds (15 seconds less that I show in last video ) the pulse motor is running on 12V 200mA input 2.4Wats.

I have another problem to solve. Because I am now controlling CAP charge and discharge to load depend on voltage ( reading with Arduino analog port ).

The problem is that the SSR relays are to slow at turn OFF. They need 10miliSeconds to turn OFF. This is too much time and I have problem when I put load with less resistance ( 1 - 10 ohm ) the CAP discharges bellow voltage I set in Arduino ( and kills my Light bulbs  ).

I have to replace SSR1 and SSR2 with Mosfets to see if this will solve the problem with turn OFF speed.

Regards JoeFR

[chalamadad]

( and kills my Light bulbs  ).

LOL! Sorry for that.