Shorting coil gives back more power

[woopy]

Hi Gyula and all

i can not open your drawing the picture encounters problem at opening.

The Hall sensor TLE 4905 is unipolar and has low ma amp draw, but my dealer  run out of stock and i had to buy the cheaper but more consumer SS443A. Both of them work on the 2 sides (one side with a north pole and the other side on a south). Just for info the TLE 4935 is bipolar but turn on on a south pole and STAYS ON  until a north pole faces it. So it is not as a reed switch which  turn on and than cut off when a pole faces it (north or south). So for fast switching i would prefer unipolar Hall sensor. I don't know if it exists a hall sensor which works as a reed switch ?

For completing my previous post, i made 3 test . where i replace the led with a resistor

1- resistor of 100 ohm  = about 2 ma  and 5 volt peak pulsed (produced shorted voltage) at all speed
2- resistor of 220 ohm =  about 1+ ma and 10 volts peak pulsed (PSV) at all speed
3- resistor 1 kohm      =  about nothing (can not read on the meter) and 20 peak pulsed (PSV) at all speed

seems to be in adequation with my red writing.

Any idea ?

hope this helps.

Laurent

[gyulasun]

Hi Laurent,

Which picture you cannot open? The link I gave to David in my reply #156
http://www.overunity.com/index.php?action=dlattach;topic=10398.0;attach=51183
directly links to the picture what I uploaded in my earlier post here (it is the same):
http://www.overunity.com/index.php?topic=10398.msg276291#msg276291

Ok on the SS443A unipolar sensor: once it works for both poles for you, no need for using another type.
Regarding the bipolar type sensors, as I wrote I have never used them, once you find they remain ON till the other pole comes, then they are not good for the task here. So no need to buy and use them.

Laurent, please consider the following:
if you used so far the two SS443A sensors which were driving the two FETs like in the schematic here:
http://www.overunity.com/index.php?topic=10398.msg276272#msg276272 and assuming you build and use a correct schematic then it means that the two FETs are always switched ON, except for the moments when any one of the Halls switches them off! A crazy situation I tried to explain in my Reply #150 (the first in this page) and I mention this just to emphesize it again. This is why an inverting circuit (represented by a 2N2222 transistor) should be used between the Hall output and the FETs input, as I showed it in my modified schematic in my Reply #150 too.

So I ask whether your results with the 3 tests shown in your last post above were received with a transistor already between the Hall and the FETs or without a transistor? 

Thanks,  Gyula

[e2matrix]

woopy,  Download the picture and rename the picture changing the extension from .jpg to .png and it will open (it's a .png file). 

[gyulasun]

I will be ordering components for this setup, so could you please draw new schematic with the 4421 driver too.

Hi joefr,

I attached a driver circuit schematic with 4421 inverting type chip, it has the same pinout like the 4422.
Here is a data sheet to both chips, Page 7 is to be studied for advice on supply bypassing and grounding.
http://www.micrel.com/_PDF/mic4421.pdf

Gyula

[woopy]

Hi gyula

Yes the picture of your reply  156 to David does not open for me.

The 3 test are made without the Transistor circuit. i use the circuit as per the  included pix, i also tried the modification you have proposed for this circuit but it did not work for me. I know that this circuit seems strange but it works as it is, i mean  with the connection between the 0 of the Hall sensor and the Source of the fets.  Doug seems to also use it in his schematic.
Finally i replicated the circuit with the transistor 2N2222A as per your recent schematic and got no succes, as if it does not short the coil at all . I will check all this tomorrow.

Anyway thanks for the input and good night

Laurent