Proof of concept - perturbing a static magnetic field

[gyulasun]

Here is a link to its data sheet:

http://www.irf.com/product-info/datasheets/data/irs2117pbf.pdf

Gyula

[bob.rennips]

The IRS2117pbf high side driver. These are a few notes that I've made concerning this driver.

1. The bootstrap diode across Vcc and Vb (pin 1 and pin8) must be an ultrafast diode. Its reverse voltage time must be less than 100ns. If you get this wrong the bootstrap capacitor will not charge up in time to produce the 10V plus over the supply voltage of the driver. i.e. to turn the mosfet on when switching say 500V you need at least 510V at the gate. The bootstrap capacitor provide the extra 10V over the 500V. The spec sheet mentions 10KF6 but stuffed if I can find out the specs on this diode. An apparent equivalent is 11DF4, reverse voltage time of 30nanoseconds.

2. Although the datasheet doesn't suggest this. I've seen designs where 1 or 2ohm resistor has been put in series with the diode to prevent high current on initial charging of the bootstrap capacitor. Seems to make sense to me.

3. Bootstrap capacitors between Vs and Vb (pin 6 and pin8) needs to be at around/at least 100nF. But if you make it too high, it won't charge up in time to give you required volts, too small and it won't be able to sustain the on charge at the mosfet gate during the required on period. I read on another board that an electrolytic should not be used for the bootstrap capacitor.

4. In other words getting these components right is the black art of electronic design! The high the frequency (or frequency range) the circuit has to cope with, the more arty you have to be!

5. Vin needs at least 10V for logic 1.

6. It has an undervoltage detect. For this reason I'm planning that all logic circuits will be connected to one battary which will happily sit at 12+ volts for many days. The high voltage (if needed) for the pulses will be provided via DC-DC convertor from another battery. This way spikes from the coils, even with diodes, will not make there way back to the battary concerned with keeping the logic going. The last thing I want is interpreting results as being anomalous when what is really happening is the driver is cutting out due to spurious transients.

Once I've got a working set of components and values I'll let you all know what they are. This may take a week or so depending on when I can sit down and do this.

If anyone else gets their first - please holler!

Cheers, Bob.

[bob.rennips]

Update on circuit

I got the mosfet driver and mosfet side of the circuit working today. I only had 12 volts switching to the two bifilar wound coils.

In short nothing unusual to report. The addition of a static magnetic field only increased the size of the output pulses by around 10%. There were no additional spikes, ringing or anything to suggest anything unusual happening.

The circuit was a compromise on the components I already had around the house. I'm going to abandon the use of the counter to control the pulse width, not enough fine tuning control. I need to be able to independantly control for each pulse train:

1. Pulse width.
2. Pulse spacing.
2. Pulse frequency.
3. Phase between pulses.

So back to the drawing board. I think my next board will be based on a Bob Boyce design.

Oh yeh, a solderable breadboard is needed for mosfets and the like due to heat and the need to get a good contact.

Image of my breadboard circuit attached.

[hartiberlin]

Hi, what frequencies did you use ? With bifilar coils you have to go over 100 Khz to see some effects, but then you will see them...

[bob.rennips]

Hi, what frequencies did you use ? With bifilar coils you have to go over 100 Khz to see some effects, but then you will see them...

The coils were wound bi-filar as per the original patent at the begninning of this thread. The idea was that two sets of pulse sequences could be put through the coils, some overlapping and some interleaved. I tried frequencies up to the 1.5MHz and as low as 500Hz. I know the patent says to use high voltage but I hoped for some smaller effect to be visible even at low voltage.

Since I started this project I've been reading up more on Bob Boyce's experiences and I can see the parallels between the patent, SM and what boyce is doing. As such I now believe that the phase between successive pulse sequences is a lot more important to the extent that one pulse may be appearing only a few 10's of nanoseconds after a pulse in another coil. i.e. You have two pulses in very rapid sequence one pulse into 1 of the bifilar coil and the other into the other part of the bi-filar coil.

I now want to take this project more towards having very fine control of the phase between two identical frequencies.