Pulse motor circuit - strange effect ?

[Slider2732]

Yep Zooty
They come out easily enough, being fridge magnet type of stuff, but it's very useful to keep it where it is.
I use that strip to attach magnets for power tests. Small neo's just clip on to the outside of the hub, with enough force to stay where they are (up to and including 12V) yet can be moved around too.
I'm not sure whether full out speed is what's needed, but that's how I set mine up.
Interestingly, with having the relay coil off center from the rotor, if you move the magnets around you can have it run in attraction or in repulsion.
An LED with suitable resistor on the output allows tuning too. You can make sure that the Hall sensor switches on and off, by watching the LED go on and go off.

Building one up now that I hope shows flexibility of build. It uses a KN2222A, a 1N4148 diode instead of the LED across Emitter and Base and a Hall sensor that came off the very fan assembly that formed the rotor.
Just 'tuning' it now. It runs perhaps quicker than the original. I had 2 square magnets and 2 long thin magnets..so broke one of the long ones in half, with the internal cutter of a pair of pliers. Many folks in the R/C community swear that ruins the magnets, but it never has for me. Got a few 6 gram planes flying around with bust apart magnets in their rudders

If nothing else...these things are very quick to build and teach us much about tuning magnets, effects of flux change, speed to power out differences and more.

[barbosi]

Congrats Slider,

I will try to add my 2 cents on your original setup. First, keep it as it is, don't modify it or use parts from it or... It will help you understanding what is going on while still working. You can try to replicate it with slight variations.

IMHO, here is what I think about your setup:
1. At core it is a Bedini circuit, or Muller if you want, the point is that is using the principle of switching on/off the coil.
2. What few realise is that the motion of magnets in proximity of coil, will induce EMF in the coil which will add to the battery voltage. The more speed the more EMF. The more current flowwing through the coil the bigger the spikes when the transistor is shut off. This is more energy stored in the cap.
3. I don't know where did you see it, or what was the inspiration of using the diode on the positive side of the coil (D1 in my drawing). But I personally think this is brilliant since it prevents the EMF induced in the coil to be "levelled/flattened" with the battery voltage. Switching the transistor OFF when it is at a max of voltage is the "secret". You may try to short the diode with a wire clip and see the effect disappearing.
4. Since you use an Avramenko plug, the LED across the transistor is not necessary anymore. Look at the schematic and see the LED is in parallel with the Avramenko's diode on the negative side. You may try removing the led and unless the led provides an obscure functionality unknown so far, the system should still work.
5. because you maintained the magnetic strip in the casing (which appear to be plastic to me), the poles from that magnetic strip might contribute to emf generation in the coil. But that you can prove or disprove only using a scope and eliminating the power from the circuit. The probes across the coil will show you how many peaks you have in one rotation.

The software I was recommending (Eagle) is also for Linux: Download Eagle - Freeware Software - Tucows
Or better yet, straight from the source: CadSoft Online: Home of the EAGLE Layout Editor
Keep up the joy while working on your project.

[Slider2732]

Thanks for those kind words and pointers Bruce

Great to know the software is for Linux too.

The diode on the coil was a natural thought, to stop the spike returning to the circuit. The LED on the transistor is simply a leftover from the Slayer Exciter circuits recently built up. My latest 'White Crow' has a 1N4148 diode instead...i'll omit it in another one and see what happens. This one i'm building will be tested when I run the 6V batt down and if it works the same, will be off and away to a family member who likes this sort of thing.
The first one was hotglued to the workbench, so now is away from the table awaiting something to mount it on..piece of board of some type.

I like anything that performs more than 1 function. That's the micro R/C aircraft side coming about again from me.
The magnetic fridge magnet stuff is well known from the aircraft work, ripping it out when rewinding motors (a brushless motor from a brushed Mabuchi cassette motor is easy, just take the back off, ditch the brushes section, wire the 3 coils to the 3 speed controller wires and the can rotates like an outrunner).
The magnetic material works very well to place the neo's, has a uniform field because it's round and continuous, serves to balance the rotor through the additional large extra weight and uses that weight for a flywheel effect when running.
Oh, a tip for when you have the right speed and power. Fix-All adhesive from Dollar Tree sticks the neo's to the hub very well


I agree about replications, just build as it was initially presented. It's not like it uses hard to find components. Any replication will be a humbling experience for me...especially if results support my findings. That people think it looks good enough or different enough to replicate is an experience !

[Slider2732]

Errrr...what's the best way to flatten a battery for repeat testings and to give an indicator of what charge was in it ?
I put the 12V PC fan on the 6V battery at 1am. The idea was to make it go flat and then run the second build overnight to test. If it did the same thing as the first circuit, it would be a replication. It's 2.30am and the darn thing is still going. The meter reads 5.8V still
I believe a 1W or 5W resistor is the way to do it but don't know for sure.

[Slider2732]

After allowing the 6V battery to discharge through the 12V PC fan, which took all night, the experiment of self charging was repeated. However, because the original circuit had been hotglued to the work table, the circuit needed to be mounted on something. I put it on a piece of card for the time being. Would the coil be in the correct spot ? As it turns out, no, not quite. A 5 minute charge from the wall adapter produced an excellent 12hr run of the circuit, but ultimately things have proven to be not quite right again yet.

So I advertised on Freecycle for a dead 6V lead acid battery.
Kindly, a woman replied asking what voltage house alarm batteries are, she had one on top of a cupboard that had sat there for a long time. I found out that at least some are 12V, 4Ah, but said any lead acid battery at this point would be fine, so would collect anyway. Picked it up last night and immediately saw something horrific - a big bulge in the side of the thing
At 11pm I began working on it with another White Crow build.
Running various tests and crudimentarily measuring with my meter as I went, the circuit rejuvenated the battery (a UB1245) to 7.9V, but with only around 20mA. Enough to light a superbright LED though...which I walked around the house with attached and decided was a success lol.
Here's where things get good though....
Looking at other related threads on the forum alerted me to this page, of John Bedini's work: New Page 1
I've been working through it, applying 'aha!' moments as I go.

The video below shows what happened when a load of a 12V light bulb was applied while charging...the current rose from 20mA to 250mA, and slowly began to light the bulb off the battery terminals.
it's a brilliant effect and worth the whole time so far with this battery just to see that happen !!!

YouTube - ‪Pulse motor - amazing current effect‬‏

I now have it sat with no return path to the circuit and with meter attached for a volt reading. Over 30 minutes, the voltage reading has risen from 7.8V to 9V at the moment. My meter forms a load of 50mA at the 50mA setting and 250mA at that setting, so it's difficult to ascertain the actual charge readings...I presume such a thing is normal for an analog meter though.
The battery is repairing slowly and i'm learning bucket loads of new information while it does it

[minoly]

Slider,
are you running the ckt as shown above?
I think that AV diode arrangement is in JB's Solid state ckt.
I've not seen that diode added to the top of the power coil however.
I smoked our last hall so I've been running this w/ a trigger coil for the time being. I'm curious of the timing for the hall, does it work in both repel and attraction mode?

That's a nifty little transistor, have you tried it w/ any others to see if you get the same effect?

I'm using the MJL21194 completely different curves than the one you have. And I don't seem to be getting the same effect. also, what diodes are you using?
,
Patrick

diode 1n4148? from vid..

[wings]

After allowing the 6V battery to discharge through the 12V PC fan, which took all night, the experiment of self charging was repeated. However, because the original circuit had been hotglued to the work table, the circuit needed to be mounted on something. I put it on a piece of card for the time being. Would the coil be in the correct spot ? As it turns out, no, not quite. A 5 minute charge from the wall adapter produced an excellent 12hr run of the circuit, but ultimately things have proven to be not quite right again yet.

So I advertised on Freecycle for a dead 6V lead acid battery.
Kindly, a woman replied asking what voltage house alarm batteries are, she had one on top of a cupboard that had sat there for a long time. I found out that at least some are 12V, 4Ah, but said any lead acid battery at this point would be fine, so would collect anyway. Picked it up last night and immediately saw something horrific - a big bulge in the side of the thing
At 11pm I began working on it with another White Crow build.
Running various tests and crudimentarily measuring with my meter as I went, the circuit rejuvenated the battery (a UB1245) to 7.9V, but with only around 20mA. Enough to light a superbright LED though...which I walked around the house with attached and decided was a success lol.
Here's where things get good though....
Looking at other related threads on the forum alerted me to this page, of John Bedini's work: New Page 1
I've been working through it, applying 'aha!' moments as I go.

The video below shows what happened when a load of a 12V light bulb was applied while charging...the current rose from 20mA to 250mA, and slowly began to light the bulb off the battery terminals.
it's a brilliant effect and worth the whole time so far with this battery just to see that happen !!!

YouTube - ‪Pulse motor - amazing current effect‬‏

I now have it sat with no return path to the circuit and with meter attached for a volt reading. Over 30 minutes, the voltage reading has risen from 7.8V to 9V at the moment. My meter forms a load of 50mA at the 50mA setting and 250mA at that setting, so it's difficult to ascertain the actual charge readings...I presume such a thing is normal for an analog meter though.
The battery is repairing slowly and i'm learning bucket loads of new information while it does it

lot of voltage

[Slider2732]

Hi Patrick and thanks wings

Glad to hear of a replication in progress
The one in the latest video is a KN2222A, purely to go to the opposite extreme as the first one for Collector output. The KN is the big meaty version it seems of the very common 2N. Wireless energy tests with my Tesla towers showed the KN2222A to outperform many others (there's a full list of tests I put together in the massive Joule Thief type thread that jonnydavro started a year or so ago). My list was of comparisons for wireless energy ability, but may well apply for this too...somehow.
I'll attach it

Some of this is counter-intuitive. You may think that a stronger Collector current would help the coil to extract any energy for firing, but, more energy means less charge efficiency. Hence, the 2SC2240 was selected, with a limit of 100mA. The coil can only use that much.
A 2N3904 'should' run well, but hasn't been tested yet. The saturation voltage is what i'll be testing later today. If an ultra low saturation of 0.3V say produces longer runs than a 0.9V, then that helps to identify the best transistors. Course, the transistor is only one aspect.

The magnets have a crucial role to play.
They are set at opposite polarities around the rotor edge.
Which direction is seen by a bi-polar Hall sensor decides which magnet reaches the coil to induce the BEMF (a theory). So, I personally refer to attraction as the magnets coming toward the coil for the Hall sensor to switch and repulsion to be going away from it. Terminology is something i'm quite rubbish with, sorry about that, new here.
Magnets should, it would seem, follow a curved placement rather than 100% fixed oppositely to each other. A magnet may point up vertically, but its opposite will be out by some degrees. Same with the other two. Such that, 2 sets of identically off centred magnets are what you have. It's all about the trigger moment as the Hall sensor switches. 4 or 6 have worked for self running. 6 run the rotor faster, but are not necessarily the best option. For some reason, the spikes aren't in phase with the plate energies of the battery (another theory). You'd think it would be more efficient, but isn't.
Pure AC would seem to be necessary too, yet i'm finding that the circuit exactly as above (replacing 1F with 6V lead acid) is working best.
Likewise, as high a rotor speed as possible is not the goal...there's a point where the power to turn the rotor is less than the power returning down the AC line...that's what you want. A very high speed normally means the magnets are being pushed harder by the coil, i.e. that it is closer to them. That forms drag when wall power is disconnected.

The diodes are again unfortunately a gleaned info scenario for me...i'll find an exact looking type and post a pic

The busted up battery of the last video has rested for the past 2 hours. It's showing 6V now which is better, but mA are still only 20mA.

[Slider2732]

* Update *

An update to a post from the dark ages, well, in terms of pulse motor theory and operation.
The original motor has undergone many weeks of careful changes, tunings, is now double diode rectified, but essentially the very same unit.

As a 4th July special, it has now put in a few runs from a 5.5V 1F supercap of around 20 minutes, greatly improving the 'standard' runs of 7 minutes or thereabouts
The supercap is the same exact one as those runs from weeks ago and these days is only filled to its rated voltage, poor thing saw quite a few runs at above the 5.5V originally.
The main difference nowadays, is that there is a load on the output, in the form of a white LED.
Here is a run from this morning, of just over 19 minutes.

YouTube - ‪19 min run on a 1F supercap - Pulse Motor‬‏

Very pleased to have it back to cool runnin'