ZPDM's Oscillator: A new?? oscillator

[ZPDM]

Posted this over at Energy Science Forum and thought I might post this here as well. Not sure if the "insert image" is working but maybe the link will.



ZPDMsOscillator_zps3ef996c3.jpg Photo by PDMaher123 | Photobucket

By way of background I was trying a short while ago to rebuild this magnetic pendulum circuit that I found in the magazine Nuts and Volts, http://nutsvolts.texterity.com/nutsv...?folio=36#pg36 . I had different transistors and a different coil and was having trouble getting it to work. I figured it was all the changes to the specs but I also at one point thought, I can't keep the pin-outs straight on transistors, and so I flipped the NPN. It still didn't work and now when the magnet came by the coil stuck to it and started humming. I felt stupid but then thought, well, that should mean the coil is oscillating, maybe Bedini has taught me something, that humming might be useful in a different scenario. So I hooked up a bridge rectifier across the coil and sure enough it was pumping out the spikes, so I went from stupid to "hey this guy's good". I first thought it would be a very particular sort of thing, but it isn't. I have run this with a PNP PN290A mated to an NPN 2n2222 or 2n4401 or 2n3904 (if memory serves). Right now I am running a Tip 42/Tip 31 set-up, it just goes. I have oscillated 10 turns or 1000 turns, straight welding rod cores and toroids, even air cores. What I am having most success with right now is toroids.

I have some toroids from Nebraska surplus with pretty high magnetic permeability I took a two inch one and put 70 winds on the primary and 900 winds on the secondary. Was getting 400+ plus volts off the secondary with a 1.5 volt input, however, after playing with it for a while I saw that for charging batteries I was getting best results off the 70 turns. So I am finding good results from just throwing 100 turns on a toroid. It seems to work a bit better than a joule thief.

The way to tune this is by changing the 1mOhm and 100K pots. For each value of the 1mOhm pot there is a best value for the 100k pot. If you open the 1 MegaOhm pot up all the way you may fry the transistors, it certainly looks to go from oscillating to wide open. If there are very few turns on the coil there is only a narrow band of resistance where the coil oscillates.

So aside from from having great fun radiantly restoring batteries and wanting to share this I am also posting because I feel a bit like Goethe's Sorcerer's Apprentice fiddling with the two potentiometers. I have no idea why this circuit works but it does so and consistently with a variety of parts. I don't have an oscilloscope and this is the first time I have really wanted one, to look at what is going on with the wave form and how it might change with changes in resistances.

I hope people replicate this circuit and if someone explains to me how it works that would be great.

ZPDM

PS There is a small chance the circuit diagram is wrong, it is the first time I have used software to diagram a circuit, I think it is right but if there are problems let me know so I can correct them.

[barbosi]

Q2 tranzistor is in "reversed biased" connection, well known acting as "negative resistor". 2N2222 was largely known to experience this, some other transistors too, but not any stock transistor will work.

In your case, Q2 is the oscillator while Q1 biases Q2 and it drives the coil with those generated pulses.

More examples:
RF Oscillator with Three Parts

Simplest LED Flasher Circuit

Or Google "2n2222 negative resistance".

[Farmhand]

Looks a bit like a "Stingo" circuit.

..

[ZPDM]

Q2 tranzistor is in "reversed biased" connection, well known acting as "negative resistor". 2N2222 was largely known to experience this, some other transistors too, but not any stock transistor will work.

In your case, Q2 is the oscillator while Q1 biases Q2 and it drives the coil with those generated pulses.

More examples:
RF Oscillator with Three Parts

Simplest LED Flasher Circuit

Or Google "2n2222 negative resistance".

Interesting I just built that simplest LED Flasher a couple weeks ago. I am not sure how negative resistance applies to that circuit, it just looks to me that when the emitter/collector junction breaks down you get a pulse, it (simple LED flasher) did work with various transistors for me. You may be correct in your analysis of what Q1 and Q2 are doing I simply don't know. The ZPDM oscillator is not a breakdown of emitter/collector junction as you would need 10+ volts to see anything. I've let an LED down to 0.7 volts with this.

There is a straight connection between point 2 (one end of the coil) on the diagram and the battery negative, so what is turning on and off is the connection between the battery positive and point 1 (the other end of the coil). This connection is either going through Q1 or Q2.

[ZPDM]

Looks a bit like a "Stingo" circuit.

..

A bit, but the PNP is reverse oriented compared to the Stingo, now that I look at it the NPN is flipped around from the Stingo as well. Thx for the reference though, had never heard of the "Stingo"

[GSM]

Hi ZPDM.

That lower transistor is NPN with its emitter towards positive, the exact opposite to what we intuitively understand as being part of the normal operating capabilities of an NPN device.

However, NPN is just that - *NPN* - 'N' junctions with respect to the 'P' base, and thus still an NPN device when connected either way round;
such that those supply rails, bias arrangements and circuit can still cause it to operate when connected with its collector as if an emitter, and the emitter as if a collector.

True, the hfe (gain) will be very much reduced, with 'transistor' action as if the device were faulty, but it will still work as a transistor !!!!!

Look at where the Q2 base is connected to - to the toroidal core via a resistor !

The toroidal field will likely be induced to oscillate via noise from adjustment of the two potentiometers, whereupon the unloaded back-EMF potential developed by that winding will likely far exceed the empowering battery voltage, and R1 adjustment will provide Q2 bias with respect to Q1.
The circuit and device capacitances/ characteristics in association with the potentiometer settings will then determine oscillation frequency.

So imagine the collector and emitter (line and arrow) connections on this ZPDM circuit diagram being swapped over for Q2 and you should immediately understand how this arrangement oscillates (conventionally).

Those accidental or deliberate pin-out reversals might well give rise to some additional control/ stabilisation effects impossible to achieve by using a *correctly* biased/ connected NPN device, but at the same time, they leave the circuit vulnerable to either back-EMF induced failure, or, bias arrangement induced failure prior to any oscillation arising at all or non-resettable latch-up.

Cheers .............. Graham.

[ewizard]

ZPDM, thanks for sharing this interesting circuit. I've got some 4 or 5" toroids I'll try this with. I've found they work in situations that smaller ones simply won't work. I can look at it with an o-scope too. Not sure I'll get to this soon with the holidays coming up but probably in a week or two.

BTW the Nuts and Volts link gives a 404 error right now... maybe just temporary? Or do you need to be subscribed?

[ewizard]

ZPDM, Three questions on your circuit:

1. Do you just have one coil wound on the toroid? So basically an inductor? Or is there a secondary also?

2. Is the schematic I've attached here correct ? It looked somewhat like you had the negative of the bridge rectifier going to one of the AC legs but I assume it is like I show in the modified schematic here.

3. You mentioned a magnet. Are you using a magnet on the toroid?

[ZPDM]

Hi ZPDM.

That lower transistor is NPN with its emitter towards positive, the exact opposite to what we intuitively understand as being part of the normal operating capabilities of an NPN device.

However, NPN is just that - *NPN* - 'N' junctions with respect to the 'P' base, and thus still an NPN device when connected either way round;
such that those supply rails, bias arrangements and circuit can still cause it to operate when connected with its collector as if an emitter, and the emitter as if a collector.

True, the hfe (gain) will be very much reduced, with 'transistor' action as if the device were faulty, but it will still work as a transistor !!!!!

Look at where the Q2 base is connected to - to the toroidal core via a resistor !

The toroidal field will likely be induced to oscillate via noise from adjustment of the two potentiometers, whereupon the unloaded back-EMF potential developed by that winding will likely far exceed the empowering battery voltage, and R1 adjustment will provide Q2 bias with respect to Q1.
The circuit and device capacitances/ characteristics in association with the potentiometer settings will then determine oscillation frequency.

So imagine the collector and emitter (line and arrow) connections on this ZPDM circuit diagram being swapped over for Q2 and you should immediately understand how this arrangement oscillates (conventionally).

Those accidental or deliberate pin-out reversals might well give rise to some additional control/ stabilisation effects impossible to achieve by using a *correctly* biased/ connected NPN device, but at the same time, they leave the circuit vulnerable to either back-EMF induced failure, or, bias arrangement induced failure prior to any oscillation arising at all or non-resettable latch-up.

Cheers .............. Graham.

Thankyou to Barbosi Farmhand and Graham for helping me start to get a handle on all this! When I reread my posts they sometimes sound a little rude it is not my intent. Graham I will need to spend some time re-reading your post, and perhaps some background reading, I have no formal background in electronics, so I understood about half of it. I will try swapping Q2 around and see what I get, however, I can tell you that, unless I am missing something, when Q2 is swapped back conventionally it is back to being identical to the Nuts and Volts hobby magazine magnetic pendulum circuit (with the exception that I replaced the 100 kOhm resistor with a 100 K pot). I need to test what you are saying, if I understood it correctly, however, I don't suspect this will self oscillate (having at one time successfully built the magnetic pendulum) nearly as easily with the NPN placed conventionally, I may be wrong there but it never oscillated (that I know of) until I flipped the NPN. This is also why I first thought it was going to be a very particular idiosyncracity (if that's a word) but it's worked with a number of transistors and they don't seem to be frying after hours of use. Thx again and give me time to mull over what you wrote.

[ZPDM]

ZPDM, Three questions on your circuit:

1. Do you just have one coil wound on the toroid? So basically an inductor? Or is there a secondary also?

2. Is the schematic I've attached here correct ? It looked somewhat like you had the negative of the bridge rectifier going to one of the AC legs but I assume it is like I show in the modified schematic here.

3. You mentioned a magnet. Are you using a magnet on the toroid?

Ewizard, you are right about the link, don't know what's up, if you just google "Nuts and Volts Magnetic Pendulum" it will bring you to a magazine page and then click the mouse icon and it will bring up the PDF with the circuit diagram (you know that circuit also might have some applicability to pulse motors as well as pendulums, I just got side tracked from it)

1. Yes just one wind, that's much of the beauty of it. If by secondary you mean like the second winding that's needed for a JT to oscillate, no. If, however you want to throw another wind on the inductor and grab some spikes off it you can. The 2' inch toroid I am using now has 70 winds and 900 30gauge secondary winds. I was getting 400+ volts off the secondary into a 450 volt electrolytic cap from a 1.5V input. I found though that for charging caps or batteries the thicker wire 70 winds (charging alone) was best, I am a little put out that I can't find anything to do with the secondary as it took me so long to wind but it does keep things simple. A couple other points, I am mainly working between 1.2 and 6.0 volts, it may be more difficult at 12V at least with my set up. And as I mentioned if there is too little resistance on the 1MOhm pot it sometimes opens up like there is a short but so far at these voltages hasn't fried a transistor.

2. Yes your revision is correct and more clear. Thx.

3. I don't use a magnet, when I was building the magnetic pendulum I had a straight welding rod core and when it accidentally grabbed the magnet I heard the coil humming. I've fiddled with it a bit with the toroids and a magnet changes the frequency but doesn't seem to have much other effect.

If you get around to replicating it and it works I'd be indebted if you put it on a scope. I've never really wanted a scope before because I figured 1) Tesla never had one and he did some cool stuff 2) for the most part all I see is people fiddling for 5 minutes with dials and then they show there is a spike off the collapsing inductor, I get it. That said I do wonder whether the 100 Kohm pot may be affecting the pulse width. Appreciate the interest and good luck with the replication.

[ewizard]

Thanks for clarifying all my questions. I'll post any results I get here including scope shots.

[barbosi]

Sorry guys to spill your toys but this is old stuff. Leo Esaki was the first, it was awarded with the Nobel Prize for Physics in 1973 and later researches in semiconductors lead to development of tunel diode. Until this was developed, HP used the reversed biased transistor for ages in their electronic research to produce pulses with rise/fall time in range of nanosec. And yeah, HP is the same company that produces cheap PC and printers nowadays for the delight of its share holders.

Hence ZPDM's Oscillator is an old oscillator! The oscilator won't bare ZPDM name on it, as it was awarded already.

But it is a great example to start learning long forgoten electronics of the past, which is not a bad idea. Bedini did it in his youth, but there is more and it is not in the school books. Bedini proved it. And just by chance ZPDM just combined fast pulses (created by the oscilator which needs no coil) with Bedini principles (which use effects in coils).

Further experiments will find eventually what biases could be used to make it deterministic. As I said, until now not all stock transistors behave the same. The name of the one who can make it (deterministic wise), it will be awarded much like Hartley, Colpitts, Clapp and others.

Regards.

[ZPDM]

Sorry guys to spill your toys but this is old stuff. Leo Esaki was the first, it was awarded with the Nobel Prize for Physics in 1973 and later researches in semiconductors lead to development of tunel diode. Until this was developed, HP used the reversed biased transistor for ages in their electronic research to produce pulses with rise/fall time in range of nanosec. And yeah, HP is the same company that produces cheap PC and printers nowadays for the delight of its share holders.

Hence ZPDM's Oscillator is an old oscillator! The oscilator won't bare ZPDM name on it, as it was awarded already.

But it is a great example to start learning long forgoten electronics of the past, which is not a bad idea. Bedini did it in his youth, but there is more and it is not in the school books. Bedini proved it. And just by chance ZPDM just combined fast pulses (created by the oscilator which needs no coil) with Bedini principles (which use effects in coils).

Further experiments will find eventually what biases could be used to make it deterministic. As I said, until now not all stock transistors behave the same. The name of the one who can make it (deterministic wise), it will be awarded much like Hartley, Colpitts, Clapp and others.

Regards.

As for myself, its coming on Christmas, no need to apologize for spilling toys. I will need to look into what you are saying, if valid then sure let's call it an Esaki style oscillator, doesn't seem like a loss to me. If HP hasn't done it then someone probably did it with sticks and stones in the 1800s. In terms of sharp transients the nanosecond sounds good which likely explains my enjoyment with charging batteries with this set-up. Its also nice to only have to worry about a single wind. I don't have the electronics background and may not have the terms right but I was looking at the 555 and the same sort of voltage splitting between two resistors seems to be used to adjust pulse width on a 555. So I still don't know exactly what, or at all, to make of this circuit, but if someone suddenly shows a simple mod that starts giving it a COP of 100 I may just start calling it the Barbosi oscillator. In all seriousness, thank you Barbosi for the interest and research I will try to look into and see if I can understand where you are coming from with it. One take I am getting at first read is that a reversed biased transistor can behave similar to a fast switching tunnel diode.

[ewizard]

Hi ZPDM, I'm looking at ordering some transistors as I think I've fried most of my 2N2222's but wanted to ask which transistors seem to be working best for you out of the ones you listed. I see the TIP 42 and 31 have a lot higher current ratings. I usually get my transistors for a very good price from a guy at Thai Shine Home Page and there's not a lot of difference in cost for the TIP versions so I'd go with those if they seem to be working best. This guy sells on eBay too but you save a little from his web site since he doesn't pay eBay fees he passes along the savings to buyers. Fast shipping too. Both his 2N2222's and TIP42 and 31 are 10 for $3.51 and that includes shipping.

[ZPDM]

Did you fry them on this circuit??!! I'd just use a tip 42 and 31 since that's what I had around but I haven't found an NPN/PNP combination that doesn't oscillate yet, just got the trannys from Radioshack.