Storing Cookies (See : http://ec.europa.eu/ipg/basics/legal/cookies/index_en.htm ) help us to bring you our services at overunity.com . If you use this website and our services you declare yourself okay with using cookies .More Infos here:
https://overunity.com/5553/privacy-policy/
If you do not agree with storing cookies, please LEAVE this website now. From the 25th of May 2018, every existing user has to accept the GDPR agreement at first login. If a user is unwilling to accept the GDPR, he should email us and request to erase his account. Many thanks for your understanding

User Menu

Custom Search

Author Topic: Mauriscivic's Pulser  (Read 15018 times)

garrypm

  • Jr. Member
  • **
  • Posts: 53
Mauriscivic's Pulser
« on: January 24, 2013, 09:32:46 PM »
Hi all,
 
Looking for advice on what type of N channel mosfets to get this simple circuit running -
 
http://www.youtube.com/watch?v=KQtBPR5GmuI
 
I am referring to the second part of the video where 2 x N channel mosfets are powered via
a coil that induces a drive via the second coil.
 
If tinselkoala can spare a few minutes and investigate, it would be most appreciated.
 
I have tried with 2 x IRF840 but can not get them to start.
 
Many Thanks, Garry

gyulasun

  • Hero Member
  • *****
  • Posts: 4117
Re: Mauriscivic's Pulser
« Reply #1 on: January 24, 2013, 11:37:50 PM »
Hi Garry,

This circuit seems to be a kind of oscillator and the winding sense of the coils is important to get positive feedback to maintain oscillations. Have you turned for instance the sensor coil (L2) 180 degree mechanically? so that its axial left hand side would be facing L1 coil instead of its right hand side, for instance?   or what is equivalent to this, keep the orientation the same as per it did not start and replace its wire endings with each other.  The IRF840s sounds good to me to use, provided they are not faulty.  It maybe also helpful to adjust the supply voltage by using a variable power supply if you happen to have one that is.   Also, the turns ratio between L1 and L2 may also count.

My understanding on this circuit is that when the switch K1 is closed, the lamp load gets the input DC voltage directly via coil L1 (this is why it is much brighter then) and both MOSFETs are 'inhibited' during this time. When you open K1, then the supply voltage can get to L1 via the Drain-Source of the right hand side MOSFET but this MOSFET can only pass current whenever BOTH MOSFETs are ON because the left hand side MOSFET is able to supply current via its Drain-Source path to the Source of the right hand side MOSFET when it is also ON. And for both MOSFETs to be ON, the sensor coil must feed their paralleled gate-source electrodes with higher than the treshold gate-source voltage which should be at least 5V peak AC coming from L2, at least a voltage this high should be induced in it by L1 to start oscillations. And the phase of this induced AC is also important to maintain oscillation.  If you happen to have MOSFETs that are so called logic level types, this means they already fully ON by a 2 or max 3V gate-source voltage level (not by 4 or 5V), this may also help when your L2 coil has a smaller number of turns.  But I do not think you would need such MOSFETs:  just try to increase coupling between L1 and L2 by inserting a common ferrite core into them if the reversing of the winding sense does not help.  The common core helps in getting a much higher induced voltage in L2 when you kick L1 via K1 with current pulses. If you do not have ferrite cores to insert, try inserting welding rods or (isolated) iron wire pieces bundled into a rod form. (Of course when the circuit does not start with the coils on a common core, you still must change the winding sense of one of the coils.)

Gyula

Groundloop

  • TPU-Elite
  • Hero Member
  • *******
  • Posts: 1736
Re: Mauriscivic's Pulser
« Reply #2 on: January 24, 2013, 11:57:31 PM »
Hi all,
 
Looking for advice on what type of N channel mosfets to get this simple circuit running -
 
http://www.youtube.com/watch?v=KQtBPR5GmuI
 
I am referring to the second part of the video where 2 x N channel mosfets are powered via
a coil that induces a drive via the second coil.
 
If tinselkoala can spare a few minutes and investigate, it would be most appreciated.
 
I have tried with 2 x IRF840 but can not get them to start.
 
Many Thanks, Garry

Hi Garry,

If you have mounted the two mosfets onto a heat sink, then first make sure that the metal
on both mosfets is insulated from the heat sink. Then try to reverse the trigger coil connections.

Attached is my drawing of the circuit. I also think that it should be possible to get this circuit
to work with only one mosfet (the Q1) because the body diode of Q2 is always conducting
the current even if the Q2 is switched on.

GL.

gyulasun

  • Hero Member
  • *****
  • Posts: 4117
Re: Mauriscivic's Pulser
« Reply #3 on: January 25, 2013, 12:12:11 AM »
Hi Garry,

If you have mounted the two mosfets onto a heat sink, then first make sure that the metal
on both mosfets is insulated from the heat sink. Then try to reverse the trigger coil connections.

Attached is my drawing of the circuit. I also think that it should be possible to get this circuit
to work with only one mosfet (the Q1) because the body diode of Q2 is always conducting
the current even if the Q2 is switched on.

GL.

Hi Alex,

Thanks for your valuable additions, I fully agree with you.

Gyula

garrypm

  • Jr. Member
  • **
  • Posts: 53
Re: Mauriscivic's Pulser
« Reply #4 on: January 25, 2013, 05:12:33 AM »
Thanks GL and Guyla,
 
I had tried all of your suggestions before I posted.
 
One thing that comes to mind is that to ensure there is enough current draw through the main coil so that it induces enough
into the trigger coil I guess.
 
Anyway, I will persist.
 
GL, thanks for the suggestion of the single mosfet - I will also give that a try.
 
Many thanks, Garry

gyulasun

  • Hero Member
  • *****
  • Posts: 4117
Re: Mauriscivic's Pulser
« Reply #5 on: January 25, 2013, 12:15:26 PM »
Hi Garry,

One test you could do if you happen to have an oscilloscope is that feed coil L1 directly via the bulb (omit MOSFET for this test) from any AC voltage source between 8-12V and couple L2 coil to it to see the induced voltage amplitude across the latter on the scope, it must be over at least 10V peak to peak so that its positive half wave, 5V peak would already trigger the gate-source electrodes of the MOSFET(s).  Do you have ferrite cores on the coils?  For AC source, you can use a normal mains transformer with 9-12V secondary coil output or a function generator with sine wave output if you happen to have. If you do not have an oscilloscope, you can use a normal AC voltmeter which is good for 50Hz.

Gyula

garrypm

  • Jr. Member
  • **
  • Posts: 53
Re: Mauriscivic's Pulser
« Reply #6 on: January 25, 2013, 12:50:01 PM »
Thanks Gyula,
 
Yes I have a scope and will setup as you suggest using 12ac
 
I had noted previously that the signal generated on the trigger coil is only a couple of volts at most - thats why I thought of using a higher wattage bulb.
 
Many thanks, Garry

garrypm

  • Jr. Member
  • **
  • Posts: 53
Re: Mauriscivic's Pulser - Looking for Tinselkoala's advice
« Reply #7 on: January 29, 2013, 11:11:46 AM »
@Tinselkoala,
 
Any chance you caould spare a couple of minutes and take a look at that video for me.
 
It may be that N-Channel IRF840 are not suitable.
 
Many Thanks. Garry

gyulasun

  • Hero Member
  • *****
  • Posts: 4117
Re: Mauriscivic's Pulser
« Reply #8 on: January 29, 2013, 07:05:19 PM »
Hi Garry,

I have redrawn Groundloop's original schematic to show how I would suggest testing your IRF840 MOSFETs, see Fig. A  below. After assembling the circuit, rotate the wiper of the potmeter to mid-position, so that it would give about +6V to the gate (with respect to the source) if the main supply voltage is 12V.  And giving this 12V to the circuit, the 12V lamp should bright almost fully because as per its data sheet it has less than 1 Ohm ON resistance.  This way you could test both IRF840s to see they are not faulty i.e. they are controllable by the potmeter:  turning down the wiper towards lower voltages the lamp should get dark at less than 2 to 4V gate-source voltage levels (from data sheet the gate-source threshold voltage range is any value between 2 to 4V). Such test can also show you whether the two IRF840s has too much difference in their threshold voltage levels or you can select the lower threshold voltage types, besides checking their normal operation.

As a next step, please assemble circuit in Fig. B and use the same potmeter at R1 and R2, leave the wiper at a position which just extinguishes the lamp in circuit Fig. A  because this bias voltage just under the threshold level will help start the oscillator in Fig. B.  Remember to check winding sense of L1 or L2 and always switch on and off and on the 12V supply voltage to give a kick, helping start oscillation.
If the oscillator still does not start, then please build circuit in Fig. C   hopefully you will have success.

rgds,  Gyula

garrypm

  • Jr. Member
  • **
  • Posts: 53
Re: Mauriscivic's Pulser
« Reply #9 on: January 29, 2013, 11:57:22 PM »
Hi Gyula,
 
Thanks for that.
 
Fig A. Setup and tested both the 840's - Result > the pot acts like a dimmer and controls the intensity of the 5 watt lamp.
 
Now to move on to Fig B.
 
Thanks, Garry

gyulasun

  • Hero Member
  • *****
  • Posts: 4117
Re: Mauriscivic's Pulser
« Reply #10 on: January 30, 2013, 12:03:52 AM »
Hi Garry,

That is ok but can you extinguish fully the bulb when the gate-source DC voltage is 2V or less?  That is important. USe a DC voltmeter to check the Vgs voltage level

Gyula

garrypm

  • Jr. Member
  • **
  • Posts: 53
Re: Mauriscivic's Pulser
« Reply #11 on: January 30, 2013, 05:24:23 AM »
Hi Gyula,
 
Have setup as you recommended plus an mA meter inline.
 
Bulb begins to glow at around 3.6v then at 3.8 or higher is seems to be full bright (same current draw as direct connect 320mA)
 
What can we deduce from this?
 
Thanks, Garry
 
p.s. I was test with the IRF840's in parallel and just using the drain of the one under test. Interesting that at full pot the circuit draws twice
as much even with the opposite drain disconnected. weird ?
 
 

gyulasun

  • Hero Member
  • *****
  • Posts: 4117
Re: Mauriscivic's Pulser
« Reply #12 on: January 30, 2013, 11:02:27 AM »

Have setup as you recommended plus an mA meter inline.
 
Bulb begins to glow at around 3.6v then at 3.8 or higher is seems to be full bright (same current draw as direct connect 320mA)
 
What can we deduce from this? 

The test shows your MOSFET behaves normally and its threshold gate-source voltage is around  +3.5V.
 
Quote

p.s. I was test with the IRF840's in parallel and just using the drain of the one under test. Interesting that at full pot the circuit draws twice
as much even with the opposite drain disconnected. weird ?

Sorry, I do not fully get this.  So you paralleled 2 IRF840s except their drains, right? On "full pot" you mean the wiper is at 12V, so the common gates and common sources get 12V, right?  Do you have the lamp as the load in the single drain which is used?
Normally, if you build the circuit as per Fig. A with your 5W, 12V bulb with a single IRF840 (or with any other power MOSFET) and the wiper of the pot is set to 12V, and then you connect a second IRF840 in parallel with the first FET, the current draw, 320mA must not change, perhaps less than a few %. In fact the current draw is influenced mainly by the lamp: its resistance is 12V/0.32A=37.5 Ohm so that a single IRF840 having say 1 Ohm ON resistance in series with it does not really count much, AND two IRF840s in parallel would have say 0.5 Ohm ON resistance which would count even less than a single IRF840. Now you found that NOT connecting one of the drains of the otherwise paralleled IRF840s draw twice as much current (640mA)?  This is weird indeed and I suggest using a "dead bug" wireing style if you are using a plug-in circuit board during the tests. OR one of the IRF840s (that has the unconnected drain pin) has a fault between its source-gate at 12V gate-source voltage level.  Or I do not know....   LOL   Do you happen to have any other type of power MOSFETs?

Gyula

TinselKoala

  • Hero Member
  • *****
  • Posts: 13958
Re: Mauriscivic's Pulser - Looking for Tinselkoala's advice
« Reply #13 on: January 30, 2013, 11:25:02 AM »
@Tinselkoala,
 
Any chance you caould spare a couple of minutes and take a look at that video for me.
 
It may be that N-Channel IRF840 are not suitable.
 
Many Thanks. Garry

I looked at the video and I spent some time this evening fooling around with the circuit.
I didn't have a pair of IRF840 mosfets, but I tried IRF830 and a bunch of different coils and transformers and could not get it to work.
Also tried IRFP450 and IRF530n, no luck.
But finally with 18N50 mosfets from a PC power supply, and one particular coil set I wound for a JT, it worked, exactly like shown in the video, but with somewhat different waveforms. It's hard to interpret what's being shown on the scope in the video. The oscillation is audible, a high-pitched whine, and shows up well on my scope.
Unfortunately I wound up blowing out all 4 of my 18N50 mosfets.... the gate spike from my coil set is too high I guess. But the interesting thing is that once it started oscillating, I could reduce the input voltage from 12V down to around 2 volts and it would continue working... the only way I could stop it was to disconnect the power supply.
Also, I found that it works with just the first mosfet oscillating, with the second one removed and a jumper from D to S (turning it into a regular one-transistor JT, I think).
I think if I do this again with mosfets, I'll probably try to incorporate gate protection zeners. I think my mosfets blew from high voltage spikes on the gates.

gyulasun

  • Hero Member
  • *****
  • Posts: 4117
Re: Mauriscivic's Pulser
« Reply #14 on: January 30, 2013, 12:09:59 PM »
Hi TK,

From earlier setups shown by mariuscivic, the active components he mostly uses are scavenged indeed from PC power supplies or from other appliances which are around.  So chances are he used MOSFET types found in PC supplies. 

Regarding the scope it seems he shows the waveforms across his L1 coil and those waveforms could be ON spikes when both his MOSFETs conduct and they momentaraly switch the battery across L1, say these are the negative spikes under the zero line and the positive spikes just seen next to the very right of the negative spikes are the inductor kick-back spikes when current is switched off in L1.  (We can see single positive spikes too but I assume those are due to false triggering.)

Did you use separate coils in loose mutual coupling?

I agree your FETs were ruined by gate-source junction "puncture" and a zener diode protection is needed.  Mariuscivic may have higher losses in his coils (together with maybe less turns ratio than you had) so his FETs did not get ruined.

Mariuscivic moved his L2 coil away from L1 by at least 3-4 cm to kill oscillations, this maybe analog to your reducing supply voltage down to 2V (though you have to reduce it it zero to kill it).  However in both cases, say when he moves L2 away by 2cm or you reduce supply voltage to say 4V or less and you both do a switch off - switch on sequence on the supply voltage it is questionable your setups have a chance to oscillate under those circumstances.  This is normal behaviour for blocking or other types of oscillators (involves loop gain and phase shift as usual).

Thanks, Gyula