Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.
Status
Not open for further replies.

David Romero

New Member
Hi guys I am trying to implement galvanic isolation for my lm555 square wave generator. I am designing a marx generator so this is essential. I am new with transformers, at the moment all I want to get is the same square wave at the secondary side of the transformer but when I run it on LTSpice the output only gives me positive and negative voltage peaks at the secondary side. Can anyone give me some advice? Am I missing any components? The square wave is at around 8kHz
I have attached the circuit and the output images. The output in GREEN is the secondary side of the transformer.
Cheers.
galvanic.jpg


wave.png
 
First, at such a low frequency you need much higher inductance in the transformer, and most likely also much larger capacitors. Second, if you want your output to be only positive, you need a dc-restore circuit. A diode like here in fig.3 should do the job **broken link removed**

Have you thought about using an optocoupler instead?
 
Yes the inductance is too low, in real life the transformer would most likely saturate and pull lots of current from the 555 overheating it.
Either a much larger trans or an opto as just mentioned is the way forwards.
If this is a true marx then the isolation standoff voltage isnt necessarily equal to the o/p voltage, just that of the first stage, unless theres a chance the o/p will flash over to the i/p in which case you'll probably need to make something fancy.
 
Please show the schematic of you Marx generator and how this circuit will connect. (more information)

Are you using the transformer to fire the first spark gap? In this case a very short pulse is not a problem.
OR
Are you going to power the Marx generator with the transformer?
 
I may be able to find you a catalog pulse transformer if you use the cursors and measure the blue pulse width.

Pulse transformers are rated by their volt-time ratings, and if the area of that pulse is within something available, then you are all set.
 
You'd be better off driving a Flyback transformer with the 555, and then using the output of the Flyback to supply power to your Marx stages.
Be careful though, there is no such thing as galvanic isolation with the voltages you can achieve with a Marx generator. Forget using the mains
to supply power to your Marx, as the first significant spark will ride back directly into the line. I killed a computer 3 rooms over from using the
mains to supply power. The sparks you see in the video below ... was powered by a 6-cell laptop battery.

Here is my 250kV Marx:


This is the kind of power you should expect from the first stage output from the flyback ... Again this is powered from a 6-cell laptop battery providing 25kV from the Flyback.

"Over driven" Electrostatic motor:
 
Last edited:
Out of interest, are you smelling any O3 when running the motor? I half expected NOx, but maybe it looks more beefy than it is. Nice job though
 
The induced voltage is proportional to the rate of change of magnetic flux. Basically a square wave will give spikes as it changes almost instantaneously. This is basically differentiation. Transformers are normally fed sine waves and the output is a cosine. Normally, high inductance will prevent the instant change and approximate to a sign wave.

Mike.
 
Out of interest, are you smelling any O3 when running the motor?
... lol, yes lots of O3 with that motor experiment. The motor measured about 7x7x12 inches. I don't know what the rpm was but if you told me it was in the neighborhood of 10,000 rpm I wouldn't be surprised. You might be able to determine the rpm from the audio just after the power supply is turned off.
 
Hi guys, sorry for my super late reply. I am working on a low power implementation of the marx generator, so I am only producing 50Volts output consisting of 2 stages. I'm using the LM555 to turn on IGBTs or thyristors to close the switches of the generator and generate the discharge. I am still having trouble with both options, when i use IGBTs the output only gives me around 20volts and when i use thyristors they won't turn off after the first switching. Here are my schematics, can anyone give me some guidance?
Cheers
 

Attachments

  • igbts.jpg
    igbts.jpg
    105.9 KB · Views: 197
  • thyristors.jpg
    thyristors.jpg
    95.7 KB · Views: 199
David,

You didn't include the ltspice file, so I can't run the sim, but I've noticed something that looks wrong.

You've used net labels A and B three places in each schematic. Using the same net name ties those nets together. While I understand you want to tie the output of your gate driver to the IGBTs, you are also tying the emitters and gates of the two IGBTs to each other.

Is that what you want to do?
 
The induced voltage is proportional to the rate of change of magnetic flux. Basically a square wave will give spikes as it changes almost instantaneously.
Sorry, but that's not how a transformer works.

The reason he sees spikes is that the magnetizing inductance is too low and the transformer shorts the input signal. You need an inductance high enough to keep the magnetizing current much lower than the signal current.
If the primary inductance is sufficient, and the leakage inductance is low, than the output will a square-wave looking the same as the input square-wave.
That's how good pulse transformers work.
This is basically differentiation. Transformers are normally fed sine waves and the output is a cosine.
Transformers do not differentiate.
For an input sinewave, the output is also a sinewave, not a cosine.
The only phase-shift is between the input voltage and the magnetizing current.
Normally, high inductance will prevent the instant change and approximate to a sign wave.
Only leakage inductance will affect the rise and fall times of a signal.
The primary magnetizing inductance has no effect on that.
 
I should have realised I was talking complete rubbish when I wrote that. I swear my brain has not worked correctly for about 3 weeks now. It started as just a cold and has taken over my brain - last week I did 3*6 and got 24!! Hopefully normal service will resume soon.

Mike.
 
Never mind Pom's, I dont think mines ever worked right.
 
ronsimpson has the right idea but you may need a disconnect from the main supply in order to ensure that the SCR's turn OFF.

....but why not use a Dickson charge pump and an H-Bridge to drive the inputs if you are only going to be running a few stages?


I just put together a video demonstrating a Dickson charge pump I designed a few years ago in action... the input is 12V with three stages, the output is 36V.... another stage and you would have close to your 50V.

 
Last edited:
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top