Help with resonant circuits / General Science and Electronics / Forums

Forums

4hv.org :: Forums :: General Science and Electronics

« Previous topic | Next topic »

Help with resonant circuits

1 2 next

Move Thread

LAN_403

...

Sun Jun 25 2006, 03:54AM

Registered Member #56 Joined: Thu Feb 09 2006, 05:02AM
Location: Southern Califorina, USA
Posts: 2445

See the 5th post, I know what I am going to need help on now...

At my internship at Ortel it looks like I might be doing a little work with lc resonant circuits... So not really knowing a whole lot about them, I decided to play around with them a little.

So at about 10pm I wrapped about 70 turns 30awg wire around a 5mm screwdriver to make my inductor. (I removed if from the screwdriver after wrapping it with tape to hold it together). And hooked it up directly to the ttl signal out of a function generator, and put my scope across the inductor. As I expected, the waveform looked like crap. Then I added a 5nf cap in parallel to it (which I had calculated to give a resonant frequency of 1mhz) and swept the frequency on the sig gen until I got the highest amplitude. This also gave the nicest sign wave, so I figured that was the resonant frequency. Since it was only about 50khz I removed turns on the inductor until I got at just about dead on 1mhz. Then I realised I had a function out on the generator that gave 0-12v out with a variable dc offset, so I hooked it up to that and was met with a nice tall (about 15v) sign wave. As I expected putting anything metallic in the inductor lowered the amplitude of the sign wave.

Then at about 11pm I grabbed a 1mhz crystal oscillator and used it to power my resonant circuit, which worked quite well. But since I wanted more power I added a 2n3904 with a 300r resistor in series with the base, open collector going to the my resistor/cap which then went to the 5v powering the oscillator. This gave me considerably more power in the resonator, and I actually was getting about 15v across it when powering the circuit with only 5. Since it was about midnight at that point I went to sleep.

Then today I tried messing around with it a bit more, and decided to have fun with my new 1mhz ac source... The first think that I did was put a full wave rectifier of uf4007's on it, a 10nf filter cap, and then a resistor box with an ammeter. The short current voltage was about 20ma, and at 10v I could get about .5ma out of it. Then I added a 1nf cap in series with the ac side of the bridge to see how much of that was rf, and got about 2ma short circuit current, and could pull about .1ma at 10v. Then I tried to see how much I could capacitively couple to myself so I took a pair of little leds (wired reverse to each other, it doesn't work with just a single led) and touched the hf side of the inductor. And was met with a dim but very noticeable green glow. Putting if directly across the inductor (with the 1nf dc blocking cap) lets them glow very noticeably. So at that point I soldered it to a aa sized lithium battery and was hoping to have made a little toy that you could set on a table and show to people (see, I am magic, I can make this light up by just touching this wire...) but found that the good ground my scope was proving essential to proper operation. So I gave up that idea and decided to take advantage of how I had wired the 2n3904 and shoved about 15v into it... The 3904 got too hot to touch after a second or to (you could see the amplitude of the output fall as it warmed up too) but I was generating about 30v across my inductor... It always draws about 100ma regardless of the input voltage. I am not quite sure why that is, I always thought that the circuit should look open except for the resistance of the coil/cap, in series with the transistor should look like a resistive load, but the current only changed about 10ma out of 100 by tripling the voltage...

In any case, is there any other stuff that you suggest I do? I am thinking of making another one also tuned at 1mhz to see if I can manage to transfer any power over, but I am thinking it would be a lost cause...

Keep in mind that the only thing powering that led is the 25pf capacitance from the scope

Wolfram

Sun Jun 25 2006, 12:34PM

Registered Member #33 Joined: Sat Feb 04 2006, 01:31PM
Location: Norway
Posts: 971

Get an AM radio and see how far away you can pick up the signal.

...

Sun Jun 25 2006, 07:22PM

Registered Member #56 Joined: Thu Feb 09 2006, 05:02AM
Location: Southern Califorina, USA
Posts: 2445

I chose 1mhz and made the amplifier open collector in hopes of being able to audiomod it and picking it up on am, but there is some mexican station transmitting from about 980-1200mhz

I can outpower it for about 3' running off 3.5v...

Conundrum

Tue Jun 27 2006, 07:45PM

Registered Member #96 Joined: Thu Feb 09 2006, 05:37PM
Location: CI, Earth
Posts: 4059

Interesting. I made a circuit a while back using a single FET and a photoflash transformer that ran an LED from 0.5V (single solar cell!)

Lambda diodes are fun to play with, iirc you can make a better one using the fet from a broken electret microphone and any P channel, or just use 2n3819/3820 pair.

-A

...

Wed Jun 28 2006, 01:25AM

Registered Member #56 Joined: Thu Feb 09 2006, 05:02AM
Location: Southern Califorina, USA
Posts: 2445

Ok, I need some help with the math's involved now...

Not having taken physics or having a real physics textbook that deals with these things is really hurting me as the internet tends to be quite ineffective for finding formulae for specific things

The circuit will be a coil with an id of 0.19", od of 0.385", 0.25" long, wound with 365 turns of ~35awg wire, so a resistance of about 10 ohms. There will be a ferromagnetic core inside the coil with an unknowns permeability, so I will have to measure the inductance when we get the core. It needs to resonate at 1khz. The hard part is that I need to find the voltage that I need to run at to get a peak current of 250ma through the inductor...

What I have so far is:
Assuming a relative permeability of 200 (for iron, just a wild guess until we get the part in) that gives me about 100mH for the inductor. So to resonate at 1khz I need a little under 300nf of capacitance.

But after that I am stuck

If the coil/cap had no resistance the circuit would act as a dead short and I would need 0v? But how do I add in that 10r resistor?

I almost want to learn how to use Pspice to model the circuit and just figure it out empirically...

ragnar

Wed Jun 28 2006, 03:02AM

Registered Member #63 Joined: Thu Feb 09 2006, 06:18AM
Location:
Posts: 1425

10-ohms? Think Q.

Sulaiman

Wed Jun 28 2006, 09:05AM

Registered Member #162 Joined: Mon Feb 13 2006, 10:25AM
Location: United Kingdom
Posts: 3140

I will attempt to decode blackplasma's cryptic gem for you.

Suppose you have 100 mH inductance of 10 Ohm dc resistance effectively in series;
at 1 kHz 100 mH = 628 Ohms (reactive)
so the "Q" will be 628/10 = 62.8 which is good enough for experimenting.

For a series resonant circuit the impedance will be 10 Ohms resistive at resonance
For a parallel resonant circuit the impedance will be 10 x Q^2 = 39,478 Ohms resistive.
So for 250 mA rms in the series resonant circuit you need to apply 2.5 V rms at resonance
Carefull! there will be 62.8 x 2.5 = 157 Vrms across the capacitor !

Now the bad news;
If the ferromagnetic core is a rod or bar in the centre of the coil
irrespective of the permeability of the material
only 2 to 3 times the inductance of an air-core is achieved
(total air path-length 1/2 to 1/3)
so 100 mH seems optimistic for your coil.
If the core is continuous (e.g. torroidal/ring core) then it will have to be iron powder,
a fairly large one (About 1/3 pound in weight) to avoid saturation/distortion.
A ferrite core would need an airgap and also be quite large (ETD49 etc.)
At 1 kHz a thin steel laminate core with an airgap would be best.
Don't even consider a solid steel core (screwdriver etc.) because of eddy-currents.

Moving up to high audio or low rf frequencies would mean more practical sized components.
My 'favourite' frequency range is 25 kHz (above audio) to 500 kHz (below AM broadcast)

Most magnetic materials manufacturers give excellent design information on their websites
I highly recommend

as a starting point.

Building/Optimising a Crystal Radio or a Tesla Coil would be a good introduction to resonant circuits.

...

Wed Jun 28 2006, 02:28PM

Registered Member #56 Joined: Thu Feb 09 2006, 05:02AM
Location: Southern Califorina, USA
Posts: 2445

I have built a tc and a cr, but still don't know jack about Q, except a few vague memories applying to chemistry

In any case, you guys are confirming my suspicions... The 'core' is tiny, about .1" in diameter, .25" long. The reason for this is that is is part of an optical switch, when you give it more than 250ma it turns off, and we need to turn it off at 1khz. Since you can't just turn a coil of wire on/off at a khz, we are adding the cap to make it resonant.

I thought that the inductance seemed a little high, but I read somewhere that a Fe core would give ~200x as much L as an air core, for air I calculate .47mH, so about 50ufd to get 1khz...

So having figured it out for an ideal core, how would you go about figuring how the non-laminated, probably saturating core?

Thanks!

Sulaiman

Wed Jun 28 2006, 05:32PM

Registered Member #162 Joined: Mon Feb 13 2006, 10:25AM
Location: United Kingdom
Posts: 3140

For switching I would not go for resonant operation
unless you NEED a sinusoidal current/magnetic field.
The 'Time Constant' (t=L/R) for your proposed coil is 47 us
so simple switching at up to about 3 kHz should be OK.
i.e. OFF = 0 V ON=2.5V
Adding 10 Ohms in series would allow 5 V operation and be twice as fast.
(using thinner wire would be cheaper and more reliable than an extra resistor.)

P.S. This is for an air-cored inductor of 0.47 mH inductance and 10 Ohms resistance

P.P.S. Any mosfet rated for >5v and >250 mA can be used
I strongly recommend an avalanche rated device
(If you don't see the word 'avalanche' in the data sheet then it isn't rated)
The stored energy is only 15 micro-Joules which is negligible at 1 kHz (15 mW)

Steve Conner

Thu Jun 29 2006, 09:56AM

Registered Member #30 Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706

In situations like that, where you want to "turn a coil on and off fast", the usual thing to do is use a higher voltage than needed, with a series resistor, like Sulaiman suggested. You sometimes see the series resistor shunted by a capacitor to give it even more of a kick, although you then have to make sure the resulting LCR circuit isn't too badly underdamped.

That takes care of the "turning on fast". For "turning off fast" you want to deliberately let the kickback develop to as high a voltage as possible, so the usual clamp diode you see in relay circuits gets replaced with a less effective snubber. You could even use an avalanche rated MOSFET and let it be its own snubber, like the previous post mentioned.

Is it a Faraday effect switch or an opto-mechanical one? If it has moving parts, you have to worry about the mechanical inertia of them, since that'll probably limit the switching speed more than the electrical time constant will.

One of our products that I worked on uses a MEMS fibre coupled optical switch to switch between two different lasers. We just bought the switch ready-made with fibre pigtails hanging out it, and a "+5v for port 1, 0v for port 2" input. It had a switching time of a millisecond or so.

1 2 next

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob