Resonance Effects For Everyone To Share

[gyula]

Hi Ren,

My understanding on the TL494 is that its duty cycle can be controlled by changing the DC level on its pin 3 wrt to the negative ground.
Here is link on a motor speed control with TL494, perhaps you have seen it but it does change the DC level at pin 3 with the 1kOhm potmeter (R6 in the link schematic).
PWM Control Speed Motor 12V By TL494 :: Electronic control circuit page 01
Of course it should control in the same way with your 10kOhm pot too. How do you mean your bread boarded circuit for duty cycle is not working well?

Maybe the duty cycle changes only 50% by adjusting the pin 3 voltage level?

The data sheet includes a hint on push-pull or single ended operation, see page 2: http://focus.ti.com/lit/ds/symlink/tl494.pdf

If you connect pin 13 to ground, the TL494 works in single ended or parallel mode (you can connect the uncomitted output collectors and emitters in parallel to increase current capability within device dissipation limit of course) and if pin 13 is tied to +5V, then comes the push-pull mode, probably with 50-100% duty on one of the output collectors/emitters and 0-50% on the other output. I think this is your problem, I am not really familiar with this IC, prefer using the CMOS version of the 555 family (LMC555C or TLC555C).

rgds, Gyula

[gotoluc]

Hi people.

Im posting this to see if anyone can help me.

The goal is to develop a PWM suitable for resonance experiments, without the need for a signal generator to drive it. I have tested out the below schematic and it works well, and I am simply trying to implement the second schematic into the first, namely the duty cycle control feature.

By combining the two on my bread board I am able to adjust duty cycle from 50% up to 100% on one of the outputs (well both Id imagine I only checked one). But I believe this wont help, what I am trying to do is adjust from 0 - 50% on time. Because the other signal is out of phase I think that anything past 50% on time will overlap the two signals?

Anyway, I picked the TL494 because I have used it before, and I found easy schematics, plus you can control dead time.

Luc, is this something Groundloop could help with? Or is there anyone else out there who has some input?

The goal is a flip flop circuit with frequency and dead time control, what it drives can come later. I am passing each output through a h11d1, to isolate the chip from the high(er) voltages that may be chosen to drive the primary. The bread board version is working well, except for duty cycle time. A different cap, or a choice of a few off pin 5 will allow frequency ranges to be changed.

Perhaps the TL494 is not the best choice, but I have used it before and it seemed to work for some of the Tesla coil builders. 300kHz is supposedly its limit, but some of them have run it at 1 mHz or so they say.

Hoping someone has some input

Hi ren,

I'm happy you are looking into this as I would love to have a circuit that could do all this but 2MHz or more would be nice also.

Anyways, we have to start somewhere. Isn't there a USB base multi channel experiment board that trough a software interface you could control all the switch time, rise time, width and so on?

Groundloop is on holidays for some weeks but if you still need help I can send him an email.

Let me know and thanks for looking into this.

Luc

[ren]

Hi Ren,

My understanding on the TL494 is that its duty cycle can be controlled by changing the DC level on its pin 3 wrt to the negative ground.
Here is link on a motor speed control with TL494, perhaps you have seen it but it does change the DC level at pin 3 with the 1kOhm potmeter (R6 in the link schematic).
PWM Control Speed Motor 12V By TL494 :: Electronic control circuit page 01
Of course it should control in the same way with your 10kOhm pot too. How do you mean your bread boarded circuit for duty cycle is not working well?

Maybe the duty cycle changes only 50% by adjusting the pin 3 voltage level?

The data sheet includes a hint on push-pull or single ended operation, see page 2: http://focus.ti.com/lit/ds/symlink/tl494.pdf

If you connect pin 13 to ground, the TL494 works in single ended or parallel mode (you can connect the uncomitted output collectors and emitters in parallel to increase current capability within device dissipation limit of course) and if pin 13 is tied to +5V, then comes the push-pull mode, probably with 50-100% duty on one of the output collectors/emitters and 0-50% on the other output. I think this is your problem, I am not really familiar with this IC, prefer using the CMOS version of the 555 family (LMC555C or TLC555C).

rgds, Gyula

Many thanks Gyula That cleared up a few things for me. I have looked at the PDF before, but unfortunately got a little lost!

The Motor speed controller schematic is pretty much the same as the one I posted above, originally based on it anyway I think. I pretty much just placed those values of pin 3 onto the first schematic I posted. Worked kinda, but like I said the channel I was scoping would only adjust from 50% on time up to 100% on, where as I would like it to adjust from 0% on to 50% on. I will scope both outputs now and see if it compensates on the other channel, or if I can remedy the situation.

Your pointers are very helpful.

@ Luc, its no problem, something Ive been wanting to do for some time now. Waiting for some parts on my window motor to be made up so got some time on my hands. USB operated one would be great, Im looking at getting a USB scope soon, similar to yours. Slowly compiling up my gadgets.

But ultimately Id llike something that is portable and only requires a DC source to run it. 2MHz plus would be great, maybe some other circuits will come out of the wood work as we experiment, anyway, up to 1MHz for now is fine to start with.

Dont worry bout hassling Groundloop until he gets back from holidays, eventually Id like to get a PCB made up that we could distribute to anyone who is interested.

Thanks again guys

[gotoluc]

But ultimately Id llike something that is portable and only requires a DC source to run it. 2MHz plus would be great, maybe some other circuits will come out of the wood work as we experiment, anyway, up to 1MHz for now is fine to start with.

Thanks again guys

Thanks Ren

Gyula mentioned that the the CMOS version of the 555 family LMC555C or TLC555C is capable of going up to 3Mhz. So I asked him to post a circuit of it that has Duty Cycle adjustment.

Luc

[rave154]

GOTO & co, for a pwm that has duty adjust see the "recycling back emf" thread

[ren]

Well what am I doing stuffing around with the TL? Lol.

Thats sounds great Luc. I managed to get the TL to work pretty good. I can adjust duty cycle from 0-50% now. See scope shots below, second one is pretty blurry, but you can see the reduced on time. With a 9 nF cap on pin 5 it can do up to 120 kHz. Funnily enough I removed the cap altogether and I can get it up to 700kHz, but Im sure this is far from ideal. My 5k pot actually decreases the hertz as it approaches its maximum value. So Im sure there are better ways to configure it.


If Guyla can post a push pull circuit based on the 555 family capable of tuning that would be awesome.


@ Rave, I'll check out the other thread.

Regards.

[gotoluc]

GOTO & co, for a pwm that has duty adjust see the "recycling back emf" thread

Thanks for your post David

the PWM that is needed here has to be able to flip flop and have duty cycle. It controls what is called an H-Bridge circuit. This is what we use to bring coils into Resonance since AC is needed. The H-Bridge circuit flips the polarity of the DC feed at every pulse so it's like two PWM's that when one is High (on) the other is Low (off) and at the next pulse the opposite happens.

Hope this helps a little to understand the difference

Luc

[gotoluc]

Looks good Ren

Have you been able to get them to flip flop? is that why you have two scope traces?

Thanks for the update

Luc

[ren]

Hi Luc,

Cheers buddy. Yeah that is the reason for the two scope signals. Couldnt do it with my old scope, cause it only has one channel. You can see that when one timing pulse is on the other is off. Each pulse is measured at the optical isolator, on the low voltage side. Pretty much just measuring the output from pins 9 and 10 of the TL494. The output transistor can then be Darlingtoned, or drive a mosfet or mosfet driver etc. So this could potentially plug straight into the H-Bridge driver you have shown earlier. Or one could build something similar with PNP and NPN transistors.

Or last but not least one could configure it like the original schematic I posted where the primary is center tapped and each end is alternatively pulsed on and off out of phase from each other.

When I clean it up and get it functioning properly I will try it like this first, with a center tapped primary over a secondary, and try and match resonant frequencies Found another TL494 circuit I will try too that JLN used to drive his MEG replication. Configured in a similar fashion. Will compare the two.

Regards

[gotoluc]

Sounds Great

Thanks Ren

Luc

[ren]

Got the adjustable pulse width sorted, along with frequency up to 220kHz. This is with 3nF cap on pin 5. Will get some pF caps tomorrow and see how fast it can go.

[erfinder]

Hello Ren,

yahoo?

[gyula]

Thanks Ren

Gyula mentioned that the the CMOS version of the 555 family LMC555C or TLC555C is capable of going up to 3Mhz. So I asked him to post a circuit of it that has Duty Cycle adjustment.

Luc

Hi Luc and all,

I have made a schematic on an adjustable duty cycle and output frequency pulse generator built with the CMOS version of the well known 555 timer IC.

The idea comes from an Electronic Design magazin article, see it here:
Printer Friendly Version

and the push-pull drive part of my schematic (which uses two gates from a hex CMOS Schmidt trigger IC) comes from member poynt99 earlier design he uploaded to overunity.com, Simple_Cheap_Low_Power_Oscillators_2.0.pdf see here:
Login

The data sheet of the LMC555CN is here:
www.national.com/ds/LM/LMC555.pdf
and the data sheet for the CD40106BC is here:
http://www.fairchildsemi.com/ds/CD/CD40106BC.pdf

(You can use the TLC555CP as well but its data sheet indicates a typical max operational frequency of 2.1MHz, cf to the typical 3MHz of the LMC555CN both in astabil mode.)

Remarks:
1) Unused pins for the CD40106 should be tied to either the positive supply or gnd, I did not indicate this in my schematic, consult the data sheet if in doubt. See also poynt99's original pdf file too.
You can use any other CMOS inverter or Schmidt trigger instead of the 40106 of course, the two gates from the hex 40106 are needed for signal separation from the timer and inverting the output to get an out-of-phase pulse too for a possible push-pull drive.
2) To get the maximum 3MHz output pulse frequency from the 555, you have to turn potmeter R3 to a short circuit position and adjust R4 towards the 100-200 Ohm position while using a 1nF capacitor for C. It is possible you have some built-in stray capacitance in your circuit board, then use 560pF or 820pF.
3) At pin 7 of the 555 there is an alternate pulse output, you can use this for either measuring the output frequency or drive something else with it, the resistor value between pins 7 and 8 is about 2.2 to 4.7kOhm, if you use this resistor (but you can omit it in case you do need that output) then it increases the total current consumption with a few mA.

I hope this circuit will be useful for the experiments.

rgds, Gyula

EDIT: I notice my schematic was reduced in size after uploading, so this explains the bad quality. Will try to upload in other format soon, sorry for this.
EDIT2 I managed to replace the bad quality with a better one now.

[ren]

many thanks Gyula

[gotoluc]

Hi Gyula,

thanks for taking the time to put this together

Looking at the circuit, if I'm not mistaken it flip flops but is making a negative pulse when it flips, so I think it would need a PNP mosfet on that side, would I not? the mosfet's I want to control are both NPN.

Any suggestions

Thanks for your help and time.

Luc