best microcontroller to produce 13Mhz pwm sine wave pulses?

Go To Last Post
67 posts / 0 new
Author
Message
#1
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

We need to generate a pwm pulses of nearly 13mhz.which microcontroller can generate with our requirement?

Last Edited: Wed. Feb 25, 2015 - 11:04 AM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

No pure sine wave with PWM.

13MHz... maybe one of these due to the clock rate.

 

http://www.pjrc.com/teensy/index...

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I'm a bit confused about the question.

You have added a RF tag, and having something with sharp edges like PWM (which it must have if it switch at that speed).

What is it you want to make (freq, power, resolution)?

 

For me it so far more sounds like a VCO controlled of a PLL (often one chip),will fit the job.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

girish.karnam007@gmail.com wrote:

We need to generate a pwm pulses of nearly 13mhz.which microcontroller can generate with our requirement?

 

A single (wrong?) number is not enough to specify PWM.

If you really meant 13MHz, the next question is what is the desired PWM step size and resolution.

Plenty of Micros can clock their PWM cells at 13MHz, not so many can generate 13MHz period PWM

That's 88ns period, so a step size of 1ns is 88 steps full scale. - if you want 13MHz period PWM, look at parts like the TI Digital Signal Controllers,with sub-ns PWM cells.

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Will the resolution and step size affects the desired sinusoidal pwm output required?

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

What is it you want to do? (is it a PA or a 1vpp clk source or ?)

If you want to make a 13MHz with PWM the PWM clk has to be 100MHz to make a descent sinewave.

 

But if that is what you need take a look at some DDS's I know that it's not PWM but it can make the sinewave.

 

But it's very hard to guess !

 

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Just a test 

It look like there are something out of sync. It looks like a double thread but it's the same thing showed twice!

 

*Yes, you are correct, if not a little late in recognising it. Atmel's Michael has acknowledged the problem and his people are working on a solution. Ross*

 

Last Edited: Tue. Feb 24, 2015 - 10:21 AM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

can u suggest any micro controller with clock frequency more than 72 Mhz and which comes with low cost development board?

What can be the optimum resolution of generating sine PWM?

And also it would be of great help if you can send me general algorithm of generating sinusoidal PWM?

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

girish.karnam007@gmail.com wrote:

can u suggest any micro controller with clock frequency more than 72 Mhz and which comes with low cost development board?

What can be the optimum resolution of generating sine PWM?

And also it would be of great help if you can send me general algorithm of generating sinusoidal PWM?

 

First, a sine wave via PWM will always have significant distortion since it is essentially a digital-to-analog conversion.

Read up on PWM and multi-stage low-bass sharp-cutoff  filters.

 

A better way is to use a microprocessor that haas a good Digital to Analog converter (DAC). But a DAC for 14MHz times a good nyquist rate is very expensive.

So first, decide how sinusoidal the waveform must be. Can it be a stair-stepped quasi-sine-wave - from a less expensive DAC?

 

Some Atmel ARM board might work. Failing that, a low cost XXX Micro board with a 168MHz CPU and a rather fast DAC for a quasi-sine wave might be a simple way.

But this is an Atmel forum so it'd be impolite of me to suggest competitors' boards. There are many, but not to be found in this AVR oriented forum.

 

Last Edited: Wed. Feb 25, 2015 - 07:13 AM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

For 13MHz sine wave (pulse implies a rectangle wave) your pwm carrier would need to be say 10 times that (130MHz) then for 8 bit resolution 256 times that. You can see we are getting numbers way too high for the average microcontroller. There are other potentially easier methods to generate a 13MHz carrier. What do you want to do with the 13MHz? Why do you need pwm? And what is the 'pulse'?

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Our reference frequency is 1.2 MHz-1.3 mhz and the carrier frequency as you said is 10 times.we kept carrier frequency as 13mhz.we need to generate 13 MHz sine pwm.we need rectangular based sine wave.we need pwm for an inverter.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

We need the pwm wave which is a sine based rectangular wave as shown in the following link(only red colour) with frequency 13mhz.that is our carrier wave and we need to generate only that,not the pure sine wave.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Sorry, you're not making much sense. Are you wanting to generate a 1.2 MHz sine wave using a 13MHz carrier? That's only around 3 bits of resolution. Please be more precise.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Yes we exactly wanted what u said just now.we want to generate 1.2 MHz sine wave using 13 MHz carrier.

Last Edited: Wed. Feb 25, 2015 - 08:48 AM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

But that means you will only have 11 samples per wave? What kind of sine wave is that going to be?

 

(answer: a fairly rough looking one!).

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

The sine wave is not of question we only want the sine based pwm.....not the final sine wave... The pwm whose width changes as per amplitude of sine wave....we will be using this sine based pwm to gate mosfets

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

What I'm talking about is the difference between something like this:

http://dlbeer.co.nz/articles/pdm-sine.png

and this

http://upload.wikimedia.org/wikipedia/commons/thumb/8/8e/PWM,_3-level.svg/350px-PWM,_3-level.svg.png

 

As you can see, if you only have a few samples in the period of the wave the resultant wave is "rough". If you have a lot of samples in the period (the "carrier" is much higher frequency than the desired wave) then the wave is much smoother.

 

But, as you say, maybe this does not matter to you?

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

do you need any dead time?

it's still hard to guess when we don't know what you want to make. You talk about a link but I don't see any

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Actually that might make some difference.... But when we simulated it in MATLAB it was fine.....anyways is a carrier frequency of more that 12 MHz possible using any microcontroller??

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

http://dlbeer.co.nz/articles/pdm-sine.png
We need this one....the red one

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

To get something like that you need to be playing something like 64, 100, 128, 200, 256.. samples per cycle. That means your sample play back rate has to be x64, x100, x128.. times faster than the desired output frequency. So if you aim for 1.2MHz you would need to be playing samples at 1.2*64MHz, 1.2*100MHz, 1.2*128MHz, etc.

 

Perhaps you want to look at using an R-2R or a real DAC rather than PWM?

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

It must be highpower 10w or more OP want to output since it needs to done with PWM

And that is why I ask for the needed deadtime.

 

if not I would use a DDS (AD98xx).

Last Edited: Wed. Feb 25, 2015 - 10:31 AM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

So the microcontrolles can't produce that range of frequency??.....and what is the maximum it can produce(Carrier frequency)..... Also let's say we use 12 mhz carrier.....assuming its fine that we end up with a rough sine wave....is it possible to produce this carrier frequency

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Also see the sine wave we will obtain will be rectified later

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I had avoided this thead because the title is talking about 13 milliHertz which is rather a different subject to the 1.2MHz sinewave that you actually want.

 

As Cliff has shown you,   a PWM frequency of 10x means a lumpy output signal.    Mind you,   if you put it through a low-pass filter it will look a lot better.

 

There are specific chips designed for switch-mode power supplies, variable frequency drives and radio receivers/transmitters.

 

1.2MHz sounds rather high for an Inverter.    But it would be very low for an RF oscillator.    What exactly do you want to do with this signal?

 

David.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

We want to give the pwm signals for gating the MOSFETs in the inverter and then we want to give this signal to an primary air cored coil(wireless transmission).from there transmission to secondary coil.then given to a rectifier finally.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

The examples you show are pdm (pulse density modulation) but you say you want pwm (pulse width modulation). As has been mentioned, you probably need a clock frequency of 1.2MHz times 256 times 10 and be able to update the pwm at 1.2MHz times 256. You'll probably need to implement it in a fpga or do it the old way with a sinewave and comparators.
I'd probably suggest for an inverter you use a quasi resonant technology for thosr frequencies rather than pwm drive.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

The Q of your air cored inductor will be fairly high.    If the resonant circuit (LC)  is tuned to 1.2MHz,   if will act like a very effective low-pass filter.

 

A regular class-C RF output stage is only driven by a short pulse.

 

I am sure that there are several Hams that can explain RF transmitters.     And App Notes that can explain wireless power transmission.    

I have no experience.

 

David.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I don't hop you live in EU or US .

If will kill all AM station miles away.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

which kind of modulation ?

I guess a square wave of 1.2 MHz or 1/3 or 1/5 .....  will do the job if your coil resonate at 1.2-1.3 MHz

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I think the OP has his wires crossed (pardon the pun)

 

Generating 13MHz PWM is one thing, the OP doesn't mention resolution at all and isn't telling us anything about what he is doing, it seems like he is purposely holding this back

 

Someone who wants 13MHz PWM is probably doing something advanced but not mentioning resolution is the sign of a n00b, think about it, you could have a fast a cycle as you wanted but the output will still be pants if you have limited actually values you can use, I have discussed this on here before, the more values you have i.e resolution then the better the harmonic current distortion, despite the fact that the voltage distortion will always be about the same!

 

Mentioning PWM and sine reference in the same sentence to says power electronics/drives/converters whatever, please tell me which device you are expecting to drive at 13MHz?, does the OP realise how hard it is to do proper HF electronics, I mean the tiniest parasitic will wreak havoc, its not fun anymore IMO

 

So no mention of resolution, wanting to do 13MHz sine reference PWM with no talk of a device to switch or any talk of hardware at all

 

And it will be rectified again later!

 

I think the OP needs to let us know a little bit more about what he is doing

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I suspect that he is doing power transfer.     A grown up version of my electric toothbrush.     That has no electrical contacts.     You just place it on the special plastic 'post' and the batteries get recharged.

 

I would guess that there will always be a trade-off between frequency, size of inductor,  switching losses.

 

David.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

girish.karnam007@gmail.com wrote:

We want to give the pwm signals for gating the MOSFETs in the inverter and then we want to give this signal to an primary air cored coil(wireless transmission).from there transmission to secondary coil.then given to a rectifier finally.

 

Still sounds like your best bet is a bottom end TI Digital Signal Controller (sub ns PWM - eg http://www.ti.com/product/tms320... ), or you might get by with a Microchip dsPIC part (DSPIC30F2011)

I don't think Atmel has anything with ns region PWM control, but other companies do.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

As David mentions, a RF output stage and you could modulate the power by doing pwm on the DC supply to the output stage. It's all about tuned circuits - just like ringing a bell.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

let us say if we reduce our carrier frequency to 6 Mhz and also reduce reference frequency correspondingly,then which micro controller we can use for generating sine PWM which is preferably low cost?

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Suitable simple MCUs, with lower performance and price, could be Infineon XMC1000 series, (5V, so could drive MOSFETS) or Atmel

ATxMega8E5 - 3v3 only, but has High resolution extension to PWM, that claims 4ns LSB

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

You don't seem to be grasping what we're telling you.
First up, is the challenge to generate a 1.2MHz sinewave or to use pwm?

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

sorry for the wrong pic here is what we need(pic attached). Also, we are flexible with using a carrier frequency of 5~6 MHz

 

Attachment(s): 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

OP needs more electronics 101 skills before undertaking this.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

probably....but if we can get back to discussion.....

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

probably....but if we can get back to discussion.....

What discussion?

 

You have totally ignored what people have said and if you can't be bothered to read whats been written ...................

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Im not ignoring anything...sorry if u felt that way....as far as resolution is considered we require a minimum resolution of 4 bits...now hardware as far as I have researched IRF820 should be able to handle that range of frequency...But if it does not we will be reducing the carrier frequency...but as of now we are concentrating to genearate the PWM pulses. If you have any other queries pls ask.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Why do you need to use pwm?

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

to gate the mosfets(full wave inverter bridge)

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

You could do the same with two mosfets and a centre tapped coil. The mosfets would be switching at a lower frequency. No pwm needed.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Lets do some basic numbers since you are ignoring basic advice.
According to your picture, your carrier is 20 times your wanted frequency. So that means your cpu has to look up a sine table and update the pwm. So at 1.2MHz, youll be looking at a update rate of 24million times a second. Lets guess at 20 machine cycles per update so that gives us 480MHz. Then the pwm clock rate is 24 million times 2^no of bits equals 24 times 16 = 148+ 240=388MHz. Maybe a Analog Devices blackfin.
Vs a couple of mosfets and a handful of discretes to make a multivibrator driving a tuned circuit. How many Watts of RF do you want?

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

But you have totally ignored my post, tell me what power device you can switch at MHz?

You will find there isn't one so what you propose isn't even possible, if there was a MHz device you wouldn't be switching it with a HVIC driver you are seriously Missing the point here your inexperience means you can't see the problem, it must be bliss!

I won't even go into how bad a waveform with 16 values will look, it would be a waste of time

Last Edited: Thu. Feb 26, 2015 - 07:40 AM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

See our project demands  the output across the inverter to be high(preferably in MHz range).Also how can we gate the mosfets using a centre tapped coil?? Pls explain. Will we be able get a sine wave through this process after inverting.??

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Again, you are ignoring what we are telling you.
How many Watts of power?
If you have basic knowlege of inverters you should know what a doubled ended configuration is. Most of the cheapy uninterruptable power supplies use this method.
I mentioned a tuned circuit - more specifically a parallel tuned circuit. You are wanting to build a RF transmitter. Marconi did it with stoneage technology but he had some knowlege.
I've given you a number of keywords to Google. Do some research.
Do you know what a multivibrator is?
Tuned circuit?

Last Edited: Thu. Feb 26, 2015 - 07:50 AM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I did not quite understand what you said on the cycles and update rate as I am very new to that. How ever I have understood that I require a device that operates at very high clock frequency. So according to you what should be the carrier frequency at which we will get a decent output.And RF watts. I think as we havnt reached that part of design yet, 3~5 Watts.

Last Edited: Thu. Feb 26, 2015 - 07:55 AM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Regarding cycles, i'm referring to the number of machine cycles of a microprocessor. What i'm trying to impart upon you is that the basic numbers are telling us we're on the bleeding edge. I'm also trying to tell you there are other methods of generating RF power without resorting to microprocessors or pwm at very high frequencies. Using pwm works well for generating a 50Hz sine wave. Also making a resonant circuit at 50Hz is a bit more difficult due to the values of inductors and capacitors. At 1.2Mhz things swing the other way. As David mentioned early on in this thread that you could use a Class C amplifier. Did you research that?
Anyway, one would think when designing an inverter one of the first questions is how much power. Luckily for us 3-5W is manageable without resorting to high voltages or currents. This means thing are less likely to explode when testing - things will generally get real hot then melt. This gives you time to turn off the power.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I try again since you as other have said you ignore the important questions.

 

So I try again

what kind of dead time do you need? (if any depend on your hardware).

 

What kind of modulation ?(Is it AM FM or not (same freq all the time) ) 

 

And to make it clear a $0.5 AVR can make what you need if no modulation and dead time from the modulator.

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Our voltage and current ratings are as follows: 5v and 0.5 amps.plz see tthat our operating current will be very much within the limit.can I email u some questions as those questions cannot be promoted in avr forum.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

We can help you but we're not going to do the work for you. You want 3-5W but only have 2.5W available, you have a problem there.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Sorry if u feel that I am ignoring the posts....didn't see few of them....here is what we want to achieve and I think it is AM....we need a triangular carrier wave and a modulating sine wave.....whenever the magnitude of sine wave is greater that carrier wave....there should be a high pulse.....I hope this clears......about dead time I will get back to you on that...as i don't know about it.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Sorry but 3~5 W....is what we would be transferring....not our required wattage....also we require less than 2w for our application..

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

And yes frequency at all time will be the same...

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Your comments make no sense. How can you get more energy out than you put in?
Did you look at the link i posted? Where did the triangle wave come from?

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Is something like this remotely connected to what you want to achieve?
http://www.ti.com/product/bq500414q

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

input wattage: 5 Watts

output we require at least 2.5 Watts

sorry for that miscommunication

and yes I am going through your posts

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

The source used is 12 v in your referred link .....Our project does not use this range of voltage.. 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Please describe what your actual project is.

 

I made a 'guess' that it was a contactless battery charger.   e.g. like my electric toothbrush.

Am I correct?

 

Of course it might be a student exercise with Proteus.     Somehow,   the original confusion between milliHertz and MegaHertz points in this direction.

 

I don't have experience with RF or power circuits.    But there are several members that do have practical knowledge.     You will get a lot more help if you describe the real purpose of your questions.

 

David.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

#62
Posted by girish.karnam00...: Thu. Feb 26, 2015 - 08:16 PM (Reply to #60)
The source used is 12 v in your referred link .....Our project does not use this range of voltage..

Errr, the concept is the same.changing the design to 5V is trivial. Sounds like you want us to design the solution for you.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

No I dont want you to design the solution....But in the link you provided what is the input....is it the DC source??....what is being amplified....i mean which signal??

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Alternative approaches would be to create the PWM sine with exactly the images posted : Use a Sawtooth and Analog Sine.

Some MHz region SMPS parts may manage this, with a DDS sine wave .

Or, you can to both a Sine LUT Table and PWM in a CPLD.

Modulation depth has not been mentioned ....

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Thanks for the reply. What range of frequencies can CPLD handle??