Post by thread:[PIC]: sinusoidal signal generation from PIC

I need to generate a pour shape (10%...20% distorsions?) sinusoidal signal with fixed amplitude and variable output frequency from 1 to 20 KHz. AFIK there are two basic outputs from PIC: PWM and R2R ladder(or DAC). About the sinus generation I've heard about Goertzel algorithm, table read and other methodes (Taylor series etc). There are also a few hardware methodes, but are even too expensive for me (direct digital syntesis chips, wien-bridge digital potentiometer controled, 8038 clones) or too complicated for the whole required frequency range (square wave, second order filtering). If you have any experience in this field I'm interested into a good suggestion for a pure and simple software sinusoidal signal generation with pic. thank you in advance, Vasile http://surducan.netfirms.com -- http://www.piclist.com hint: The PICList is archived three different ways. See http://www.piclist.com/#archives for details.

I had to design a SINE-WAVE generator with fixed, precise frequencies and low distortion for a test system. I used an 8-bit counter, an EPROM, and a D/A converter. Simply drove the counter at the right frequency from the processor. Needed very little filtering (R-C only). But I think that these new flash PICs could do it with tables if the frequency is that low. --Bob At 01:44 PM 8/7/2003 +0300, you wrote: {Quote hidden}>I need to generate a pour shape (10%...20% distorsions?) >sinusoidal signal with fixed amplitude and variable output frequency >from 1 to 20 KHz. AFIK there are two basic outputs from PIC: >PWM and R2R ladder(or DAC). About the sinus generation I've heard about >Goertzel algorithm, table read and other methodes (Taylor series etc). >There are also a few hardware methodes, but are even too expensive >for me (direct digital syntesis chips, wien-bridge digital potentiometer >controled, 8038 clones) or too complicated for the whole required >frequency range (square wave, second order filtering). >If you have any experience in this field I'm interested into a good >suggestion for a pure and simple software sinusoidal signal generation >with pic. > >thank you in advance, >Vasile >http://surducan.netfirms.com > >-- >http://www.piclist.com hint: The PICList is archived three different >ways. See http://www.piclist.com/#archives for details. --------------- NOTICE 1. This account can accept email & attachments up to 10M in size. 2. Federal Monitors: At request of client, some attachments are encrypted. Please DO NOT delay traffic; please reply with credentials for password. -------------- -- http://www.piclist.com hint: The PICList is archived three different ways. See http://www.piclist.com/#archives for details.

Hi Non PIC: Try the XR2206 from Exar(.com) PIC: Read MC App note: AN655 Quentin -- http://www.piclist.com hint: The PICList is archived three different ways. See http://www.piclist.com/#archives for details.

Have a look at the electronics page on my website at ... http://homepage.ntlworld.com/moonshadow/New_Folder/electronics.htm It shows how to get sine from PWM and square wave with just two components. A 4 MHZ 16F877 can generate 2R2 waveforms up to about 87 kHz if you open the software loop. It is also described on the web site. If you generate the wave points with a loop the maximum frequency drops to about 7 kHz. Can I ask what you are using it for ?. John Kent {Original Message removed}

> I need to generate a pour shape (10%...20% distorsions?) > sinusoidal signal with fixed amplitude and variable output frequency > from 1 to 20 KHz. AFIK there are two basic outputs from PIC: > PWM and R2R ladder(or DAC). About the sinus generation I've heard about > Goertzel algorithm, table read and other methodes (Taylor series etc). > There are also a few hardware methodes, but are even too expensive > for me (direct digital syntesis chips, wien-bridge digital potentiometer > controled, 8038 clones) or too complicated for the whole required > frequency range (square wave, second order filtering). > If you have any experience in this field I'm interested into a good > suggestion for a pure and simple software sinusoidal signal generation > with pic. I had to do something similar once where the frequency was low but needed to be very accurately controlled. One complete sine wave was stored in a 512 entry table of 8 bit values. The CCP module in PWM mode was used at the fastest possible speed that still yielded 8 bit resolution. A new value was loaded every CCP interrupt (every 64 cycles), then the value computed for the next interrupt. Frequency was adjusted by setting the table increment per CCP interrupt. To get high frequency resolution and accuracy, this value had 16 fraction bits. You don't need to do interpolation between table values at that table configuration, so the table index was just the 9 integer bits of the 9.16 table index accumulator. The fraction bits accumulated the fractional table entry offset, which occasionally caused a carry into the integer part as the accumulated fractional error exceeded 1. With a little more addressing math you can use a 1/4 cycle table. I had the room, so I used a full cycle to make the code easier. With a 20MHz PIC, you can get 8 bit PWM resolution at 78KHz PWM frequency. That's about 4 times your highest output frequency, so the output filter will need to be more than a couple or Rs and Cs, depending on how much amplitude variation due to frequency you can live with at the high end of the frequency range. A PIC18 at 40MHz gives you an additional factor of 2 headroom between your highest frequency and the PWM frequency. One problem with this method is that at high frequencies when the table increment is a substantial portion of a cycle, you will get significant distortion that will look like amplitude modulation at the beat frequency. In theory, the output filter can get rid of this, but in practise that will be nearly impossible. Another way of looking at it is that the beat frequency is always there, but the intermodulation products get stronger and move closer to the desired frequency as the desired frequency approaches half the PWM frequency. ***************************************************************** Embed Inc, embedded system specialists in Littleton Massachusetts (978) 742-9014, http://www.embedinc.com -- http://www.piclist.com hint: The PICList is archived three different ways. See http://www.piclist.com/#archives for details.

> I need to generate a pour shape (10%...20% distorsions?) > sinusoidal signal with fixed amplitude and variable output frequency > from 1 to 20 KHz. Check the magical sine waves from Don Lancaster, use the older (fixed-interval) version rather than the newer version. Wouter van Ooijen -- ------------------------------------------- Van Ooijen Technische Informatica: http://www.voti.nl consultancy, development, PICmicro products -- http://www.piclist.com hint: The PICList is archived three different ways. See http://www.piclist.com/#archives for details.

It's been awhile since I read any of it, but Don Lancaster did a huge amount of sinewave generation research a few years back. Try here: http://www.tinaja.com/magsn01.asp Good reading, weather you can use it or not. Denny {Original Message removed}

> I need to generate a pour shape (10%...20% distorsions?) > sinusoidal signal with fixed amplitude and variable output frequency > from 1 to 20 KHz. I am working on a programmable signal generator that will produce audio frequency with 1Hz tuning steps. The idea is to simulate a NCO (numerical controlled oscillator) in software. Here are some links on similar projects: http://www.chipcenter.com/embapps/emba057.htm AVR (not PIC): http://www.myplace.nu/avr/minidds/ www.geocities.com/leon_heller/minidds.html http://www.qsl.net/zl1bpu/micro/SIGGEN/Siggen.htm Hope this helps, Adi -- http://www.piclist.com hint: The PICList is archived three different ways. See http://www.piclist.com/#archives for details.

DDS using only a PIC & 8-bit DAC works great for me. I use a 1024-entry full-cycle lookup table. What are the range of frequencies you are looking at? Depending on that, you might also be able to build an internal triangle/ramp for a 1-bit DAC in the PIC, and then use active lowpasses to get your analog waveform. Cheers, Ishaan {Original Message removed}

Vasile Surducan wrote: ... > Do you know more ? Did anybody mention PWM calculated "on-fly" (both period and duty cycle) according to A/D input? Mike. -- http://www.piclist.com#nomail Going offline? Don't AutoReply us! email spam_OUTlistservTakeThisOuTmitvma.mit.edu with SET PICList DIGEST in the body