Post by thread:[EE]: Using PICs for simple audio tasks

I'm interested in sound generation right at the moment. The specific task is building a metronome that has a decent sounding, high volume "TOCK" sound. But I figure since there are quite a few issues in terms of generating audio, that I may as well examine them for this project. My thoughts: 1) My goal is to use some form of PWM for output. Presuming that I want to generate 4000 8 bit samples/sec, would using each sample as the duty cycle be sufficient. By this I mean if I have three successive samples of 64,128, and 192, that for the first PWM cycle I drive the output high for 62 uS, then low for 187 uS, a hi/low cycle of 125/125 uS, then a cycle of 187/62 uS? 2) Is it possible to configure the hardware PWM so that it changes the duty cycle each cycle? 3) Where would I start in terms of the values for a low pass filter for the PWM output? 4) I want to drive a cheap 8 ohm speaker. I've seen a couple of push/pull transistor amplifiers, but it seemed they were digital in nature. Since the output of the low pass filter is an analog value should I be looking at a transistor amp? op-amp? LM386? 5) And finally the metronome question: Any pointers on how to generate a quality TOCK sound? Thanks for any help you can offer. BAJ -- http://www.piclist.com hint: The PICList is archived three different ways. See http://www.piclist.com/#archives for details.

BAJ, It is possible to take a raw .wav file and playing it back by simply writing successive values straight to the PWM register at the sample frequency. When I did this (it was a Philips 80c552 chip, but same principles apply), I just used an RC filter and then AC coupled it into an audio amp. I ignored the header info at the start of the .wav file and played it like it was audio data, the result was just a slight click at the start of the sound. I set the PWM frequency as high as it would go. When not producing a sound set the PWM output to 50% (2.5V), as the data in the wave file is also biased at "half way". On my second attempt I used the PWM signal to directly control a MOSFET H bridge driving a loudspeaker. This was quite a bit louder (which is what I wanted). Hope this gives you some ideas to experiment with. Brent Brown Electronic Design Solutions 16 English Street Hamilton, New Zealand Ph/fax: +64 7 849 0069 Mobile/text: 025 334 069 eMail: spam_OUTbrent.brownTakeThisOuTclear.net.nz -- http://www.piclist.com hint: The PICList is archived three different ways. See http://www.piclist.com/#archives for details.

On Wed, Oct 17, 2001 at 08:46:29AM +1300, Brent Brown wrote: > BAJ, > > It is possible to take a raw .wav file and playing it back by simply > writing successive values straight to the PWM register at the > sample frequency. That's why I love this list. A quick, great, easy to implement suggestion. > > When I did this (it was a Philips 80c552 chip, but same principles > apply), I just used an RC filter and then AC coupled it into an audio > amp. I ignored the header info at the start of the .wav file and > played it like it was audio data, the result was just a slight click at > the start of the sound. I set the PWM frequency as high as it would > go. When not producing a sound set the PWM output to 50% (2.5V), > as the data in the wave file is also biased at "half way". Understood. After a quick read of the Midrange manual, I realize that the highest speed I'll be able to run the PWM is 19.5Khz with a 20 Mhz Fosc. The reason is that you start to lose resolution as you crank up the frequency. Since the PWM is 10 bits and I'm not real thrilled at having to fight with splitting the sample to store the lower two bits in the config register. So I plan to zero out the lower two bits and store the sample in the upper 8 bits. > > On my second attempt I used the PWM signal to directly control a > MOSFET H bridge driving a loudspeaker. This was quite a bit louder > (which is what I wanted). I think I'm going to try an LM386 for starters. The circuit here: http://www.reprise.com/host/circuits/lm386.asp Seems simple enough. Thanks for the info. BAJ -- http://www.piclist.com hint: The PICList is archived three different ways. See http://www.piclist.com/#archives for details.

----- Original Message ----- From: Byron A Jeff <.....byronKILLspam@spam@CC.GATECH.EDU> > Since the PWM is 10 bits and I'm not real thrilled at having to fight with > splitting the sample to store the lower two bits in the config register. So > I plan to zero out the lower two bits and store the sample in the upper 8 bits. Can anyone explain why uC people decided to put those last two bits in an odd place? Even register banking would be simpler than stuffing the extra bits where they belong in the current implementation. Register banking is how they chose to address the last two bits in A to D so it looks inconsistent from that angle. Jeff -- http://www.piclist.com hint: The PICList is archived three different ways. See http://www.piclist.com/#archives for details.

Byron A Jeff wrote: > > I'm interested in sound generation right at the moment. The specific task is > building a metronome that has a decent sounding, high volume "TOCK" sound. > But I figure since there are quite a few issues in terms of generating audio, > 5) And finally the metronome question: Any pointers on how to generate a > quality TOCK sound? Ever attached a 555 timer with a square wave output to a speaker?? You get a real good tock sound. :o) If you output a single square wave pulse, driven by one transistor, to the speaker through a 0.1uF capacitor you will get a decent loud tock. Then you can simply adjust the length of the pulse and this will adjust the volume of the tock, over a pretty good range too if I remember right. -Roman -- http://www.piclist.com#nomail Going offline? Don't AutoReply us! email listservKILLspammitvma.mit.edu with SET PICList DIGEST in the body

> 1) My goal is to use some form of PWM for output. Presuming that I want to > generate 4000 8 bit samples/sec, would using each sample as the duty cycle > be sufficient. By this I mean if I have three successive samples of 64,128, > and 192, that for the first PWM cycle I drive the output high for 62 uS, > then low for 187 uS, a hi/low cycle of 125/125 uS, then a cycle of 187/62 uS? I haven't checked your numbers, but that's what PWM is all about. For PIC PWM the period remains constant and the duty cycle is changed. You low pass filter the resulting signal to recover the "average", which is the audio signal. It's a cheap man's D/A. It works fine as long as the PWM frequency is several times the highest frequency you want. This allows it to be filtered out easily. I've done this several times. For example code, see my HAL PIC project at http://www.embedinc.com/pic/hal.htm. The PWM output is updated in the interrupt routine in the HAL_INTR.ASPIC module. > 2) Is it possible to configure the hardware PWM so that it changes the duty > cycle each cycle? Yes, you can have the CCP interrupt every PWM period. This allows you the opportunity to update the PWM output value. > 3) Where would I start in terms of the values for a low pass filter for the > PWM output? Run the PWM as fast as you can to still get the desired output resolution. Hopefully this is several times higher than the highest desired audio output frequency. Usually a 2 or 3 pole RC filter just above your output frequency limit is all you need. This isn't rocket science - just do the math. > 4) I want to drive a cheap 8 ohm speaker. I've seen a couple of push/pull > transistor amplifiers, but it seemed they were digital in nature. Since the > output of the low pass filter is an analog value should I be looking at > a transistor amp? op-amp? LM386? I use an op amp driving 4 bipolar transistors in the HAL project. This works very well for low quality sound. > 5) And finally the metronome question: Any pointers on how to generate a > quality TOCK sound? This is usually done with a few cycles of a low frequency wave. You need to experiment. Try something like 3 cycles of 50Hz, but there are lots of choices for different effects. Simulate it on a PC with a sound card. ******************************************************************** Olin Lathrop, embedded systems consultant in Littleton Massachusetts (978) 742-9014, .....olinKILLspam.....embedinc.com, http://www.embedinc.com -- http://www.piclist.com#nomail Going offline? Don't AutoReply us! email EraseMElistservspam_OUTTakeThisOuTmitvma.mit.edu with SET PICList DIGEST in the body

> When I did this (it was a Philips 80c552 chip, but same principles > apply), I just used an RC filter and then AC coupled it into an audio > amp. I ignored the header info at the start of the .wav file and > played it like it was audio data, the result was just a slight click at > the start of the sound. I set the PWM frequency as high as it would > go. When not producing a sound set the PWM output to 50% (2.5V), > as the data in the wave file is also biased at "half way". Note that this WAV file hack only works if the WAV data is 8 bit monophonic. In general this is a bad idea - don't try this at home. Reading a WAV file properly is really quite easy. ******************************************************************** Olin Lathrop, embedded systems consultant in Littleton Massachusetts (978) 742-9014, olinspam_OUTembedinc.com, http://www.embedinc.com -- http://www.piclist.com#nomail Going offline? Don't AutoReply us! email @spam@listservKILLspammitvma.mit.edu with SET PICList DIGEST in the body

On Wed, 17 Oct 2001 12:24:00 -0400, Philip Galanter wrote: >I think this is the case, but just to be clear... > >To play sampled sound using the PWM technique, you just do a simple >linear mapping of the waveshape amplitudes (the samples) into the >range 0-255 for the pulse width. > >Right? As long as your integrating capacitor is fed by a constant current source and you use the PWM output only to discharge the cap and not source current into it, but rather, let the constant current source charge it. Should be nice and linear that way. There may be simpler ways to do this though... Dan > >thanks, Philip > >-- >http://www.piclist.com#nomail Going offline? Don't AutoReply us! >email RemoveMElistservTakeThisOuTmitvma.mit.edu with SET PICList DIGEST in the body > > > -- http://www.piclist.com#nomail Going offline? Don't AutoReply us! email spamBeGonelistservspamBeGonemitvma.mit.edu with SET PICList DIGEST in the body

> > When I did this (it was a Philips 80c552 chip, but same principles > > apply), I just used an RC filter and then AC coupled it into an > > audio amp. I ignored the header info at the start of the .wav file > > and played it like it was audio data, the result was just a slight > > click at the start of the sound. I set the PWM frequency as high as > > it would go. When not producing a sound set the PWM output to 50% > > (2.5V), as the data in the wave file is also biased at "half way". > > Note that this WAV file hack only works if the WAV data is 8 bit > monophonic. In general this is a bad idea - don't try this at home. > Reading a WAV file properly is really quite easy. Yes, 8 bits mono. No, it's a good idea - do try it at home. It is rough and ready but the point was it's really easy to try and then you can develop it further. Like you say it shouldn't be too hard to use the PIC to decode the wave file header. Brent Brown Electronic Design Solutions 16 English Street Hamilton, New Zealand Ph/fax: +64 7 849 0069 Mobile/text: 025 334 069 eMail: TakeThisOuTbrent.brownEraseMEspam_OUTclear.net.nz -- http://www.piclist.com#nomail Going offline? Don't AutoReply us! email RemoveMElistservTakeThisOuTmitvma.mit.edu with SET PICList DIGEST in the body

> Yes, 8 bits mono. No, it's a good idea - do try it at home. It is rough > and ready but the point was it's really easy to try and then you can > develop it further. Like you say it shouldn't be too hard to use the > PIC to decode the wave file header. I meant have the host properly interpret the WAV file, then send the converted content to the PIC. WAV files are RIFF files, which are pretty straight forward to interpret. Write a routine once that reads them and you can use it for all future projects. Just dumping raw data may work under the right conditions, but will come back and bite you sooner or later. ******************************************************************** Olin Lathrop, embedded systems consultant in Littleton Massachusetts (978) 742-9014, olinEraseME.....embedinc.com, http://www.embedinc.com -- http://www.piclist.com#nomail Going offline? Don't AutoReply us! email EraseMElistservmitvma.mit.edu with SET PICList DIGEST in the body

On Wed, Oct 17, 2001 at 06:03:58PM +1000, Roman Black wrote: > Byron A Jeff wrote: > > > > I'm interested in sound generation right at the moment. The specific task is > > building a metronome that has a decent sounding, high volume "TOCK" sound. > > But I figure since there are quite a few issues in terms of generating audio, > > > 5) And finally the metronome question: Any pointers on how to generate a > > quality TOCK sound? > > > Ever attached a 555 timer with a square wave > output to a speaker?? You get a real good tock > sound. :o) I took 2 minutes to test dumping a cap into a small speaker. I wasn't real impressed. > > If you output a single square wave pulse, driven > by one transistor, to the speaker through a 0.1uF > capacitor you will get a decent loud tock. Then you > can simply adjust the length of the pulse and this > will adjust the volume of the tock, over a pretty > good range too if I remember right. I'll test it. Thanks, BAJ -- http://www.piclist.com#nomail Going offline? Don't AutoReply us! email RemoveMElistservEraseMEEraseMEmitvma.mit.edu with SET PICList DIGEST in the body

At 07:01 PM 10/17/01 -0400, Byron A Jeff wrote: {Quote hidden}>On Wed, Oct 17, 2001 at 06:03:58PM +1000, Roman Black wrote: > > Byron A Jeff wrote: > > > > > > I'm interested in sound generation right at the moment. The specific > task is > > > building a metronome that has a decent sounding, high volume "TOCK" > sound. > > > But I figure since there are quite a few issues in terms of > generating audio, > > > > > 5) And finally the metronome question: Any pointers on how to generate a > > > quality TOCK sound? > > > > > > Ever attached a 555 timer with a square wave > > output to a speaker?? You get a real good tock > > sound. :o) > >I took 2 minutes to test dumping a cap into a small speaker. I wasn't real >impressed. One of my early pyro projects were 'bang simulators' used for doing rehearsals with actors and such: they used little 4" speakers in a plastic case. The actual bang is created by discharging a 10 uF cap charged to ~50 Vdc (six 9V batteries in series) into the speaker. The discharge mechanism is a MOSFET driven by a monostable timer. Its loud but doesn't sound anything like a metronome. dwayne Dwayne Reid <RemoveMEdwaynerspam_OUTKILLspamplanet.eon.net> Trinity Electronics Systems Ltd Edmonton, AB, CANADA (780) 489-3199 voice (780) 487-6397 fax Celebrating 17 years of Engineering Innovation (1984 - 2001) * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * Do NOT send unsolicited commercial email to this email address. This message neither grants consent to receive unsolicited commercial email nor is intended to solicit commercial email. -- http://www.piclist.com hint: The list server can filter out subtopics (like ads or off topics) for you. See http://www.piclist.com/#topics