digital guitar tuner, avr based

[didier]

hi, while searching i found this link:
http://www.myplace.nu/avr/gtuner/

does anyone here have some experience with programming avr?
and maybe in particular this little project?

[nelson]

I would like to know too.

Would be great if someone had the necessary hardware and could program a few chips and sell them off at a wee bit of profit. 

[The Tone God]

Took a quick look at it. The project looks like it was created a few years ago so its showing it's age.

Firstly it is using a slightly out of date / mature/ EOL'd uC. You can still get that uC or get a uC that is many times more powerful/featureful for more then half the cost. It would be better suited to a newer processor. Not that big of a deal as the software is fairly portable.

Secondly speaking of software the code will not compile using today's compiler. It is using some depricited functions. Not a big deal once again really as the code is pretty simple. It could be re-code fairly easily/quickly.

Thirdly, by doing some math and tweaking the software the circuit can be simplified to not require the crystal making for easier construction.

Andrew

[MR COFFEE]

Anything LESS than a crystal-controlled timebase for any guitar tuner will be an unmitigated disaster that will drift with temperature mercilessly and drive any musician with half an ear within range friggin' nuts...

The rest of the comments are right on target. Guitar tuner software isn't that hard to write and the problems with tuners are usually related to cutting corners in the analog aspects of the circuit design (AGC, Filtering, cheesy vibration pickups, etc.)

I agree - it could be a nice idea for a diy project!

[nelson]

So who wants to load the code into some chips and sell them off at a reasonable price?

[The Tone God]

Anything LESS than a crystal-controlled timebase for any guitar tuner will be an unmitigated disaster that will drift with temperature mercilessly and drive any musician with half an ear within range friggin' nuts...

For some reason my edit after my post didn't stick concerning this. The processors can be calibrated and save the average DIY the need of finding the right crystal. On the other hand a crystal would make the DIYers life easier by not requiring the calibration and possible recalibration but it still suffers some stability problems although far less then the internal clock. There are other clock sources that are stable and temp compensated that would be alittle more appropriate IMHO but once again these are specialty parts for the average DIYer.

So who wants to load the code into some chips and sell them off at a reasonable price?

Hmm...I would be willing to update the code and post the required files for programing including a precompiled hex. I would also then be forced to publish the updated code as per GPL, I would have prefered a BSD licence but oh well, for others to use. I'm not really interested in shipping pre-loaded processors unless someone like Smallbear were interested in a batch to sell to the community.

I actually did some work on my own tuner design that had a line of LEDs that swung back and forth. I also did some work on one with a analog output that could be used to drive the brightness of an light or a VU meter. Something more natural for the user. Oh then there was my take on a strobe tuner. Ah so little time.

Andrew

[Transmogrifox]

I just ordered a bunch of crystals from Mouser and it was no more of a pain than ordering resistors--and for using them, I don't see that being any more difficult than soldering on a capacitor.  I would agree with any reply that points out the fact that most DIY stompboxaholics don't have a nice supply of crystals hanging around.  Popping a capacitor and resistor of any given value out of the drawer is much more likely.  I wouldn't be very convinced that the calibration thing would be acceptably stable for a guitar tuner.

[Peter Snowberg]

Don't count on a barely temperature compensated RC oscillator which was "calibrated" to an uncalibrated source, to be anywhere near good enough for tuning an instrument. A good crystal is the only way to go there.

I would skip the DIY tuner concept and just buy a StroboPick. It's my all time favorite tuner. It's cheap, it's got unique advantages, and it's pretty cool to use for creating a "drone" with a delay pedal.

If anybody wants to do any updating, may I suggest the ATtiny13 as a target?

[The Tone God]

Don't count on a barely temperature compensated RC oscillator which was "calibrated" to an uncalibrated source, to be anywhere near good enough for tuning an instrument. A good crystal is the only way to go there.

I was only suggesting the internal clock calibration, which can be accurate in short term if the person does it correctly, for the bedroom plunker or someone who wanted to play around with this. I would expect that if someone were to sell preprogrammed IC that they would included the clock source with the uC.

I would skip the DIY tuner concept and just buy a StroboPick. It's my all time favorite tuner. It's cheap, it's got unique advantages, and it's pretty cool to use for creating a "drone" with a delay pedal.

I played with as strobe tuner. I might offer that to some distributer but I belive the maker of the StroboPick doesn't mind selling kits too with the board an all so DIYers might want to look in that. Although my strobe tuner work in a slightly different way.

If anybody wants to do any updating, may I suggest the ATtiny13 as a target?

That was the plan if this were to go through.

Andrew

[MetalGuy]

Several months ago I was looking for one too. I found couple of old unmaintained MCU based projects and even tried one of them without any success.
I'll be glad to find a DIY project and/or somebody offering preprogrammed and tested MCUs for it. Some may ask why build a tuner if you can get one for 10 bucks or so.
I would  build them into some of my amplifiers and pedals for example. Unfortunately the 10 buck ones are not very suitable for this.

[BN]

Well, found out about this http://www.elecdesign.com/Articles/ArticleID/6264/6264.html just the other day. If my memory doesn't descieve me it uses the same principle as the projet mentioned above even though slightly more sophisticated. Hopefully I will give it a try rahter soon, but replacing the Motorola MCU with an ATMega8. If I do I'll let you know.

Here's the schematic: http://www.elecdesign.com/Files/29/6264/Figure_01.gif 

[The Tone God]

It is basically the same circuit operation only a more sophisticated analog section that gives a cleaner pulse. To be honest Jesper's audio circuit is fine. I re-wrote Jesper's code from scratch awhile back to update it as well as port it to the Tiny13. It work well enough I thought. For DIYers is it alot more simpler.

Andrew

[BN]

To be honest Jesper's audio circuit is fine. I re-wrote Jesper's code from scratch awhile back to update it as well as port it to the Tiny13. It work well enough I thought. For DIYers is it alot more simpler.

So you don't think it's worth the extra effort? That suites me, I'm rather lazy... 

I was thinking about using the analog part from elecdesign.com but using Jasper's code, at least to begin with. But maybe with some other features...

[The Tone God]

So you don't think it's worth the extra effort? That suites me, I'm rather lazy... 

I was thinking about using the analog part from elecdesign.com but using Jasper's code, at least to begin with. But maybe with some other features...

I tried his analog section and it worked so there is nothing to complain about it. The other version is just a little more fancier but this is digital which allows from some laziness.

Remember that Jesper's code needs to be updated if to be used as is.

Andrew

[Nikolay]

another tuner : strobo tuner
http://www.gtlab.net/gtlab2/filemanager/userpix/diz/strobo_tune.pdf

but file is in russian

[The Tone God]

Meh...

http://www.diystompboxes.com/smfforum/index.php?topic=52811.msg409638#msg409638

Andrew

[BN]

Well, here I am again.

I was doubting myself for awhile whether I was going anywhere with this project or not.
Why such delay? Long story short: Left home -> Bought Mac without RS232 -> Sold stk500 -> Bought AVRISP MkII -> Had to set that thingie up to work with Mac OSX. Phew...

I started to port all those obsolete commands in the code to those now supported by avr-libc and got it to compile without errors. Hence I've decided that my program works and since then I've been experimenting a little bit with the analog input circuits. I took a look at both the simple and the fancy one through a scope and found out that the fancy one seems to be the better. That is if you remove the 100 Hz LP-filter. But I've only had the opportunity to test the simple circuit together with the Mega8 yet. It made the LEDs blink but not in an satisfying way.

From what I saw on the scope (and on the led I hooked up to the output of the circuit via a voltage follower) I would guess that my problem is that the circuit trigger the AVR rather arbitrarily as soon as the guitar signal fades just a little in amplitude. Do you have any other suggestions?

Anyway, I think I'm going to try the fancy version as well and if that won't do the trick I'm thinking about using the built in analog comparator as some kind of "through zero"-detector.

This is my status so far. I'll report back later.

/Björn