Author Topic: Oscilator question  (Read 3673 times)

G. Hoffman

Oscilator question
« on: December 29, 2009, 02:48:55 PM »
I'm working on a MIDI switched looper (THIS ONE), and while I'm waiting for the few parts I'm missing from Mouser, I want to bread board it.  So, I understand it needs a 4MHz oscillator to work with other MIDI devices, but can I substitute a slower one (32.7kHz, or there abouts) for the bread board, just to see if it is working?  Or, can I make some other substitute until Mouser arrives next week?  I've looked at all the local stores, and none of them carry anything but resistors and LED's anymore.


Gabriel

cloudscapes

Re: Oscilator question
« Reply #1 on: December 29, 2009, 03:28:06 PM »
No, unless you can modify the code to make it aware it's being driven by the different value oscillator. And even then, I don't know if 125 times slower will be fast enough to perform the same tasks.

EDIT: I see the source code isn't published, so no. Definatelly no.
« Last Edit: December 29, 2009, 03:30:53 PM by cloudscapes »
~~~~~~~~~~~~~~~~~~~~~~
{DIY blog}
{www.dronecloud.org}

G. Hoffman

Re: Oscillator question
« Reply #2 on: December 29, 2009, 04:38:14 PM »
Well bummer.  What about a transistor oscillator or some such?  Or should I just see if I can raid one of the left over computer parts I have laying around?  Or do I just have to wait?  (I HATE waiting!!!!!)


Gabriel
« Last Edit: December 29, 2009, 04:40:19 PM by G. Hoffman »

ExpAnonColin

Re: Oscilator question
« Reply #3 on: December 29, 2009, 05:26:36 PM »
If you can wire up an oscillator that provides a reliable, square 4mhz clock signal you could use that.  Most oscillator circuits won't have that kind of bandwidth though.  And, at least for AVR's, you need to modify the microcontroller's fuses to tell it to look for an external clock rather than a crystal.

If you hate waiting, pay for next day air!  Or move to silicon valley.

-Colin

PRR

Re: Oscilator question
« Reply #4 on: December 29, 2009, 06:53:22 PM »
Actually, the oscillator is in there, but it needs a 4MHz crystal.

If you had a 4MHz oscillator, instead of a crystal, you could force the internal oscillator. Read PIC16C84A docs.

"Note that you must use a crystal with a frequency of exact 4MHz, otherwise the midi output does not work."

Also: this seems to be the main CPU clock. I do not know the internal technology. If it has -any- dynamic registers, it may run 4MHz or 0.4MHz, but may not "hold a thought" between clocks at 0.04MHz (38KHz). (Long-long ago I had reason to under-under-clock CPUs, and knew which ones would and wouldn't.... )

Also the oscillator for a small ~~4MHz crystal may not work right with a much limper 38KHz crystal. (String the right-side of your piano with double-bass strings....)

But heck. What else are you doing today? Drop the 38K in. Waiiiit. Supposedly the brain will self-test and count 0-9. If this takes 10 seconds at 4MHz, it will take 1,000 seconds at 0.04MHz. It may count 0-9 at one digit a minute, or it may not. Better watch some paint drying to keep boredom at bay.

Are those 2N3906 drawn wrong or what? They may even "work" this way, but it dont look right.

> Or move to silicon valley

I don't think there's any Silicon made in Silicon Valley any more. When I need stuff fast, they air-ship from Singapore.

G. Hoffman

Re: Oscilator question
« Reply #5 on: December 29, 2009, 07:16:07 PM »
If you can wire up an oscillator that provides a reliable, square 4mhz clock signal you could use that.  Most oscillator circuits won't have that kind of bandwidth though.  And, at least for AVR's, you need to modify the microcontroller's fuses to tell it to look for an external clock rather than a crystal.

If you hate waiting, pay for next day air!  Or move to silicon valley.

-Colin



Well, obviously, but I'm poor!  I mean look, I build guitars for a "living".  Money is always an issue here.

I suppose I could move up north to Thief River Falls, and then I could pickup Digi-Key orders on my bicycle!


Gabriel

G. Hoffman

Re: Oscilator question
« Reply #6 on: December 29, 2009, 07:33:56 PM »
Are those 2N3906 drawn wrong or what? They may even "work" this way, but it dont look right.


There are a bunch of issues with the remade schematic and layout.  I'm starting over with the original schematic, and even there I'm changing some things after the 74HC573's - I'm only using four relays for loops, and I'm using a 4016 or a 4066 (depends on what I have on hand) to switch four amp functions (three of which are driven by relays, come to think of it :icon_confused:).  Well, two amp functions, and two for the internal/Leslie speaker switch.

ExpAnonColin

Re: Oscilator question
« Reply #7 on: December 29, 2009, 09:54:34 PM »
> Or move to silicon valley

I don't think there's any Silicon made in Silicon Valley any more. When I need stuff fast, they air-ship from Singapore.

Indeed, there's definitely not, but that's not why I suggested it - silicon valley is one of the few places in the country where most electronics parts are a short car ride away (there or thief river falls, or mansfield texas for mouser)

-Colin