zvex lofi loop junky - hows that work?

[ethrbunny]

Anyone know what goes on in a 20 second looper? Is there anyway to mod a GGG PT-80 or similar to get storage like that? Maybe a schematic somewhere?

[The Tone God]

Its probably based on a voice recorder chip like those used in answering machines or stupied singing fish.

I don't think you can mod a PT-80 to do that.

Andrew

[Peter Snowberg]

I believe the Lo-Fi Loop Junkie is built around a "chipcorder" device from ISD (which was bought by Winbond).

The technology uses what are normally digital EEPROM cells to store analog values. The chip fabrication process is more suited to digital info, thus the Lo-Fi aspect.

http://www.winbond-usa.com/

The best (read as most flexible) control of these chips is done via microcontroller, but I don't know if that's the route that Zachary took with his design.

A PT-80 cannot do such things because the PT-80 is based on a much smaller amount of RAM to save the audio.

[DaveTV]

Based on the specs for the Lo-Fi Loop Junky my guess is he's using the ISD1420 chip. The ISD1400 series of chips are actually very well-suited for easy pushbutton control compared to some of the more recent chips in the ISD family. You could build a pretty simple looping circuit using this chip, but it's the modulation features that give the ZVex pedal its little special something. I've got an ISD1420 lying around that I was hoping to experiment with when I find the time.

[zachary vex]

it's a winbond 1420 with proprietary control circuitry.

[bobbletrox]

http://www.winbond-usa.com/products/isd_products/chipcorder/datasheets/1400/ISD1400.pdf

That's the datasheet.  It has an example circuit (simple!) for connecting a mic and speaker.

[Torchy]

it's a winbond 1420 with proprietary control circuitry.

... and thats what I respect about ZV - answers a question and seeds another to help us along. Cheers guy  :lol:

[ethrbunny]

TY ZVex! I appreciate your response (and all the others too).

[ethrbunny]

BTW ZV - I spent quite a while watching videos on your site last night. You have built some great things!

[The Tone God]

I play with that family of ICs before. I built a more sophicsticated recorder that has four recording banks that can be called up on demand. It has loop as well. Minimal logic used. Very simple.

Andrew

P.S. I wasn't kidding about the singing fish either.

[Samuel]

Yeah you're not kidding, those things are stupid

[ExpAnonColin]

I've played with a 2532 and got mixed results.  I think that the 1400's are more user friendly for us DIYers.  I talked to someone who fooled around with the XCLK pin and the results sounded very interesting.  Once I dig myself out of this hole I will try to get around to trying them again.

-Colin

[The Tone God]

Yeah there are other familes of those ICs which are much better for DIY stuff. I might even do an article on it.

Yeah you're not kidding, those things are stupid

I hacked one of those. They use one of these ICs to store the audio with a power amp driving the speaker. It has motors to articulate the movements like the head, mouth, and tail. Everything is controled and sequenced by a PIC. By reprogramming the PIC and recorder IC you can make it do different "things". hehehe.

Andrew

[Mike Burgundy]

Please do tell. Both on the IC's and the fish

[The Tone God]

Please do tell. Both on the IC's and the fish

While Mr. Vex just use the 20sec version there are other ICs that can hold a lot more time. On some of the familes you can directly access what part of the memory to start at. This is how things like answering machines store multiple messages. Its also how the fish works. They load up all the clips into the IC at the factory and have the PIC call up each clip's starting address. The only tough part is getting the PIC to control the motors in sequence with the audio.

For my recorder I used a 2min version which I broke up into four 30sec banks. I use the same trick of calling the starting address but I use very basic logic stuff. No mircoprocessors although using one could add a processor for some more features.

If theres enough interest I might write an article of how I did things.

Andrew

[Sic]

andrew, thats sounds really really interesting, im sure all of us would love to hit goodwill and buy a cheapfish for like 2 bucks... LOL

i can see the news now "SUDDEN CRAZE STRIKES GOODWILL:: SINGING FISH GOING LIKE HOTCAKES!!"

lol

[The Tone God]

Sometimes I get silly gifts so I get my revenage by hack them. Some of them have even been turned into guitar effects.

Oh it uses a Atmel processor not a PIC. I did this a long time ago with some help so I forget the details. Here are some sites for kicks:

http://entertainment.howstuffworks.com/singing-fish.htm

http://www.mit.edu/~vona/bass/bass-details-billy.html

http://www.mit.edu/~vona/bass/bass.html

Andrew

[ethrbunny]

Question about the 1400 - on the "sample schematic" they show the chip driving a 16 ohm speaker. Does this chip really put out that kind of power? with 5V? ?? Or does this mean it can go from there to a standard audio amplifier? like a cd player or tape deck...

The same schematic shows a microphone in - what would be done differently to use a guitar as the input? "Not much" would be the answer I'd guess at. Both would deliver a similar (unamplified) power level.

per this page: http://www.winbond-usa.com/products/isd_products/chipcorder/datasheets/1400/ISD1400.pdf

[gurutzeta]

I built a pedal with the ISD2560. Here is my web page:

http://humano.ya.com/guruzeta/Guitarra/dirtysam/

It is in spanish (sorry about that), but schematics are in english:
Digital design:
http://humano.ya.com/guruzeta/Guitarra/dirtysam/digital.html
Analog Design:
http://humano.ya.com/guruzeta/Guitarra/dirtysam/analog.html
Power Supply:
http://humano.ya.com/guruzeta/Guitarra/dirtysam/power.html

The schem. is missing a pull down resistor in the inFX.

I planned to translate it, but I am very busy lately

[Benny]

These look cool but... what about the finite life span of the EEPROM storage in these chips?  100,000 write cycles sounds like a lot but that works out to about 23 days of continuous 20-second loops.  Plenty for an answering machine but well under the expected life span of a pedal.  Maybe the write cycle specification is very conservative?  How do they behave as they die, is it a sudden event or does the sound quality degrade gradually as storage cells go bad?